Universidad Politécnica de Madrid

Escuela Técnica Superior de Arquitectura de Madrid

Departamento de Construcción y Tecnología Arquitectónicas

Impresión digital 3D.

Diseño y fabricación digital de superficies
continuas

Tesis Doctoral

Luis Borunda, Architect

Arquitecto

Director de Tesis

Dr. Jesús Anaya Díaz

Doctorado en Construcción y Tecnología Arquitectónicas (R.D.99/2011)

2020
 Tribunal nombrado por el Sr. Rector Magfco. de la Universidad Politécnica de
Madrid, el día...............de.............................de 20....

Presidente:

Vocal:

Vocal:

Vocal:

Secretario:

Suplente:

Suplente:

Realizado el acto de defensa y lectura de la Tesis el día..........de........................de 20….

en la E.T.S.I. /Facultad....................................................

Calificación .....................................................

EL PRESIDENTE LOS VOCALES

EL SECRETARIO
 Universidad Politécnica de Madrid

Escuela Técnica Superior de Arquitectura de Madrid

Departamento de Construcción y Tecnología Arquitectónicas

IMPRESIÓN DIGITAL 3D. DISEÑO Y FABRICACIÓN DIGITAL DE SUPERFICIES CONTINUAS

DIGITAL 3D PRINTING. DESIGN & FABRICATION OF CONTINUOUS SURFACES

Doctoral Thesis by

Luis Borunda, Architect

Thesis Director

Jesús Anaya Díaz, PhD.

2020
 Committee:
Dr Manuel Blanco Lage

Committee Director.
School of Architecture Director, Universidad Politécnica de Madrid, Spain

Dr Maria Josefa Cassinello Plaza

Department of Architecture Construction and Technology, Universidad Politécnica de Madrid, Spain

Dr David López López

Department of Architecture, Universidad Politécnica de Cataluña, Spain

Dr Marta Malé-Alemany

Department of Architecture, AUAS -Amsterdam University of Applied Sciences, Netherlands

Research Stay Director:

Dr Daniel Cardoso

Assistant professor and Co-chair, Graduate Programs, Carnegie Mellon University, USA

International Mention Doctoral Research Reviewers:

Dr Richard Blythe

Dean of the College of Architecture and Urban Studies, Virginia Tech, USA

Dr Marcelo Stamm

School of Architecture, Virginia Tech, USA

Dr Gilberto Velázquez

Universidad Autónoma de Ciudad Juárez, México

Dr Jaime Gómez Gómez

Universidad de Guadalajara, México

 Digital 3d printing
design & fabrication of continuous surfaces

Luis Borunda - December 2020

Summary
Current advances in construction automation and specially in large-scale additive manufacturing
proves the enormous potential for robots in architecture. The construction industry is on the brink of
transformation. Although construction automation is not new, recent advances in computation make
the question of what it means to design and build with the new technology an open call for design-
ers to restructure fundamental aspects of how we think, design and build. Likewise, it is essential
to improve the efficiency in building processes to fulfill the ever-growing demand of resilient and
sustainable building environments, indispensable to achieve the sustainable development goals. Au-
tomation in construction is unique in its potential to reproduce highly complex structures, present-
ing novel prospects on how to confront construction industry challenges society faces today. Novel
design models and fabrication techniques are key to this purpose.

Recent advances in computational design and in robotic manufacturing have proven to increase the
construction efficiency, productivity, and performance of large-scale membranes. To advance one
step into the question of how one turns rapid prototyping techniques into large-scale 3D printing
forms and structures, this research presents computational methods of design and robotic construc-
tion of cellular membranes.

The morphological model is based in a system of forces in equilibrium, inspired in the knowledge of
geometry, material economy, construction rationalization, audacity and many other qualities of shell
designers. This research contributes to the advancement of construction automation and robotic
additive manufacturing offering meso-structure configuration by discrete design methods suitable
for robotic fused deposition of spatial lattices and their large-scale architectural implementation in
automated manufacturing of shell structures.

For the generalization of a design and fabrication method of lightweight fibrous shell structures ca-
pable of being manufactured by robots on-site, first, this research explores how form can be digitally
created by emulating a given static system of forces in space. The computational design methods
and additive manufacture techniques are tested in the construction of complex surfaces.

Secondly, inspired in the complex mechanical behavior of cancellous bone, this research explores
the design space by addition of discrete units that conform continuous trabecular-like lattices, defin-
ing their geometry, limitations, opportunities for optimization and mechanical characteristics.

Finally, this research explores methods of configuration of cellular structures by application of ex-
ternal forces of precompression that extend automated additive manufacturing applications to large-
scale, on-site or prefabricated membrane construction.
 Acknowledgements
Financial support of Consejo Nacional de Ciencia y Tecnología de México research
scholarship, and international research stay support from Consejo Social de la Uni-
versidad Politécnica de Madrid

Academic support of Laboratorio Nacional de Vivienda y Comunidades Sustentables

in Mexico.

Virginia Tech Research Scholar program in Blacksburg, Virginia

Carnegie Mellon University Research Short Term Scholar and Research Collabora-
tion with Manuel Rodriguez Ladrón de Guevara in Pittsburgh

Virginia Tech

D Richard Blythe, Dean of the College of Architecture and Urban Studies, Virginia
Tech

Enric Ruiz-Geli Global Forum on Urban and Regional Resilience

D Marcelo Stamm. School of Architecture

La Caixa Banking Foundation

Mila Moskalenko Cloud 9

Research Stay directors Jeremy Ficca and Dr Daniel Cardoso, Carnegie Mellon
University

School of Architecture Design fabrication Lab and School of Computer Science at

Carnegie Mellon University

Insight, Dr Ramesh Krishnamurti, Dr Daragh Byrne

Josh Bard, Mary-Lou Arscott

Dr Manuel Blanco Lage

Dr María Josefa Cassinello Plaza

Dr David López López

Dr Marta Malé-Alemany

Dr José Luis García del Castillo

Dr Jaime Gómez Gómez

Fabrication laboratories Laboratories

Gianluca Pugliese, LowPoly and WASP Iberia

Daniel Garcia and Ignacio Prieto at Istituto Europeo di Design IED

Research and Demonstration Facility at School of Architecture, Virginia Tech

Thank you

Jesús Anaya your support is the keystone of this work.

Marta Banach for your relentless support, Manuel Rodriguez for our journey ,
Pavel Aguilar for your kind and infinite help. Enric Ruiz’Geli and Olga Subirós,
with whom I learned to imagine. Gilberto Velazquez for your guidance all along this
work. Juan Carlos Sancho, Soledad Madridejos with whom I learned the practice
of architecture. My friends and colleagues at Cloud 9 and SMAO teams. Marcelo
Stamm. Eugenia Muscio, Sebastian Almeida. My family, inspiration, beauty, and
Gloria.

This work was finished thanks to all your support.

 Contents
1 INTRODUCTION................................................................................................................................. 19
1.1 Organization of the Thesis........................................................................................................... 19
1.2 Justification.................................................................................................................................... 21
1.3 Hypothesis........................................................................................................................................ 23
1.4 Goals................................................................................................................................................ 25
1.5 Methodology.................................................................................................................................. 27
1.5.1 Timeline.................................................................................................................................... 29
1.5.2 Expected results based on the methodology..................................................................... 31
1.5.3 Challenges identified............................................................................................................ 33
1.5.4 Research framework.............................................................................................................. 33
1.5.5 Phases breakdown.................................................................................................................. 35
1.6 State of the Art.............................................................................................................................. 41
1.6.1 Digitization of construction............................................................................................... 45
1.6.2 Study on the concept of digital.......................................................................................... 65
1.6.3 Additive manufacturing in architecture............................................................................ 97
1.6.4 Membrane design and optimization.....................................................................................113
1.6.5 Resolution in architecture................................................................................................ 123
1.6.6 Hierarchical differentiation biomimicry model.............................................................. 133
1.6.7 Cellular Solids, Lattice Structures and Fibrous Materials......................................... 137
1.6.8 Computational Thinking for Large scale Additive Manufacturing............................. 145
1.7 Comprehensive morphological principle.................................................................................. 153
1.7.1 Structure of the Thesis............................................................................................................ 157

2 FREE ORIENTED ADDITIVE MANUFACTURING................................................................. 161

2.1 Customizing 3D Print processes................................................................................................. 163
2.1.1 Collaboration at Carnegie Mellon University................................................................ 165
2.1.2 Robotic Additive Manufacturing....................................................................................... 165
2.1.3 Spatial additive Manufacturing...............................................................................................
2.2 Generalization.............................................................................................................................. 171
2.2.1 Materials................................................................................................................................ 173
2.2.2 Computational methods....................................................................................................... 203
2.2.3 Application of spatial additive manufacturing to architectural-scale shells........... 205
2.2.4 Case Study............................................................................................................................. 213
2.3 Case study discussion................................................................................................................... 217
2.3.1 Capacity of automation........................................................................................................ 219
2.3.2 Form efficiency..................................................................................................................... 221
2.3.3 Information handling, optimzation and complexity in computational design............223
3 DISCRETE MODELLING COMPUTATIONAL DESIGN........................................................ 227
3.1 Informational techniques in architecture.............................................................................. 231
3.1.1 Design Scripting tools for particle tectonics of AM..................................................... 233
3.2 Multi-Resolution Based Methodology.................................................................................... 239
3.2.1 Fabrication overview............................................................................................................ 241
3.2.2 Data structure...................................................................................................................... 243
3.2.3 Continuity by concatenation of individual units with distinct properties............... 231
3.3 Ulna bone case study................................................................................................................... 261
3.3.1 A new digital design pipeline based on multi-resolution discrete design method.... 261
3.3.2 Case study discussion.......................................................................................................... 263
3.3.3 Discrete design..................................................................................................................... 265

4 GEOMETRICAL CHARACTERIZATION AND MECHANICAL GRADING..................... 271

4.1 Towards Robotically fabricated composite building structure.
Large-scale fibrous additive manufacturing................................................................................ 275
4.2 Cellular membrane configuration for fabrication................................................................ 281
4.2.1 Informed shell...................................................................................................................... 285
4.2.2 Linear and non-linear characterization tools for structural analysis. ................... 287
4.2.3 Custom Infill Method: from FEM to g-Code................................................................... 289
4.2.4 Cellular membranes printability........................................................................................ 293
4.3 AM construction based in hierarchical biomechanics of bone............................................ 299
4.3.1 Fibrous arrangements applied to large-scale additive manufacturing....................... 301
4.3.2 Density variation, trabecular architecture and material configuration................... 307
4.3.3 Space filling polyhedra....................................................................................................... 323
4.4 Varying the elastic modulus in a continuous extrusion........................................................339
4.4.1 Variable solid fraction........................................................................................................ 341
4.4.2 Variable trabecular architecture properties.................................................................. 343
4.4.3 Variable material properties............................................................................................... 343
4.5 Hierchical membrane: large-scale application....................................................................... 349
4.5.1 CFV Cloud Living Lab 7 Robotic AM................................................................................. 351
4.5.2 Host Centers......................................................................................................................... 351
4.5.3 Hierarchical membrane construction by robotic means................................................ 355
5 PRE-STRESS FORCES, TOWARDS LARGE SCALE ADDITIVE MANUFACTURING

OF CELLULAR MEMBRANES......................................................................................................... 365

5.1 Prestress fibrous structures...................................................................................................... 371
5.1.1 Digital manufacture of large-scale continuous surfaces............................................. 373
5.1.2 Stress, Resolution and Manufacture................................................................................ 375
5.1.3 Change of scale.................................................................................................................... 381
5.1.4 A novel construction model............................................................................................... 385
5.2 Shaping by force........................................................................................................................... 387
5.2.1 Linear tendons in ruled surfaces. Subdomains pre-stress.............................................. 393
5.2.2 Curvilinear tendons in freeform shapes. Principal stress lines definition................ 395
5.2.3 Comparative study................................................................................................................. 395
5.3 Results....................................................................................................................................... 403

6 CONTRIBUTIONS............................................................................................................................. 419
6.1 Conclusion and current limitations......................................................................................... 425
6.2 Future works................................................................................................................................ 427

7 REFERENCES.................................................................................................................................... 432

8 APPENDICES...................................................................................................................................... 515

Spatial additive manufacturing electromechanical signal configure for marlin firmware

Spatial additive manufacturing electromechanical signal configure manual application
Multi resolution membrane discretization method and tessellation target configure
 List of Abbreviations and Acronyms
AM Additive Manufacturing

CAD Computer Aided Design

CAM Computer Aided Manufacturing

CONACYT Consejo Nacional de Ciencia y Tecnología de México

CMU Carnegie Mellon University

CNC Computer Numerical Control

DM Discrete Methods

FDM Fused Deposition Modeling

FEM Finite Element Methods

FFF Fused Filament Fabrication

FOAM Free Oriented Additive Manufacturing

IED Istituto Europeo di Design

LNVCS Laboratorio Nacional de Vivienda y Comunidades Sustentables de México

MIT Massachusetts Institute of Technology

SDG Sustainable Development Goals

STL Stereolithography file format

VT Virginia Tech
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 1 Continuously fabricated membrane concept for shell structure construction proposal.
Virginia Tech CFV Living Lab #7 Cloud project. Architecture by Cloud 9. Render by Playtime
Membrane design and computational configuration for Additive Manufacturing by the author based on this research

16
17
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 2 Elipsoid membrane segment, 3D printed model

Based on discrete architectural design and computation method for additive manufacturing for complex surfaces.
Fabrication technique based in fused deposition modelling
(Borunda, Ladrónd de Guevara, Anaya 2019 CHECKyear)

18
 1. Introduction
Organization of the Thesis

1. Introduction

1.1 Organization of the Thesis

The thesis is organized in 6 chapters compiling research developed in 3 phases:

• State of the technology of construction automation by 3D printing

and studies in continuous surfaces developed in the Universidad Politéc-
nica de Madrid, Spain

• Review and making of computation methods and custom robotic fab-

rication technique developed in a research stay in the Design Fabrication
Lab at Carnegie Mellon University in Pittsburgh, USA along with a col-
laboration with WASP Iberia print lab in Madrid, Spain

• Large-scale application of the hypothesis and methods developed in

this research for on-site robotic construction of a continuous membrane for
the central pavilion of CaixaForum Museum in Valencia, Spain, being cur-
rently under development at the Global Forum for Urban and Regional Re-
silience of the Polytechnic University of Virginia in Blacksburg, Virginia

Concepts and the research framework conditions are presented in chapter 1.

The necessary mechanical conditions and strategies of technique are estab-
lished through the creation of a novel 3D printed technique capable of large-
scale on-site and off-site production of complex shapes, presented in chapter 2.
Chapters 3 through 4 establish computational modelling conditions necessary
to handle design to production workflows by discrete data constructions. Here,
design is presented as curated organizations of data and the parameters of form
configured by data. Chapter 4 establishes algorithmic design principles of me-
chanically differentiated cellular membranes based on results of non-linear
static analysis internal stress distribution and compiles 3D printing studies in
lattice-like geometrical arrangements and parameters capable of constructing
additively structures of both synclastic and anticlastic shapes. Chapter 5 ex-
plores the computation and application of external forces of precompression
as driver of cellular monocoque form design for large-scale continuous and
discontinuous construction. Chapter 6 compiles key takeaways, outcomes, and
contributions.

19
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 3 Nurbs surface segment for UV isocurve transition to discrete design

Own work

20
 1. Introduction
Justification

1.2 Justification
Automated means of construction present new tools for the future challenges
of the industry. The advancements in productivity possible through automa-
tion can positively impact our capacity to build safe resilient and sustainable
environments.

The construction industry is facing an imminent transformation (World Eco-

nomic Forum [WEF] 2016). The digitization of construction is expressly re-
lated to waste reduction, safer work conditions, and mechanized repeatability
of operations and demands the development of novel computational models of
design and fabrication necessary to exploit the potential efficiency and mass
customization capabilities recently available to other industries by automation.

Robotic construction proves enormous potential of increasing the productivity.

Yet, the use of automated means typically remains very limited despite being
greatly reliant in mechanical operations (Gharbia et al., 2020).

Current studies of advancement in construction automation and computational

design of complex structure present significant advances and a growing inter-
est in research fields, due to the computational power and material tools newly
available.

Novel construction techniques based in digital tools (Oxman, R., 2017) are
significantly contributing to the advancement of the field. Technology is shap-
ing and evolving a complex relationship between theory, models of design and
process of fabrication and architecture is a field leading this transformation
(Carpo, 2017).

The relationships between digitization of design, material fabrication coding,

digital fabrication and machine represent both a new knowledge and new skills
(García de Soto et al., July 2018). This research reviews the high performance
of computational designs, and its capacity to interpret multiple scales, from
material composition to large-scale assemblies.

To provide further insight in the transformation of the construction industry

by automated means, the research studies possibilities of creation of synclas-
tic and anticlastic form by three dimensional tessellations produced by spatial
printing with aims to establish parametric principles necessary to configure
large-scale structural membranes designed for robotic additive manufacturing.

Fig. 1. 4 Scale studies of dif-

ferentiated lattices for synclas-
tic membrane segment
Own work

21
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

22
 1. Introduction
Hypothesis

1.3 Hypothesis
The construction of complex shapes in contemporary architectures by rigid
surfaces with complex geometries requires a constructive definition with light-
weight materials, high moldability and high strength, whose configuration as
structural shells can be solved by self-supporting construction elements, pro-
duced with fibrous structures and whose technique allows continuous mem-
brane structures, conceptualizing a comprehensive theory of design and fab-
rication of membranes that has the potential to increase productivity in the
Architecture, Engineering and Construction industry.

Robotic Manufacturing and automation enable unprecedented construction

precision. Its recent advances, combined with computational design meth-
ods, introduce significant innovations in construction technology. The forms
of contemporary architecture of high complexity designed using information
technologies require an algorithmic definition of their geometry, and general-
ization of construction methods for their computational design and additive
manufacturing.

Discrete architectural design and computation for additive manufacturing of

shell structures benefit from the current advances in computational design the-
ory and an ever-increasing computation power. On one hand, it is possible by
current computational power to produce a digitally designed continuous mem-
brane with capacity of automation for which a comprehensive understanding
of fundaments of structural, material, computational design and fabrication
resolutions and optimization is required. On the other, additive manufacturing
applied to the field of construction offers a relevant and wide filed of research.
Additive Manufacturing opens the door to a new generation materiality.

This research explores how the digital era in which we are immersed influences
our construction technique influences the culture of practices in contemporary
architecture. It studies how the digitization of the process of design in com-
bination with the digital transformation of the physical world can extend the
role of the architect to approach current challenges in building construction.
This research aims to define surface conceptualization configuration criteria
based on shape performance, mechanical behavior, constraints inherent to the
fabrication tools, material composition and prestress load capacity. Methods
of discrete architectural design and computation for additive manufacturing of
shell structures based in biomechanics and biomimicry of cortical and trabecu-
lar bone structures are presented.

23
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 5 Concept for on-ste continuous Robotic Aditive Manufacturing. Differentiated spatial tessellations based on
stress performance
Own work

24
 1. Introduction
Goals

1.4 Goals
The main goal of this research is to define algorithmic design methods and ad-
ditive manufacturing techniques for the construction of complex forms whose
geometrical and mechanical characterization allow the large-scale construc-
tion of continuous structural surfaces. Specifically, the research focuses on hi-
erarchical material optimization and geometrical characterization of cellular
structures based in principles of biomechanics.

The research aimed to produce, apply, and evaluate a theoretical, computation-

al, and manufacturing framework of contemporary geometries in architecture.
The objectives of each phase of the research are:

• To design, configure, manufacture and evaluate custom robotic ad-

ditive manufacturing methods to compare different production methods;
to propose a scale change that current construction methods based on con-
crete cannot reach and consider a continuous robotic construction system
if the robot can produce the surface while maintaining a stable structure
during the printing sequence.

• To establish principles for additive manufacturing of complex ligth-

weight membranes defined either mathematically or free form.

• To establish the parameters of modelling space and forms cellularly

to draw three dimensional tessellations and geometries susceptible of addi-
tive manufacturing, exploring the potential of 3D printing in building and
construction.

• To interpret geometry through the process of 3D printing defining a

geometrical and structural typology based in the material and fabrication
systems.

• To review the capacity of algorithmic design to enable the formula-

tion of forces in space that determine a form in equilibrium.

• To identify the necessary parameters to include the mechanical be-

havior information in the process of additively configuring a membrane
structure.

• To examine the methods presented and review the logic of digital de-
sign and manufacture process of stabilization of forces in space that result
in the configuration of a digital form ideated from the internal tensional
distribution system after given loads and the introduction of external forces
of precompression.

25
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 7 Methodological framework

Own elaboration based on (Ladron de Guevara, Borunda, et al. 2019)

26
 1. Introduction
Methodology

1.5 Methodology
With the interest of reviewing the capacity of additive Manufacturing variables
to configure complex free-form surfaces, prefabricated or continuous, the re-
search focuses on biomechanically influenced surface modelling and fabrica-
tion based in trabecular and cortical structures of bone biomechanics to create
shell structures for architecture.

My interest in the field of large-scale membrane construction was first fos-

tered through my Master’s studies in performative architecture at the Kungliga
Tekniska Högskolan School of Architecture, were Professor Ulrika Karlsson
and Professor Marcelyn Gow introduced me to the field of performative de-
sign, endogenous growth in form through geometrical organizations that shape
form by generative design and the tools for parametric design of complex sur-
faces. This experience launched my professional practice with Enric Ruiz-Geli
at Cloud 9 (Ruiz-Geli, Enric and his team at Cloud 9, 2017) in Barcelona
focusing in digital design, smart construction and new materials. Later, in my
participation as architect of Sancho Madridejos Architecture Office I was in-
troduced to the study of exogenous forces that shape and fold surfaces. When
establishing my practice in Madrid, Glitch Design, I leaned towards specializ-
ing in computational design and the technical development of complex forms.

This research is funded by the Consejo Nacional de Ciencia y Tecnología de

México (CONACYT) for the advancement of sustainable construction meth-
ods that can be applied to strategic areas of development in Mexico. This de-
termined the initial decision to analyze emergent technologies of automation
in construction and their applicability in Mexico.

The capacity to build with robots is key to reach the expected increase in pro-
ductivity, sustainability and quality of the built environment possible by auto-
mation. Current studies in construction automation focus significantly on the
emerging field of Additive Manufacturing as a promising option for improving
quality and productivity in the construction industry, while allowing a certain
level of decentralization.

Additive manufacturing in architecture offered two key takeaways that deter-

mined the next steps: move from rapid prototyping to functional additive man-
ufacturing, the capacity of large-scale additive manufacturing and the capacity
of taking instruments out of the controlled parameters of a fabrication lab.

Jesus Anaya, Director of this Thesis, and I determined complex surfaces pro-
vide the ideal field of study to test how digital design, robotic tools and new
materials configure optimized form configurations. His scientific rigor and

27
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 8 Assembly of glass fiber and polyurethan core shere segment surface. Architecture by Cloud 9
Photo by the author
Fig. 1. 9 Glass fiber double curvature retaining wall. Architecture by Cloud 9
Photo by Jordi Alcala

28
 1. Introduction
Methodology

technical expertise in structures, architecture and digital design influenced

every step towards the goal of large-scale functional structures that traces a
continuous criteria and road map starting from the study of geometry, force,
novel tools, and materials towards an original morphological concept and con-
tribution in the field of 3D printing:

To define the methodology, we formulated the following questions:

• What characterizes additive manufacturing forms?

• Which surfaces are susceptible of being manufactured by large-scale

Additive Manufacturing?

1.5.1. Timeline
In 2017 and 2018 we performed experimental tests on 3D printed and mea-
sured their mechanical characteristics:

• An investigation of the links and influence that novel manufactur-

ing techniques have over form, particularly focusing on the relationship
of stress and deformation of complex infills and shapes manufactured by
additive techniques.

• A mechanical analysis of 3D printed trabecular-based tokens in the

Material Analysis Laboratory of the Department of Mechanical Engineer-
ing, Universidad Politécnica de Madrid, Spain.

During 2018 and 2019 we distinguished and studied several types of tessella-
tions and woven-like geometrical organizations. In particular:

• We identified Triply Periodic Minimal Surfaces and Space Filing

Polyhedral arrays as paths to creating isotropic and anisotropic membrane
infills.

• During my research stay and Thesis collaboration in Carnegie Mellon

University in Pittsburgh, USA, I had the opportunity to establish criteria for
building cellular structures with novel additive manufacturing technique
with robotic systems.

• We formulated complex 3D infills based on vector location and orien-

tation, and test their manufacturability in 3 axis gantry systems and 6-7 axis
robotic arms. (Borunda, Luis et al., April 2019).
Fig. 1. 10 Manual fiber and
• A collaboration with the Laboratorio Nacional de Vivienda y Comu- resin application of structural
membrane segments. project by
nidades Sustentables de Mexico, a research lab of the CONACYT network
Cloud 9, photos by Jordi Alcala
was carried out. and by the author

29
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 11 Ellipsoid shape inflatable formwork

Cloud 9,2013. Photo by Jordi Alcala
Fig. 1. 12 Membrane formwork for concrete paste injection
Cloud 9, 2013. Photo by Jordi Alcala
Fig. 2. 1 Pasadena Pavilion
Cloud 9, 2013. Photo by Jordi Alcala

30
 1. Introduction
Methodology

• Large-scale tests of layer-based and spatial printing methods at WASP

Iberia, one of the industry leaders in large scale additive manufacturing,
and at Istituto Europeo di Design in Madrid, Spain.

During 2019 and 2020 we attained knowledge of the geometrical limitations

of fabrication techniques and studied the angular criteria of neighboring units
of tessellations, computational methods, multiresolution algorithm findings.
Results were presented in several conferences (Borunda, Luis et al., 2019; Bo-
runda, Luis et al., April 2019; Borunda, Luis Ricardo et al., 2018; Ladrón de
Guevara et al., May 2019) .

We are currently working towards:

• Classification of the constructability of additive manufacturing shell

membranes over varying modules of elasticity.

• Prestressing technique for large-scale functional structures

• Characteristics and criteria of prefabricated solutions and continuous

on-site fabrication studies.

• Cloud Skin Project, prototype and construction of a large scale cel-

lular membrane based on this reserach.

1.5.2. Expected results based on the methodology

The study of resolution based on computational design and free oriented ad-
ditive manufacturing of discrete fibrous components bonded into continuous
complex forms of variable capacity and locally tailored performance allows to
determine geometrical principles and fabrication constraints of fibrous archi-
tectures.

To verify the hypothesis:

• We develop custom instruments and formulate the principles of the

technical behavior of custom deposition processes. We study and present
digital 3D printing principles for continuous surfaces, namely, the digitiza-
tion of form by discrete methods and the generalization of lattice typolo- Fig. 1. 13 Digitial design and
custom structural analysis of
gies. a complex topology concrete
membrane. Engineering by
• Assuming a theoretical biology model of bone biomechanics and the BAC, project by Cloud 9, 2013
capcity of transferring principles to membrane architecture, we formulate
the parameters that define a hierarchically differentiated tessellations by
Fig. 1. 14 Membrane fabrica-
variations in volume fraction and fiber architecture. We define geometric tion, complex formwork. En-
principles of tessellations that can induce variable modulus of elasticity gineering by BAC, project by
Cloud 9, 2013

31
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 15 Surface principal stress line calculation with custom algorithmic design modeled in Rhino. Stress line computation in Grass-
hopper and Karamba plugin.
Own work based on Stress Line Additive Manufacturing methods (CITAR TAM)

32
 1. Introduction
Methodology

and strength in membranes.

• We were able to conclude the influence of geometry in the mechani-

cal capacity and strength of a cellular membrane, its materialization, which
involves the configurability by 3D printing mechanisms and the digital de-
sign workflow of intricate fibrous arrangements.

• We can formulate the internal structure of shells to dissipate external

loads optimally and introduce precompression mechanisms to extend the
capacity of configuration of large-scale structures. As a conclusion, com-
plex surfaces can be determined by the forces that shape it.

1.5.3. Challenges identified

Early in the research, the question and intuition that determined bounds and
scope of the methodology came from the state of the technology review: how
far in scale and functional structure production can the commercial 3D printer
systems go? Problems raised includeProblems raised by the methodology in-
clude:

• The identification mechanisms that can successfully produce large-

scale functional pieces.

• The verification of the relationships between fibrous arrengement

density, cellular membrane sturt architecture, stress trajectories along the
membrane and overal membrane strenght.

• The geometrical interpretation of the process of deposition.

1.5.4. Research framework

• The concept of fibrous structure designates aligned or randomly posi-

tioned networks of thread or filament buildup and tissue that forms a com-
plex structure.

• The concept of cellular denotes porosity, cellular materials are com-

posed of a solid fraction below 30%.

• The concept of digital 3D print denotes the material configuration of

a digital form additvely concatenating discret in space to make a continu-
ous shape. Generally 3D printing involves an approximation of a shape by
stacking layers, Digital 3D printing involves the physical, explicit imple-
mentation of a digitally configured model.

33
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

34
 1. Introduction
Methodology

1.5.5. Phases breakdown

Phase 1 is centered on the review of various systems in the field of large-scale
additive manufacture, its projection in a global panorama and its direct and
indirect effect on the development of this research. The state of the technology
and review in construction automation of additive manufacturing focused on
visiting research and innovation centers and interview industry experts from
different countries in Europe and in the US.

From this review we reached the conclusion that a first challenge of 3D print-
ing efficiently is to overcome limitations of an originally small scale and slow
process of fabrication.

We center the research in the study of the geometry that can be scalable with
the goal of delimiting the field of experimentation and documentation.

We study different algorithms for the generation of scalable infills triply peri-
odic surfaces, stochastic matrices and branching-type organizations in Rhinoc-
eros 3D software, Rhino script and Grasshopper software.

We implemented linear static and non-linear static analysis workflows. The

static analysis information is used to determine optimal arrangements of mate-
rial and the extensive layer-wise 3D printing testing allows to verify the appli-
cability of the geometrical arrangements in large scale configurations.

The first thing that caught our attention is how complex cellular systems pro-
vide enormous benefit in mechanical strength. We studied the geometrical and
mechanical characteristics of Triply Minimal surfaces studies and Spatial Lat-
tices and set digital design and manufacture principles of membrane with lo-
cally differentiated elastic modulus (Borunda, Luis et al., 2019).

At this stage, it was necessary to create a novel morphological concept and

project basis that centralizes form configuration starting from a static system
of forces in space. This concept is centered in a program that creates a continu-
ous Finite Element Analysis, geometrical configuration, and fabrication code
workflow to successfully 3D print cellular structures and it is based in growth
and biomechanics theoretical models at a large scale.

We approximated the cellular composition in theoretical biology models with

3D fibrous arrangements. We then verified the validity of this approximation
testing the mechanical performance under compression of a series of fibrous
arrangements. The destructive mechanical tests were performed by the Depart-
ment of Mechanical Engineering of the Universidad Politécnica de Madrid, led
by Rafael Claramunt and Marta Muñoz.

35
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 16 Carnegie Mellon Univesrity, School of Architecture Design Fabrication laboratory. Studies in robotic addi-
tive manufacturing with compact extrustion system and custom built End of Arm Tooling.
Research stay and thesis collaboration with Manuel Ladrón de Guevara at Master in Advance Archtiectural Design
program directed by Jeremy Ficca.
(Borunda, Ladrón de Guevara, 2019-D, 6:00)

36
The result of initial large scale 3D printing studies demanded explorations on
continuous membrane forms of anti-buckling patterns with light weight, po-
rous, space filling geometries to reduce fabrication time and improve mechani-
cal behavior. This led to the necessary development of custom instruments,
leading to the development of a custom fabrication technique and computation
method, described in chapter 2-3.

From these studies we conclude a possible path to large scale is to break the
principle of layer-wise 3D printing with robotic arms.

Phase 2 is developed in collaboration with the Master in Advanced Architec-

ture Design program of the School of Architecture of Carnegie Mellon Univer-
sity in joint leadership with the student Manuel Rodriguez Ladrón de Guevara,
Jeremy Ficca, Daniel Cardoso, Mary Lou Ascott and Josh Bard (Ladrón de
Guevara & Ficca, 2018).

The research collaboration involves the development of custom materials for

algorithmic design an instrument of robotic deposition that will allow to verify
the hypothesis of this research. Together, both Universities produce custom ex-
trusion tools to improve manufacturing performance and adaptation of tools to
different robotic manufacturing platforms. We developed new Python-Grass-
hopper Computational Models based in variable resolutions. Furthermore, we
were able to test 3D printing case studies of lattices and published the results
in several papers.

The collaboration with Manuel Rodriguez Ladrón de Guevara at Carnegie

Mellon University results in the concept of ‘Multi-resolution’, a design meth-
odology that scrutinizes direct information from the user to perform a com-
pletely tailored architecture.

Phase 3 involves digital characterization and formulation of principles of cel-

lular membrane construction with 3D printing. We apply the custom instru-
ments and materials to digitally design and computation for robotic additive
manufacturing of fibrous structures.

We measured the influence of force in trabecular arrangements through the

investigation. We were able to comprehend their manufacturability and the
increase in percentage of fibers to accrue resistance. The final aspects of the
research conclude in the necessary studies to prove hypothesis based on the
definition of form as a combination of forces in space and the development of
prestressing concepts for large-scale potential applications of the methods and
techniques presented.

An holistic process flow is designed for additive manufacturing of multi-

37
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 18 Testing filament extrusion accross multiple systems, layer-wise and spatial deposition,
WASP 2040 commercial 3D printer, 3d print custom toolpaths in ABB 6640 robot arm with custom extrusion equipment
Own work in collaboration with Manuel Rodriguez Ladrón de Guevara

38
 1. Introduction
Methodology

resolution surfaces suitable for use as architectural components. This holistic

process is compounded by input parameters, computational morphogenesis
design, topology optimization, FEM analysis, feedback loops, and robotic fab-
rication.

Phase 4 involves a comprehensive group of design-for-fabrication methods to

incorporate additive manufacturing limitations, material qualities, algorithmic
design, and optimization within a continuous workflow is presented.

Finally, the geometry research is centered on the analysis and manufacturing

of a 3D printed pavilion influenced by biomimetic properties of the trabecular
and cortical bone fibrous structure. Based on the initial investigation on how to
scale robotic fabrication processes and maximize material efficiency, 3D print-
ing cellular membranes present the most efficient material outcome coupling
the high level of complexity possible that robotic deposition achieves with the
high mechanical capacity of digitally configured fibrous structures.

The studies of shape by force result in the intricate membranes configurations

of hierarchical material organization. To advance the problem of scale, the fi-
nal studies are centered in the configuration of workflows that induce external
forces of precompression. Those forces configure efficient trabecular arrange-
ments that are capable of dissipating stress in such form that maximizes the
potential mechanical performance of a digital material formulation.

Key findings of this research are applied to the design of a prottype and the
future construction of an on-site robotic fabrication of a closed cellular mem-
brane led by the Global Forum on Urban and Regional Resilience at the Col-
lege of Architecture and Urban Studies, Virginia Tech.

The Cloud Skin will be constructed in the cente of Agora building of the
Museo de las Artes y las Ciencias of Valencia, Spain, as part of the CaixaFo-
rum Valencia museum, designed by Enric Ruiz-Geli - Cloud 9, winner of the
international cometition (Enric ruiz-geli.).

39
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 19 Shamballa Housing proposal for sustainable settlements. Semisinusoidal deposition along circular axis, de-
hydration of clay materials and rice industry waste insulation in cavitites, WASP 2019..
Shamballa project based on polar Infinity crane system
WASP, 2018

40
 1 Introduction
State of the Art

1.6 State of the Art

Additive Manufacturing offers a fabrication method of highly sophisticated
form of unparallel design freedom (Achillas et al., 2015) highly complex me-
chanical behavior (Reynolds et al., Sep 2017) capable of merging data and
material (Willmann et al., 2013) to integrate various functions (Flores et al.,
2019; Loh et al., 2018).

At the architectural scale, additive manufacturing poses a novel production

scenario through the use of robots in construction (Gramazio & Kohler, 2014)
the automation of construction processes (Gharbia et al., 2020) with promis-
ing cost reductions related to distribution, customization, and transport to site.
Robotic manufacturing methods for architecture (Bock, 2008) open interdisci-
plinary research fields around computational design and engineering that de-
mand the identification of the construction elements susceptible of automation
(García de Soto et al., 2018).

In a plausible coming scenario of robotic ubiquity (Bock, 2015; Melenbrink et

al., ; World Economic Forum, 2016) additive manufacturing poses a challeng-
ing research agenda that demands original form design, on-site and prefabri-
cated production methodologies (Buchli et al., 2018; Chea et al., 2020; Chen
et al., 2018; Labonnote et al., 2016; Olsson et al., May 2019; Tay et al., 2017;
Wu, P. et al., 2016).

Additive manufacturing requires the identification of original geometrical pro-

posals that comply with conditions of deposition and with the high mechanical
capacity, incremental growth and assembly capacity, essential for their func-
tional application in contemporary architecture.

To further comprehend the potential and limitations of additive manufacturing

in construction it is first required a parametrizable geometrical definition, to-
gether with its constructability definition. With construction and design being
incrementally digitized (Oxman, R. & Oxman, 2014), numerical tools impact
every level of production of architecture, and that this revolution completely
affects the way in which the building of our time are thought, developed and
made (Malé, 2016).

Additive manufacturing consists of progressive adding of layers of material

through numerical control to build a structure to create a global shape (Ngo et
al., 2018). This process of materializing a given form by placing matter where
there was not before is fundamentally different from forming and subtract-
ing processes (Malé, 2016). Additive manufacturing is ideal for prototyping,

41
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 21 Rolex Learning Center Freeform Concrete Shell

Lausanne. Architect: SANAA, Engineer: Bollinger+Grohmann. Structural Base Concept SAPS / Sasaki and Partners
(Walmer, Knippers 2015)
Fig. 1. 20 Frei Otto, Soap Film Model for the development of the chalice-shaped column, Stuttgart 21, 1997
Reproduced with permission of GTA Archives, ETH Zurich

42
 1 Introduction
State of the Art

especially for manufacturing complex topologies which are otherwise too ex-
pensive or near impossible to achieve with traditional manufacturing methods
(Ngo et al., 2018).

Recent trends have registered that digitization, advanced automation and new
technologies and materials (Gao et al., 2015) are beginning to impact the con-
struction industry demonstrating productivity benefits to the construction of
complex shapes (Hack et al., 2020).

Shell structures composed of double curvature continuous surfaces have long

captured the attention of architects and engineers due to their capacity to work
by shape (Bradshaw et al., 2002), but face significant challenges in their con-
struction as a result of the non-linear nature of their shape. Computational de-
sign and additive construction technologies offer opportunities that reconsider
traditional membrane manufacturing techniques and present novel approaches
to their design, construction, and engineering.

Architects F. Otto, B. Fuller, R. Le Ricolais have shaped membrane structural

morphologies of great beauty and efficiency. Currently, K. Sejima, T. Ito and
M. Sasaki put forth groundbreaking methods of continuous surface parametri-
zation and construction. Bock Research Group is leading the development of
design engineering and construction methods of ultra-thin vaults using novel
digital tools (Shell structures for architecture

form finding and optimization2014; López et al., Sep 2014).

Cellular membranes are very lightweight structures composed 70% or more

of “void”. In R. Le Ricolais’ words “the art of structure is where to put the
wholes”.

Following F. Otto’s view of nature and minimal surface efficiency, B. Fuller’s

discretization of membranes, M. Sasaki’s parametrization of membranes and
stress in Flux Structure Theory, this research converges current developments
in architecture towards robotic construction of stress influenced, lightweight,
membranes.

The state-of-the-art examines:

• Digitization of the construction industry, specifically, additive manu-

facturing applied in large scale robotic construction

• Discrete design methods for a digital stereotomy

• Membrane design and optimization, specifically, non-linear mechani-

cal characterization of cellular membranes

43
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 22 Exponential growth of robotic technologies for on-site building construction

Additive manucafturing is the most frequently rapidly developing technqiue.
Own elaboration based on (Gharbia et al 2020)

44
 1 Introduction
State of the Art

• Cellular Membrane fabrication with spatial additive manufacturing

• Configuration of complex geometric fibrous arrangements based in

biomechanics

• On-site and prefabrication methods of construction automation

1.6.1. Digitization of construction

The digitization of construction involves the automation of manual operations
in construction (Bock, 2015).

Digital fabrication and digital design are some of the key aspects of construc-
tion automation. Current digital fabrication methods can be categorized as
formation, subtraction and additive techniques executed by machines through
coded and repeatable instructions (Malé 2016).

Today, construction automation is reaching the industry (World Economic Fo-

rum, 2016) and shows significant potential for increased productivity thanks to
decades of experimentation.

Digitization in various manufacturing industries has shown great increase in

their productivity in recent years. However, in the case of construction its ra-
tionalization and digitization still require the development of automated and
robotic systems. The construction industry has not yet developed a significant
deployment of robotic media due to its inherent complexity, to variable envi-
ronments, and the uniqueness of each job execution (Melenbrink et al., 2020).

The capacity to build on-site is key to reach the expected increase in productiv-
ity, sustainability, and quality of the build environment possible by automation.
The literature review by Gharbia, Chang-Richards et al. (Gharbia et al., 2020)
finds that the most frequently applied technologies are: additive manufactur-
ing, automated installation, automated robotic assembly, autonomous robotic
assembly and robotic bricklaying, among the most cited.

In situ robotic fabrication, automated concrete spraying, autonomous spraying,

distributed robotic construction, fused filament fabrication, printing technol-
ogy and unevenness recognition, received less attention.
Fig. 1. 23 Topologically op-
In general, the most frequently studied application is additive manufacturing. timized Aeronautical Compo-
nents, AirBus, 2016 (Imprimir
The predicted increase in world population requires rethinking current con- el mundo, 2017).
Parts designed to replace exist-
struction practices to reduce the negative impact derived from construction. ing components. Additive man-
The role of designers in the construction industry is changing thanks to current ufactured allowed to reduce
their weight by 1/3 in compar-
informational tools. sion with mechanized parts.
Photo by the author

45
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

IN-AM-01-Rev. 1
Waste and material optimization

IN-AM-01-Rev. 1
Formwork and complex shapes formation

El refuerzo de los bordes que van a estar en contacto con el suelo se hará con
las piezas fuera de útil.

6.2 Laminado de Piezas.

Prefabrication,
La estructura transport
Las solapasse
de las
hapiezas 1and
y 2 seassembly
dividido pre-taladrarán
en en el taller
cuatro piezas a diámetro 2
simétricas 3 mm
a y2 con la
avellanado para tornillo DIN7991 de diámetro 5 mm. En ambas piezas se taladrarán
numeración que se indica:
las solapas longitudinales con un paso de 120 mm. En la pieza 1 la solapa del plano de
simetría con un paso de 50 mm.

F. Solana 18 11/04/2014

Cada pieza se debe laminar de acuerdo al plano según la tabla de apilado. Se

prestará atención especial a los detalles pues en ellos se referencia la posición de las
solapas que están acotadas con respecto al contorno teórico de pieza dibujado sobre el
útil. Las solapas saldrán de la pieza laminada sobre el útil.

Engineering of a shell structure for the architecture firm Cloud 9, Polyurethan & fiberglass shell,
Fig. 1. 25 Fiber reinforce regions plans
Fig. 1. 26 Dobly curved formworl
Fig. 1. 27 Components assembly and fiber joints of continuous semi-spherical membrane.
Currently the process of construction of complex forms tends to be heavily dependent on artisanal means of industrial-
ized and mechanized parts. The model on which the processes are based is digital and the process of construction un-
digitizes the prodcut, acting very much like a fax machine that destroys the digital information embedded in a digital
documen or email. Oportunities of productivity increase in shell construction by on-site robotic11/04/2014
manufacturing.
F. Solana 17
Florencio Solana, aerospace engineer and Enric Ruiz Geli, architect, 2013

46
 1. Introduction
Methodology

14
12 10 15
04

One of the highest priorities in the sector is delivering sustainable solutions

that face habitat challenges and evolving through the digitization of construc-
tion, commonly referred as Construction 4.0 (Wang, M. et al., 2020).
13
05

09 08

Autonomous construction systems could easily adapt to site-specific condi-

tions and constraint, as well as contribute to sustainable construction in gen- 06

eral. 02

Approaches to the problem of establishing Numerical Control (CNC) algo-

07
11

rithms in robot construction is not something recent (Haas, Skibniewski et al.

transporte a Aiguablava, SES FALUGUES 8,
BEGUR GIRONA.

1995, Warszawski 1984). Yet, some of the latest advances in automation in

01

manufacturing, automotive, and aeronautical industries, alongside the devel-

opment of novel computational design methods, are contributing significantly
to the use of robots in construction (Gramazio & Kohler, 2014; Lynn, Greg &
Foster, 2011; Menges & Ahlquist, 2011; Sheil, 2012).

Advances in industrial robotics that expand the functional capabilities of addi-

tive manufacturing processes (Bhatt et al., 2020) pose a scenario in which the
mass production of personalized designs (Carpo, 2012a) with complex quali-
ties (Dillenburger & Hansmeyer, Jul 2013; Downey, 2012; Mitchell, William
J., 2005) is becoming increasingly viable.

The resulting reduction in construction waste, the significant customization

potentials, the improvement in the ecological footprint, and the improvement
of working conditions in hazardous environments are among the main po-
tentials of construction automation to achieve the Sustainable Development
Goals. (Sachs, 2012)

This research focuses on additive manufacturing of continuous surfaces due to

its potential to build light-weight membranes of high mechanical capacity in
architectural scale and large format. The evaluation of viable applications of
additive manufacturing derives from the decision of applying this technique
to cellular membrane construction, founded on the review of ‘what’ and ‘how’
additive manufacturing is being successfully applied in construction. Fig. 1. 28 Transport evalua-
tion of prefabricated shell seg-
These are some of the most important aspects held in consideration: ments.

• What are the types of robotic additive manufacturing technologies Fig. 1. 29 Prefabrication of fi-
ber glass components.
reported in the literature
Fig. 1. 30 Assembly of fiber
• What are the types off-site and on-site applications of additive manu- glass components. Engineering
and design of complex fibrous
facturing
membrane formwork for con-
crete shell formation,.2013 Au-
• Which building components are susceptible of being constructed with
thor: Florencio Solana, aero-
additive manufacturing effectively space engineer and Enric Ruiz
Geli, architect, 2013.

47
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 31 Diagram of Robotic Construction in Carnegie Mellon Univesity

(Ladrón de Guevara,, 2018)

48
 Introduction
Scope organization and structure of the introduction

• Why build membranes by additive manufacturing

• What benefits could be achieved by using the identified robotic tech-

nologies

• What are the existing challenges and gaps toward their broader adop-
tion in the industry

1.6.1.1. Influence of digital technologies in construction

and design
Digitization and automation have profoundly transformed the way we design.
Interestingly enough, computer aided manufacturing (CAM) preceded com-
puter aided design (CAD) (Cardoso, 2012). Since the advent of CAM, decades
of transformations put construction and design in a mature state of transfor-
mation (World Economic Forum, 2016). The new computer aided tools call
for significant transformations to the pre-digital design workflows, and to the
challenges that designers face.

R. Oxman and R. Oxman (2014) established the theoretical bases and traced
design thinking after evaluating the developing theories of design in the digital
age.

M. Malé highlights several key developments in design practices and digitiza-

tion architecture:

• CAD advancements in architectural representation in the 60s

• New Capacities for Visualization in three dimensions in the 80s

• New Instruments for design Thinking in the 90s

• Current CAM integration for construction

Progressing from digital blueprint production, innovation of the display me-

dium, the advent of fully digital paperless architecture studios, and today the
transformation of the way we build.

Conceptualizing the digital transformation (Carpo, 2012b; Lynn, Greg, 2013a; Fig. 1. 32 Digitized constrution
site.
Negroponte, 1996), architecture has progressively exhibited (Jencks & Kropf, Site laser scan. Figerglass re-
2006; Kipnis, 1993) more complex designs and relationships (Jencks, 1997) taining wall construction in
Casa de la Creatividad. Project
of post-industrial (Lynn, Greg, 1993b; Rifkin, 2015) emergent or bottom up by Cloud 9, scan by Tecmolde,
(Hensel et al., 2004; Johnson, 2002; Weinstock, 2004) generative and paramet- 2013
ric characteristics (Burry, 1996; Schumacher, P., 2019).

Since architects like Greg Lynn or Karl Chu started to design based in the logic Fig. 1. 33 Digital design envi-
of computation, specifically with the current developments of particularization ronment
Own work

49
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

ADDITIVE MANUFACTURING TECHNOLOGIES

Bulk ASTM Category Power Source Example Printed Material

Technologies
Pulverized Powder Bed Fusion Laser Beam Selective Laser Sintering Polyamids
Electron Beam Selective Laser Melting atomized metal
powder
Pulverized Directed Energy Deposi- Laser Beam Laser Engineered Net Molten metal
tion Shaping,
Electronic Beam Welding
Solid-Paste fim Sheet Lamination Laser Beam Laminated Object Manu- Plastic, metal and
facturing ceramic sheets
Liquid VAT Photopolymerisation Ultraviolet Laser Stereolitography P h o t o p o l y m e r,
ceramics
Liquid Material Jetting Thermal Energy Polyjet Photopolymer
Photocuring and wax
Pulverized Binder Jetting Thermal Energy 3D Printing Polymer (plaster,
(Indirect Inkjet Printing) resin), ceramic,
metal powder
Solid - Paste Material Extrusion Thermal Energy Fused Deposition Mod- Te r m o p l a s t i c s ,
eling or Fused Filament ceramics, slurries
Fabrication and Contour and metal pastes
Crafting

TABLE. 1 Additive Manufacturing Technologies

The American Society for Testing and Materials (ASTM) group formulated 7 categories of Additive Manufacturing.
According to ASTM F42 “Additive Manufacturing” refers to the process of adding material layer by layer to an object
to create the whole object. FDM is the technique most commonly used in large scale application processes due to the
capacity and limitations of handling the bulk material and the ability to apply efficiently power sources iunvolved..
Own work based on GAO,. (Male, 2016; Wu, P., et al. 2016 ; Woensel, Oirschot et al 2018)

50
 1. Introduction
Methodology

in design and 3D printing in fabrication, M. Carpo argues that architects and

designers are leading the contemporary understanding of the digital (Carpo,
2017).

Digital design involves data manipulation and the process typically extends
outside a single modeling software and source. The digital “drawing” is com-
posed of code and offers the possibility of indexing, accessing, organizing, and
characterizing the model as well as incorporating information. With informa-
tional tools, such as digital fabrication, “the gap of imaginable and feasible has
reduced” (Mitchell, William, 1999), as demonstrated in the outer skin designed
for Gehry’s Guggenheim in Bilbao.

The Guggenheim and the Yokohama Terminal by. A. Zaera and F. Moussavi
projects are two axiomatic examples of the formal expression of informational
designs. However, their structure is closer to a mere system of trusses.

“It’s just their shapes that are nonlinear, and essentially there is no relation
between the nonlinear analysis and the generation of their structural forms”
(Sasaki, Mutsurō et al., 2007).

In ‘Morphogenesis of Flux Structure’ M. Sasaki presents an entirely different

theory of form.

In our interpretation of the Shape Design by Sensitivity Analysis theory, there

are two key aspects, presented in M. Sasaki’s sequential projects, to which we
have dedicated special interest in this research.

• The membrane of the Kakamigahara Municipal Funeral Hall, de-

signed by T. Ito, M. Sasaki presented a concept based on calculating stabil-
ity by discretizing the membrane in large sections and reinforcing with un-
differentiated orthogonal rebar meshes in the general flowing surface that
create enormous stress concentrations in the boundaries. The need to create
highly customized solid capitels for transfering loads to the ground de-
mands highly complex formworks and requires highly sophisticated forms
and techniques from marine shipbuilding.

• The Rolex Learning Center membrane, designed by K. Sejima and

R. Nishizawa, M. Sasaki presents a digital theory of design that influence
the structure itself, informing shape, performance and materials with new
discrete computational methods. Instead, for the flowing membrane, Sasaki
Fig. 1. 34 3 axis gantry sys-
presents a concept of surface originated in an analytical model constrained
tem. Commercial Fused Depo-
in it’s boundarys. K. Sejima would press and pull varying the height co- sition Modeling 3D printer.
ordinate of a two dimensional array of points until the analytical model Fig. 1. 35 7 axis robotic sys-
tem.
presents a transformation iteration of minimum potential energy iteration Own work

51
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 37 Materials: AARB 6640, Design Fabrication Lab, Carnegie Mellon University
Photo by the author
Fig. 1. 36 Toolpath testing in 3D modeled environment
Own work

52
 1. Introduction
State of the art

ensuring equilibrium. The pull and push of Flux structure Sensitivity Anali-
sis define stress parameters that inform the design of the membrane, and
Sasaki translates this pull into the criteria of prestress. The stress lines are
easily computed so that rebar mesh follow the complex form. The array of
points translates efficiently to the formwork, but still creates an enormous
problem of manufacturability and presents challenging wastes. In design,
the implementation of the Flux theory of the form depends on the digitiza-
tion of the membrane through point sampling. In fabrication, the materi-
alization process required highly complex formworks of highly artisanal
skilled labour.

Flux is a continuous calculation and analysis of the digital form. Similarly, this
research aims to study the geometrical digital configuration of a given form to
propose a non-artisan method of manufacture.

The value of this research is to establish the capacity of the geometric configu-
ration for its material expression.

1.6.1.2. Digital 3D printing

Robots and digital fabrication in architecture offer a possible alternative where
the digital model is directly transferred to a material equivalent advancing the
performative aspect (Kolarevic, 2003; Moussavi & Kubo, 2008) and tecton-
ics (Aranda et al., 2005; Iwamoto, 2009; Reiser & Umemoto, 2006) of digital
design further.

To build robotically drawing plans are redundant, challenging the modern no-
tion and role of the architectural drawing established by L. B. Alberti (Hauck
et al., 2017). Computers iterative logic and generation of solution spaces that
converge in a valid answer seems opposite to the one of the industrial age
based in modern science. Computers logic demands the acceptance of simula-
tion search solutions as true (Carpo, 2019a). To build robotically, designers
Fig. 1. 38 On-ste robotics.
must continue the cross-pollination from aerospace, automotive, boatbuilding Bringing Robotics to the Con-
(Lynn, Greg & Foster, 2011) reaching new skills and presenting new methods struction Site to Manufacture
the Future. Credits: IAAC team
(García de Soto et al., July 2018). (Concept, Coordination, De-
sign, Extrusion, Material &
The substantial effort directed at assessing challenges of automating construc- Sensors), Tecnalia Team, Nou-
tion processes and parts, particularly in 3D printing, highlight the importance mena Team (Drone develop-
ment, Data collection).
of research in the field (Chen et al., 2018; Wu, P. et al., 2016). Considering (Izard, et al., 2017; IAAC,
M. Carpo’s simulation science hypothesis as true, this research aims to build 2017)

upon the notions of Morphogenesis of Flux structure (Sasaki, Mutsurō et al., Fig. 1. 39 3MT 3D printer Pel-
let extruder by WASP. Photo by
2007) and robotic manufacturing advances to present a concept of digital 3D the author

53
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

M104 S200 ; set temperature

G28 ; home all axes
G21 ; set units to millimeters
G90 ; use absolute coordinates
G92 E0 ; reset extruder count
G1 X27.64 Y26.58 E2.049

Fig. 1. 40 gCode script extract; Typical 3D print code.

The extract shows minimum code to operate: setting, heating, move and extrude. CNC operates weaving target coordi-
nates and electromechanical operations
Own work

54
 1 Introduction
State of the Art

printing.

Digital 3D printing infers the existence of a digitized equivalent model to

which the fabrication technique is applied. It infers the existence of intelli-
gible data (Ackoff, Russell L., 1999) in the model to be printed and presents
a continuous method of analytical data processing from raw data into usable
form, into instruction (Ackoff, Russell, 1974) centering form creation in the
electronic and software. It demands the particularization process of structuring
a model into individual indexes of information and the configuration of the
digital pulses that materialize the units of information.

1.6.1.3. Construction automation concepts

R. Ackoff describes the emergence of the post-industrial era (digital) in the
development of three key technologies: capacity to record, capacity to manipu-
late and capacity to transmit symbols electronically, technologies of informa-
tion made possible to mechanize mental work, to automate. “Automation is
what the postindustrial revolution is all about” (Ackoff, Russell, 1974).

Digital fabrication appears with the recent possibility of linking digital designs
with their physical production using computer-controlled tools, descendants
of the first programmable loom that transferred “inscriptions to protocol” with
sequences of punched boards engineered in 1877 by J. M. Jacquard (Cardoso,
2012) and the first Numerical Control Mill developed in 1952 by R. Kregg in
collaboration with MIT (Gershenfeld, 2012).

The robot arm and the development of novel large-scale applications of 3D

printing present new tools for reconsidering the digital design and fabrication
of complex membranes.

3D printing is a challenging but promising new production techniques that has

received a lot of attention in recent years. It is also one of the current focus
points within Construction 4.0, which is a term used to refer to the digitization
of the construction industry. 3D printing is set to revolutionize the construction
industry yielding unparalleled aesthetics, better quality control, lower cost, and
a reduction of the construction time.

The construction of complex parts is becoming more and more feasible. The
use of numerical control mechanisms demonstrates to be an effective way to
increase productivity in construction industry. Additive manufacturing is de-
veloping rapidly and the use of robotic arms and custom 3 axis gantry mecha-
nisms is increasingly being tested. A specialization of deposition mechanisms
is significantly advancing robotic construction field. This research looks into
how to make flexible and efficient use of the numerical control mechanisms

55
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

56
 1 Introduction
State of the Art

with studies in both 6 axis and 3 axis mechanisms. The study of linear axis of
movement for increasing the scale of the deposition field is also studied in this
research, both for prefabrication and on-site applications.

Tools of robotic manufacture are becoming increasingly accessible to design-

ers due to their advances in other industries. Previously, robotic construction
required the specialization of electromechanical systems, very complex sys-
tems for performing specific tasks (Skibniewski, 1996; Warszawski, 1984;
Warszawski & Navon, 1998).

Early large systems of automation like The SMART System by Shimzu Manu-
facturing (Yamazaki & Maeda, 1998) focused on integrating several special-
ized procedures in a single movable or fixed construction platform or “working
floor”. In its introduction in Nagoya, Japan the SMART System displayed sig-
nificant increase in productivity, up to 70% reductions of waste and potential
savings in time of fabrication (Bock, 2008) but were very costly and difficult
to implement.

The automotive industry presents a different solution to automation. The ad-

vent of the robotic arm allows to separate specialized tooling, called ‘End of
Arm Tooling’, from numerical positioning. Prepackaging the inverse kinemat-
ics calculation behind robotic arms movement simplifies the application and
allowed designers to easily use the systems. In a parallel way, the advent of
low-cost 3D printing methods and applications and prepackaging into stan-
dardized firmware of simple scripting tools such as gCode, emerge from the
release of patents of industry precursors, such as Fused Filament Fabrication,
known more commonly as Fused Deposition Modeling, by Stratasys one of the
leading 3D print industries.

The robotic mechanism can be specialized to carry out a single task as was
common for instruments in pioneering systems or divided into a general-pur-
pose mechanism and a separate tool mechanism. Current construction robots
can be classified into robots for a single task and multiple tasks robots based on
the number of construction activities to be performed on-site.

Each type of machine can operate in different code languages, but they have
in common the operation through Digital Outputs and Digital Inputs. These
inputs allow the coordination of the different machines by interpreting low
voltage digital signals to drive specialized mechanisms.

The ‘End of Arm Tooling’, a key aspect in digital design for robotic manufac-
turing enables ability to link electromechanical systems through digital I/O
signals. This allows a processor to organize, detect and act the instructions

57
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

58
 1 Introduction
State of the Art

for the numerical control. Numerical Control (NC) machines combine a mo-
torized manoeuvring tool and often a motorized manoeuvring platform, both
controlled by a computer core, according to specific input instructions.

At present, any type of code that actuate a group of electromagnetic mecha-

nism to perform a set of tasks includes the instructions for all operations in a
linear way.

Industrial CNC routers or consumer 3D printers work essentially in the same

way but with three degrees of freedom for tool positioning. Again, coded in-
structions deliver instructions via low voltage digital I / O to general and spe-
cific systems, a group of motors for position, and a specialized tool for material
handling.

Since any component may require the use of several different tools, modern
machines often combine multiple tools in a single “cell.” A wide range of
unique robotics initiatives are currently being investigated and successfully
applied in construction. Several robotic mechanisms are reviewed. How does
this technology influence contemporary architecture? And particularly, what
kinds of changes does it introduce in the way we design, and we build it?

Digital design tools allow a technical definition of complex surfaces and high-
ly precise robotic manufacturing techniques make possible their construction.
Current practices in construction of non-standard elements can greatly benefit
from the determination of frameworks of robotic construction. This research
brings together computational design, robotics, and architecture for the pur-
pose of developing innovative fabrication processes and new methods of con-
struction of designed and engineered complex surfaces.

Digitally fabricated architecture is planned, evaluated, and optimized through

simulation in virtual environments and feedback loops during the design phase,
understanding construction as an integral part of the design (Willmann, Gram-
azio et al. 2013). Once a digital model is defined, information on geometry and
material grades becomes easily accessible and facilitates the interpretation for
instantiation of specific manufacturing scenarios.

1.6.1.4. Simulation and optimization, current Digital

design for manufacturing practices
New technologies of computer-aided design (CAD) and computer-aided
manufacturing (CAM) are challenging traditional design practices (Menges,
Ahlquist 2011) developing a body of research and practice in digital architec-
ture (Oxman, R., Oxman 2014).

By connecting computational design with manufacturing, CAD is no longer an

59
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

60
 1 Introduction
State of the Art

analog equivalent to “computer-aided drafting”, but becomes a main thrust, a

true computerized “assistant” of design intent (Negroponte 1975) as originally
conceived (Cardoso Llach 2015).

Current digital design practices involve not only a geometric approach to the
digital object as in the previous CAD, but also a true digital construction of
the object (Kolarevic 2003). We can define, simulate, optimize and virtually
constitute a digital concept for its subsequent manufacture digital.

The digitization of design and construction involves the transformation of ana-

log “design and build” drawings for manufacturing and production, into the
implementation of “CAD-CAM” workflows of digital design and automated
production. This digital-based work methodology is fundamentally alien to
industrialization.

The production of architecture through the manipulation of typologies emerges

from the previous industrialization and consequently, requires an update to
incorporate a revolution in the methodology posed by the informational one in
which we are immersed (Rifkin 2014).

The typology raises a scenario of “industrializable” repetition of models (Rossi

1982), the digital model is an algorithm, and it raises a personalization sce-
nario, based on families and instantiations (Lynn, Kelly 1999).

The iterative nature of generative design and the data driven nature of para-
metric design can be conceptualized in iterative simulation process to reach a
design solution space of optimal performance based on goals.

An increasing interest in researchers suggest simulation is as a new type of

science (Carpo, 2017; Chronis et al., 2017; Gunal, 2019) different from the
scientific approach of repeatability of methods, and focused on the logic of
computation based in iteration, fundamentally different from empirical models
with no end results imposed by the designer according to a top down rules.

For performative methods of design, the building performance becomes a prin-

cipal axis of design. Simulation tools, developed for computationally intensive
generative and iterative solution space, create virtually endless possibilities
where the designer draws a continuous flow of decisions, in order to narrow
down the search space towards a detailed solution by learning from prior virtu-
ally developed cases.

On design, systematical and computational thinking evaluate interdependence

between parameters of performance. As the practice of architecture changes,
so do the tools architects employ. At the same time, as new media and digital
tools play a stronger role in design, a deep understanding of the digital process

61
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

62
 1 Introduction
State of the Art

become even more necessary.

1.6.1.5. A transformation of the architect’s role

The informational revolution in which we are immersed is reaching the materi-
al world through digital fabrication and the digital transformation of the physi-
cal world (Gershenfeld 2008). The processor is to the digital revolution what
the steam engine is to the industrial (Gaja I Díaz, 2003). This transformation
raises a series of questions: how does this technology transform contemporary
building practices? What kinds of changes are introduced in the way we think,
develop, and build architecture?

Digital technology has contributed to the emergence of new

roles for the designer according to the nature of his interaction
with the media. Information has become a “new material” for
the designer. These developments are supporting new roles for
designers including the designer as a tool builder. The appli-
cation of computational science and programming languages
make of the designer a “digital toolmaker” (Oxman, R., 2006)

The diagram at the upper part shows at a very basic level, the role of the ar-
chitect until the 21st century. An architect generally plays a centralized role in
the process as a, quote and quote, guru, who by analog ways tries to design in
order to respond to a problem, and the fabrication happens at the end of the
production line. The architect manages the site but does not have the owner-
ship of the fabrication.

Thinking architecture with computational logic has mainly one reason: the in-
troduction of programming languages in design. Architecture uses, for the first
time, tools typical of computer science.

Programming languages involve data and procedural abstractions (Colburn &

Shute, 2007). With programming, the shape can be replaced by data in the
design process, that is we no longer manipulate the geometric representation
of the shape, but data structures where the geometry representation is a conse-
quence of the design of algorithms.

Geometry is used as interpretation of the subjacent data that the computational

designer manipulates, this has a huge impact. For the first time, the architect
can learn to think in computational terms and make algorithmically decisions.

With computational methods the architect can design digital tools that manipu-
late information and procedural models. The digital tool put the designer in a
role of information manager, and flow organizer and positions software as the
basis of the architectural project process. The product of computational design

63
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Representation of a teapot

Blueprint geometric ab- Structured set of samples

straction of a teapot (Ot- follow a consecutive in-
toson 1938) dex order

bSpline mathematical-
representation of outer
perimeter for computer
graphics Digitized teapot. Imple-
mentation of a teapot
model with 145,800
structured samples

NURBS digital repre-

sentation of outer pe-
rimeter

Computational model
Voxel based digial Digital teapot inter-
teapot pretation where each
sample is informed
with data

Fig. 1. 41 Top to bottom: Patent model for teapot, first use of bspline calculation for digitization of tje surface of the
Utah Teapot model by Martin Newel (1975), digitization of the outer perimeter of a Teapot in Rhino by the thesis author,
digitization of the anatomy of a teapot by discrete units “voxels” based on (Male, 2016) Digitization of a teapot model,
from analog abstractions with properties embodied in the model, to representational (formal statement of the model)
towards explicit digitazed models by digital implementation (actual digital construction of the model) .

64
 1 Introduction
State of the Art

methods is therefore reusable, automated, and a way to code architecture. The

capacity of design automation and fabrication provides a novel vision to assess
sustainable solutions to challenges of scarcity and lack of quality of the built
environment (IEAG, 2014; Sachs, 2012).

Additive manufacturing via decentralization poses a different productive sce-

nario for the construction sector. Digitization of construction and manufacture
can increase productivity, and reduce costs related to the distribution and trans-
portation of components to the construction site. The role of architects in this
scenario significantly changes to give way to design by algorithms, capable of
mass customizable instantiation.

1.6.2. Study on the concept of digital

The use of programming languages in design opens previously unexplored
and unworkable territories. Their application in architecture is the product of
the computational technological moment in which we are currently immersed
(Gaja I Díaz, 2003).

In the 90s, thanks to the popularization of Computer Aided Design software,

coupled with the development of Bezier curves, visualization and computa-
tion, and to their capacity to determine curvilinear forms by calculus (Lynn,
Greg, 1993a), the curved and complex architectures where more attainable
than ever before. The growth of digital fabrication by subtraction, addition
and formable techniques, through Computer Aided Manufacture made curves
simpler to produce at no extra cost.

The moment the curve reaches the same status as the line, curvilinearity found
prosperous applications in building construction and curvilinearity began to
formalize a new language. When in the 90s the language of continuity and
seamless gathered all the attention of architectural speculation and research
and established its manufacturing standard at the CNC. Today, the language
of the discrete, of fragments and particles seems to have taken relevance in
the world of research, in part due to the growing popularity of 3D printers and
digital fabrication by additive processes (Carpo 2017).

As expressed by Oxman, R (Oxman, R., 2006) “topological design may be

seen as characterizing the first formal statements to accommodate the new
complexity of non-linear, networked conditions, and to depart from the more
static and topologically deterministic logic and design methodologies of the
previous generation”.

This research builds on the notions of model present in modern and contem-

65
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

66
 1 Introduction
State of the Art

porary architecture:

• Pre digital notion of typology (Rossi et al., 2013)

• Notion of family (Lynn, Greg, 1999) in parametric or generative de-

sign (Burry, 1996; Schumacher, P., 2019)

• Animation in dynamic design trademark of the first digital turn (Car-

po, 2012b) continuous curvilinearity.

This research presents a digital model as an explicit construct of digital con-

figuration based on discrete methods and programming, and uses the cognitive
concept of a model and abstraction of software engineering and computer sci-
ences (Barzel, R., & Barr, A. H. 2013). In a computational view the progres-
sion of a cognitive model is based on the abstraction from implementation to
representation, that is to say, the process of removing physical, spatial or tem-
poral details (Colburn & Shute, 2007) from the concrete and explicit, similar
to a process of generalization.

One main contribution of this research is the introduction of a process that

characterize, conceptualize, and originate a form. This research presents
computational abstract, implicit generalized model similar to the pixel but in
three dimensions. During this process, the designer can create complex forms
through the aggregation of individual units of architectural information. We
called this unit Arxels (presented in chapter 2).

This process of characterization allows the designer to create form explicitly,

not implicitly. Moreover, introduce non-linear and complex information and
relation along the parts that digitally construct a form through an informed
digital stereotomy.

Due to the serialized nature of addition of discrete members, the stereotomy

nature of computational processes as are discrete methods, such as Finite Ele-
ment Methods and Computational Fluid Dynamics, are fit to be fabricated with
numerical control mechanism of additive manufacturing when structured in
ordered matrices.

The architect traditionally designs in a three-dimensional continuous space.

This exploration involves a virtual environment where space is a digital con-
struct, in which we reproduce objects with physical properties. The coordinate
is not only a geometric reference, but also an information index, an ordered set
of data that we can access, with x, y, z coordinates (Ladrón de Guevara et al.
2019). This index can have as much information as the digital tool can handle.
(Sprecher & Ahrens, 2016).

67
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 42 Simulation workflow for positive feedback loop, iterative results towards optimization of
multi-objective parameters
(Aguilar, Borunda Pardal 2019)

68
 1 Introduction
State of the Art

1.6.2.1. Digital form and digital structure

As stated by R. L. Ackoff (Ackoff, Russell L., 1999), organized data creates
information and organized information creates knowledge. The digital frame-
work of design allows the manipulation of abstract symbols to form data, that
is to say create information.

Neil Leach argues that ‘while there is clearly a practice of designing that in-
volves the use of digital tools, there is no product as such that might be de-
scribed as digital’ (Leach, 2015). Digital design and fabrication tools might
have a specific type of design, but they could also be used for objects which
are not ‘digital’ (Jimenez Garcia, Retsin, 2015). However, M. Carpo identifies
the digital character of a design method, arguing for the intrinsically discrete
nature of computational processes (Carpo, 2014).

Another approach is given by MIT professor N. Gershenfeld, who distinguish-

es between analog and digital organizations (Gershenfeld et al., 2014). He
draws a parallel to the way how data is organized. Analog means continuous,
while digital data means discrete.

In an analog system, a piece of matter has infinite connection possibilities,

whereas a digital structures have discrete connections, and void (Gershenfeld
et al., 2014; Jenett et al., 2019) robotic systems demand the application of the
discrete logic of computation to create mechanically continuous surfaces.

The relevance of the theory of architecture and professional practice is, due to
this digitization process, again evaluated in relation to how additive manufac-
turing is applied in architecture. Between precursor theories about this design
paradigm shift is the P. Eisenman’s work “Opposition 6”, which explains the
transfer of a frame humanist theorist based on the opposition “form and func-
tion” to modernism as “displacement of man as the center”, establishing the
bases of study of the evolution of the form in itself, and not as a function of
humanism. Fig. 1. 44 Radiation capture of
solid surfaces.
In continuity with these studies, G. Lynn worked form as “families” and con- Source: (Aguilar, Borunda Par-
dal 2019)
ceptualize the design territory emerging for architects in the 1999 manifesto
“Animate Form”.
Fig. 1. 43 Parametric tools of
Aspects of information, force, motion, and time, which have been long eluded design, Design bounds setting
rules of the design space. In
by architectural description due to their vague essence, they may now be ob- this case, a radius of flexibile
jects of experimentation supplanting traditional tools of accuracy and static toolpath aims to optimize the
spatial structure for creating as
with gradient tools, flexible envelopes, temporal flows, and forces (Lynn 1999).
much solar radiation capture.
Diagrams by the Author based
In this manifesto, G. Lynn describes the idea of informed landscape as a model
on (Aguilar, Borunda, Pardal
2019)

69
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Stimulus Message Message Response

Receptor Control Aparatus Effector

Feedback

search space

Fig. 1. 45 Digital model design by Positive feedback loop

Fig. 1. 46 Locally taylored properties along change in thickness. Section optimized for maximimizing Radiation pen-
etration and air filtration (Aguilar, Borunda, Pardal, 2019).
Fig. 1. 47 Algorithmic implementation based on General System Theory, Simple Feedback Scheme by Ludwig von
Bertalanffy in (Menges, Alquist, 2011).

70
 1 Introduction
State of the Art

for design: a landscape is a site affected by historical energy flows and move-
ment along its surface. These forces are made evident in a form’s geological
development through the movements that have occurred on its surface. The
gradual geological evolution shows how the landscape is informed by effects
of erosion, manifesting the physical representation of a virtual story. These
slow transformation processes result in motion-oriented shapes.

Aspects of force can now be methodically implemented within the origin of

form with simulation tools (Gunal & Karatas, 2019). Performative frameworks
of design present emerging relevance in practice and invite to rethink digitally
advances of the pioneers of complex form construction with space grid struc-
tures (Borrego, 1972) such as the emblematic discretization of the sphere and
construction patent on geodesic structures and laminar geodesic domes by B.
Fuller (Buckminster Fuller, 1965a; Buckminster Fuller, 1965b) or the steel lat-
tice structures of V. Shukov (Edemskaya & Agkathidis, Sep 2016)

Shoei Yoh’s Prefectura Gymnasium project is among the first example of a

membrane structure entirely developed in a digital sense according to G. Lynn
(Lynn, Greg, 2013b) creating a complex discretization lattice based in numeri-
cal sheets (Lynn, Greg, 2013b; Yoh & Lynn, 1993).

B. Kolarevic (Kolarevic, Aug 2001) is among the first to present the emergence
of digital design and manufacture techniques in a framework of performative
architecture that differentiate the notions of parametricisim, algorithmic design
and generative design.

These theories pointed to the opportunity to create a responsive architecture

through the use of digital tools, a fact that had already been demonstrated
physically through the use of models by architects such like F. Otto (Boller &
Schwartz, 2020)and A. Gaudi (Burry, 1996). This research aims to collect in-
formation from analytical model to create mechanically complex performative
designs, who pioneered performative design with the use of analog models and
tools. The configuration of complex, mechanically continuous, performative
membranes shaped by force can be discretely manufactured with advanced ro-
Fig. 1. 48 Functionally Grad-
botic techniques of spatial additive manufacturing. Performative frameworks ed Material, Radiation maxi-
demand the definition of a model of force and structure. mization case: Paste Extrusion
application.
1.6.2.2. Performative Design Fig. 1. 49 Slat 3D model.
By incorporating the material
“Performance oriented design” also called morpho-ecological design is a de- quality constraints in 3D mod-
sign process based on the integral relationship between generation of shape, elling, the virtual prototype de-
picts with a high precision the
material behavior and capacity, manufacture, and assembly. In the case of expected 3D printed outcome.
architecture, it includes environmental modulation and a spatial conditioner Source: Aguilar, Borunda, Par-
dal, 2019

71
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

72
 1 Introduction
State of the Art

arranged in such a way that it provides a heterogeneous space. (Menges and

Hensel, 2013).

The seminal work by F. Moussavi and A. Zaera, Yokohama Port Terminal was
described by M. Sasaki as a combination of industrial based Domino system
by Le Corbusier and V. Shukovs Radio Tower steel lattice, conformed by stan-
dard truss construction system (Sasaki, Mutsurō et al., 2007).

This research aims instead to configure a structure digitally, shape a system of

forces in equilibrium into a form.

Precursors of the digital configuration of a membrane include S. Yoh’s Prefec-

tura Gymnasium digitization of point members by computational calculations.
F. Otto, the only true precursor of parametricism according to P. Schumacher
(Schumacher, P., 2019), and Ove Arup applied for Bundesgartenschau Mul-
tihalle in Mannheim (Addis, 2014) a clever digitization process of the scale-
dependent model of a membrane surveyed through photogrammetry that to
coordinate data from manual and analytical models.

In the case of F. Otto, the study of biological systems’ natural formations and
experiments resulted in constructive descriptions of a variety of structures,
among them “minimal surfaces” forms that were developed with a series of
analogous techniques such as models of hanging cables, stretched textiles and
the use of soap films.

F. Otto’s use and interpretation of models in varying conditions resulted in

early parametric designs solutions.

Digital design tools give rise to new design methods. M. Burry defines para-
metric design as “based on the rules and references governed by geometry and
therefore provide the designer with the syntax to create an unlimited number of
morphologically distinct versions of the same design template” (Burry, 1996).

Parametric tools linked to structural analysis inform the conceptual design and
form definition from inception through optimization (Holzer et al., 2007). Cur-
rent digital manufacture tools allow to extend this optimization to the construc-
tion process and create structures of great complexity.

R . Oxman, (Oxman, R., Sep 2015) defined material design fabrication as a

computational informing process that enhances tectonic relationships between
structure and material within the logic of fabrication technologies. This marks
a transition towards the integration of computationally informed processes in
design. Beyond the first classical integrations of Computer Aided Design and
Manufacturing the continued evolution of digital fabrication media has pro-
duced a new symbiosis between technology and design.

73
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 50 Simulation workflow for positive feedback loop, iterative results towards optimization of multi-objective
parameters
(Aguilar, Borunda Pardal, 2019)
Fig. 1. 51 Iterative design + simulation to fabrication workflow.
Performance and constructability is continuously tested. Visual review of coordinates and numerical control move-
ments for testing printability in NC Viewer software (CITE NCVIEWER)
Diagrams by the Author based on (Aguilar, Borunda, Pardal, 2019)

74
 1 Introduction
State of the Art

The evolution of the field of Material Fabrication Design (Oxman, N. & Louis-
Rosenberg, 2007; Oxman, R., 2005; Oxman, R., Sep 2015) is the result of the
continuous independent development of each three fields: material, fabrica-
tion, and design. The development of computational media and digital tech-
nologies is enabling new processes of design integration and collaboration due
to the growing inter-relationship between design and technology. As a result,
the convergence of architecture, engineering, and materialization technologies
is creating a new material praxis in Architecture. Design is now informed by
structural and materialization processes (Iwamoto, 2009).

The unprecedent freedom of design, the will of exploring form as material

(Picon, 2020) and abstract projections (Bachelard, 1965; Didi-Huberman,
2000; Rajchman, 2000) and the techno-social transformation (Bauman, 2002;
Sterling, 2005) of a post-industrial economy and manufacture turn (Anderson,
2012; Rifkin, 2015; Rosa et al., 2017) shape the expression of architecture in
the digital age.

After the incorporation of computing to the field of design not only new archi-
tectural forms are possible and novel properties become accessible. Synthetic
senses, responsiveness, and morphological adaptability are express degrees of
artificial intelligence. Cloud 9 studio in Barcelona has been incorporating these
tools in their designs in such a way that prototype, case study and final model
are the same. In the project for elBulli Foundation, sensors connected with
trees control the building’s energy use, endowing architecture with meanings
synthetics. Aiguablava Case Study project explores construction by concrete
injection, made possible due to the logic of analysis and parametric design.
The prototype is gradually becoming an expression of architecture, and the
information that an architectural element embraces more complex and adapted
to local requirements.

Digital manufacturing is part of virtually every product that we manufacture,

which entails an adjustment of the method. The chip is to the digital revolu-
tion what the steam engine is to the industrial revolution (Gaja I Díaz, 2003).
The key is to understand the potential of transformation, not a translation of
the same principles of our current system. The primary application of digital
manufacturing in the developed world is manifested as personalized design
to foster what makes us “unique” but in the rest of the world, due to its abil-
ity to agglomerate collective knowledge (Diez, 2012), digital manufacturing
involves the possibility to interpret the complexity of local problems (Gersh-
enfeld, 2012; Gershenfeld et al., 2017). The revolution of mass fabrication into Fig. 1. 52 Digital model de-
sign by Positive feedback loop
mass customization offers a potentially more sustainable and efficient scenario (Aguilar, Borunda Pardal,
2019)

75
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 53 Robotic Fabrication Laboratory , 2010-2016

Research laboratory for large-scale robotic fabrication.
Credits: Gramazio Kohler Research, ETH Zurich. Collaborators: Tobias Bonwetsch (project lead), Michael Lyren-
mann. Selected experts: Bachmann Engineering AG. Selected contractors: Guedel AG, ABB Schweiz AG. Sponsors:
Guedel AG, ABB Schweiz AG
Rreproduced with permission of Gramazio Kohler Research, ETH Zurich, Switzerland

76
 1 Introduction
State of the Art

of production.

1.6.2.3. Mass Fabrication

At the beginning of the 20th century the efforts of mass construction where
largely based on the industrial capacity to produce identical copies.

The application of industrialization in architecture can be found in visionary

projects of the best architects of the last century. B. Fuller designed the vision-
ary projects ‘Dymaxion House’ in 1929 and Dymaxion bathroom in 1926, both
pre-manufactured. The project was never accepted by the market since the
product “lacks interest in the idiosyncrasies of any given context and which is
projected as a prototype for series production” (Framptom 1997)

Efforts of mass fabrication of the 1950s and 1960s prototypes and complete
buildings where largely influenced by developing technologies in compos-
ites. Eames Case Study House, Nagakim Capsule Tower, House of the Future
Monsanto, Alison and Peter Smithson’s House of the Future, Moshe Safdie’s
Habitat 67, among others, are relevant examples that help to understand the po-
tential industrialization of housing as well as the challenges it poses. Their lack
of success can be attributed to poor composite design, but the convergence be-
tween computational power, digital design, and additive manufacturing bring
new attention to mass construction. The digitization of industry is shifting the
principles of mechanical work in mass construction, towards computational
work in mass customization.

(Giedion, 1969) suggests to the power of the machine in modern times yet not
from the point of view of the factory adaptability but discussing the assembly
line. The use of informational tools in construction and design involves infor-
mation processing within assembly line. Post-industrial workflows automate
processes, mental or mechanic (Ackoff, Russell, 1974) .

According to F. Gaja, the societal transformations that rise with new infor-
mational tools can be compared to the transformations of the first industrial
revolution (Gaja I Díaz, 2003). Then, the invention of mechanized work with
steam engines transformed every part of the process of making. C. Shannon’s
information theory, computation microprocessors and automation set the basis
of a new industrial revolution that shows potential to confront the climate crisis
in which we are immersed (Gaja I Díaz, 2003).

Several authors approximate new science of build construction closer to new

way of making of subsequent industrial and post-industrial advances in manu-
facture (Bock, 2008; Bock, 2015; Gershenfeld, 2007; Rifkin, 2015). Going
even further Rifkin (Rifkin, 2015) speculates about new economical scenarios

77
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

[8]
[8]

Fig. 1. 54 New skill in AEC. Data in Architecture and changing roles of Designers for Digital Manufacturing work-
flows (top). Data intensive parametric arcitectural design is coupled with robotic fabrication (below)
Own elaboration based on (Ladrón de Guevara, 2018)

78
 1 Introduction
State of the Art

of zero waste, non-standard, abundance-based economies that will strongly

influence, and to a certain part, transform the way in which we make and con-
sume.

New technologies and methods bring several design challenges among which
stand out the possibility of generating complex geometries resistant to starting
from components and if these geometries are really capable of conforming
differentiated spaces, where identity plays a key role. If the gap technological
is too broad, the new methods are not implemented (Huxley, 2012). The capac-
ity adapts to reproduce locally and be indefinitely edited is a crucial aspect of
mass customization in construction. The inquiry of identity and idiosyncrasy
rises as a parameter applicable to the design. Alienation can occur with in-
novative materials and processes. N. Negroponte illustrates an entirely new
world of the digital with the concepts of multi-media, meta-data, precursors
of streaming, with the fundamental unit of the bit, what emerging, ever-more
digital generations look like, based in immaterial information processed by
chips and transistors replacing the canonical materials of the first and second
industrial revolution steam, steel and concrete.

In our informational society, the digitization of our mechanization processes

that shape lifestyles, ethics and aesthetics are based on information (Talbott,
Apr 2006). Aesthetic complexity and differentiation relate to individual and
cultural constructions (Rosenbauer, 1947); complex aesthetics are a crucial
part of empathetic relationships between people and their built environment.
As a consequence, the implementation of industrialization in housing demands
the management of complex conditions.

Computational tools are increasingly allowing to process data and complexity.

This has a very significant impact in the architectural concept of form. The
model of a type tuple is reinterpreted as iterations of a family (Lynn, Greg,
2008), to produce “objectiles” rather than models, (Cache, 1995) forms based
in relativeness by algorithmic familiarity. Currently, designing without compu-
tational tools is rather unlikely.

The application of new tools proper to the digital age question traditional tools
and aesthetics (Talbott, Apr 2006) as there is an emerging conceptualization
of novel aesthetics. Architect N. Negroponte in the recapitulation of publica-
tions A digital world explains how “each generation that comes will be more
digital” which would mean an experience of the virtual and abstract gradually
less dehumanizing, or in other words, an experience of the virtual and abstract
more human (Negroponte, 1996).

M. Carpo explains that this is because the design professions are now coming

79
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

design

fabrication

simulation

Fig. 1. 55 Rhinoceros 3D / Grasshopper parametric design interface. Diagram of methods for continuous design -
simulation - fabrication workflow.
Own work

80
 1 Introduction
State of the Art

to terms with a new kind of digital tools they have adopted, no longer tools for
making but tools for thinking. In the early 1990s the design professions were
the first to intuit and interpret the new technical logic of the digital age: digital
mass-customization (the use of digital tools to mass-produce variations at no
extra cost) has already changed the way we produce and consume almost ev-
erything, and the same technology applied to commerce at large is now herald-
ing a new society without scale, a flat marginal cost society where bigger mar-
kets will not make anything cheaper. But today, the unprecedented power of
computation also favors a new kind of science where prediction can be based
on sheer information retrieval and form finding by simulation and optimization
can replace deduction from mathematical formulas (Carpo, 2017).

Current construction methods emerged during the industrial revolution, intro-

ducing standardized components of construction made of the new materials
that became available such as metal, and plastics. The recent developments of
new construction techniques and digital tools make possible the serial produc-
tion of non-standard components.

1.6.2.4. Mass Customization

According to (Kieran & Timberlake, 2003) mass customization has its origin
in the computer industries’ proposal to customize to-order personal computers
by components. This illustrates perfectly the potential adaptation of architec-
ture to democratize design through mass customization hence target lacking
infrastructural problems of the build environment (Gaja I Díaz, 2003).

For this reason, large-scale additive manufacturing advances are frequently

directed towards producing inhabitable space, especially aimed towards mass
construction of personalized characteristics (Branch technology 3D prints
building walls with world’s largest freeform 3D printer - launches 3D printed
home competition.2015)

Today, the evolution of 3D printing, the scope and current potential of access
to internet (IEAG, 2014) and the low cost of computerizing processes have
made it possible, through open design, sustainable development is accelerated
through the application of “open source appropriate technologies” (Pearce et
al., 2010; Pearce, 2012), defined as the usefulness of technologies easily and
economically available to communities used to supply needs in accordance
with cultural, economic, educational and environmental constraints.
Fig. 1. 56 Robotics Labora-
Parametric design template introduces the possibility to generate schemes that tory at the University of Cali-
fornia at Los Angeles, USA.
can be instantiated in unique solutions, informed by conditions of each place, Fig. 1. 57 WASP Headquar-
need and economy (Schumacher, E., 2010). To build these forms it is neces- ters in Massa Lombarda, Italy.
Photos by the author.

81
 First Big Data crisis in History

Impresión digital 3d, diseño y fabricación digital de superficies continuas

Digital 3d printing, design and fabrication of continuous surfaces

Today’s Big Data in architecture

[ 40 ]

[7]

Fig. 1. 58 Central role of data along the workflow in iterative processes or computational design
(Ladrón de Guevara, 2018)

82
 1 Introduction
State of the Art

sary to identify parametrizable construction elements, capable of digital fab-

rication, and to make a geometric proposal optimized to meet the assembly
conditions, mechanical capacity and growth capacity necessary for use as a
construction system of building in large-scale.

The 3D printing of construction elements allows the distribution of digitized

architecture processes through peer-to-peer manufacturing so that anyone with
access to information and technologies of additive manufacturing can access
to the design.

K. Kelly (Kelly, 2016) explains the imminent transformation by digitization of

the industry. Similarly, in architecture, we are required to manage this amount
of data (Bellinger et al., 2004) by the use of algorithms that sort, cull and curate
the excess of information. Designers need to take a step back and examine the
data systemically (Ackoff, Russell, 1974).

• Mass customization requires the development of algorithmic models

of design that digitize information in intelligible ways, ordering data struc-
tures.

With digital design processes, distributed automated construction infrastruc-

tures can empower the community, offering tools to materialize architectural
elements.

Material extrusion using CNCs, 3D printers and robotic arms has one point in
common that is relevant to mass customization and decentralization: a code
with coordinates in space and the corresponding instructions can recreate a
given form with the instruments and materials of deposition. The main prob-
lem for implementation in construction identified is that robotic processes are
more easily applied in laboratory conditions, where a method are controlled
and easily repeated.

In contrast, complex large-scale applications such as construction have proven

to be a significant challenge (Kohler 2014). Yet, the body of research and appli-
cation of large-scale construction robotics is increasing exponentially (Chen,
García de Soto et al. 2018).

To the question asked by Discover Magazine in 2003 of what digital fabrica-

tion could be useful for, N. Gershenfeld, director of the Center for Bits and
Atoms at MIT, replied: “there is a surprising need for emerging technologies
in many of the least developed places on the planet. While our needs are rela- Fig. 1. 59 Kamer Maker, large
format 3D printer and detail
tively well resolved, there are billions of people on the planet for whom not. of thermoplastic extrusion sys-
Your problems do not need gradual improvements in current technologies, but tem..
(DUS, 2016).
a revolution.” (Discover Magazine, ) Photos by the author.

83
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 60 3D printed Urban Cabin by DUS architects, Amsterdam, 2016.

Fig. 1. 61 Modular component. DUS canal house layerwise application
Fig. 1. 62 Click system for assembly of 3D printed components
(DUS, 2016). Photos by the author.

84
 1 Introduction
State of the Art

1.6.2.5. Sustainability
According to the World Economic Forum (World Economic Forum, 2016) the
digital revolution of the last half century has made the world a much smaller
place – opening markets, creating new products and goods, and facilitating
rapid industrialization in emerging economies. The pace of innovation, with
improved communication and increased productivity, has been remarkable in
many sectors, including health sciences, education and consumer goods. In the
automobile industry, for example, robotics, computerized design and a host of
other technical and work process innovations have helped to create a global
industry that is now more productive and cost-effective, and increasingly envi-
ronmentally friendly and sustainable.

According to Gharbia et al. (2020) “the need for more affordable housing as
well as social, transportation, and utility infrastructure has placed the industry
under a societal obligation to transform. The industry has vast potential, how-
ever, for improving productivity and efficiency, thanks to the potential offered
by digitization, innovative technologies, and new construction techniques”.

The productivity gains by using industrial robots over the past decades have
helped to open up new application area of robotics. Precursors to robotic pro-
cesses like Maeda (1994) have already managed to reduce up to 70% of con-
struction waste and around 50% of labor on-site (Bock 2008). Waste reduction
is critical to achieving the Sustainable Development Goal.

Since the ecologic impact of digital manufacturing processes are negligible

compared to the footprint of materials processing, any savings in the use of
materials corresponds to a reduction in the overall environmental impact of
construction (Agustí-Juan, Habert 2017). Furthermore, the use of such digital
fabrication techniques shows significant improvements in current construction
practices and the potential to address various Sustainable Development Goals,
particularly on issues related to:

• Efficient use of material

• Safety and specialization in construction

• Management of complex environments

We are around 7,200 million inhabitants, and by 2100 it is estimated that we

will be 11,000 million (Gerland et al., 2014). The construction industry has a
duty to resolve the already pressing challenge of providing housing for the 1.5
billion more inhabitants to come. This is a priority in international political
agendas (Cuchí A, 2014). Precarious is the most common characteristic of the

85
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 63 Pablo López Luz, Vista Aérea de la Ciudad de México, XIII (Aerial View of Mexico
City, XIII), 2006
inkjet print; 23 x 23 in. (58.42 x 58.42 cm)
San Francisco Museum of Modern Art, Accessions Committee Fund purchase
© Pablo López Luz
photograph: Don Ross
Reproduction rigths provided by The San Francisco Museum of Modern Art (SFMOMA)

86
 1 Introduction
State of the Art

build environment.

Besides the homes that will be required, 50% of the world’s population lives
in homes with some kind of deficiency, whether they are houses in marginal
conditions, built with fragile materials, in irregular settlements, in unsanitary
conditions or located in vulnerable areas with limited access to infrastructure,
equipment and services. (Rooms 2016)

The Universal Declaration of Human Rights in the General Assembly of Na-

tions of October 12th, 1948 in its article 25 proclaims that everyone has the
right to an adequate standard of living that ensures you, as well as your family,
health and well-being and especially food, clothing and housing. However, as
of 2014, 880 million people still inhabit urban slums, representing 30% of the
world’s urban population and although is 9% less than in 2000, (SDG Report
2016) as indicated in the Goals of Sustainable Development included in the
new Habitat III urban agenda, and the Paris Agreement on Climate Change, it
is a priority to improve the conditions of habitat of almost a third of urban areas
to make them inclusive, safe, resilient and sustainable.

According to the World Bank’s housing program “Housing for All by 2030”
300 million new homes will be required by 2030 when 93% of adults do not
have access to formal housing finance today. The marginal situation in which
a high percentage of the population lives makes this a fundamental issue in the
action agenda of national and local governments.

The informal city is, if things do not change, the city of the future for the great-
er part of humanity (Gaja 2006). The production of low-cost housing alterna-
tives with materials and technologies produced from a vision of sustainability
to generate affordable, dignified and durable housing is a critical demand of the
construction industry.

1.6.2.6. Potential of Building Automation for Sustainable

Development Goals
To analyze the potential impacts of construction automation on the Sustain-
able Development Goals (SDGs), three significant improvements that would
be made with digital manufacturing over current construction practices were
considered: efficient use of materials, safety and specialization in construction,
and managing complex environments.

For each of these improvements, ODS were associated, with which a greater
link is identified to analyze the opportunities and challenges that arise when the
use of robotic construction technologies increases considerably.

Regarding efficient use of materials, SDG 12 seeks to guarantee sustainable

87
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 64 Lack of quality in the built environment. Sustainable development goal 11. Sustainable cities and commu-
nitities. https://sdgs.un.org/es/goals

88
 1 Introduction
State of the Art

consumption and production patterns (UN 2015). Some goals of this objective
are mentioned below, with opportunities and challenges for the improvement
in the efficiency of the use of materials that come with robotic construction.

SDG Target 12.5 focuses on significantly reducing waste generation through

prevention, reduction, recycling and reuse activities. In this goal, there is the
opportunity to considerably reduce the generation of waste because the robotic
construction process is more precise than manual construction in calculating
the quantities of materials that are going to be required in a given process and
construction times. In each process they are determined more precisely so that
there are fewer losses of materials that are sensitive to the passage of time,
such as concrete.

SDG Target 12.6 encourages companies, especially large and transnational

corporations, to adopt sustainable practices and incorporate sustainability in-
formation into their reporting cycle. Agustí-Juan and Habert (2017) carried
out an environmental evaluation of digital manufacturing and found that the
sustainability of the projects depends mainly on the production of construction
materials Specifically, the impacts of digital manufacturing were negligible
compared to the impacts of manufacturing processes for building materials.
This highlights the importance of considering environmental aspects in the
manufacture of materials to be used in automated construction processes and
saving materials in digital manufacturing projects compared to conventional
construction to reduce environmental impacts on the construction industry.

The most accepted methodology for sustainability reports of construction

processes and in the manufacture of construction materials is the Life Cycle
Analysis (Ding 2014 cited by Agustí-Juan and Habert 2017), so it is recom-
mended to use this methodology when evaluating the environmental impacts
of traditional and automated building systems.

Safety and specialization in construction - SDG 8 promotes inclusive and sus-

tainable economic growth, full and productive employment and decent work
for all (UN 2015). Some goals of this objective are mentioned below, with
opportunities and challenges presented by the improvement in safety and spe-
cialization in construction that are achieved through digital manufacturing in
the construction industry.

SDG Target 8.2 seeks to achieve higher levels of economic productivity through
diversification, technological modernization and innovation, including by fo-
Fig. 1. 65 Sustainable devel-
cusing on high value-added and labor-intensive sectors. The construction in- opment goals potentially im-
dustry is one of the sectors with the highest added value and with an intensive proved by construction auto-
mation. https://sdgs.un.org/es/
use of labor and of the sectors that have had decrease or lack of growth in their goals

89
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

90
 1 Introduction
State of the Art

productivity in recent decades, where other industries have taken advantage of

automation to improve their productivity, such as automotive manufacturing
for example (Bock 2015). García de Soto et al. (2018) found that productivity
is higher when robotic construction methods are used in complex structures,
which allows greater specialization in construction by reducing the need for
more labor for the realization of complex shapes or structures due to function-
ality issues or aesthetics.

SDG Target 8.3 promotes development-oriented policies that support produc-

tive activities, the creation of decent jobs, entrepreneurship, creativity and in-
novation, and encourage the formalization and growth of micro, small and
medium-sized enterprises, including through access to financial services. With
the exponential growth taking place in construction automation research and
field-testing advances, it is expected that there will be a considerable decrease
in the demand for labor in the construction industry. In developing countries,
where construction is one of the labor-intensive sectors (McKinsey Global In-
stitute 2012), personnel training policies should be implemented so that they
can actively participate in robotic construction projects, which is where you
will have more job creation. In developed countries, the construction industry
is not very attractive to young people, which represents a challenge to meet the
demand for labor (Sisson 2018). In some European countries and in Japan, the
aging of the population is causing a deficit of construction workers who require
significant body strength from their workers, making robotic construction a
necessity. The challenges and opportunities of automated construction depend
on the economic, labor and demographic context of each country or region.

Management of complex environments SDG 11, is focused on making cities

and human settlements inclusive, safe, resilient and sustainable. Here are some
of the goals of this objective and how the management of complex robotic con-
struction environments can be applied to improve access to housing and basic
services, as well as to have a more inclusive and sustainable urbanization.

SDG Target 11.1 seeks to ensure access to adequate, safe, and affordable hous-
ing and basic services and to improve slums and goal 11.3 seeks to increase
inclusive and sustainable urbanization. These two goals are related to access
to housing and sustainable urbanization with access to basic services. Urban
environments are complex, so there is no standardized solution for the con-
struction of housing that helps increase the percentage of the population that
has access to decent housing with adequate, safe and affordable basic services,
in addition to the cultural aspects that must be taken into account (Rosenbauer,
1947).

91
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 66 MiniBuilders small robots printing large-scale structures.

Fig. 1. 67 Electromecanic componet follows previosly deposited layers of a dobly curved shell.
Credits: Shihui Jin, Stuart Maggs, Dori Sadan, Cristina Nan, led by IAAC faculty:
Saša Jokić and Petr Novikov. Reproduced with the permission of Dori Sadan.
Fig. 1. 68 Infinity crane 3D printer. Concept for a coordinated group of automated polar mechanisms of deposition.
Fig. 1. 69 Shamballa house prototype, 3D printed with infinity crane, presentation in Massa Lombarda, Italy, 2018
Project by WASP, photo by the author

92
 1 Introduction
State of the Art

Complex shapes by robotic fabrication. Complex large-scale applications such

as construction have proven to be a challenge for digital fabrication (Kohler
2014), but the body of research and application of large-scale construction
robotics is increasing exponentially (Chen, García de Soto et al. 2018) with
which there will be significant progress to reduce the housing deficit in mar-
ginal neighborhoods. Currently there is a boom in technology companies fo-
cused on printing houses in 3D ranging from printing emergency shelters in
less than 30 minutes with special polymers (https://www.ls2n.fr/) printing a
complete house of 55 to 74 square meters in 24 hours with concrete based for-
mulated material (https://www.iconbuild.com/), printing a clay house in less
than 10 days using local construction materials from construction-site (https://
www.3dwasp.com/en/3d-printing-architecture/), up to an entire multi-story
residence or villa in 45 days (http://www.winsun3d.com/) which shows that
the future of automated construction is already shifting from research and de-
velopment to implementation, which is expected to become massive in the
coming years (Carpenter 2019).

Goal 11.5 seeks to reduce the number of deaths, people affected, and economic
losses due to disasters, and goal 11.6 focuses on reducing the environmental
impact of cities. The robotic construction uses materials specially formulated
for each application to be carried out, taking care of the structural aspects and
resistance to possible environmental impacts in addition to the modeling that
is carried out simulating different situations before starting the construction,
so that the houses and buildings that result from robotic construction are more
resilient than those made by self-construction or other means that are economi-
cal for the low-income population, who are the most affected by high-impact
natural phenomena (hurricanes, floods, earthquakes, etc.). In automated con-
struction there is a more efficient use of materials and less waste is generated,
thus significantly reducing the environmental impact of the construction in-
dustry in cities.

At the present time, it continues to be of strategic importance study and as-

sess the advancement of design and manufacturing technologies, keep watch-
ing how the determining factors of architectural production are overcome and
made more flexible traditional and, fundamentally, analyze how manufactur-
Fig. 1. 70 Large Scale Polar
ing developments interact with advances in the field of architectural design CNC mechanism
through convergence and integration of design and manufacturing (Malé 2014) Ceramic extruder
for the conception and implementation of increasingly sustainable architec- Fig. 1. 71 Local clay materials
and rice waste recycling for po-
tures that have a decisive impact on the built environment. lar extruder mixture for Infinity
Crane 3D printer
Construction 4.0 and the digitization of building works pose a paradigm shift. WASP, 2019
Photo by the author

93
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

94
 1 Introduction
State of the Art

The change presents new challenges and opportunities to address with new
computational tools some of the problems related to the ecological crisis in
which we find ourselves immersed. To address the objectives of sustainable
development, it is essential to take advantage of the potential increase in pro-
ductivity in which the digitization of construction affects through the imple-
mentation of new computational design methodologies and new digital manu-
facturing techniques.

Construction automation is an emerging field that encompasses different disci-

plines, and requires discussing the traditional work of the architect to include
in the range of his capabilities some computational tools that address the nature
of digital design, the virtual characterization of the form and the indexing of
the information.

In addition, recycled plastics are increasingly being applied in additively man-

ufactured construction since the development of composites with adequate
fluidity. Using thermoplastic can reduce the amount of cement and associated
CO2 emissions required for a given load with a volume fraction non-cementi-
tious can compound this advantage. The polymeric phase could be in principle
made from recycled material, cutting the carbon footprint (Salazar et al., 2020).

Digital manufacturing and construction automation have a positive impact on

the buildingg process, particularly in terms of the efficient use of material, the
safety and specialization of construction work, and the ability to manipulate
and execute complex geometries. 3 Sustainable Development Goals are identi-
fied which have a significant impact.

● Efficient use of materials - SDG 12: targets 12.5 and 12.6

● Safety and specialization in construction - SDG 8: targets 8.2 and 8.3

● Management of complex environments - SDG 11: targets 11.1 and 11.5

1.6.3. Additive manufacturing in architecture

Additive manufacturing, commonly referred to as 3D printing, represents
a new paradigm of digital manufacturing due to high flexibility to produce
complex geometries at low cost, high functional features and minimal waste
production. Architects and designers widely apply additive manufacturing for
prototype or model production due to its numerous benefits. The last years
have seen many successful efforts in applying additive manufacturing systems
to architectural scales. Pioneering efforts of transferring the technique to con-
struction were presented by R. Buswell (Buswell et al., 2006; Buswell et al.,
Jan 2005), E. Dini (Cesaretti et al., 2014; Gardiner, 2011) and B. Koshnevis

95
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 72 12 meter high gantry system for large scale additive manufacturing, WASP

96
 1 Introduction
State of the Art

(Khoshnevis, 2004; Khoshnevis et al., 2006). Since then, the field of additive
manufacturing in construction is growing exponentially. However, most sys-
tems are based in academic research and are considered experimental.

The possibility of optimizing the deposition of material in additive manufactur-

ing is particularly relevant due to the reduction of material expense, manufac-
turing time and precision, characteristics that make the robotic additive manu-
facturing technique ideal to produce architectural-scale structural surfaces.

In 2009 the American Society for Testing and Materials, ASTM defines Ad-
ditive Manufacturing (AM) as “Process of joining materials to make objects
from models digital 3D, generally layer by layer, contrary to the processes of
subtractive manufacturing. Synonyms: additive manufacturing, additive pro-
cesses, additive techniques, additive layer manufacturing, layer manufacturing
and free-form manufacturing”.

Layer based 3D printing at architectural scales is studied mainly with cementi-

tious materials or straw reinforced mud, FDM. Additive Manufacturing is a
revolutionary technology due to its versatility and flexibility to produce cus-
tom engineered designs and advances in material deposition provide unique
opportunities to produce complex, informed components. It is first required to
define algorithmic construction methods and identify geometries susceptible
of robotically augmented manufacturing.

A recent activity standardization of ASTM and ISO brings these together pro-
cesses in 7 categories. As in other industries (Ngo et al., 2018), the most com-
monly applied in architecture are based on Material Extrusion.

In Material Extrusion, a ductile material is extruded through a nozzle, to form

multilayer models. The research is based on exploring the potential of this low-
cost due process in relation to the resistance obtained.

This type of manufacturing allows to create objects of a complexity geometric

that has no equivalent in the other digital manufacturing processes that do not
apply many of the conditions imposed on subtractive processes and formative
as the angles of milling or folding, respectively (Malé, 2016).

Generally, there is a gantry frame that holds the extrusion mechanism where,
as in commercial 3D printers, the extrusion mechanism is positioned by driv-
ing a set of “stepper” type motors. There are also advances in lighter and more
scalable positioning systems, such as those made up of cables, or custom sys-
tems (Izard et al., 2017; Izard et al., 2018).

The influential material extrusion open-source project RepRap proposes a 3D Fig. 1. 73 Infrastructure as-
sessment in Ciudad Juárez
printer, according to its authors, capable of replicating itself. The open-source Photos by the author

97
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 74 Large scale Gantry CNC system commercialized by D-Shape up to a size of 12x12x10
(Cesaretti, Dini et al. 2014). Printed with permission of Enrico Dini

The Radiolaria, printed using D-Shape SLS printer, the first industrial-scale 3D printer,
(Cesaretti, Dini et al. 2014). Printed with permission of Enrico Dini

98
 1 Introduction
State of the Art

3D printer enables enthusiasts and professionals improve the current perfor-

mance of this technology through the process of ‘crowdsourcing design’ where
a community openly participates in the development and improvement of the
technique.

This has, for example, made it possible to transform a commercial system into
a mobile housing component fabrication system called Kamer Maker from
Dutch “room maker”. In essence it uses the same mechanisms. It has been
designed and manufactured with the intention of printing the first house. Ac-
cording to the authors, the Kamer Maker may produce a cubic meter segment
of a house that will be used as formwork less than one day. The forms produced
with standard thermoplastics can be easily recycled if any deformations during
print happen. This 3D printer basically enlarges the fused deposition modeling
in scale and pioneers the use of printed components on-site. Still the applica-
tion remains challenging for functional components production that require no
other material than the deposited material itself.

At the architectural meso and macro scales a set of different strategies is re-
quired to create large structures, where printing time and materials are critical.
As the construction scale increases, increases the volume of material flow and
forces the design into a new process. (Buswell et al., 2007)

Also, the application of large-scale 3D printing to housing manufacturing has

received particular interest. The Italian company WASP (WASP, ) has recently
published a clay wall 3D printing system, developed for the mass production
of low-cost housing with a reduced ecological footprint.

In WASP’s Shamballa project, the extrusion system is powered by a pump

whose power reaches the maximum value of 800 kW. It is associated with a
pressure sensor to monitor the incoming flow.

A curved planimetric shape has been given to the structure which improves
structural resistance. WASP house prototype exemplifies the potentials of local
materials use, zero waste, decentralized manufacturing system, and complex
performative geometries. The geometry of the house is highly performative:

• The cross section of the wall is, instead, alveolar formed by an inter-
nal circumference, two sinusoidal curves and two cosinusoidal curves with Fig. 1. 75 New mayerials and
axes of the circular curves. The first sinusoid and the first cosinusoid have techniques open exploration to
new geometries
the same amplitude and frequency and share the same axis, the radius of Own work
which is greater than that of the internal circumference by an amount equal
to the semi-amplitude of the two curves.
Fig. 1. 100 Ceramic 3D print-
ed gyroid
• The second sinusoid and the second cosinusoid have the same fre-
Own work

99
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 77 On-ste robotics. Cable deposition system presented at Construmat 2017, Barcelona.
Fig. 1. 78 On-site robotics. Wall detail. Reproduced with the permission of Noumena
(IAAC, 2017) IAAC team (Concept, Coordination, Design, Extrusion, Material & Sensors), Tecnalia Team (Cable Ro-
bot) , Noumena Team (Drone development, Data collection).
Fig. 1. 76 3D printed complex surface by discrete plastic componetns.
(Imprimir el mundo, 2017). Photo by the author

100
 1 Introduction
State of the Art

quency as the first two curves, but a lower amplitude; the axis shared by
the two curves has a circular shape with a radius equal to the radius of the
first axis increased by the semi-amplitude of the axis of the second pair of
curves. The sinusoidal trend of the curves is also present longitudinally,
made by varying the amplitude of the planimetric sinusoids along the verti-
cal axis.

• The overtures created in between sinusoidal waves are filled with

insulating local rice industry waste. The overtures and air circulation cre-
ates channels of ventillation that allow the structure to dehidrate to solidify
faster. Internal cavities can also create better thermal performance (Chronis
et al., 2017).

Most large-scale 3D printers have Marlin-type firmware, which provides a

common computational language and can print from gCode files, or also from
stereolithography “.stl” files, processing in this case with cutting software such
as SLic3r, Cura, Simplify 3d. The layer-by-layer manufacturing paradigm, due
to its similarity to 3D printing, is efficient and easy to implement, other custom
printing mechanisms however require a more sophisticated extrusion system
movement. These systems require more complex space positioning and gener-
ally require custom gCode writing.

The main constraints of additive manufacturing of architectural elements are

that of scalability and complexity of the construction environment. Arguably,
the future of construction might rely on the ubiquity of robots and automated
processes. Workflows for the intricacy of the building construction process
must be developed.

Coupled with more productive construction workflows, the uncanny preci-

sion of the robot affords for contemporary form to deal with great mechanical
complexity, locally differentiated architectural qualities and locally engineered
performance conceivable through information technologies and computational
design.

Robotic Additive Manufacturing presents unique capabilities of realization of

such designs. Fig. 1. 79 Prototypying scale
extruder.
In general, a major limitation of applying FDM technologies to construction Fig. 1. 80 Kamer Maker Large
automation is the time it takes to produce large-scale builds. Its application in scale extruder. DUS architects.
Fig. 1. 81 Larges-scale 3D
architecture has been quickly growing (Bock, 2015; Malé, 2016) since the pub- printing thickened layer. DUS:
lic release of patents of additive manufacturing techniques (Ngo et al., 2018). FDM systems based on same
power source principles, elec-
tromechanical and manufac-
ture. (DUS, 2020)
Photos by the author

101
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 82 Trabeculae Pavilion is a 3D printed lattice structure at construction scale.

3D printed with a farm of WASP printers, assembled with hidden joint. Bio-inspired geometry and design
(Naboni 2019)
Photograph: Gabriele Seghizzi
Reproduced with permission of Roberto Naboni

102
 1 Introduction
State of the Art

1.6.3.1. Fabrication Techniques: current areas of

advancement
Although robotic additive manufacturing originally used concrete as main
printing material, technical advances are enabling the application of other ma-
terials such as thermoplastics at the building scale. reviews seminal studies
of 3D deposition of thermoplastics as a base material. Not limited by com-
mon properties of thermoplastics, they present geometrical and computational
methods to produce large-scale membranes.

M. Malé (Malé, 2016) and other projects from the Institute of Advanced Archi-
tecture of Catalonia ( IAAC ) show the potential of performative and digital
design, extending the limits of the design space and moving towards a terri-
tory of post-parametricisim, including generative and morphological design,
and digital tectonics of architectural forms. A group of students and faculty
at IAAC (Maxim Tomash, 2014; The Institute for Advanced Architecture of
Catalonia, (IAAC), ) presented an elegant and clever mobile solution for the
formation of continuous layers by mini builders soft angular movements in
a clamp-like mechanism. Robotic locomotive systems are a current area of
research, the systems demonstrates scalability of robotic collaboration (Asfari
et al., Apr 2018; Feng, Z. et al., 2020; Kayser et al., Aug 2018; Yuan, P. & Yan,
2020). Zhang, Li et al. (Zhang et al., 2018) Present the automated collaborative
3D printing of a segment wall built by a group of synchronized mobile robot
arms.

Moving towards mechanisms of production of large components, this research

collaborated with WASP Iberia to define important steps to scaling solutions.
WASP Iberia has manufactured some of the most advanced deposition mecha-
nisms for large-scale additive manufacturing. They also focus on challenges
of habitat, sustainability and renewal materials, providing new tools to review
ever standing challenges such as the lack of building construction.

Large scale effective and fast printing is not a trivial task. The study of geom-
etry, porosity, void and continuity is fundamental. Current research is widely
focused on layer-wise techniques, optimizing mechanical performance, such
as in stress informed additive manufacturing techniques (Naboni et al., 2019;
Naboni et al., 2020).

However, advances in non-layer wise techniques that involve robotic deposi-

tion are increasingly being developed (Wu, P. et al., 2016). Cheung (Cheung,
2012) presents complex tessellation arrangements of variable orientations that
conform digital cellular solids using a non-layer wise technique.

103
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Structural analysis and verification of the overall lattice structure: up (Sofistik -Von Mises stresses for a dead load.
Image source: Valentin Koslowski; down (Ansys) -a) Structure's boundary conditions; b) Max stresses in frames
(axial+flexural) (KPa); c) Flexural over total stress ratio. Image source: Francesco Marelli

104
 1 Introduction
State of the Art

A digital form is explicitly personalised for specific functions, increasing po-

tential optimization, applying materials in the most efficient configuration pos-
sible, therefore reducing waste. This research studies the digital design itself,
in optimal geometric formulations that embed fabrication constraints, to fur-
ther optimize material distribution and robotic toolpath production.

The reductio of material use not only increases the performance of a compo-
nent, optimizing material use can reduce the time required for production. This
is as the principal area of opportunity for significant improvement.

The capacity to structure the matrix of indices that define the internal constitu-
ents of a form harnesses the potential of informing the deposition process with
a locally differentiated digital characterization, enabling direct incorporation
of performance, simulation and feedback loop results into the manufacture of
form at the material level (Fleck et al., 2010).

However, robust computational methods specific to additive manufacturing

techniques at a meso- to large-scale are not limited to optimization and en-
gineering roles. The focus of this research builds on current computational
methods and fabrication techniques for the manufacture of functional graded
form (Aguilar et al., 2020), founded in a performative and aesthetic inquiry of
form design and materialization.

With the advent of computation and simulation technologies (Moreno & Moli-
na, 2019; Tankova & Da Silva, 2020), current research shifted towards topolo-
gy optimization, on-site applications, with the help of advanced computational
tools. Computational power is changing drastically the focus, limitations, and
interest of additive manufacturing in architecture.

Lattice structure of thin struts tessellations are very difficult to print with stan-
dard methods.

Robotic deposition requires novel computational design methods and fabri-

cation techniques impulse the emerging body of research of Spatial additive
manufacturing and wireprint (Mueller et al., Oct 2014) in architecture.

The interest in spatial additive manufacturing stems from its capacity to create
very light structures. Recently Branch Technology (Branch Technology, 2020)
presented CFAB cellular construction automation processes and applications
of prefabricated spatial additive manufacturing and concrete composites in-
tended for complex surfaces formations. Their patented technology, Additive
manufacturing of building and other structures (Boyd, 2019), is being applied
in largescale membrane construction, which verifies the application of poly-
mer tessellations as functional structures (Pasquarelli et al., Sep 2017). Im-

105
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 4. 1 Architectural plans and Bim modeling of complex spatial structure.

Architect: S-MAO Architecture Office; Structural Engineer; Alejandro Bernabeu; Construction Supervision; Cosmos
Engineering & Consultants. El Cairo, 2017. (S-MAO, 2020). Reproduced with permission of S-MAO

106
 1 Introduction
State of the Art

proving construction and deposition time through topology optimization, ma-

terial economy and robotic collaboration fabrication methods for on-site and
prefabricated is key to successful implementation in the industry.

This research moves towards building-scale 3D printing upon knowledge in

spatial additive manufacturing, a subfield that studies the optimization of print-
ing processes by discretizing areas of interest (Mueller et al., Oct 2014) while
isolating a main principle of 3D printing: stacking layers. Universities and
industry are exploring this potential extensively.

1.6.3.2. Digital design of spatial structures

Spatial frames and structures are of interest to this research because of their
capacity to configure continuous complex forms with individual segments.
Governing geometrical rules of tessellations capable of producing complex
mechanical behaviors are studied. Certain key developments and projects
of architecture that span from the beginning of the 20th century can provide
principles for the discretization and the parametric definition of membranes.
Biomimicry based structures and industrial based structures alike present out-
standing mechanical properties. Current computational tools allow to optimize
further and extend such structures to a level of unprecedent complexity with
the use of additive manufacturing.

The geometrical characteristics of spatial structures determine their strength,

structural engineering advances have influenced theoretical biology advances
and vice versa. Geometric principles and their mechanical qualities, due to
their abstract nature, can be transferred across applications and scales (Addis,
2014).

Mainly present in architecture through Platonic solids until the end of the 20th
century when it took on value in terms of the envelope, Space grid structures
(Borrego, 1972), structured, periodically repeated, or individually determined
can be efficiently differentiated to perform optimally, varying strut thickness
and orientation while working mechanically continuous.

The papers of Fuller and Le Ricolais before and after the Second World War
were fundamental for the reflection and understanding of repetitive spatial
structures. In 1951 B. Fuller developed the geodesic dome, which he patent-
ed, based on research on the projection of a hexagonal mesh onto a sphere.
What led to a great advance in the conception of spatial structures (Schueller
& Heck, 1983).

Current architecture practice works increasingly towards more variable spa-

tial structures construction based in complex parametric models (Weinstock,

107
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 83 Cultural Center Alburouj, detail of spatial frame structure.

The lattice design is differentiated to optimize materials based on stress. The lattice was welded on-site following the
3D design and drawings plans specifiying different types of joints. The concept design required a parametric workflow
crossing multiple platforms interpreted with a polygonal array of 3D polylines: Rhinoceros 3D for lattice design,
Grasshopper and Geometry Gym plugin for building a BIM Model, TEKLA for BIM analysis and structural engineer-
ing. Architect: S-MAO Architecture Office; Structural Engineer; Alejandro Bernabeu; Construction Supervision; Cos-
mos Engineering & Consultants. El Cairo, 2017. (S-MAO, 2020). Reproduced with permission of S-MAO

108
 1 Introduction
State of the Art

2004). New computational design methods allow to control space structures at

unprecedent resolution.

When designing the geometrical constitution of complex lattices, a fiber net-

work is redefined working together and forming a larger and complex system
that responds to a geometric pattern. It is what could be defined as an artificial
arrangement of fibers

Patterns appear in all-natural systems and the frequency and occurrence of cer-
tain geometric patterns (particularly triangles, pentagons, and spirals) in many
different organizations across vastly divergent scales (Weinstock, 2004). These
patterns are generic; therefore, it can be said that the hierarchical ecologies are
essentially geometric.

In nature, there is a morphogenetic search for efficiency reducing the material

only to what is necessary as well as burning the optimal sections (Gibson, Lor-
na J., 2012) resulting in a mechanical optimum. The search results in geometri-
cal arrangements whose section adapts to stress, which can be locally different,
but all are capable of resisting the same stress to which they are subjected.
The forms of nature are determined as a relationship between physical laws
and the most optimal geometry for the efforts to which it is subjected. F. Otto
introduced the concept of “Form Finding” for example in his Munich Olympic
Arena, the production of physical models was required to engineer the mem-
brane geometry to produce a stable tensional system (Otto & Rasch, 1996).

Look in back to the past century, there were various visionary designers such
as A. Gaudi, H. Isler, F. Otto, F. Candela and V. Shukhov using design meth-
ods comparable to today’s computational design, operating in a precomputa-
tion parametric set-up (Addis, 2014; Bletzinger & Ekkehard, 2014; Boller &
Schwartz, 2020). They applied mathematical algorithms, physical models and
principles observed in nature to develop pioneering buildings.

A new stage of lattice structure development is presented in Finding Form Fig. 4. 2 BIM modeling in
(Schumacher, P., 2019) where Frei Otto explains his way of modelling and TEKLA software.
Fig. 4. 3 Complex spatial lat-
calculating grid surfaces (Winslow 2014). Otto’s writings are particularly in- tice for outdoor stage.
teresting because he started working with lattice structures at the dawn of the Differentiatoin of members to
recreate a spherical, hyperbo-
era of computational engineering and was one of the first to have collaborated loid and eliptic spatial struc-
with computational designers in realizing complex lattice structure projects. ture.
Architect: S-MAO Architecture
As a result of the use in recent decade of computational technologies by en- Office; Structural Engineer;
Alejandro Bernabeu; Con-
gineers and architects becoming increasingly common. Traditionally the in- struction Supervision; Cosmos
dustrialization of units and struts pre-determined grid-shell and spatial frames Engineering & Consultants. El
Cairo, 2017. (S-MAO, 2020).
shape. Computationally driven grid-shell structures methods of design and Reproduced with permission of
S-MAO

109
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 84 Frei Otto, Institute for Leightweight Structures, Stuttgart, 1964.

Reproduced with permission ot GTA Archives, ETH Zürich.

110
 1 Introduction
State of the Art

engineering have been developed allowing formal freedom and infinite geo-
metric complexity (Pasquarelli et al., Sep 2017; Winslow, 2014). At the same
time, mass customization is making non identical construction units more af-
fordable. The construction cost of high complexity projects is still consider-
ably higher than conventional projects (García de Soto et al., 2018; Wu, P. et
al., 2016)

Additive manufacturing is already changing the model of design and construc-

tion of contemporary architecture at all levels. The quality and Accessibility to
this technology is progressing rapidly. It is in the hands of the designer Take
advantage of the potential that additive manufacturing offers.

Membranes can conform highly efficient structures, nonetheless, additive man-

ufacturing brings opportunities to increase strength and reduce weight by mak-
ing possible the construction of more complex forms, like patterned surfaces
(Rivera, 2020). With additive processes we can reinterpret and build grid-shell
and shell design at the level of the individual member and three dimensionally.

This research builds upon the science of mathematical shape and form finding
design of shells to present a computational based cellular membrane design
that are free form or parametric, mechanically continuous through a pixelated
gradation controlling each member that conforms a cellular structure. Grid-
shell membrane design and optimization (Linkwitz, 2014) can apply additive
manufacturing technologies to produce very complex and highly performative
complex cellular structures (Boyd, 2019; Cheung, 2012).

1.6.4. Membrane design and optimization

Recent advances in computational design and construction automation exem-
plify potential application of spatial additive manufacturing in the production
of digitally designed complex surface and membranes (Ney & Adriaenssens,
2014).

As stated by M. Sasaki “ In ordinary structural analysis, the structural frame

is assigned some kind of shape with predetermined coordinates, and then it
follows the process of calculating the deformation and stress. By contrast, in
shape analysis, the target values for stress and deformation are used to define
the most desirable structural form. It could be called inverse analysis, that is Fig. 4. 4 Dome Construction
spherical shape based on Paten
the opposite of ordinary analysis”. (Sasaki, Mutsurō et al., 2007) Number (Fuller, 1965).
Fig. 4. 5 Standard Joint. 7 dif-
This research interprets the contributions in Kakamigahara Municipal Funeral ferent types of members where
Hall and Rolex Learning Center projects as cases to improve the materializa- used in the construction. Con-
struction by Pacific Domes,
tion of complex membranes (Sasaki, Mutsuro, 2014). 2020.
Photos by the Author

111
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Interpretation of Sensitivity Analysis in this

research:
1. External forces are applied in search of
maximum potential energy configuration.
2. This can inform the position of optimum
pre-compresssion
3. The method relies on manipulations in z
axis, so the boundary in x and y axis is limited.
4. Final z coordiante can define the membrane
formwork location in a distributed network of
points in space

Fig. 1. 85 Modification of shape in Theory: Shape Design By Sensitiviy Analysis and the 3D Extended ESO Method.
(Sasaki, et al. 2007, p.104)
Fig. 1. 86 Analysing the Algoritht in Theory: Shape Design By Sensitiviy Analysis and the 3D Extended ESO
(Sasaki, et at.,2007, p. 105)

112
 1 Introduction
State of the Art

• In Kakamigahara Municipal Funeral Hall, the surface is discretized

as a group of plate, makes the structural behavior be conceived, modelled,
fabricated and structurally perform in local segments. This allows the
structural steel mesh reinforcement to not be a complex reinforcement but
a simple steel mesh. The more difficult part occurs in the structural joint
where many steel rods converge and merge into a pillar with a sense of
continuity. The geometry itself can only be built by marine carpenters due
to its inherent complexity. Additive manufacturing becomes thus an opti-
mal source for fabricating (García de Soto et al., 2018). The density of steel
mesh can be achieved by more or less traditional and craftmanship means,
so Sasaki presents a different solution for the knots, based on a more solid
concrete solutions.

The discretization of the surfaces generates sections, patterns of cut. a bril-

liant solution comes from the fact that the patterns for construction can be
either more easily determined linearly with straight segments of a ruled sur-
face, or curved segments scalable with the use of pantographs. The robotic
fabrication paradigm, due to its uncanny precision is an optimal resource of
fabrication for complex geometric segments in construction of membranes,
and current challenges reside in the development of novel computational
design methods aligned with the and to the constraints of the fabrication
processes and the digital nature of the 4.0 Construction paradigm.

• In Rolex Learning by SANAA Nishisawa Sejima, the determining

factors and outcomes of value in the production of the complex surface is
the establishment of criteria for a concept of the digital surface in conjunc-
tion to the interpretation of the a concept of structural equilibrium, based
on the fabrication of the geometry. The geometry, it’s shape and parametric
values, informed and defined the ultimate state of equilibrium and digital
conceptualization.

Two Main directions emerge. A combination of the aforementioned con-

cepts coalesce in a geometrical configuration to resolve the equilibrium
state. The intersection of form, geometry, performance, digitization, and
Fig. 1. 87 Crematorium in Ka-
structural performance establishes an analytical model for a configuration kamigahara.
of plates, in lieu of establishing a configuration physical geometric unique. Fig. 1. 88 Construction of
formwork and supports for con-
The research aims to establish a parameter for determined form, to look for crete shell.
equilibrium. Morphogenesis of the concept
of flux structure. Kakamigaha-
The maximum potential energy stablishes the state of equilibrium, and ar- ra Crematorium with Toyo Ito.
(Sasaki, et at.,2007)

113
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

114
 1 Introduction
State of the Art

eas of maximum kinetic energy are resolved to optimize shapes for maxi-
mizing potential energy by establishing a set of functions of equilibrium
capacity.

A solution space of shape in equilibrium is determined by Flux morphogen-

esis analytical model. The stability boundary allows form finding resistant
to flexural stress. The key element resides in the fact that transformation
forces to search for maximum potential energy shapes, that is of maximum
stability, will become the same areas of prestressing the shape.

The problem of equilibrium thus resolves as a problem of geometrical

configuration. In order to interpret the larger scale, an equilibrium system
will transform complex type of stresses into simple (tensile – compression)
structural stress systems.

With Bezier’s mathematical model of curves, particularly Non-Uniform

Rational B-Spline type, the positions for determining the rebar shape are
known and allow for aligning mesh wires within the limits of the surface.
Once Sasaki determines the optimized potential energy at any given point,
a stable shape is analytically determined.

Precision and Complexity are fundamental novel aspects of the process of

design, the global shape is “shaped” by applying forces, without changing
the topology of the membrane. This is possible thanks to the capability of
processing systems that can compute all given sections independently and
continuously, which before was impossible.

“Seeking the optimum structural form of a free-curved surface shell, we take

an unknown quantity of design parameters as coordinates in the Z axis, then
try to minimise the objective function of the strain energy C (Z,U,(z)) under a
variety of design criteria. This is pursued through nonlinear iterations within
a computer. In contrast to purely formal manipulations of Yokohama, which
disregard material characteristics.” (Sasaki, Mutsurō et al., 2007)

The interpretation of the Flux theory of form provides the guidelines of form
definition by forces in equilibrium (Pugnale et al., 2014; Sasaki, Mutsuro,
2014).

Biology represents a valuable model to advance membrane design, two mem- Fig. 1. 89 Construction of
formwork and supports for con-
brane optimization considerations are of particular interest. Ideal shells, op- crete shell.
timized for a determined load case, can be extremely susceptible to varying Fig. 1. 90 Jack down process.
Removal of supports.
loading conditions (Bletzinger & Ekkehard, 2014). Grid shell parametrization Morphogenesis of the concept
and optimization of membranes offers a discrete constitution alternative, based of flux structure. Kakamigaha-
ra Crematorium with Toyo Ito.
in individual cells (Winslow, 2014) (Sasaki, et at.,2007)

115
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 91 Mesh Mould, ETH Zurich, 2012-2014

ETH Research Project

Credits: Gramazio Kohler Research, ETH Zurich.

In cooperation with: Physical Chemistry of Building Materials group (Prof. Dr. Robert J. Flatt), Institute for Building
Materials, ETH Zurich; Collaborators: Norman Hack (project lead), Willi Viktor Lauer; Industry partner: Sika Tech-
nology AG, SwitzerlandGuedel AG, ABB Schweiz AG
Rreproduced with permission of Gramazio Kohler Research, ETH Zurich, Switzerland

116
 1 Introduction
State of the Art

Biology creates a different strategy of optimization by creating remodelling of

highly complex interconnected fibrous structures. Efficiently redundant theo-
retical biology models in fibrous structures can influence positively the design
of complex membrane design.

The research aims to set a methodology of digital design that characterizes the
structure three dimensionally and in cloud point samples, instead of stating a
clear boundary and a study of potential energy based on heigh transformation
in points that define the membrane geometry.

Today digital design provides a new tool for generating complex surfaces with
optimized material use. Robotic fabrication allows a precise and optimized
material configuration of complex surfaces (Waimer & Knippers, 2015). Free-
form design coupled with structure calculation and simulation software can
analyze characterize, prior to construction, intricate three dimensional lattices
and computationally determine the type of fiber, density of fibrous arrange-
ments and its extrusion thickness necessary to incrementally build a mem-
brane.

The artificial arrangement of fibers by custom additive manufacturing allows

the automated creation of continuously mechanical complex surfaces with dis-
crete units.

• The characterization of a membrane is conformed of spatial lattices,

individual struts in different directions of space that act mechanically con-
tinuous and efficiently distribute the loads based on stress trajectories.

• This geometry is based on porous cellular solids and cellular mem-

branes due to their exceptional capacity to create light weight structures.

1.6.4.1. Computational methods

Standard 3D printing process distinguishes two major parts: the outer perim-
eter or boundary of the object, and the interior of the object or infill. Recent ef-
forts focus on topological optimization of the global shape. Parameters of both,
the outer layer and infill, can be modified within an STL generator software.

Commercially available slicing software include undifferentiated 3D infills,

interesting for their better mechanical properties as compared to continuous
vertical extrusion of patterns in 2D solutions. Newer developments in slicing
software include variations in density and the possibility of manually introduc-
ing infill density differentiation by varying the percentage of solid and empty
parts within volumetric thresholds defined by the solid meshes interior to the
global geometry.

117
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

118
 1 Introduction
State of the Art

However, although one can choose between a limited set of options, the depo-
sition process itself is not considered as part of the design process. Custom-
izing the deposition process (Kulkarni et al., 2000) allows the design criteria to
be implemented at a material performance level further enabling control over
geometry.

This research adapts to a larger scale design and manufacture methods that
define material efficient geometries based in computing stress flows such as the
Stress Line Additive Manufacturing “SLAM” technique presented by (Tam,
K. & Mueller, 2017) The geometry generation must be coordinated with cus-
tom toolpath creation to optimize fiber deposition and geometrical assemblies.
FEM influenced micro and meso scale studies and fiber orientation provide
promising structural improvements (Daynes et al., 2017; Gospill et al., 2017;
Khurana et al., 2020; Kubalak et al., Jul 2019).

The computational methods presented in this research are influenced largely

by nature inspired design methods presented by N. Oxman and William Mitch-
ell in (Oxman, N., 2010), non-linear simulation processes that inform addi-
tive manufacturing (Chronis et al., 2017) and gradient characterization and
voxel approximation to 3D printed present in (Oxman, N., 2011). The teams
of researchers at Stratasys, MIT and Delft University presented a voxel-based
material property mapping workflow.

The process of discretization inferred in voxelization creates a digital equiva-

lent of a given form, where individual units can be characterized separately to
create locally differentiated, functionally graded forms (Loh & Pei, ; Loh et
al., 2018).

Among the most influential researches in construction robotics (Gharbia et al.,

2020) is the work of researchers at ETH in Zurich, who presented the differen-
tiation potential of robotically manufactured spatial lattices (Hack et al., May
2014), presenting both a novel application of spatial additive manufacturing
wire print (Mueller et al., Oct 2014) and an innovative methodological ap-
proach to complex surface design and fabrication.

The aforementioned works have advanced the use of extrusion 3D printing and
discrete methods in design, and is incrementally being applied to architecture
due to the computational logic it shares with the serial process of additive
manufacturing (Retsin & Jimenez, Oct 2016).

Although the classical theories of continuum and discrete me-

chanics can approximate the stress field reasonably well for any
given structural member, the final design of such member rarely

119
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 92 Robotic extrusion at Nagami

Robotic Spatial Printing by discrete methots. A discrete approach to spatial printing uses a single toolpath
fragment as basic unit for computation (Jimenez Garcia, et al. 2017).

Fig. 1. 93 Discrete Computation Lab - Discrete software interface by Jimenez Garcia, et al. (2017).
Reproduced with permission of the Manuel Jimenez Garcia

120
 1 Introduction
State of the Art

reflects the principles of optimized inner forces given a certain

material as many fabrications constraints have to be taken into
account like formwork or casting and consolidation, if such con-
straints are removed by the 3D printing process, existing nu-
merical methods can be sided to optimize and directly influence
the material utilization within a given structure with respect to
given loading, boundary condition and other performance ori-
ented goal (acoustical, thermal, etc). (Podroužek et al., 2019)

Coupled with discrete design methods, Finite Element Analysis based models
can effectively simulate the internal and external, non-linear, distribution of
forces in complex forms (Borunda, Luis et al., April 2019) and present hierar-
chically differentiated cellular membrane that can be 3D printed.

The application of Finite Element Methods provides guidelines to more ef-

fective fibre deposition of large-scale cellular membranes that demand new
computational workflows. This research presents a computational method that
aims to bridge cellular membrane production and static analysis models to cre-
ate shapes in architecture determined by forces in equilibrium.

For this purpose, the computational methods presented build upon the ideas
of voxelization, local differentiation, and material grading discussed. This re-
search utilizes SimScale (CFD, FEA, and thermal simulation software in the
cloud.) and ParaView (Ayachit, 2020) software to process the Finite Element
Methods non-linear static analysis. ParaView allows to process and access
point samples data through Python programming language, easily transferring
data from Simscale to Rhinoceros 3D and to 3D print code through CSV data
sheets.

SimScale is an open source, computer-aided engineering program that allows

the static analysis of structures on a matrix of stress points, discretizes mini-
mum units to establish the deformation according to those stresses.

ParaView is defined on its website as: An open-source multiplatform data vi-

sualization and analysis application. ParaView users can quickly create visu-
alizations to analyse their data using qualitative and quantitative techniques.
Data exploration can be done interactively in 3D or programmatically using
ParaView’s batch processing capabilities.

Stress influenced design and bio-inspired configurations result in highly intri-

Fig. 1. 94 Discrete
Computation Lab -VoxelChair
v1.0 VoxelChair (Jimenez Gar-
cia, et al. 2017)
Reproduced with permission of
the Manuel Jimenez Garcia

121
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

122
 1 Introduction
State of the Art

cate configurations with a significant amount of associated information, meth-

ods of data management are key to balance computational power and map
areas of interest.

1.6.5. Resolution in architecture

“Complexity in architecture is an ever-growing area of develop-

ment with every model and style being increasingly able to ac-
commodate information, curvature and free form, but with expo-
nentially growing computation power, are we hitting a milestone
of information management and free form design in architec-
ture?” (Schumacher, P., 2019)

Resolution is defined as the amount of data in a given area. Its application in

architecture is relatively new since conventional design processes do not oper-
ate on such raw data. Instead, architects manipulate geometries. In other fields,
for example, imaging, the term resolution reflects the detail an image holds.
Digital images measure resolution by its number of pixels, each containing a
masked color or grayscale value as data. Historically, resolution has been an
issue in communication systems, as the transmission of images entails data
conversion from the transmitting station to the receiving station.

In architecture, resolution has been explored by architects Dillenburger and

Hansmeyer, who define it as the number of voxels per volume, looking at the
maximum resolution at which an additive manufacturing method can fabricate
(Dillenburger & Hansmeyer, Jul 2013). In this research, however, resolution
is the quantity of information at a localized area of a discretized 3-dimentional
shape. It is important to note that a shape designed by a conventional design
model cannot be measured by its resolution, for instance, when geometry is
manipulated using a modeling software. A resolution-based methodology is
necessarily driven by manipulation of data using programming tools.

Today’s computational technologies offer the possibility to directly design in

3- dimensions without recourse to encoding and decoding shapes via 2-dimen-
sional drawings. Furthermore, programming languages provide a means of
constructing shapes through raw data without geometrical manipulation. Prior
to the use of algorithms, the design process was purely intuitive, and designers
normally made decisions based on what they perceived from visual feedback
(OXMAN Educating the designerly thinker).

Technological advancement often comes with a period of exploration in which

architects learn how to implement them (Carpo, 2017). In the late 1980s, with
the introduction of CAD software, manual drafting was substituted by digital

123
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 95 Cloud Pergola for the Venice Architecture Biennale. 40,000 voxel cells printed (Vimeo 53:23)
(PENDING PERMISSION) By Alisa Andrasek reproduced with permission of the author

124
 1 Introduction
State of the Art

drafting; during the 1990s, software developed for industries such as film– Au-
todesk’s 3D Studio, or aeronautics– Dassault System’s’ CATIA, were appro-
priated. Nonetheless, the essential paradigm remained the same: manipulation
of geometry and modification based on visual feedback. On the fabrication
side, the combination of CAD/CAM technologies created a seamless connec-
tion between design and manufacturing, which gave architects the possibility
to re-gain their lost status as master builders (Cardoso, 2012).

Parametric tools, such as Grasshopper and Dynamo, have generalized visual

scripting, augmenting the design process by the insertion of procedural logic
and data management, facilitating contemporary avant-garde architectural
styles such as, parametricism (P. Schumacher, “Parametricism). However,
such tools neither, necessarily, address design in terms of discreteness nor
resolution.

Discrete methods are normally used by analytical frameworks, for instance, fi-
nite element method (FEM), a numerical technique for solving problems such
as structural analysis, heat transfer or fluid flow. An object is discretized into
smaller units that are studied separately, yet as part of an integrative process.
Karamba 3D by Clemens Preisinger or STAAD.Pro by Bentley, are examples
of force distribution solvers. While these analyses can be integrated in the de-
sign process, there is no such generalized discrete design methodology.

This research provides a step towards a new methodology of design, based

on multiresolution adjustments, using a discretization technique, which crafts
raw data, without manipulating geometry. Resolution measured by the quantity
of data in a given area is natural to this method. A localized area comprises a
cluster of neighboring fragments, with information containing numerical and
textual values, such as coordinates, material information, location, context
awareness and Boolean values. Shape is understood as multiple dictionaries of
information, which define its resolutions levels. In essence, this data construes
discrete geometrical units informed by the fabrication process.

Additive fabrication technology needs a certain amount of data and machine

time, does not involve reusable cast, mold, or stamp, and does not require any
voxel-generated volume to be identical to any other, regardless of scale or
size. The marginal cost of a voxel is always the same, no matter how many
we print. In this way, 3D printing brings the logic of customization from the
macro to the micro scale of production of physical matter, and at previously
unimagined levels of complexity and granularity (Carpo, 2016). Utilization of
additive manufacturing methods is clearly ideal when designing with different
resolutions.

125
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 96 Cloud Pergola for the Venice Architecture Biennale. 40,000 voxel cells printed (Vimeo 53:23)
(PENDING PERMISSION) By Alisa Andrasek reproduced with permission of the author

126
 1 Introduction
State of the Art

1.6.5.1. Design Complexity

The possible solutions available to the designer are increased with the geo-
metric freedom that additive manufacturing offers. Thereby, digital design and
optimization process become more important. (Carpo, 2016) challenges the
idea of parametric streamlining, articulation and goals of an aesthetic continu-
ity (lack of void) as a style of the digital age and traces this methodological
framework’s end, questioning the resources, energy waste, and disconnection
from the essence of computation in which the information age is based, Bool-
ean logic, discrete binary encoding, and continual iteration.

An unsystematized system for complexity in architecture evolved fundamen-

tally to excess. This research aims to schematize the use of data for performa-
tive architecture continuing the intuition of a first approach to the digital of
authors like (Kolarevic, 2003) and (Burry, 1996).

The notion of resolution as a driver in design fields is relatively new —it is

linked to a recent computational revolution in which brand-new science be-
gets a new way of thinking, and to the advent of additive fabrication such as
3D printing. The following designers include the word “resolution” in their
descriptions. Voxel Chair and Curvoxels by M. Jimenez and G. Retsin, use a
discrete approach specifically to directly design for a 3D Spatial printing fab-
rication (Retsin & Jimenez, Oct 2016; Soler et al., Nov 2017). However, their
design approach is not driven by concepts of resolution and is rather a medium
for a specific fabrication method. The Computational Chair Design Studies by
Phillipe Morel of EZCT Architecture & Design were among the earliest dem-
onstrations of discrete approaches.

Interestingly, a Finite Element Analysis (FEA) framework is used as a design

tool. Digital Grotesque, by Dillenburger and Hansmeyer, is the first work to
tackle resolution as a driver in design and fabrication. Their work is a manifes-
tation of exuberance and a celebration of computational power (Dillenburger
& Hansmeyer, 2014; Dillenburger & Hansmeyer, Jul 2013). While resolution
is their main motive, their project looks at the highest resolution possible, rath-
er than taking resolution as a platform for multiple inquiry.

Recently, A. Andrasek showcased the Croatian National Pavilion, Cloud Per-

gola composed of 3D printed 40,000 voxels, at the Venice Biennale (Andrasek,
). A large-scale robotic spatial printing workpiece is designed using algorithms
for multi-agent systems MAS. The deposition process of the pergola had to be
optimized for a timely deposition, reduced from the original scope of 80,000
voxels. A main aspect A. Andrasek points out is the demand of artificial in-
telligent solutions to automate lattice construction, allowing the robotic tool

127
 Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Making of EDAG Light Cocoon concept car

EDAG, 2013
Making of EDAG Light Cocoon concept car
EDAG, 2013

3D, diseño y fabricación digital de superficies continuas | Luis Borunda |

3D, diseño y fabricación digital de superficies continuas | Luis Borunda |

Fig. 1. 97 EDAG Light Cocoon - The Making Of. “EDAG Light Cocoon” is the concept vehicle from the interna-
tionally active engineering service provider EDAG, which was mainly manufactured using the 3D printing process.
EDAG, March 2015.
Fig. 1. 98 EDAG Genesis one piece body structured monocoque consists of a sandwich body that joins individual
struts with outer solid layers.
Fig. 1. 99 2014 Geneva Motor Show. “EDAG Genesis” is the concept vehicle from the international engineering ser-
vice provider EDAG. The concept car shows the potential and design possibilities through the generative manufactur-
ing process. The exterior of the concept car is based on bionic patterns of a turtle. EDAG, March 2014.
Reproduced with the permission of EDAG „Copyright EDAG Engineering GmbH“.

128
 1 Introduction
State of the Art

to sense points in space, which will increase the efficiency and geometric
freedom in constructing spatial lattices (RMIT Architecture & Urban Design,
2018). Her algorithm computes active discrete elements whose behavior is de-
termined by a collection of rules. Although the resolution is the same across
the piece, we can see the potential in designing through discrete approaches:
this strategy yields a high performance at structural, material wastage and aes-
thetic capacities.

In other fields, common use of discrete methods includes volumetric imaging

in medicine, representation of terrain in games and simulations, and granular
flows and rock mechanics in engineering. A crucial stage in volume graphics
is the synthesis of voxel-represented objects. This stage, called voxelization, is
concerned with converting geometric objects from their continuous geometric
representation into a set of voxels that approximates the object (Kaufman et
al., 1993).

1.6.5.2. Optimization
Outlining historical viewpoints, F. Otto presents the connection between per-
formance and form with buildings of high intricacy. He applies tessellations
to create double-curved surfaces using non standardized elements. Parametric
design and additive manufacturing compliment very well each other for the
purpose of creating non-standardized elements. The engineering and optimiza-
tion of tessellations becomes essential to harness the potential of the combina-
tion of these two tools.

The challenge of accommodating multiple functions in the natural world is

most often overcome by growing structural patterns in multiple scales to pro-
vide for various functions. The spongy bone, for instance, is designed as a
cellular solid at the mesoscale and as a fibrous composite at the microscale.

Cellular structures and fiber textures provide for high weight to volume ratios
while maintaining very strong structures.

EDAG Genesis Utility Incorporating these manufacturing processes and the

qualities of Composites materials studied previously, we establish a case of
analytical study with examples on a medium scale, the automobile. An ex-
ample of material edited and generated with methodologies parametric and 3D
printing of sandwich membranes is the vehicle of the future Genesis by EDAG
Geneva Engine. This sandwich solution relies on one sheet on top of another.

Material Based Computation and digital fabrication introduces the procedures

of engineering anisotropy for 3D printing pieces of gradually differentiated
materials for optimal performance.

129
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 100 ICD Research Buildings / Prototypes. Bundesgartenschau Heilbronn 2019, Germany.
Fig. 1. 101 Development Process images by ICD/ITKE University of Stuttgart. Engineering methods - Global model,
digital environment merging different types of information: (left) detailing; (middle) finite element analysis (red in-
dicates zones of higher stress); (right) abstract geometrical description. Fiber depositoin is optimized according to
surface stress differentiation to improve mechanical performance.

Project Credits: Institute for Computational Design and Construction; Institute of Building Structures and Structural
Design; FIBR GMBH; in (Bodea, et al., 2020). © ICD/ ITKE, University of Stutgart.
Reproduced with the permission of ICD/ITKE University of Stuttgart.

130
 1. Introduction
State of the Art

Many biological structures and tissues are hierarchically built-up fibers. Cel-
lular materials are composed of a solid fraction below 30 %. Fiber geometry
is the most influential aspect of their strength. In this research, methods pro-
posed optimize volume fraction and architecture of fibrous structures is based
in variable resolution algorithmic design to trace segments of stress threshold
in a membrane and map different types of space filling structures to adjust the
theoretical biomimicry model, to the scale of architecture.

This results in a mechanical re-assessment of the structure to perform a lo-

cal change of density geometrical characterization and mechanical grading for
large-scale lattice additive manufacturing of shells.

Variable resolution algorithms can trace segments of stress threshold and map
different types of space filling structures to adjust the theoretical biomimicry
hierarchical model, to the scale of architecture. This results in a mechanical
re-assessment of the structure to perform a local change of density geometrical
characterization and mechanical grading for large-scale lattice additive manu-
facturing of shells.

Formal research currently studies three types of scales of behavior optimiza-

tion, the macro scale, the mesoscale and the micro scale. The relationships
between each scale are essential to interpret behaviors on a macro scale, which
is the scale to study in Architecture and the composition of the materials used
in obtaining complex formal structures, or micro scale.

1.6.6. Hierarchical differentiation biomimicry

model
Biomimicry is innovation inspired by nature. Nature provides infinite exam-
ples of efficient fibrous arrangements of exceptional mechanical properties
(Fratzl & Weinkamer, 2007) that can be considered to mechanically behave as
cellular solids (Gibson, Lorna & Ashby, 1997), from which the particular inter-
est of this thesis is the structural lattices found in the nature, and particularly in
cancellous bone. (Gibson, Lorna, 1985).

Architecture has a long tradition of looking at natural formations for inspira-

tion. Learning from the natural world, Biomimicry conscious emulation of life
Fig. 1. 102 Interior impression
“genius” is taking the design principles and learning something from it. The of the BUGA Fibre Pavilion’s
identification underlying geometrical principles behind complex assemblies Characteristic fibrous node
configurations: the dome struc-
and their additive manufacture is a current area of research (Pelanconi & Or- ture exhibits five axes of sym-
tona, Sep 1, 2020) metry. © ICD/ ITKE, Univer-
sity of Stutgart.
The power of mechanical work by shape can be applied to increase efficiency. Reproduced with the permis-
sion of ICD/ITKE University of
In the realm of biology, form structure and performance are intertwined parts of Stuttgart.

131
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 103 Digital representation of a biomimetic double-layered surface

https://andres.harris.cl/
Copyright © 2020 Andres Harris - London.
Reproduced with permission of Andres Harris

132
 1 Introduction
State of the Art

one coherent development. Precursors into the morphogenetic and algorithmic

study of the natural world can be traced back to J. W. Goethe’s “The Metamor-
phosis of Plants” published in 1790 (Goethe, 2009) and D’Arcy Thompson’s
“On Growth and Form” section on the Theory of Transformations, published
in 1943 (Thompson, 1992) presented influential studies of underlying prin-
ciples of complex natural assemblies on morphogenesis of form in the natural
world that led to hypotheses of growth form by algorithmic generalizations.

E. Haeckel’s studies of “Art Forms in Nature” influenced architects such as Z.

Hadid, A. Gaudi and F. Ghery, and currently large-scale additive manufactur-
ing of complex structures based on Radiolaria identified by E. Haeckel has
been successfully fabricated by E. Dini (Jungck et al., ).

N. Oxman in her Doctoral thesis interprets the structural changes in the vari-
ous scales of natural wood fibers (Oxman & Mitchell 2010) and concludes
highlighting the characteristics of heterogeneity, anisotropy and hierarchical
structure in natural fibers at the material level. Its formal adaptability to struc-
tural level dependent on external conditions and finally at the system level how
the behavior of the fibers adapts to external agents.

Among numerous examples of cellular structures, and properties of cancellous

bone, this research looks at its hierarchical structure and variation of strength,
stiffness, and modulus of elasticity along variations of density, architecture and
material as a role model to challenge current design models of lightweight and
structurally efficient architecture.

Architects and researchers Andres Harris and M. Pawlyn have identified ben-
efits of transferring concepts from bird skull tissue (Dumont, 2010; Novitskaya
et al., 2017) to architectural applications to configure lightweight optimized
membranes.

More specifically, we focus on the internal structure, namely, cancellous bone

and lamellar bone tissue in bird skulls Fig. Andres Harris, where individual
bone cells are varying in density as a function of the strain and in orientation
aligning with the main stress (Zippel, 1992) to withstand tensile and compres-
sive forces and therefore reducing their weight.

1.6.7. Cellular Solids, Lattice Structures and

Fibrous Materials Fig. 1. 104 Structure Optimisa-
tion_Cancellous Bone Tissues:
The detailed understanding of the biological model is the first requirement High Strength to Weight ratio.
https://andres.harris.cl/
Copyright © 2020 Andres Har-
ris - London
Reproduced with permission of
Andres Harris

133
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 105 3D-Printed bird skull model by Exploration Architecture and Idil Yucel.
Image by @kellyhillphotography Reproduced with permission of Michael Pawlyn

134
 1 Introduction
State of the Art

towards the goal of controlling hierarchical fibrous configurations. The origin

of Wolff’s trajectory hypothesis can be traced to Culman’s short, cantilevered
beam (Culmann, 1866).

In biology you often find very complex structures by placing material exactly
where it needs to be (Dumont, 2010). This process emerges the hypothesis that
bone structure is remodeled following stress trajectories (Zippel, 1992).

Wolf law of bone transformation is summarized in:

1. optimization of strength with respect to weight
2. alignment of trabeculae with principal stress direction
3. self-regulation of structure by cells responding to me-
chanical stimulus
(Skedros & Baucom, 2007) found this to be only applicable to trabeculae for-
mation under simple stress, contesting the generalization concept presented by
Wolff (Rüttimann, 1992).

The anisotropic nature of complex fibrous arrangements like cancellous bone

makes it both difficult to model and to analyze, laborious manual stress line
computation and comparison with bone sections was needed. Wolff extrapo-
lated K. Culman’s (Wolff et al., 1870) method to create graphical abstractions
based in perpendicular arcs to simplify the mechanical work in trabeculae
yielding inconsistent results. Yet, the advances of simulation by finite elements
the analysis are advancing our capability to understand such complex struc-
tures. With better computational models, architects and industry researchers
can study potential applications of the principles behind the efficiency of bone.

Although the trajectorial hypothesis presented by Wolff’s is still a matter of

discussion (Skedros & Baucom, 2007), advances in this field determine that
mechanical properties in fact depend on volume fraction, novel finite elements
analysis methods that efficiently the role and geometry if cancellous bone and
discrete abstractions of trabeculae formation, role of orientation or density in
bone strength provide essential insight in understanding fibrous structures.

Currently, critics (Skedros & Baucom, 2007) and supporters (Keaveny et al.,
2001) of the trajectorial hypothesis have come to agree that the basis of bone
strength has found to be primarily:

• •Density
Fig. 1. 106 Detail of 3D print-
• •Thickness and orientation of struts ed soft seat. (Imprimir el mun-
do, 2017). Photo by the author.
• •Material
Fig. 1. 107 Detail of Biomim-
icry; 3D Printed soft seat.
(Lilian van Daal, 2014.)

135
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 108 von Meyers (1867) composite illustration shows the Culmann crane and sections of various human bones
with stylized arching trabecular patterns. Reproduced with permission of Walter de Gruyter, Berlin, Germany. 1992.
Text chapter by Ruttimann In Wolffs Law and Connective Tissue Regulation. p. 15. Fig. 1

136
 1 Introduction
State of the Art

in that order of importance, as stated by (Keaveny et al., 2001). This research

will focus on transferring the order and criteria of bone strength principles to
digital design strategies for fibrous membranes.

As suggested by bone strength analysis, this research mainly focuses in the

first two:
• Application of notions of trabeculae density variation (hier-
archization) in principal stress volume fractions
• Deposition of trabeculae architecture aligned with principal
stress lines (Gibson, Lorna, 1985) to cellular membrane construction,
particularly to membranes prestressed under simple compression.
Materials are tested considering current developments, recognizing a trend
of sophistication of fused deposition instruments and compositions, fibrous
(Kwon et al., Aug 2018), metal (Tankova & Da Silva, 2020), ceramic and ce-
mentitious (Borg et al., 2018; Mechtcherine et al., 2019; Wangler et al., 2016;
Wangler et al., 2019).

The mechanical differential is achieved manipulating internal geometry

to create a shape of locally controlled modulus of elasticity and stiffness in
membranes that result in elevated material resource efficiency. Traditionally
architectural design subordinates material and fabrication parameters to a
preconceived form, this workflow is an example of a biomimetic approach
(Achim, Jan 2007; La Magna et al., 2013; Naboni et al., 2019; Parascho et al.,
2014) simultaneously originating form, and characterizing it by simulation,
materialization by robotic construction, all tested and optimized in a continu-
ous virtual workflow.

Nature produces cellular structures up to 1000 times the material volume in

a solid object by optimizing its fibrous arrangement. Computational design
offers the possibility to create more complex structural typologies (Knippers,
2013) and additive manufacturing is making it much easier to mimic that level
of complexity and efficiency. Birds skulls are a great example of complex and
very efficient, light-weight shell structure.

The simultaneous force configuration, design for fabrication and character-

ization workflow demands to design cellular structures in bone biomechanics
with the constraints of manufacture in mind and to integrate the parameters of
manufacture to the design process.

We use computational design to create a version of the biological model and

Fig. 1. 109 Detail of Biomim-
created tokens of segments and scaled versions of cellular membranes present- icry; 3D Printed lattices. Lil-
ed in chapter 4. The result is a design concept oriented towards generalizing ian van Daal 2014 in (Imprimir
el mundo, 2017). Photo by the
the hierarchical configuration of membranes. author.

137
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 110 Composition of Graphic Statics stress trajectories in structures (Cullman 1866) and the approximation of
bone stress trajectories by J. Wolff (1870).

Advances in structural engineering influence biomechanical studies in theoretical biology.

The structural engineer K. Culmann illustrates spanning trajectories in the line section of a bar in Die graphische
Statik; the graphical stress trajectory curves of a beam is abstracted as two systems of curves press together from both
sides, and determines that stress trajectory curves intersect at right angles (Culmann 1866, fig. 107, p. 236).
Wolff hypothesis of stress trajectory in bone illustrate the apparent similarity between the cancellous bone architecture
and the his graphical interpretation of a system of stress trajectory curves.
(Wolff 1870)Reproduced with the permission of Springer Nature. Copyright © 1870, Springer Nature

138
 1 Introduction
State of the Art

• Abstract in geometrical patterns the way it actually grows in the bio-

logical model by controlling volume fraction, fiber architecture and mate-
rial differentiation.

• Implement as controlled fibrous arrengements of interconected struts,

layers, and ties

• Create an understanding of the expected characteristics of the fibrous

tissue engineered through simulation and mechanical tests.

• The explicit application can be overall incredibly light weight and

very structurally efficient

Experimental testing provided proof of concept that a single material (chapter

through a complex toolpath fiber deposition is capable of behaving completely
different mechanically and to be varied at a meso-scale with impact on the
macro-scale and global shape. The variation of geometries of deposition and
“three dimensional infills” gives insight into the capabilities of additive manu-
facturing to structure mechanically characterized elements only by varying the
serialized deposition of fiber orientation (micro), strut or thread density and
porosity (meso) and shape organization (macro) within the continuous mate-
rial system.

The goal is to determine the ability of these variables to configure complex

free-form surfaces. Of this experimentation comes from the relationship be-
tween material and structure.

Very often, these fibers are not positioned randomly in space but are aligned .

This property provides the basis for more complex structure and directionality
in a variety of tissues. For example, in bone, this leads to increased mechanical
strength Fig. 1. 111 Principal stress
trahectories calculation for 3D
In the proposed workflow trabecular and cortical algorithmic adaptations of printing optimization.
density and architecture are modeled, to study membrane functional adaptation Cantilever beam and beam
under 3 point loading. SRed
to mechanical loading. trajectories correspond to
compressive loads while blue
Common finite element methods used to simulate fibrous structures are have trajectories correspond to ten-
inconsistencies and are computationally demanding (Wang, Z. & Mondry, sile loads (Khurana, Simpson,
Frecker 2020).
2005) a non-continuum model is critical to adapt to available computational
power. Modern Finite Element Analy-
sis experimental evaluations
The combination custom instruments FOAM technique and Multi-resolution verify aspects of K. Culmann’s
theory of graphical method to
presented in chapter 2 are applied to the configuration of membranes. Chap- draw force trajectories for very
ter 4 presents the characterization bone-inspired cellular structures influenced simple load conditions. Other
conditions yield in significantly
with stress information form FEM analysis to create the volume-based non- more complex results.

139
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 112 Skedros and Baucom (2007) analyszed trajectorial stress theory interpretations by different authors demonstrates the lack
of consensus; this demonstrates the elusive challenge of generalizing the mechanical behavior of complex fibrous structure. Current
computational tools are improving our understanding of mechanical properties of complex fibrous structures.

Fig. 1. 113 Articulation of anisotropic materials: Bone section, principal stress and streamline integration.
Today, architecture and engineering are transferring current advances in theoretical biology with biomimicry strategies.
Innochain research at the Institute of Architecture at the University of Applied Arts Vienna: Simulation of anisotropic materials
Author Cristoph Hermann. Industrial partners: Cloud 9 and Bluhmer Lehmann
Reproduced with permission of Cristoph Hermann

140
 1 Introduction
State of the Art

continuum functional adaptation of membranes applied to architecture.

The characterization involves performing analytical studies of spatial struc-

tures, biomechanics principles of cancellous and cortical bone approximation
with Triply Periodic Minimal Surfaces and space filling polyhedral arrays that
will conform the core of a cellular membrane.

This way, Finite Element Analysis provides the basis of a form in equilibrium
as the result of the analysis of forces in space. This fact demands to consider
the resulting form as “valid answer” in a solution space (Carpo, 2015), based
in the iterative processes of simulation. The value of this effort resides in the
free form membrane finding based on interpreting a form structurally, result of
the discretization process that creates digital - physical analog.

The digital properties of the geometrical configuration shall be tested to the

proper design conditions of components modeled by extrusion fused deposi-
tion.

1.6.7.1. Design considerations for components fabricated

by extrusion.
The substantial difference from additive manufacturing in general is that tech-
nologies of digital fabrication by subtraction and training are, de facto, the re-
sult of automating ways of manufacturing that already existed before the com-
puter. They reproduce ways of making that can also be done by hand (Malé,
2016).

The Additive manufacturing works conventionally by solidifying layers in

stacking order, according to the data sent from a file. In definitively, it re-
sponds to a form of materialization new and typical of an environment digital.
Typically, the results exhibit up to 70% of the strength of its injection molding
comparable counterpart (Ahn et al., 2002).

3D printed tokens always exhibit more strength in the direction of the fiber,
and bonding between layers tends to be more fragile. The anisotropy observed
in products made by 3d printing are generally artifacts of the manufacturing
process rather than the process deliberate design (Ahn et al 2002). Geometric
studies are fundamental in the future development of digital manufacturing.

Even for isotropic materials, the stiffness and resistance threshold to the length
of the deposited filament is usually greater than that of the cross section. Spa-
tial lattices and cellular solids are not built into layer-wise solid components, Fig. 1. 114 Code extract (AN-
NEX I) in a typical computa-
and do not respond to the generalized anisotropy of common 3D print tokens. tional design interphase.inter-
prets and processes mechanical
Pattern design can precisely influence stress distribution and deformation in and geometric data.
Own work

141
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Global shape is discretized for non-linear Finite Element Analysis

1.
and sorted into pressure quartils

Q 3-4 High strain

3-4 Q2 Mid strain

Q1 Low strain

Material deposition is tailored for density variation and infill toolpath

2. differentiation according to strain

Fig. 1. 115 VTK cell discretization for finite element analysis and Open Foam processing software
Fig. 1. 116 Stress distribution in compless shape, tertiles computation.
Fig. 1. 117 Functionally graded design and Finite Element Analysis influenced 3D printing process for improved me-
chanical behavior, improved stiffnes by engineering digital material distribution with local variations of density
(Borunda, Ladrón de Guevara, Anaya 2019)

142
 1 Introduction
State of the Art

orthotropic, anisotropic or isotropic matrices (Berrio et al., 2019; Martínez et

al., 2017; Martínez et al., 2018).

The deposition therefore can yield varied fiber functionality or anisotropy

without the need to introduce physical changes to the properties of the material
composition of the extruded filament. This process of Material geometric opti-
mization is currently the most accessible approach. The manufacturing process
of constructive elements by deposition of filaments demands a geometric study
at the “mesoscale” product scale to generate fibrous textures that provide an
efficient ratio of weight to volume while maintaining high structural strength.

It also demands a study and finite element simulation of structural behavior

where the Print resolution (DPI X&Y), geometry and material composition are
as follows. determining factor.

This conceptual change in the materialization of the constructive element in-

volves the passing of the vector used in other processes of representation and
previous manufacturing, voxel, volume resolution unit, allows to design manu-
facture any object with unprecedented freedom.

Following the morphogenetic logics of the trabecular bone, a functionally

graded cellular structure was generated where topology, porosity and geomet-
ric orientation of the lattice structure responds to the external loading con-
ditions. A combination of solid layer wise arrangements and spatial arrange-
ments allow flex and stiffening variations in modulus of elasticity and strength.
Considerations of the 3D printing process apply accordingly.

1.6.8. Computational Thinking for Large scale

Additive Manufacturing
While first digital turn architecture gravitate around seamless curvilinearity,
event space and animation (Lynn, Greg, 1999), contemporary digital design
begins to align to the discrete nature of computation (Carpo, 2016), and pres-
ents form as a procedural design (Dillenburger & Hansmeyer, 2014; Dillen-
burger & Hansmeyer, Jul 2013), complex relational construct (Jencks, 1997),
digital stereotomy.

Computer science offers procedural abstraction and data abstraction tech-

niques (Colburn & Shute, 2007) to efficiently handle complexity. Systematic
algorithmic design thinking is key for automation. Construction robotics can
introduce the unparalleled level of efficiency of complex form with the defini-
Fig. 1. 118 Custom toolpath
tion of algorithmic procedures (Menges & Ahlquist, 2011). based on stress threshold 3D
print
Nature provides fruitful examples of algorithmic complexity, displayed in the (Borunda, Ladrón de Guevara,
Anaya 2019)

143
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Geometry, manufacture and mechanical behavior studies based on the hierarchical differentiaton of
fibrous arrangements
1. Infill density
2. Boundary and infill architecture
3. Materials
To attain constructability of fucntional fibrous arrangements at large scale

Fig. 1. 119 Scale, geometry and mechanical characterization studies.

Based on the constrain of scaling up a significantly slow process and due to the lack of technical criteria for produc-
ing functional components at large scale, the initial considerations for determining membrane design and deposition
parameters are focused at optimizing material use, fabrication time, cost, and mechanical performance.
1. To what degree shape and porous infill affect the mechanical properties of a membrane?
2. What geometries can configure fibrous arrangements and be manufactured?
3. What materials can be applied by extrusion in cellular membranes?
Own work
Carbon fiber reinforced filament gyroid infill and compression test in collaboration with Gianluca Pugliese at Wasp
Iberia, R. Claramunt and M. Muñoz at the Materials lab, Department of Mechancial Engineering, UPM

144
 1 Introduction
State of the Art

wide interest and influence of current robotic construction research (Bechert et

al., Jul 2018; Felbrich et al., 2018; Hensel et al., 2010; Krieg et al., 2014; La
Magna et al., 2013).

Morphogenetic and theoretical biology strategies help create algorithmic prin-

ciples of complex form generation (Achim, Jan 2007). Goethe Morphogenesis
is described as the biological process that causes an organism to develop a
certain shape It is a concept that arises from the observation of nature and that
it is possible to apply it in other disciplines.

Fuller referred to this concept that he called Synergetics (Buckminster Fuller

& Loeb, 1982) and the transformations “nature’s coordinate system”. Interest
of the research is in interpreting geometry and determining principles of how
to approximate geometry based in the process of deposition and 3D printing.

Geometry and structure are linked, a goal of the research is to interpret the
significance of the resulting form, to determine what is the set of possible solu-
tions, thus the structural and form. Interpret geometry through the process of
3D printing for a geometry and structure typology based in the materials.

1.6.8.1.1 Challenges: Large-scale and On-site fabrication

Fused Deposition Methods has demonstrated its applicability as a technology
for small scale fabrication. Variations of normal fused deposition methods that
consist in a layer-by-layer approach are those consisting in extruding filaments
in the space, in which the plastic is molten and solidified in the air connected
normally by nodes. This approach is an alternative that potentially permits the
building of lighter structures at a reduced cost and time (Mueller, Inn et al,
2014).

Even though significant advances on large-scale 3D printing have occurred in

the last years, the technical factors of implementation for practice (Davila et
al., 2019) and the large-scale architectural application for construction remains
among the main challenging areas for advancement (Chen et al., 2018; Ngo
et al., 2018; Olsson et al., May 2019; Pessoa & Guimarães, 2020; Shah et al.,
2019; Wangler et al., 2016; World Economic Forum, 2016). The complexity of
the construction-site, the lack of collaborative strategies for safety (Wu), the
variable environment that the site presents, the economic models to precede
technology transfer are among the more relevant challenges

Currently, robots are not yet able to deal with complex assembly task (Chea
et al., 2020). For successful on-site implementation, robotic construction de-
mands the development of both robot friendly designs and robotic tooling that
improve the efficiency and applicability of robots on site (Chea et al., 2020).

145
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

146
 1 Introduction
State of the Art

For this reason, this research necessarily explores and develops robot friendly
design and custom robotic additive manufacturing techniques in variable in-
frastructure conditions.

Thermoplastic materials offer plausible routes towards scaling up the process,

as demonstrated by ongoing applications.

Extrusion of thermoplastic polymers offers viable prospect of large format

construction than other additive manufacturing techniques, which tend to have
higher resolution but require more expensive, difficult to scale means of energy
and less manageable bulk material to deposit materials. Moreover, thermoplas-
tic extrusion can be scaled up by increasing nozzle diameter and using robotic
printers, whereas photopolymer deposition rates and printing volumes remain
more limited, despite recent progress.

To increase build efficiency and, as escape the lab are other main challenged
identified (Gharbia et al., 2020). Therefore, a new methodological framework
that creates a morphological continuity that links design, simulation and fab-
rication in large scale on site additive manufacturing. This research introduces
a novel Free Oriented Additive Manufacturing “FOAM” technique that adapts
to more complex infrastructure conditions and breaks the horizontal stacking
principle of standard 3D printing.

Coupled with powerful simulation and optimization tools, complex toolpath

algorithmic for lattice structures design present an important and current field
of research.

The research reviews possibilities of large-scale on-site fabrication through

case studies construction of functional membranes applying the methods and
techniques developed.

Although robotics and construction automation research has been part of the
academic and industrial effort since early 1980’s (Skibniewski, 1996), technol-
ogy transfer from academic research to industry is another of the main chal-
lenge identified (Gharbia et al., 2020; Wang, M. et al., 2020).

Connecting academic research with industry partners for on-site deployment

of innovative techniques is key to the development and harnessing of augment

147
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 120 Cellular membrane composed of 80,000 blocks.

Virginia Tech CFV Living Lab #7 Cloud project. Architecture by Cloud 9.
Membrane design and computational configuration for Additive Manufacturing by the author based on this research

148
 1 Introduction
State of the Art

in productivity available to the AEC by automation.

Results of this research are tested and implemented within a new collabora-
tion path to create innovation with industry partners called Living Lab, part of
The Global Forum on Urban and Regional Resilience’s Project in the Virginia
Polytechnic Institute and State University, Virginia Tech (Virginia Tech, ) .

Living Labs provides a fundamental example of industrial – academic joint

innovation efforts in the field.

149
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

CURVES
INPUT

4
1 (X0,Y0,Z0)
2 (X0,Y0,Z0) COORDINATE
3 (X0,Y0,Z0) EXTRACTION
4 (X0,Y0,Z0)
...
ELSE IF 2
GATE 0 0
Identify
coordinate 4 END 2 2 2
by Type coordinate
TYPE index item [-1]
travelling

3 REPEATED
coordinate
TYPE item nearby
knot previous item
4
2 NEW
3
TYPE
coordinate
item not nearby 3 3 3 3
mid-air previous item

2
1
TYPE
BASE
1 if Z<tolerance
2 2 2 2
bed
Collate with
Python
0 INVALID 1 1 1 1 1 1
coordinate
TYPE index item [1]
None without support

Merge into
Gcode
GCODE OUTPUT

Fig. 1. 121 Custom additive manufacturing categorization.

General considerations for algorithmic configuraction of custom toolpaths based on vector direction and coordinate
location. The 3D printing of three dimensional lattices is based on the concatenation of 4 types of points. This principle
allows the robotic manufacturing of architectural cellular membranes
Own work

150
 1 Introduction
Comprehensive morphological principle

1.7 Comprehensive morphological

principle
The need to configure complex cellular membranes demands the conception
of a comprehensive simulation-fabrication morphological method where dif-
ferent skills and disciplines converge to successfully present a model of hierar-
chically engineered continuous surfaces that follow a system of endogenic and
exogenic forces in equilibrium and obey the principles 3D printing.

For the development of an abstraction model that links form configuration to

a comprehensive workflow of design-simulation-3D printing and to create a
digital 3D printing model:

• Current large scale additive manufacturing techniques demand the

configuration of custom instruments of layer-wise and spatial deposition.

• Digitally configured membranes are based in the logic of computa-

tion. A logic that is fundamentally different form pre-digital models in the
demand of creating virtual representations of varying levels of abstraction
and resolution from explicit to implicit models which demand the develop-
ment of implementation methods that result in a digital stereotomy.

• This establishes a methodological framework for digital 3D printing

(3D printing based in particlizing a form and concatenating commands of
deposition) that can be applied to the construction of stress informed mem-
branes of very complex hierarchical organization.

The implementation model that capable of following trajectorial hypothesis of

stress in complex cellular arrangements demands a large-scale assessment of
geometries that follow the principles of digital 3D printing.

• Geometry is the most influential aspect of strenght found in very

lightweight architectures found in biological composites.

• The analysis, identification and generalization of geometries and

methods to produce hierarchical fiibrous arrangements that comply with
the several limiting constraints of a robotic technique of 3D printing de-
mand:

first, the development of algorithms that bridge mechanical analysis infor- Fig. 1. 122 Open and closed
mation with discrete data structures; free form cellular membranes
derived from stress trajectory
second, algorithms that bridge the discrete data structures and geometrical and additive manufacturing pa-
rameters.
configurations capable of originiating a mechanically continuous mem- Own work

151
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 1. 123 Continuously fabricated membrane concept for shell structure construction proposal.
Virginia Tech CFV Living Lab #7 Cloud project. Architecture by Cloud 9. Render by Playtime
Membrane design and computational configuration for Additive Manufacturing by the author based on this research

152
 1 Introduction
Comprehensive morphological principle

brane through membrane design and optimization with non-linear mechan-

ical characterization

The resulting membrane form configuration workflow is capable of creating

density variations and trabeculae architecture variations based in endogenic to
create complex geometric fibrous arrangements based in biomechanics.

Large scale additive manufacturing is identified as the initial challenge in the

literature review. The second next challenge identified is that of creating new
methods for on-site and prefabricated robotic construction.

• Cellular Membrane fabrication with spatial additive manufacturing

shows potential to create very lightweight structures that are capable of
construction automation but demand the evaluation of scale constraints
which require high mechanical strenght associated with anti-blucking ca-
pacity to maintain membrane work by shape.

• Exogenic forces, linked with the engogenic stress trajectroy influence

in optimal distribution of trabeculae architecture and volume fraction can
introduce forces of precompression to assure that the membrane whole will
work under simple compression stress, avoids buckling and utilizes mate-
rial at its most efficient condition.

• The effort of this research will therefore focus on presenting a method

to shape a membrane configuring the forces in equilibrium that define it.

153
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Digital 3D Printing custom materials

and instruments developed

[ 15 ]

Chapter 2 presents a custom Chapter 3 presents custom

fabrication technique computaitonal methods

Process applied to ceullular mem-

branes configuration. Methods of
discrete arcthiectural design and
computation for additive manufactur-
ing of shell structures.

Chapter 4 presents geometrical studies and case studies for

stress based membrane configuration

Principles for 3D printing continuous

membranes

Chapter 5 presents force based mem-

brane design and robotic AM

Fig. 1. 124 Thesis structure and goals.

The main goal of this research is to identify and sablish principles of digital 3D printing membranes in equilibirum,
along the additive process local particle deposition and in global monocoque, shaped by a system of forces in space.
Chapters 2 and 3 present a digital 3D printing process based on the discretization of a given form. Applying the digital
3D printing process to the design of membranes, chapter 4 uses the indexed discretization to hierarchically characterize
a fibrous structure. Finally chapter 4 provides insight into the capacity of materializing, on-site and by prefabrication,
large scale cellular membrane structures originated by forces in equilibrium.
Own work

154
 1 Introduction
Comprehensive morphological principle

1.7.1. Structure of the Thesis

In the development of a novel additive manufacturing application, the first
section of the research focuses on construction, to assess the limits of a new
construction technique that will determine and constrain next steps, to assure
the constructability of the design methods.

The methods therefore evaluated in the process of abstraction, inverse to im-

plementation.

• Construction: Free Oriented Additive Manufacturing Digital Fabri-

cation Technique Tests non-horizontal continuous spatial 3D printing of
discrete cellular units in new and existing infrastructures.

• Computation: Differentiation-based Computational Design: Digi-

tal Multi-Resolution. Optimization: Engineering Anisotropy by 3D Grid
Transformation and FEM mapping

• Conceptualization: Digital stereotomy model of differentiated archi-

tectural qualities contests the notion of digital design(Lynn) of curvilinear
continuity (Schumacher), revaluates performative continuity (Kovaleric)
and builds upon the notion of discrete digital form (Leach, Oxman, Ger-
shenfeld) development of a novel additive manufacturing application and
material (Oxman, Fleck) through the implementation of units of architec-
ture computation.

155
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 2 Free Oriented Additive Manufacturing technique. Design Fabrication Laboratory, CMU, Pittsburgh 2018.
Own work

156
157
 Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

System Detailing Fabrication Data Rob

Spatial system Arxel definition. Intelli- Robotic spatial printing consists of a Robot
that allows the gent unit that overlaps series of toolpaths and geometry that Proce
interaction an information from the with some sorting algorithms make a Exter
intelligent unit user, geometry and ma- file ready to print. Mater
(arxel) with a terial.
given medium. Geometry - Angle of printing
- Module Orientation
- Reachabiliity check
- Collision check
- Multimaterial check
GH Python GH Python GH Python - HAL Rob

Feedback loop

em Detailing Fabrication Data Robot Control

system Arxel definition. Intelli- Robotic spatial printing consists of a Robot + Tool Data
ws the gent unit that overlaps series of toolpaths and geometry that Procedure Loop
ion an information from the with some sorting algorithms make a External tool communication
nt unit user, geometry and ma- file ready to print. Material selection
with a terial.
edium. Geometry - Angle of printing
- Module Orientation
- Reachabiliity check
- Collision check
- Multimaterial check
ython GH Python GH Python - HAL Robotstudio - Controller

Feedback loop

Fig. 2. 3 Workflow for robotic construction of custom 3D pritn processes

Digital fabrication workflows allow the introduction of data and simulation routines. A system establishes the frame-
work of parameters (variables, constants and algorithmic procesess) creating a design solution space. Detailing it-
erates geometric solutions in the design space. The geometry is abstracted as data dictionaries and lists. The robot
controller naturally reads code and produces geometric results without need of blueprints, drawings or other represen-
tational means. This research establishes a positivre feedback loop between simulation and fabrication data to optimize
the process of deposition.

158
 2. Free oriented additive manufacturing

2. Free Oriented Additive

Manufacturing

Additive Manufacturing—AM—has grown steadily, finding numerous appli-

cations across all types of industries since the advent of 3D printing in the
mid-1980s. Among the numerous additive manufacturing methods, fused de-
position modeling, which is generally based on polymer materials, has become
the most widely used due to its lower cost, ease of use and versatility. Current
research in additive manufacturing contributes to increase the number of print-
ing materials and to achieve a better quality in the overall process (Ngo et al.,
2018). However, for architectural applications, 3D printers are yet to overcome
challenges of scale and of on-site adaptability.

At a building scale, and without defining the specifics of the machine, 3D

printing consists of a gantry system of numerical control deposition, usually
orienting the printing end perpendicular to a surface that receives the printing
matter. The complex nature of construction sites might demand more flexible
or adaptive additive manufacturing systems. Industrial robotic arms have prov-
en to be successful in adapting to such complex conditions and are excellent
candidates for in-situ fabrication. Research on robotic additive manufacturing
has recently demonstrated potential in large-scale construction (Labonnote et
al., 2016).

This chapter presents a novel large-scale additive manufacturing application

of situated robotic fused deposition modeling of thermoplastic lattice struc-
tures called Free-Oriented Additive Manufacturing—FOAM. Such technique
permits variations in the spatial conditions, deposition direction, and geomet-
ric patterns in order to adapt to complex infrastructure settings. FOAM, thus,
breaks the conventional layer-by-layer stacking principle and the need of verti-
cal printings.

We report the limitations and potential of 3D printing of thermoplastic lattices

on 6 degrees of freedom—DoF— machines. Learning from other industries
that have harnessed the potentials of automation, this research aims to improve
productivity and safety in the construction site. We also present a study of
suitable digital forms that constitute minimal discrete units of printing, called
Arxel. We conclude by discussing about the arousal of new aesthetics and sen-
sibilities in architectural design, nurtured by constantly evolving digital turns
(Carpo, 2019b).

159
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

160
 2. Free oriented additive manufacturing
Customizing 3D Print processes

2.1 Customizing 3D Print processes

The process of 3D printing is founded on the principle of vertically stacking
layer upon layer of a given fluid material, typically thermoplastic, deposited
through a numerically controlled mechanism. Generally, FDM systems work
by applying pressure to extrude fused thermoplastic and its subsequent vitrifi-
cation. A digital 3D model is fed into a specific software for 3D printing, which
slices the model into a series of stacked horizontal layers. At each layer level,
its area is filled with a continuous curve, generally for rigidity purposes, that
follow predefined filling patterns, generally raster, zigzag, contour, continuous
or hybrid patterns, called tool-path. The overall task of translating a 3D model
to a series of continuous toolpaths is called path planning.

The research presents a novel robotically augmented manufacturing technique

for 3d printing cellular and lattice-like structures the time necessary to produce
layer-wise structure significantly limits the application additive manufacturing
in larger scales; cellular structures are of interest because of their efficient use
of material to produce lightweight optimal weight to strength ratio. However,
common 3D printing techniques follow conventional layer-by-layer stacking
foundation, where a horizontal support is always needed, requiring temporal
supporting material, rendering it less suitable for these types of structures.

This section explores the application of Spatial FDM in variable infrastructure

conditions to break the standard horizontal stacking principle and introduces a
novel Free Oriented Additive Manufacturing (FOAM) technique that adapts to
more complex infrastructure conditions and produce cellular structures.

This research has principally two major contributions: the break of the con-
straint of necessarily printing along the negative Z axis, thus, addressing the
challenge of gravity in the printing process, and the compacity of the design of
the end effector tool, which maximizes reachability and robotic motion to print
over complex infrastructures.

FOAM provides insight into the possible applications of implementing extend-

ed additive manufacturing techniques using commercial equipment available.
Particularly, it implements a flexible printing orientation method, suitable for
novel spatial FDM applications, where minimum equipment dimensions al-
low to maximize the capabilities of geometric production. FOAM requires the
study of two key factors: the physical tool and material, and the computational
design and tool-path workflow specifically contextualized for the former fac-
tor.

161
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 4 ABB 6640 roboy arm. Design Fabrication Lab at Carnegie Mellon University
Photo by the author

162
 2. Free oriented additive manufacturing
Customizing 3D Print processes

2.1.1. Collaboration at Carnegie Mellon

University
This section presents my collaboration with Manuel Rodriguez Ladrón de
Guevara, who was developing his master’s thesis at the Master in Advanced
Architectural Design —MAAD— at the School of Architecture at Carnegie
Mellon University. This section covers the robotic 3D printing tool develop-
ment and thesis collaboration at the Design Fabrication lab at Carnegie Mellon
University. The establishment of a communication platform for the results of
research and collaboration can be found at the following link: https://multi-
resolution.wordpress.com/

(REF ANNEX)

2.1.2. Robotic Additive Manufacturing

Conventional FDM techniques require a continuous thermoplastic polymer
filament to 3D print layers of material following a stacking principle of print-
ing on top of a previous layer.

While this technique has proved to be useful as a design tool at small scale,
it presents problems to large-scale prototyping and construction field applica-
tions, such as inferior mechanical properties and anisotropic behavior. Large-
scale manufacturing not only requires the abandonment of 3D printers and
small robotic arms, but also the development of an end effector tool that con-
siders thicker material printing and optimized tool-pathing, to avoid material
waste, high costs, and time-consumption manufacturing processes (García de
Soto et al., 2018; Ngo et al., 2018).

Most importantly, this challenges conventional design thinking methods wide-

ly shared across designers, since new informed workflows are to be developed Fig. 2. 6 3D printer essentials.
Arduino activates a Solid State
to design and fabricate with such technique. Relay draw current to a Heat
Block and activates a stepper
One of the main differences between commercial 3D printing and robotic motor to drive filament through
3D printing is the increase in the number of axis of movement, from three the Heat Block. EOAT design
to six axis Interestingly, the benefits of the shift towards robotic fabrication
are not yet considered and implemented in design processes; Branch Technol-
Fig. 2. 5 Arduino wiring mock-
ogy implements a RFDM technique slightly tilting the nozzle when required, up for feed - flow system can be
however, no major challenge in changing the orientation of the base is made. tested independently.

Robotic arms augment the scale of fabrication; for instance, Hack et al [3]
Fig. 2. 7 3D printer essentials.
challenge the stacking principle of 3D printers. However, current state-of-the- Arduino activates a Solid State
art does not alter an important paradigm: the constraint of printing locking the Relay draw current to a Heat
Block and activates a stepper
nozzle along the negative Z axis. This unfolds a larger problem at construction motor to drive filament through
the Heat Block.

163
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

void loop() {

digitalWrite(pinLED, 0);

delay(1000);

digitalWrite(pinLED, 100);

duration = pulseIn(pinVal, HIGH);

Serial.println(duration);

if(duration > 5000){

engine = true;

pulseIn(pinVal, HIGH);

casoA();

Fig. 2. 8 Schematic describint the Digital Input / Digital Output communication between the source: (robot) and 3D
printing device processor (arduino board) to activate an AC stepper motor that will feed the filament through a heated
nozzle for spatial printing deposition..

164
 2. Free oriented additive manufacturing
Customizing 3D Print processes

scale: the impossibility of in-situ printing unless the printing piece is necessar-
ily anchored on the ground.

Consider a 3D printed large-scale piece, such as a 4.9’ x 6.5’ ceiling panel.

Current available techniques would need to fabricate it on the ground, orient-
ing the tool along the negative Z axis, break the piece down into transport-
able units, implementing secondary assembling parts, and finally mounting the
piece. We show, through three case studies, an automatic technique for in-situ
robotic printing, with potential to skip logistic and assembling parts, consider-
ably decreasing costs and time consumption of construction.

With the arrival of the 6-axis robotic arm and the generalized inverse kinemat-
ics of a system capable of reaching points in space through pre-engineered post
processors, the applicability of robots is significantly extended. A key develop-
ment is the separation of the mechanism that enacts automated actions from the
mechanisms that reach points in space. The interface is based in an End of Arm
Tooling mechanism capable of hosting all particularities of the robotic action
that allowed easier applications development.

The studies presented apply both Gantry 3 axis systems and 6 DOF systems
and review their different capabilities. This type of mechanization in construc-
tion is possible due to two key innovation in the field, computer aided design
and computer aided manufacturing. The studies focus on bridging digitally (for
example without the use of plans) the two through a software application that
algorithmically resolves form through simulation virtual prototyping and con-
nects the native digital design directly to its manufacture.

Fig. 2. 9 Low voltage signals

and pneumatics coupling in
Note 1. Extract of custom Arduino code for extrusion, full code in Annexes. End of Arm Tool activated by
In this piece of code, a loop defines constant reading of low voltage signals. If the Source: (robot)
signal is within a certain lenght threshold (around 5000 milisecons) a specific routine
and action is activated. This is the basis of the custom 3D printer design. The electronic
building block of Spatial 3D printing is the digital signal that activates different types
of pulses to communicate between mechanisms and to interphase with physical com-
puting. The movement code that determines the numerical control of the robot and the
positioning of the extruder End of Arm Tooling will control electronic mechanisms
of extrusion for the correct coordination of movement, extrusion and cooling system Fig. 2. 10 End of Arm Tooling.
variables. Photo ATI

165
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

166
 2. Free oriented additive manufacturing
Customizing 3D Print processes

2.1.3. Spatial additive Manufacturing

Current advances in robotically augmented additive manufacturing have re-
cently proved to significantly reduce cost and time of construction of complex
shapes (García de Soto et al., 2018). Spatial FDM is a method by which mol-
ten polymer sections are configured in a spatial pattern that has demonstrated
capability of fabricating functional products with optimized performance (Liu,
Li and Li, 2018) and reduce fabrication time of prototype prints by differentiat-
ing the local value of a printed specimen along extrusion process (Mueller et
al., 2014).

Advances in automation construction, deconstruction and re-customization

provide tools for reconsidering existing construction practices (Bock, 2015).
Novel computational design methods and fabrication techniques impulse the
emerging body of research of Spatial FDM at an architecture scale (Hack et al.,
2014; Garcia, 2017; Retsin and Jiménez García, 2017).

Novel algorithmic design methods and spatial additive manufacturing tech-

niques are required for large-scale manufacture of structurally continuous
complex surfaces.

Building on the notion of serializing the solutions to continuous deposition

constraints (Oxman, N. et al., 2012; Oxman, N. et al., Sep 2011; Soler et al.,
Nov 2017), we develop and test different computational methods for design
and fabrication based on local differentiation in order to:

• Achieve a wider understanding of the inherent constraints of our tech-

nique.

• Present the geometrical possibilities of introducing such technique in

an automated construction workflow.

• Illustrate the applicability of additive manufacturing (AM) in con-

struction.

The computational approach to manipulate individual units of information

(Sprecher & Ahrens, 2016) struts in cellular structures differentiating nodes
and serializing the solutions for successful additive manufacturing are either
discrete by voxelization or serialized by algorithmic growth (Richards & Mar-
tyn Amos, May 14, 2015; Soler et al., Nov 2017; Sprecher & Ahrens, 2016;
Wu, P. et al., 2016; Wu, R. et al., 2016). Discrete computational algorithms,
because of their defined boundaries, are suitable to all additive manufacturing
production scales.

167
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

CURVES
INPUT

4 4
1 (X0,Y0,Z0)
2 (X0,Y0,Z0) COORDINATE
3 (X0,Y0,Z0) EXTRACTION
4 (X0,Y0,Z0)
...
ELSE IF 2
GATE 0 0
Identify
coordinate 4 END 2 2 2 4
by Type coordinate
TYPE index item [-1]
travelling

3 REPEATED
coordinate
TYPE item nearby
knot previous item
4 4
2 NEW
3
TYPE
coordinate
item not nearby 3 3 3 3
mid-air previous item

2
1
TYPE
BASE
1 if Z<tolerance
2 2 2 2
bed
Collate with
Python
0 INVALID 1 1 1 1 1 1 1
TYPE
coordinate
index item [1]
4
None without support

Merge into
Gcode
GCODE OUTPUT

Fig. 2. 11 Spatial Printing generalization

Own work based on (Borunda, Ladrón de Guevara, Anaya 2019)

168
 2. Free oriented additive manufacturing
Generalization

2.2 Generalization
Spatial Fused Deposition (SFD), also referred as wire-print or spatial printing,
is a technique in which molten thermoplastic gets extruded and vitrified in
space almost instantaneously, so that a network of threads can be arranged in
different self-supporting configurations, responding, generally, to either lattice
or cellular typologies. The difference with conventional printing techniques
reside in the violation of the layer-by-layer deposition principle that additive
manufacturing commonly uses.

SFD has demonstrated the potential of optimizing functional products (García

de Soto et al., 2018), reducing fabrication time for rapid prototyping. Recently,
SFD has also demonstrated its capability of being applied at any orientation,
without the misgiving of potential deformations due to gravitational forces, in
a technique that uses a robotic arm, called Free-Oriented Additive Manufactur-
ing. This last feature is due to the lightweight properties of thermoplastics, and
it is driven by an algorithm that adjusts the printing parameters to any given
orientation.

Spatial additive manufacturing consists of four stages (Liu, Shuting et al.,

2018): horizontal, upwards, downwards and dangling. The main factors in-
volved in this process are deposition speed, extrusion flow, cooling system,
and collisions check. Spatial 3D printing does not rely singularly on the use
of a robotic arm; it can be also applied in commercial FDM instruments. 3D
printers can be altered to produce spatially printed designs facilitating the gen-
eralization of this technique. However, programming skills are needed, as the
generation of the code file that contains the instructions for the machine cannot
be based solely on using commercial slicing software.

A key finding is the printing procedure can be generalized into different opera-
tions:

i. bed extrusion,

ii. knot,

iii. mid-air, or

iv. end of line.

For SFD, the model is not sliced but rebuilt by following some of the afore-
mentioned typologies.
Fig. 2. 12 Design of a Roboti-
cally Manufactured Partition
Discretized into one continues
linear toolpath.

169
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Thermoplastic
pellets bulk
material

Thermoplastic
filament bulck
material

Infitite screw
setup in heated
barrel for direct
feed

Nema 17 motor
in Mk8 Type
setup for direct
feed

Induction heat
block

Heat block with

ceramic heat
cartridge

Typical FDM setup for Pellet extrusion Typical FDM setup for Filament extrusion

Fig. 2. 13 Typcal FDM systems. Nozzle constitution basic elements in two most common feed systems, pellet and fila-
ment, and table of thermoplastics applied in this research
Photos by the author. 3MT 3D printer Pellet extruder by WASP Iberia
Fig. 2. 14 Characteristics of typical thermoplastics used in FDM, tested in this research
Own work

170
 2. Free oriented additive manufacturing
Generalization

2.2.1. Materials
The system acts as a portable 3D printer, inspired in the simplicity and versatil-
ity of 3D Doodlers and positioned by a robotic arm instead of a gantry system.
The compact extrusion system that integrates all components at the end of the
arm tool (EOAT). A spool of 2.75mm diameter thermoplastic filament is fixed
at the EOAT and fed into an MK8 type extruder.

2.2.1.1. Printing Materials

Pellets as material source is found to be more common for printing at a robotic
scale, however, the FOAM technique would not have been successful using
a pellet-based system, as it would require a fixed vertical cannister to feed
the nozzle. An alternative strategy would be to create a pumping system that
would pump the material up to the EOAT, but this would have augmented
the cost of the tool prototype and also the time implementing the mechanical
functionality.

Instead, we use commercial 3mm filament rolls that can produce up to 4mm
thickness of printed material, depending on the triplet time-feed rate-velocity
parameters. For instance, if the velocity of the robot is set to 3mm/s, and the
feed-rate is set to 1/8 steps per second, the system would generate thicker ma-
terial than fixing feed-rate ratio but speeding up the robot motion. Tested ther-
moplastic filament rolls include PLA, PLA-PRO, ABS and PETG. While ABS
and PETG offer higher capabilities than PLA such as lightweight properties
(1.04, 1.23 and 1.24g/cm3 respectively), major elongation at break (22, 228
and 8%), and major impact strength (19, 8 and 4Kj/m) we found that PLA pres-
ents the best results as it maintains a successful melting point at a wider range
of temperatures, from 180 to 220 degrees Celsius —however, PLA does not
vitrify as quickly as ABS or other higher temperature thermoplastics. TABLE

FOAM benefits from a generous robotic reach, which scales up design op-
portunities and makes this technique freely and adaptable to reach any point
within the robotic work-cell. This makes printing optimal for flexibility to 3D
print on geometrically complex cases. FOAM was tested on a 6-axis 6640
ABB robotic arm sitting on a 6-meter length track, proving its viability through
different scenarios.

Applying spatial printing on commercial 3D printers is becoming widely pop- Fig. 2. 15 Dimensions of two
main nozzles used in this re-
ular. It is a cost-effective technique that offers materially efficient rapid proto- search.,
typing [7], skipping any hardship that the use of robotic arms might imply. In Minimum footprint nozzle with
ceramic heatblock. Nichrome
this research we used delta type 3D printers as their motors move faster along based heat block with ABS 3D
the Z axis than cartesian printers. The following section is divided into two printing protection and incor-
porated air canalization

171
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

HOT END SYSTEM Fuse system main characteristics

Heat block and flow design

typical FDM nozzle for layerwise

deposition

standard heat break,

heat sink and cooling

retract area melt zone

standard heat block

small scale flow nozzle

Thicker filament feed

Bowden or direct feed systems

threaded coupling
vitrification systyem Extended melting zone for retract
movements

Extended heat-block for flow augment

Reduced foot print to minimize

self intersection

Fig. 2. 16 3D print instrument genral components and adjustment, nozzle characteristics

Own work

172
 2. Free oriented additive manufacturing
Generalization

main parts, the first relates to hardware adjustments; the second, to software
considerations.

2.2.1.2. Hardware considerations

Although technically it is not strictly necessary to alter any major component
of a 3D printer, there are adjustments that yield a variety of improved results.
Air cooling is a main factor of spatial printing, as such, having air-hoses or air-
adaptors that point precisely at the tip of the nozzle is recommended. Hot-end
and nozzles have a major impact on print quality. This research uses bespoke
nozzles that adjust the printing thickness to the research needs. Custom nozzle
for FD delivery). The diameter of the aperture of the nozzle directly determines
the thickness of the printed extrusion, whereas the morphology of the hot-end
constraints the angle at which a lattice or cellular structure can be printed.
Thus, the narrower the hot-end the less constrained the print is. This affects
precisely the heating source and the thermal sensor. We tested multiple op-
tions, ranging from conventional 12V ceramic cartridges inserted into an alu-
minium block to a 22-gauge nichrome wire wrapping an M6 threaded barrel.
More sophisticated options such as a bespoke mechanized all-in-one hot-end
offer a higher precision and printing quality.

2.2.1.3. Software Considerations

Commercial 3D printing software cannot be used with this technique as they
rely on layer-wise deposition, a common constraint that SFD violates. Regard-
less of which algorithm construes a design typology from object to a suitable
print, for the success of this technique the g-Code generator should convert
XYZ coordinates to machine instructions. To achieve successful results, an
algorithm should weave moving commands, extrusion flow-rate, waiting times
at discontinuities, and cooling control. Due to a highly complex nest of instruc-
tions, and avoid potential downstream problems, we recommend the use of a
free source online viewer software that acts as a bridge between the generated
g-Code and the fabrication step.

2.2.1.4. Exploration of electromechanical means of

deposition
Understanding the logic and mechanism before designing and fabricating a
tool is efficiently effective. During a big part at the beginning of the research,
Fig. 2. 18 3D print tests in com-
the acquisition of smaller scale 3D printing machines provided with enough mercial 3D printer
knowledge for future fabrication steps. Prior research demonstrated that hack-
Fig. 2. 17 3D print tests in mod-
ing regular 3D printers results in faster printing process, speeding up the fabri- ified 3D printer to increment
cation time up to 10 times compared to traditional layer-based printing (Muel- extrusion thickness and nozzle
penetration

173
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 19 Modification to commercial 3D printer nozzles with extended tip attachments to extend heat block and mini-
mize footprint (top), Custom aluminum nozzle heatblock designed (bottom)
Own work

The heatblock introduces the necessary enery to melt thermoplastic as it is extruded, molten strut thermoplastic di-
ameter depends on the capacity to distribute heat correctly along a barrel mainting the fuse temperature. A correct
temperature read and correct heat incorporation is fundamental for the correct melting and clog-free deposition. 10
different heatblock systems and nozzles where custom fabricated and tested. The best results tend towards compact and
minimum footproint systems, with Direct Drive feed prefered (stepper mottor directly attached to barrel vs Bowden
(stepper motor connected via TPFE tube)

174
 2. Free oriented additive manufacturing
Generalization

ler, Im et al, 2014).

All necessary testing in order to learn about time and speed ratios was done on
a Hatchbox Alpha desktop 3D printer and a WASP 3040 desktop 3D printer
—a printer that activates the nozzle using six vertically actuated arms. Besides
testing, these machines fabricated components, such as nozzle protectors for
the robotic tool.

Some wireframe structures were tested and prototyped using a 3D printing pen.
We analyzed flow rates, anchor points over printed nodes and motion rates.

2.2.1.4.1 Design for fabrication

We found that hardware explorations inform how to approach to design from
an engineering perspective. Under a computational craft perspective, digital
design requires the embedding of potential constraints in the process of mak-
ing into the design. To create digitally, the act of design is fundamentally trans-
formed to a coordinated multidisciplinary effort.

Elements fabricated by extrusion methods are affected by post-printing con-

tractions due to material properties. Printed nodes in contact with the print-
ing bed tend to remain adhered, leading to potential problems of non-uniform
shrinking.

Printing nodes are those points in the space that act as joints or articulations
in the printing mesh. They can be thought as punctual connections between
linear elements —like articulations in metallic trusses. Such nodes are classi-
fied according to their functionality in the global structure: foundation nodes in
contact with the printing bed, structural nodes that offer support to neighboring
printing material, directional nodes that change the direction of printing and
end-of-segment nodes, where a discontinuity in printing happens (Borunda,
Ladrón de Guevara et al. 2019). Points in the lattice that do not fall into any of
these categories are labeled as not 3D printable and do not comply with stack-
ing requirements. Fig. 2. 21 3D printed custom
cooling ring design
The printing proceeds as described in Fig. 4, the (x, y, z) print-head mobility
for PE is the most common setup and the extrusion may be continuous or dis- Fig. 2. 22 Nozzle studies for
1.75mm filament with 3D pen,
continuous. Discontinuous extrusion is better for design freedom, as it affords MK8 standard extruder for
more flexibility to the process. 2.85mm and WASP pellet ex-
truder system.
Other parameters are constant flow between points in the space, waiting times
upon reaching a break in the continuity of a line (make a turn), and to avoid un- Fig. 2. 20 Simple extrusion sys-
tem with nichrome wire coiled
desired creep material depositions and controlling the dragging phenomenon to 6mm insulated threaded bar-
of previous deposited threads that generally happen when reaching to corners. rel to reduce nozzle thickness
and maximize geometric flex-
ibility.

175
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 23 Extruder system development for meso-scale testing

176
 2. Free oriented additive manufacturing
Generalization

Form is accomplished by encoding these operations for its fabrication within a

set of instructions that orchestrate the electromechanical pieces of equipment.
Discretizing the given shape into a list of concatenated coordinates is an es-
sential part of the design process. Initially for this research, Silkworm plugin
for Grasshopper in Rhinoceros 3D software and custom python gCode parser
(Fig. 5) are used for creating custom gCode with a set of basic 3D printer pa-
rameters.

2.2.1.5. FOAM Description

We maximize the tool’s compacity in order to achieve as much freedom in ori-
entation and movement as possible to reach complex existing infrastructures.
A compact unit is designed such that all needed components for printing are in-
tegrated at the EOAT, avoiding externalizing the thermoplastic polymer feed-
Fig. 2. 24 Basic barrel system
ing mechanism. Attenuation was achieved by elongating the distance between for nichrome-wire based heat-
the nozzle and the electronic box with a non-rigid 2” x 12” x 0.12” aluminium block
plate. The configuration of this element makes it bend under collision, protect-
ing the nozzle from break. Single or double-curved surfaces are inherently
more prone to jeopardize the tool than planar surfaces, as the algorithm that
defines the toolpath over the digital model should absorb a tolerance of 0.04”
with respect to the physical model. For the simplicity of this research, no scan-
ning tools or proximity sensors were employed.

2.2.1.5.1 End of the Arm Tool (EOAT) and thermoplastic

polymer extrusion system
This research utilizes components from commercial 3D printers to make a
cost-effective tool. A regular 1.75mm diameter MK8 extruder is hacked to
adapt an M6 threaded barrel, used to feed a 3mm diameter thermoplastic fila-
ment through its interior. The barrel is wrapped with a 0.64mm diameter Ni-
chrome wire and length of 800mm, achieving a resistance of 3.5 ohms to pro-
vide 12VDC 60W. This is enough to keep the nozzle at a melting temperature
—up to 300C— within the range of type of thermoplastics tested (PLA, PLA+,
ABS, PETG). The Nichrome wire wraps a barrel in its total length to ensure a
steady distribution of heat. A 0.33 mm thickness stainless-steel wire has also
been tested. To get the same resistance of 3.5 ohms, a length of 1500mm was
needed.

To protect the nozzle from the ambient temperature, ceramic-fiber insulation

and an ABS printed cover are needed. This piece holds four air-hoses that blow
pressurized air pointing at the tip of the nozzle. Best results are obtained by
keeping a distance of 250mm to 400mm between the air hose and the tip of the Fig. 2. 25 Advanced aluminum
barrel for ceramic based heat-
nozzle. An electronic box contains Arduino boards to control temperature and block

177
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 26 Free Oriented Additive Manufacturing final End of Arm Tool (EOAT) Design.

178
 2. Free oriented additive manufacturing
Generalization

feed flowrate. The logic voltage of the motor driver varies from 3 to 5.5v, hav-
ing a maximum current per phase of 2A if pressurized cooling air is provided,
or 1A otherwise.

Robotically Augmented Additive Manufacturing equipment consists normally

of a material extruder system and a cooling system. The equipment is designed
as a compact unit with all components directly attached at the EOAT to maxi-
mize freedom of movement.

2.2.1.6. Physical constraints

The process starts by setting the temperature of the nozzle at a temperature
that melts the filament. The robotic arm then moves the nozzle following a
given toolpath. Generally, straight segments are chosen as curvilinear elements
would require other technique to solidify the material in such shape. At every
polyline discontinuity, there is a wait time to keep the geometry as stiff as pos-
sible. The direction and taper of angles are directly constrained by the angle of
the nozzle. An important point during the fabrication process is the transition
of molten plastic to rigid state. A cooling system is needed to integrate the two
states. We use the internal pneumatic system of the robotic arm to inject air at
a desired pressure. We design a four-tube air injection and situate them at 30
mm distance from the Tool Center Point.

Each point of discontinuity in a given structure presents different wait time

scenarios, responding to the node’s classification, so that it ensures proper
bonding with already printed material. Furthermore, flow-rate adjustments are
needed to respond to tensile efforts that that are caused by the nozzle when
moving.

One of the challenges of this technique is to avoid potential collisions with ma-
terial that has already been printed. We employ some simple strategies that en-
sure successful results, such as offsetting the Z coordinates at the last point of
each segment. A more robust approach is to apply a backtracking algorithm in
the design stage which can calculate whether a collision might occur. Robotic
spatial printing requires the synchronization of robotic motion, stepper motor
feed-rate, necessary wait times at points of discontinuity, and cooling system.
An overall of 3 mm/s speed combined with an 1/8 motor step was proven
successful. The overall fidelity between the digital and physical models was
maintained at a ~93%, measuring this parameter based on overall deformation,
Fig. 2. 27 End of Arm Tool.
regardless the orientation of the object. Photo for 3D Spatial 3D print-
ing, bulk material (top), control
2.2.1.6.1 Robotic Relative reach mechanisms (middle), extru-
sion mechanism (bottom)
This section explains the relation between the end effector tool and the robot
Own work

179
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 28 Qualitative analysis of parameters for optimal coordination of Vitrification Mechanism, Speed of deposition /
Flow Rate, Feed Rate and Heating Temperatures relative to the spatial lattice segment typology

180
 2. Free oriented additive manufacturing
Generalization

on which is mounted. present a static setting, adding the material holder and
feeding system as a third element besides the robotic arm and the printing
tool. Alternative configurations have explored adding this element on the ro-
botic arm itself . Willingly, to expand the reach of the robotic arm and liberate
it from potential collisions within the work-cell, we add the material holder
and feeding system to the EOAT. Using this setting, the robot can be oriented
in any direction, only limited by the internal axis configurations of the robot
itself, being capable to print directly over non-horizontal and non-planar infra-
structures.

A side effect of this configuration is that the replacement of the filament spool
breaks an ideally continuous process, however, at the time of writing, no com-
plete automated printing techniques free the human from the manufacturing
process.

2.2.1.6.2 Robotic Absolute reach

As a consequence, the robot can use the entire work-cell’s volume without
further constraints. This potentially contributes to printing over existing infra-
structures, since the robotic arm is able to move over a track. Most importantly,
the robot’s position and motion do not depend on a fixed location of the feed-
ing system. Thus, the printable area increases substantially, bringing closer
scenarios in which a robot could move freely around certain infrastructures.

2.2.1.7. Digital and Analog Communications

2.2.1.7.1 PID, cooling system, and stepper synchronization

One of the critical factors that intercede in a consistent extrusion is the control
of temperature. For this, a proportional – integral – derivative (PID) controller
(or three term controller) is used to get a desired temperature. This is a widely
used mechanism used in industrial control systems. A PID controller continu-
ously calculates an error value e(t) as the difference between a desired setpoint
(SP) and a measured process variable (PV) and applies a correction based on
proportional, integral, and derivative terms.

An Arduino outputs a signal that is converted through a metal-oxide-semicon-

ductor field-effect transistor (MOSFET), which is a type of field-effect transis-
tor (FET). It has an insulated gate whose voltage determines the conductivity
of the device. In order words, it is used as a gate that controls the amount of
power in Watts that is loaded at the output wire. Other mechanism can be
used, such as a solid-state relay (SSR). This is an electronic switching device
that has no moving parts (it does not wear out, to the opposite of electrome-

181
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fuse and vitrification system general limitation

Collision Printable

Printing process and geometrical limitation

Fig. 2. 29 Self Collision principle. Critical geometric contraint for correct spatial deposition.
Fig. 2. 30 Custom 2.85 fillament extruder with nichrome wire 3D printed cap and multidirection cooling system
(Ladron de Guevara, et al. 2019)

182
 2. Free oriented additive manufacturing
Generalization

chanical relays). When using an SSR, the control signal must be coupled to the
controlled circuit in a way which provides galvanic isolation between the two
circuits. The type of SSR used in this research utilizes optical coupling. When
it switches, on a photo-sensitive diode turns on a back-to-back MOSFET to
switch the load. After testing both methods, we found the SSR to yield more
stable results.

2.2.1.7.2 Arduino-robot communication

The coordination and synchronization of the functions of the robot and the
Arduino require a communication device. This is done through a DSQC 651,
which is a circuit board normally mounted inside the robot controller, although
it could also be mounted in an external I/O module. The I/O (inputs/outputs)
unit handles digital and analog communication between the robot system and
any external systems (more information can be found in the 3HAC020676
ABB IRB6400TR official manuals).

• We use digital inputs due to time constraints. The communication be-

tween robot and Arduino only happens in one direction, from the robot to
the Arduino. The wiring is made through an ATI QC-110 plate. Particularly,
the interchange of information between both machines is directly related
to the geometrical information at both scales, meso and micro scale. The
robot has information about the location of each target and the stepper mo-
tor should respond to the direction of the printing path. For instance, if it
is printing upwards, the velocity of the stepper motor should be slightly
slower than the motion of the robot to tense the segment. On the contrary,
printing downwards might require quicker steps (Retsin and Jiménez Gar-
cía, 2017). The robot emits electrical pulses with different lengths that are
related to the target position that correspond to the geometrical pattern. The
Arduino then receives this pulse and interprets it accordingly actuating the
motor. It is critical to work with ranges instead of fixed values, as naturally,
this communication is noisy enough to make the transmission not legible.
In this research, successful values found to eliminate all noise range in the
10000 milliseconds. The Arduino processor circuit schematics that control
the stepper motor.

Custom electromagnetic commands take a certain pattern of D/O pulses and

are required for specific type of strut segments (upwards, downwards, dangling
or travelling) based on the vector orientation (direction pf start position and

183
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

184
 2. Free oriented additive manufacturing
Generalization

end point) and type of segment in the lattice according to the role of the end
point of the segment (mid air, knot, bed, end of extrusion, points). For example:

• Mid Air segment

slow extrusion is set for dangling geometries pointing downwards, with

low feed rate,

quick extrusion for dangling goemetries pointing upwards, with high feed
rate

• Joint extrusion

reaching a joint with lower Z coordiante: retract for preparing next down-
wards segment

reaching a joint with higher Z coordiante and increasing flow for preparing
subsequent upwards segment

• Bed extrusion

[0] printing the first layer. Z coordinate within a tolerance distance to base

[-1] finalizing first layer, final index element in list of coordinates

• End of “3D polyline” extrusion

stop

travelling

retract

Approximating next 3D polyline

185
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 31 Arduino setup coordinating motor actions and temperature reading

Basic Schematic of a 3D printer system used for this research. Heat block system in the left, Processor and motor cool-
ing mechanisms in the center, Feed - Flow system in the right. Mosfet transistors allow a gradient current draw, Solid
State Relays instead provide a discrete current draw and accumulate less heat.

186
 2. Free oriented additive manufacturing
Generalization

2.2.1.8. Tool components

2.2.1.8.1 Feeding system and Hot End

All operations are based in one Arduino microcontroller to operate the feeding
system and one Arduino microcontroller to operate a custom fabricated hot
end. A regular 1.75mm diameter MK8 extruder is hacked so it can work for
3mm diameter filaments. The MK8 has the following characteristics:

v. Extrusion nozzle: 0.3mm

vi. Material of Print: 1.75mm PLA/ABS

vii. Flow rate of Nozzle: about 24cc / h

viii. Sports shaft speed: 40mm/s

ix. Voltage of heating nozzle: 12V

x. Thermistor: 100K NTC

xi. Operating voltage of cooling fan: 12V

xii. Heating rods: 6mm, 12V, 30W

xiii. Net weight: 450g

xiv. Normal working temperature: 190°- 230°

The melting point of the print media is compounded by a print nozzle and a
heater mounted in an aluminum block. The hot end consists in regular com-
mercialized aluminum barrels drilled with a slightly bigger drill bit than the
orifice, so it removes the Teflon tube that lives inside and breaks the capped
end of the barrel. This allows to have a diameter of ~3mm, enough to make
the 2.85mm filament pass through. The barrel is wrapped with Nichrome or
stainless-steel wire, with 0.64mm diameter that provides with 12VDC 60W.
This is enough to keep the nozzle to a melting temperature within the range of
type of thermoplastics tested (PLA, PLA+, ABS, PETG). We create a longer
nozzle by joining two commercial barrels with a nut. This has the downside of
not having the wire wrapped at the areas where the nut threads the barrel. This
could be failsafe by wrapping the nut, although it won’t provide with the same
temperature at this area. This load is then charged on a 0.64mm Nichrome
wire, which in order to work at 60W, it has to have a length of 800mm, hav-
ing a resistance of 3.5 ohms, which is enough to achieve temperatures up to
~300C. This wire wraps a barrel in its total length to ensure a steady distribu-
tion of heat. We tested stainless steel wire instead of Nichrome. This requires

187
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 32 Temperature control and cooling coordination tests. correct coordination of fuse, flow and vitrification sys-
tems
Fig. 2. 33 High temperature creating burnt segments of filament
Own work

188
 2. Free oriented additive manufacturing
Generalization

the wire to have 0.33mm thickness, but 1500mm length was needed to achieve
the same resistance.

To keep the temperature stable, we need to isolate the barrel from the cool
high-pressurized air from the robot’s inner cooling system. Even though the
end of the air hoses points towards the tip of the nozzle, the barrel is exposed
to severe temperature drops. To avoid such problem, we wrap the barrel with
Kapton tape, a type of polyamide tape that supports high temperatures. Like-
wise, the heating block is protected with heat-resistant ceramic fiber.

We then protect the insulation layers from the cooling system with a wind-
screen. This windscreen is designed to support the 4 Teflon tubes that provide
cool air coming from the pneumatic system. Numerous iterations have been
done in this last apparatus, all of them, printed in regular 3D printers using
ABS thermoplastic, as it supports higher temperatures than PLA. A distance
between the air hoses and tip of the nozzle between 250mm to 500mm provide
best results.

2.2.1.8.2 A4988 Micro stepper motor

An Arduino Uno controls the feeding system. As described above, an MK-8
commercial stepper motor was used to feed the nozzle. The Arduino sends
different signals that are translated into on, off, different ratio steps/time and
direction of the motor steps (counter-clock wise, or clock wise).

We use an A4988 micro-stepping driver which includes a built-in translator

for easy operation. This means that it is possible to control the stepper motor
with 2 pins from this controller, one manages rotation direction and the other
controls the steps.

The driver has five different step resolutions: full step, half, quarter, eight and
sixteenth step. Also, it has a potentiometer for adjusting the current output,
over-temperature thermal shutdown and crossover-current protection.

Its logic voltage ranges from 3 to 5.5V and the maximum current per phase
is 2A if good addition cooling is provided or 1A continuous current per phase
without heat sink or cooling. The eight-step resolution has demonstrated to
work more accurately.

The control of the feeding system is a negotiation between the driver resolution
and the delay on the steps loop in the code of the Arduino board.

189
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 34 Nichrome coil resistance heating with minimum nozzle taper size and distribution of heat around stainless
steel barrel isolated with kapton tape.
Own work

190
 2. Free oriented additive manufacturing
Generalization

2.2.1.9. Tool operation

2.2.1.9.1 Thermal energy system

A controlled thermal energy source is required to fuse the thermoplastic to a
liquid state. One of the critical factors that intercede in a consistent extrusion is
the control of heat outcome to maintain a heat block at the temperature range
at which the polymer melts. For this, a Proportional – integral – derivative
controller (PID) is used to incrementally approximate the desired temperature.
A PID controller continuously calculates an error value e(t) as the difference
between a desired setpoint (SP) and a measured process variable (PV) and ap-
plies a correction based on proportional, integral, and derivative terms.

This research has explored two different methods that are used as PID control-
lers. The first one is a standardized commercial PID. Normally, these types of
controllers use a thermocouple sensor (type K) to measure the temperature of
the nozzle. The second and most used approach in this research takes an Ar-
duino Uno with a PID algorithm coded in it. This approach works better with
thermistors.

The signal that the Arduino outputs is converted through a metal-oxide-semi-

conductor field-effect transistor (MOSFET), which is a type of field-effect
transistor (FET). It has an insulated gate; whose voltage determines the con-
ductivity of the device. In order words, it is used as a gate that controls the
amount of power (Watts) that is loaded at the output wire. Other mechanism
can be used, such as a solid-state relay (SSR). This is an electronic switching
device that has no moving parts (it does not wear out, to the opposite of elec-
tromechanical relays). Using a solid-state relay (SSR), the control signal must
be coupled to the controlled circuit in a way which provides galvanic isolation
between the two circuits. The type of SSR used in this research utilizes optical
coupling. When it switches on a photo-sensitive diode turns on a back-to-back
MOSFET to switch the load.

2.2.1.9.2 Temperature measurement

We attach a temperature sensor —thermistor— adjacent to the barrel wrapped
with Nichrome wire. A thermistor is a resistor that changes value (non-linearly)
based on the temperature. The type of thermistor used is an NTC or negative
temperature coefficient. NTC thermistors decrease resistance as temperature
rises. Due to the fact that microcontrollers do not have a built-in resistance-
meter, but a voltage reader (known as an analog-digital-converter), converting
resistance into a voltage is required. Adding another resistor and connecting
them in series will work, so when the resistance changes, the voltage changes

191
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Vitrification system
Cooling and air flow design

23.02 24.76 26.15

23.53 23.53 23.53
14.94

15.86

15.86

24.35
37.00

37.00
26.21
26.16 26.16 26.53 27.68 23.52
10.70
32.00

37.00

37.00

70.73

70.73

Fig. 2. 35 Tests of cooling and filament airborne vitrification systems

Fig. 2. 36 The air flow for correct filament vitrification is paramount in spatial 3D printed. Diagram depicts different
typologies tested with radically different results.

192
 2. Free oriented additive manufacturing
Generalization

too, according to a simple voltage-divider equation. We recommend to keep

one resistor fixed.

For instance, if the fixed resistor is 10k and the variable resistor is R, the volt-
age output (Vo) is:

Vo = R / (R + 10K) * Vcc

Where Vcc is the power supply voltage (3.3V or 5V)

It is connected to a microcontroller. When measuring a voltage (Vi) into an

Arduino ADC, the result is a float number.

ADC value = Vi * 1023 / Varef

Combining the two (Vo = Vi):

ADC value = R / (R + 10K) * Vcc * 1023 / Varef

If Vcc (logic voltage) is the same as the ARef, analog reference voltage, the
values cancel out.

ADC value = R / (R + 10K) * 1023

Finally, R (the unknown resistance) is the result:

R = 10K / (1023/ADC - 1)

However, Arduino boards interferences may vary the result, reducing signifi-
cantly the accuracy. An implemented solution is to use the 3.3V voltage pin
as an analog reference. The 5V power supply comes from a computer’s USB,
making the signal noisier (as more than one task in parallel is being sent) than
the 3.3V power supply (it goes through a secondary filter or regulator stage).
An alternative is to take more readings and average them. This is especially
useful as some readings fluctuate to outliers’ peaks, out of a natural range.

In order to convert resistance to temperature, a simplified B parameter equa-

tion of the Steinhart-Hart equation is used where:

To: 25C = 298.15 K (room temperature),

B: 3950 (Coefficient of thermistor),

Ro: 10Kohm (Resistance at room temp)

2.2.1.9.3 Feeding system

Another crucial system in order have consistency while printing is the motor
that feeds the filament into the nozzle. The flow rates the stepper provides must
be in sync with the motion of the robotic arm. The toolpath segments must
correspond to a different feeding speeds depending on the direction of the tool-

193
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 38 Qualitative analysis of parameters for optimal coordination of Vitrification Mechanism, Speed of deposition /
Flow Rate, Feed Rate and Heating Temperatures relative to the spatial lattice segment typology

Fig. 2. 37 Initial tests exhibit that slow speed of deposition and flow rate are necessary for thermoplastics that fuse at
low temperatures. The difference in temperature affects thermoplastics that fuse at higher temperatures to the benefit
of vitrification but become increasingly more dificult to fuse to a non tempered surface

194
 2. Free oriented additive manufacturing
Generalization

path. For instance, if a horizontal segment is being printed, and it represents

the connection with the base or prior printed segments, the motion of the robot
should decrease while the ratio steps/time of the stepper motor should increase.
When printing vertically, the motion of the robot should be slightly bigger than
the feed rate, as it would create tension in the segment, avoiding sagging ef-
fects. When printing the end of a vertical or diagonal segment, the robot must
wait to provide with enough time cooling the filament in order to continue.
Depending on the situation, the stepper must continue feeding the filament
at a small rate or should stop the extrusion (if the current node is the end of a
continuous segment or the following segment is shorter than the previous one).

2.2.1.9.4 Vitrification: cooling system

Fused filament vitrification (molten polymer transition to a solid state) is the
third crucial aspect during the fabrication process. A cooling system is needed
to be able to make this process happen at the same time as the printing is hap-
pening. For this system, we use the pneumatic system the robot has internally
to inject air at a desired pressure. This variable becomes really important to
master as minor changes produce major consequences, leading to undesired
low fidelity geometries with the original design.

Vitrification and welding processes present high sensitivity towards hardware,

and after extensive iterative prototyping, we found that four air injection tubes
with pressure regulation connected to a 3D printed nozzle directed to the fused
filament at 30 mm distance from the TCP (Tool Center Point) work efficiently.
The 3D printed nozzle protects the hot end of cooling prior to deposition to
avoid the obstruction by solid filament at the nozzle throat.

Cooling systems in spatial 3D printing methods are one of the crucial com-
ponents to consider when designing this part of the tool. In this version of
the tool, an ABS 3D printed ring-shaped object has been designed in order to
hold the PTFE tubes coming from the pneumatic system of the robot through
Digital Outputs. Due to the fact that the projected air points at the tip of the
nozzle, a secondary layer of protection needs to be considered in order to pro-
tect the heated barrel from the air. A conic shape is incorporated to the ring as
a windshield element.

Numerous iterations have been done in this last apparatus, all of them, printed
in regular 3D printers using ABS thermoplastic, as it supports higher tempera-
tures than PLA. Results have shown that the distance between the air hoses and
tip of the nozzle are best between 250mm to 500mm .

195
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 39 The digitization of build space allows to define collaboration operations. The coordenate systems are continu-
ous and precision is milimetric. Robotic collaboration test printing a continuos lattice along 5 directions.

196
 2. Free oriented additive manufacturing
Generalization

2.2.1.9.5 Tool Orientation

The tool is taught such that the Z elongation of the tool definition is perpen-
dicular to the ground, being the ground the Work Object base, normally, with
positive Z pointing up, following the right-hand rule. For large-scale proto-
types and testing, the tool has been inclined when working with more than one
work object, and normally when the tip of the nozzle had to touch ground, in
order to avoid self-intersections. For each of the three case studies, the extruder
has been positioned perpendicular to the work object, however, in parallel on-
going study cases, we are testing the fabrication tilting the tool.

2.2.1.9.6 Robot toolpath

To finish the computation process, a robotic fabrication toolpath needs to be
coded in order to make the geometry printable. In regular 3D printing process-
es with standard 3D printers, the fabrication file is first generated by convert-
ing the model into an STL file format (an abbreviation of Stereolithography),
which describes a given geometry in a raw, unstructured triangulated surface
by the unit normal and vertices of the triangles using a three-dimensional Car-
tesian coordinate system. A generator —normally open-source software such
as Slic3r— receives a given STL file and generates a G-Code file (i.e., the
instruction language used by 3D printers) for the production of the 3D mod-
eled object. Robotic toolpaths require the custom engineering of deposition.
The robotic deposition is fundamentally the act of coordinating a list of x,y,z
points of movement with electromagnetic commands of deposition. A full list
of points will approximate a layer by layer or spatially produced form.

Our fabrication methodology involves robotic fabrication, making the genera-

tion of the fabrication file take an alternative way. Specifically, the research in
this thesis used the ABB IRB 6640 robotic arm, situated at the digital fabrica-
tion lab (dFAB), at the School of Architecture at Carnegie Mellon University..

ABB uses a high-level programming language called RAPID. The fabrication

code is a combination between Arduino functions and ABB robot functions
(this generally entails motion instructions such as moveAbsJ, moveJ or moveL
commands, and inputs instructions that controls features such as pneumatic
system, electric system, and any added system a robot might have). For this
research, only pneumatic and electric system are used. Various methods can be
taken in order to write a module that contains procedures and definitions that
actuate the robot.

1. On-line programming: requires the use of an IRC5 Controller (or Fig. 2. 40 Virtual environment
tests in Grasshopper HAL plu-
teach pendant). It is the most popular method of robot programming. Accord- gin to allow simulation and
multi-robot collaboration.

197
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 41 System depends on finding a previous point to print upon. A more robust deposition procedure and lattice
structure allows some tolerance in finding points to print upon, either by repetition or by grating larger areas of contact

198
 2. Free oriented additive manufacturing
Generalization

ing to the British Automation and Robot Association (BARA), over 90% of
robots are programmed using this method. The logic of this program can be
generated using either a menu-based system or simply using a text editor. The
main characteristics of this method is the means by the robot is taught the po-
sitional data.

2. Off-line programming: this method is also known as simulation. It

ensures that advanced control algorithms are operating correctly before mov-
ing them onto a real robot. It is also used to improved efficiency, as it allows
a much faster re-configurability condition. Similar to the way in which CAD Fig. 2. 42 Continuoslyadapting
direction on extrusion. FOAM
systems are being used to generate numerical control programs for milling applied over a non-planar sur-
machines it is also possible to program robots from CAD data. face.

This research uses both methods with clear distinct purposes. The on-line
method is used to teach tools and work objects. Setting the home position as
a house keeping procedure is also taught here. For the rest of the operations,
an off-line approach is used. There are a variety of software that can be used
to off-line program robots. ABB has its own software, RobotStudio, which is
highly recommendable, as it is highly reliable when checking the kinematic
solvers and collisions. However, a perhaps easier alternative is the use of HAL
plugin for Grasshopper, Rhinoceros. We need to configure the data of the ro-
bot, track, as well as tool, work objects, with the addition of adding meshes in
some components to have a higher fidelity and reliability in simulation. HAL
does the engineering behind the scenes to solve the robotic configurations and
provides the fabrication file in RAPID language. While this could be sent di-
rectly to the controller, bypassing RobotStudio software is extremely discour-
aged, as sometimes HAL does not detect internal collisions nor configurations.

Fused Deposition Methods have demonstrated its usability as a technology for

small scale fabrication. Variations of normal layer-by-layer fused deposition
methods consist in extruding filaments in the space, in which the plastic is mol-
ten and solidified in the air. This approach is an alternative that potentially per-
mits the building of lighter structures at a reduced cost and time (Mueller, Inn
et al, 2014). Examples of this approach has been primarily studied at the ETH,
Gramazio Kohler Research or at The Bartlett School of Architecture, UCL.

This thesis explores the fabrication through this method, using different type of
materials, such as PLA, PLA +, PETG or ABS. Thermoplastics are commonly
used in this method as they require minimal heating energy in order to melt
(Table 1).
Fig. 2. 43 Continuoslyadapting
direction on extrusion. FOAM
applied over a non-planar sur-
face.

199
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 44 Continuosly adapting direction on extrusion. FOAM applied over a non-planar surface.

Fig. 2. 45 The compact design of the EOAT allosws fiber deposition in varying directions and onto varying bed surface.
Tool operating within a 80cm space between vertical walls.

200
 2. Free oriented additive manufacturing
Generalization

2.2.1.9.7 Spatial extrusion

Although some large-scale tests have been tested, for each of the study cases,
short rectilinear segments are chosen in order to design and fabricate. The pro-
cess starts by setting the temperature of the nozzle at a temperature that melts
the filament. The robotic arm moves the nozzle following a given toolpath.
While this technique allows a high degree of freedom of movement, there are
some constrains. For instance, printing vertical from top to bottom is not pos-
sible as the nozzle would collide with the filament printed. The direction and
taper of angles are directly constrained by the angle of the nozzle. For study
case I and II, the toolpath follows a 2D version of the octet-truss patented by
Buckminster Fuller in 1961 (Ashby, Deshpande and Fleck, 2001). This is a
triangle-pattern that forms layers of trusses. In order to add a higher structural
strength and continuity at the nodes, the triangular structure is broken at its
vertices, adding short horizontal lines that provide a larger supporting area for
upper layers. The logic of the pattern would be the following: lower horizontal
segment – diagonally upwards segment – horizontal higher segment – diago-
nally downwards segment.

2.2.2. Computational methods

2.2.2.1. Design
We tested two different algorithms for the design of the workpieces. The first
one is a top-down approach. A volume is discretized into smaller cubical units
or voxels forming a cell-like structure. An underlying multi-size 3D grid, simi-
lar to a customized oct-tree algorithm, determines whether a voxel lies inside
or outside the volume. We design a set of different polylines at each voxel
responding to design criteria. The algorithm computes the overall set of indi-
vidual geometries or units, in response to parameters such as opacity, structural
behavior and material properties. The principal part of the algorithm is to en-
sure that every unit is printable, and there are not any collisions with already
printed neighbors. A backtracking algorithm is implemented to find a printable
solution.

The second design method consists on a bottom-up lattice-based computa-

tional approach which takes non-uniform rational basis splines (NURBS) as
a base geometry. These two splines, say, A and B, are laid next to the other as
if forming the two sides of a wall’s section. The algorithm creates a NURBS
surface out of each spline to further extract a designed lattice structure based
on the repetition of the following pattern: short supporting segment – diagonal
downwards segment – short supporting segment – diagonal upwards segment.

201
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

ANNEX II - GEOMETRIC MANUFACTURABIL- ANNEX III -DIGITAL INPUTS AND OUTPUTS

ITY CONFIGURATION & DESIGN CODE ELECTROMECHANIC COMMANDS CODE
def isValidFloor(self, (i,j,k)):
#return True
testGeo = self.geos[i][j][k]
floorLst=[]
floorPointLst=[]
if testGeo == None or i ==0:
return True
if i > 0:
#APPEND ALL FLOOR CRVS TO STUDY
floorLst.append(g.geos[i-1][j][k])
#ORTHOGONAL FLOORS
if j != 0: floorLst.append(self.geos[i-1][j-1][k])
if j != rows-1: floorLst.append(self.geos[i-1][j+1][k])
if k != 0: floorLst.append(self.geos[i-1][j][k-1])
if k != cols-1: floorLst.append(self.geos[i-1][j][k+1])
#DIAGONAL FLOORS
if k != cols-1 and j != rows-1:
floorLst.append(self.geos[i-1][j+1][k+1])
if k != 0 and j != rows-1:
floorLst.append(self.geos[i-1][j+1][k-1])
if k != 0 and j != 0:
floorLst.append(self.geos[i-1][j-1][k-1])
if k != cols-1 and j != 0:
floorLst.append(self.geos[i-1][j-1][k+1])

for geo in floorLst:

if geo != None:
for crv in geo:
if type(crv) == type(line):
stpoint=rs.CurveStartPoint(crv)
endpoint=rs.CurveEndPoint(crv)
floorPointLst.append(stpoint)
floorPointLst.append(endpoint)

for crvTest in testGeo:

if crvTest != None and type(crvTest) == type(line):
#CHECK AGAIN SO GENERAL LOOP (AFTER OPACITIES) ALSO VALID TEST
stPointsTest= rs.CurveStartPoint(crvTest)
for floorPoint in floorPointLst:
if stPointsTest == floorPoint:
return True
return False

def isValidRowsCols(self, (i,j,k)):

#return True
testGeo = self.geos[i][j][k]
rowsColsLst=[]
if testGeo == None :
return True
#STUDY PREVIOUS THUS “PRINTED” COL, ROW AND COLROW
if j > 0: rowsColsLst.append(self.geos[i][j-1][k])
if k > 0: rowsColsLst.append(self.geos[i][j][k-1])
if j > 0 and k > 0:
rowsColsLst.append(self.geos[i][j-1][k-1])
#including diagonal +col+row in case of corner intersectino
if k != cols-1 and j > 0 :
rowsColsLst.append(self.geos[i][j-1][k+1])
#EVALUATE INTERSECTIONS OF NOZZLE WITH PRINTED CRVS

for geo in rowsColsLst:

if geo!= None:
pointsCollision=[]
testPoints=[]
for crv in geo:
if type(crv) == type(line):
points = rs.CurveEditPoints(crv)
for point in points:
pointsCollision.append(point)

for crvTest in testGeo:

if type(crvTest) == type(line):
points = rs.CurveEditPoints(crvTest)
for point in points:
testPoints.append(point)
for testPoint in testPoints:
upperPt=testPoint[0],testPoint[1],testPoint[2]+resolution*zShift

for point in pointsCollision:

test = rs.Distance(upperPt,point)< (resolution-2)*zShift

if test:
return False
return True

ANNEX 1 - ROBOTIC TOOLPATH CREATION CODE BASED

ON LATTICE TYPE CLASSIFICATION
ANNEX 1 - PLACE PSEUDOCODES

def isValidFloor(self, (i,j,k)):

#return True
testGeo = self.geos[i][j][k]
RAPID CODE

Fig. 2. 46 Code extract.

Deposition mechanism Arduino code extract for deposition based on point typology and direction vector. The code
reads Digital Inputs from Source: (robot) to coordinate electromechanic elements for filament deposition activating
two systems depending on geometry to produce: feed- flow system, and heati system. Cooling systems and positioning
of nozzles are controlled by the Source: (robot). Code described in Annex.
Own work

202
 2. Free oriented additive manufacturing
Generalization

The algorithm then weaves the two lattices connecting the nodes of the lattice
A with the nodes of the lattice B, giving different densities upon request.

2.2.2.2. Fabrication
We recommend a tectonic approach that considers the affordances and limi-
tations of the material and physical tool when designing for this technique.
However, anticipating constrains does not ensure a clean and failsafe fabri-
cation process. A feedback loop between design and fabrication processes is
implemented to overcome such unforeseen problems. The main challenge in
spatial printing is that the nozzle has to respect already printed material in
order to prevent collisions. For instance, in the algorithm used for the cell
structure, a detector for collisions should be checked between voxels. A simple
strategy for lattice structures is to implement an offset in the Z direction, higher
than the height of the voxel is given at the first and last points of the polyline
at each unit. This ensures that the nozzle never collides with already printed
geometries during non-printing motions.

The taper of the nozzle constraints the design pattern. For instance, two vertical
segments must have a separation that equals the width of the nozzle. Note that
printing downwards become impossible when the slant of the print is steeper
than the taper of the nozzle.

We use a plugin for Grasshopper called HAL Robotics, in which the work cell
is digitalized, and the robotic arm specifications introduced. A Python com-
ponent weaves the standard robotic commands such as type of motion, axis
configurations or speed with the digital outputs that the microprocessor needs
in order to output different orders to the printing tool. All the required compu-
tation, both in the design and fabrication parts are written in Python.

2.2.3. Application of spatial additive

manufacturing to architectural-scale shells
The manufacturing by digital methods is intimately linked to custom tool de-
velopment. Robotic means of construction have significantly evolved from ini-
tial speculations (Bock, 2008), particularly in the direction of a generalization
of kinematics in providing more efficient solutions for applications.

Highly specialized methods are difficult to develop and are regularly prepared
for a specific application.

Construction systems like (Maeda, May 1994; Yamazaki & Maeda, 1998)
show a promising augment of productivity, but only at a great cost. A common
current trend is to look towards the robotic arm – EOAT combination.

203
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

End Effector Equipment

Equipment element Units
Nozzle opening 0.3 mm
Material 2.8 mm
Nozzle Flow Rate About 50cc/h
Speed 5- 40 mm /s
Heating block Stainless Steel Barrel
Heating System 3.5 ohm Nichrome Wire resistance
30 W, 12V DC
Temperature Control 100K NTC Thermistor
Arduino PID control closed loop
Cooling System
Heating Rods None
Net Weight
Working Temperature 190 ºC – 210’ ºC

204
 2. Free oriented additive manufacturing
Generalization

The kinematics together with an extensive set of libraries make the robotic
arm concept ideal for the simpler application of robotic means in construction,
but the benefit of bypassing the development phase comes with the penalty of
scale. Limited to the work cell area of the robot, large-scale operations will
require novel techniques to be developed.

Currently, a second trend can be identified in the fact that the emerging sci-
ence of construction automation has mainly inherited logics of the more robust
and more readily available manufacturing techniques of the automobile, naval,
military, and other industries like such (Lynn, Greg & Foster, 2011).

FOAM carries a custom EOAT tool that provides capacity of additive produc-
tion at a larger scale than current commercial means of manufacture. Coupled
with a 7th axis of movement, the production space is virtually extended indefi-
nitely in a linear manner. Challenges of applications, particularly outside of the
laboratory, remain unresolved. Efforts in robotic displacement in surface by
teams of mobile robotics present a change in paradigm of print while in move-
ment with great potential (Helm et al., Oct 2012; Zeng et al., Dec 2019). Ef-
forts in assembly of distributed mechanisms linked by virtual space and design
orchestrate automated production by the more traditional method of scaffold
(WASP, ).

A promising solution of smaller distributed robotic additive manufacturing that

uses previously deposited material as scaffold and structure has been tested
at small prototype scale. This research builds upon aforementioned strategies
with a use of the mathematical and form finding capacity of shell structures;
shells are structures defined by three dimensional surfaces, with the particu-
larity of having one dimension meaningfully smaller compared to the other
(Bradshaw et al., 2002).

When shells are design with minimized bending forces, shell structures work
at simple axial compression, or tension (membranes). Coupling the scaleless
(pre-scaled processes and pre-scaled geometries) property of structures at pure
tension such as certain shells and membranes, and spatial structures, this re-
search will study the capacity of shell production (in principle for the 2 main
dimensions) in sets of linear batches of spaceframe like trabecular arrange-
ments (for the body or mass), both continuously on-site and discrete off-site in
the following chapters.

2.2.3.2.1 Coding Form - methods for improving Performance by

Geometry
Geometry is the most influential part in determining the strength of a cellular
membrane (Keaveny et al., 2001; Torres et al., 2019).

205
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 47 Design Process for Lattice Structures.FOAM test with double curvature surface

206
 2. Free oriented additive manufacturing
Generalization

Architecture materials and ultra-lightweight structures of high stiffness and

strength in derived from bone inspired cellular structure have demonstrated
superior mechanical properties through geometry rather than composition in
small and meso applications (Ashley 2019). Certain geometries that conform
porous assemblies such as Triply Periodic Minimal Surfaces (Podroužek et
al., 2019) and spatial polyhedral (Ashley 2019) show exceptional mechanical
properties due to their complex shapes (Qin et al., 2017), and are constructable
by additive manufacturing.

This research studies advances in understanding microarchitecture materials

(Fleck et al., 2010) and cellular structures such as foams (Zhu et al., 2018) and
lattice structures transferred to large scale membrane design. The purpose of
developing this technique is to provide an instrument capable of fiber deposi-
tion of variable orientation, flow and feed to create architected trabecular struts
that conform cellular membranes.

Given the large-scale of the robotic arm, discrete and continuous computa-
tional models where tested successfully by 3d printing 3mm thickness a ther-
moplastic matrix onto a 1m by 0.50m formwork with gradual angle variations
of 35, 30 and 25 degrees respectively , testing the possibility of automating
variably oriented robotic additive manufacturing practices.

Additive manufacturing techniques applied to the production of fibrous ma-

trices for material differentiation and engineered anisotropy can define local
performance differentiations, fundamental in designing distribution of forces
and monocoque behaviour. Spatial additive manufacturing has numerous ap-
plications such as tendon pre-stressing sheaths, light transmission, porosity,
and fibrous reinforcement. The larger print space in industrial systems enables
the continuous spatial deposition of engineered reinforcements according to
specific load conditions. The digital model makes positioning accurate

In order for a standard fan cooling system to vitrify filament deposition accu-
rately, a speed of 1500 mm/min for straight segments and a speed of 800 mm/
min for curved segments is implemented. More sophisticated cooling systems
such as vortex tubes may significantly speed up this process.

With six degrees of freedom (DoF), the robotic arm positions and actuates the
End of Arm Tooling (EOAT), and the machining tool performs a specific task.
Communication between the robot and the EOAT mechanism requires a digital
or mechanical drive. The digital actuation is carried out by means of the in-
terpretation in a processor of weak signals programmed within the deposition
path of the robot. The processor interprets the weak signals by actuating each
of the EOAT’s electromechanical components. The image shows the custom

207
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 48 If fused, spatial lattices and 3D print builds are suceptible of being printed regardless of gravity by adjusting
flow parameters based on directo of deposition.

208
 2. Free oriented additive manufacturing
Generalization

extrusion system and the physical computing weak signal communication dia-
gram developed for the Arduino platform.

The signal emitted by the robot is interpreted by the Arduino through the dif-
ferentiation of lengths to activate extrusion equipment with weak signals of the
Pulse Wave Modulated Signal (PWM) type.

Robotic collaboration is possible through the implementation of a general

coordinate system, so that robotic mechanisms share the same virtual space
(Ladrón de Guevara et al. 2019).

Fig. 2. 49 Design of a Roboti-

cally Manufactured Partition
Discretized into one continues
linear toolpath.

209
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 50 Free Oriented Additive Maufacturing by seven axis robotic deposition. Reserach at Carnegie Mellon Univesr-
ity in collaboration with the Master of Advanced architectural Design and the Design Fabrication Laboratory.

210
 2. Free oriented additive manufacturing
Generalization

2.2.4. Case Study

Two main case studies have been tested to accommodate both structures. While
a cell structure is tested by conventional 3D Spatial printing principle, fixing
the nozzle at the Z axis, a lattice structure is tested at different orientations us-
ing two configurations: implementing a multi-robot workflow, where a second
robotic arm moves the work object, and implementing in-situ over an existing
ceiling infrastructure at 1:1 scale, rotating the horizontal plane 180 degrees,
and achieving full automation.

2.2.4.1. Lattice Structure

Lattice structures are generally organized by rows of repeated patterns. We
first print a cell structure, and its results informed the design of the pattern for
the lattice structure. For instance, an equilateral-triangular pattern has angles
of 60 degrees, which, while being within a printable range, it presents some
drawbacks such as lack of horizontal support for upper rows. This presents
associated risks like a decrease in the height of each row, since acute angles
are prone to bow towards the direction of printing due to tensile strength of the
nozzle pulling towards the opposite direction. A workaround is to introduce
small horizontal segments at the vertices, such that any 60-degree angle would
be converted into a 120-degree obtuse angle.

We test the behavior of the material at a smaller scale, by using two synchro-
nized robotic arms so that one of the arms fixes an MDF board at a given orien-
tation. The robotic arm holding the printing tool starts printing the lattice until
the robot holding the board changes its position and orientation. The printing
robot readjusts its position to continue printing from the last point printed at
the previous position. This test proves two major breakthroughs:

- The algorithm adjusts the speed of the flowrate and applies a retrac-
tion force that depends on the orientation of the tool. This immediately coun-
terweights the gravity force in the printing process.

- The same lattice structure has been tested over horizontal, vertical
and four other different orientations. With the right parameters, no difference
in material quality nor geometry deformations are observed.

This outcome makes us believe that spatial printing is breaking established

constraints that delay its application into the construction industry. Improve-
ment in the design machine and fabrication algorithms help resolve fabrication
hurdles that are natural to FDM. Fig. 2. 51 Continuoslyadapting
direction on extrusion. FOAM
The following case study is informed by prior printing tests. We create a fully applied over a non-planar sur-
face.

211
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 52 Movement diagram. Laeral view of Custom fabricated extrusion equipment connected to EOAT. Compact End
of Arm Tool maximized depositiion capacity.
Source: Author

212
 2. Free oriented additive manufacturing
Generalization

automated set-up using an ABB IRB 6640 arm, with a maximum reach of
2.55m and a wrist torque of 1324 Nm. The workpiece bonding box’s dimen-
sions are 2x1x0.3m, fabricated as a single piece, having its base on an MDF
board attached to a metallic beam structure on the ceiling at a 4.20m height
from the floor in the dFAB, at Carnegie Mellon University. The robotic arm sits
on a 6m track, facilitating its reach.

The lattice structure is divided into three main parts. The main patterned rows,
a continuous cord that lies over each patterned row, and diagonals that weave
and connect the nodes of each side of the lattice. The robot is then actuated so
it varies its motion velocity depending on these three main parts and also on
the grammar of the pattern, following the same instructions:

- Vertical downwards segment, air, retract, wait time 1

- Short horizontal segment, air, wait time 2

- Diagonal upwards segment, air, wait time 3

- Short horizontal segment, no air, wait time. 2

- Diagonal downwards segment, air, retract, wait time. 4

The piece was completed in 40h, printing each patterned row in 18 minutes,
each continuous cord in 8 minutes, and each weaving sequence in 35 minutes.
Different wait times are implemented depending on the geometrical part, and
the start and stop of the injected air is also crucial in this process. The workflow
is automated, and the human interaction occurs when changing the spool and
when pausing, reacting the work.

Although performance results conclude that 3d printing is capable of produc-

ing functional light complex cellular assemblies, these initial tests require very
large times to build so high resolution and “non-informed” deposition tool path
are assessed.

The first stage of the research concludes that it is necessary to break the con-
straint of horizontal stacking principal to achieve reasonable printing times for
large build so the effect of orienting the extrusion direction on non-layer spatial
fused deposition modeling is studied in next stages.

The possibility of adapting the extrusion technique to bond to non-additive

manufacture elements such as existing infrastructures provides greater flex-
ibility of design, better print times and retrofitting capabilities. A novel Ad-
ditive Manufacturing application of Fused Deposition Modeling on varying
spatial conditions, varying deposition direction and varying geometry to adapt
to complex infrastructure conditions is developed.

213
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 53 Profile of robotic manufacturing device, 3D print operation in auto mode. The robotic mechanisms operates
automatically and at safe distance from supervision post
Source: Author

214
 2. Free oriented additive manufacturing
Case study discussion

2.3 Case study discussion

Free Oriented Additive Manufacturing (FOAM) provides a fabrication tech-
nique for architectural speculation of previously unexploited spaces, providing
the means for reconfiguring and bonding qualities to new or existing infra-
structures. It has the potential to produce and mass-customized complex forms
of bespoke performance.

FOAM presents a first step to define a framework of bonding and application

of functionally graded components in architecture, computational methods
of resolution-based discrete units’ concatenation and the constructability by
FOAM continuous spatial extrusion basis for an efficient production alterna-
tive of 3D Printed cellular and lattice-like functional continuous structures.

The proposed technique was tested under four case studies, each one using a
different combinatory of resolution-based computational design methods and
fabrication workflow setups of the same robotic cell.

These case studies demonstrate the conception of detaching 3D printing pro-

cesses to horizontal work beds for reaching functional print time, explain its
fabrication process, test over existing surfaces, and conclude by evaluating the
robotic implementation in additive manufacturing applications.

Research computational design development and robotically augmented ad-

ditive manufacturing are temporarily being developed at Carnegie Mellon
University with the collaboration of Master in Advanced Architecture student
Manuel Rodriguez Ladrón de Guevara under the supervision of professor Dan-
iel Cardoso Llach and professor Jeremy Ficca.

Our main takeaways from this case study are:

- FOAM helps overcome the scalability problem of 3D printing by en-

abling in-situ fabrication, reducing logistic and assembly costs.

- FOAM can print large single structures that might support greater
overhangs due to the lightweight nature of thermoplastic polymers.

The impact of additive manufacturing in the building industry shifts design

to ever more computational design approaches. In order to explore additive
manufacturing capabilities in complex concrete shell structures, we applied
two methods in three different processes.

Commercial 3D printers have proved a reliable complex formwork fabrication

method, allowing for extra detail and precision to be embedded in the concrete
cast. This fact opens opportunities to research surface performance enhance-
ment through superficial patterning.

215
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

[ 36 ]

Fig. 2. 54 Repeatability of operations regardless of orientation depens on the fidelity of the digital model, toolpath simu-
lation and correct lattice formation in space. Although highly tolerant, the sistem of threads depend on the vitrification
of previous points and an error in serial operations results in failure.
Source: Author

216
 2. Free oriented additive manufacturing
Case study discussion

Industrial 3D printers and scalable gantry systems provide successful tools for
engineered reinforcement matrices deposition. Concrete casting has signifi-
cantly shifted filament positioning for thicknesses of 1.0mm and 1.7mm diam-
eters in the initial case studies where we used 3D printers, and 3.0mm diameter
in the robotic arm case study. FOAM provides a significant performance ben-
efit through nozzle orientation enabling a wider application; however, it is less
scalable than available gantry systems.

Challenges such efficient geometric constraint and toolpath computation meth-

ods, quick spatial deposition, robot mobility and construction environment
awareness are identified as areas of future research for developing a precise
and functional construction method.

These set of techniques introduce novel construction workflows in concrete

construction.

AM provides tools for construction automation of continuous structural skins

through spatial deposition, pre-computed stress behavior and intricate rein-
forcement systems.

2.3.1. Capacity of automation

Although the initial steps of design are time intensive, once parameters are
coordinated the laboratory setting that consists of repetition of commands is
significantly simple. The same settings in a non-laboratory environment —
building site— prove to be an immense challenge to overcome. The capacity
of automation of the technique proposed relies fundamentally on either the
capacity of repetition which can be associated with physical tolerance for error,
or the capacity of sensing.

This research presents a novel FDM technique called FOAM. The major con-
tribution of this technique is the ability to print in-situ over existing non-neces-
sarily planar or horizontal infrastructures. This potentially reduces logistic and
assembly costs in construction field. It also opens up new design methodolo-
gies in practice and academic institutions. The use of robots in pedagogical
frames potentiate new aesthetic sensibilities that are difficult to assess at the
time of writing. We hypothesize that this is due to a lack of precedents amidst
a phenomenon of investigating new science with old machines (Carpo, 2017).
Even though the results are successful in terms of design thinking and novel
fabrication techniques, there are still some drawbacks present in the FOAM
process.

Time cost is still inefficient. This might be overcome by reducing wait times at

217
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

218
 2. Free oriented additive manufacturing
Case study discussion

each node and speeding up the robotic motion. This is not a straight-forward
task and more extensive research is needed. Thermoplastic polymers, while
being lightweight and cost-effective, still have a deficient ecological print, and
present a lack of tensile strength and anisotropy is a general problem in the 3D
printing industry. The incorporation of reinforced plastics or metals (Tam and
Mueller, 2017; Tay et al., 2017; Mitchell et al., 2018; Ngo et al., 2018) would
be a step forward for the application of the FOAM technique into the construc-
tion field.

The automation of the design-to-fabrication workflow also is at an early stage:

additive manufacturing is increasingly being applied in many industrial sectors
for its flexibility and ability to produce complex designs at no extra cost. To
comply with the relative slowness of the process, design and manufacturing
workflows have to be optimized to reduce the time of fabrication, amount of
material used, and the risk of errors in the entire production chain.

2.3.2. Form efficiency

The technique presented demonstrates both potential in production of com-
plex shapes and potential in accommodating an optimization based on positive
feedback loops within the design and simulation processes. There are limits in
the design presented by the limitation of the tool. The nozzle should be opti-
mized to avoid collisions with already printed material, for instance, a thinner
nozzle has a major printable angle range than a thicker nozzle.

As a result, a 6-axis degree of freedom machine can reach any point within the
robotic work-cell, with any given orientation, making the printing optimal for
the adaptability to 3D print on complex geometric settings. The authors of this
research have previously tested similar tool-path algorithms on normal hori-
zontal settings where the referenced parameters were tested building on top of
another research works (Liu, Li and Li, 2018). During these contrasting tests,
variables such as general orientation, motion velocity, stepper feed-rate, wait
times and temperature range are also changed to find an optimal solution. We
have found that, maintaining the same set of variables but the general EOAT
orientation, the result of the printed structure differs insignificantly from the
digital model in for shapes (in equilibrium) (that maintain stability along the
additive fabrication process) due to accuracy of the numerical deposition. Free
Oriented Additive Manufacturing provides a fabrication technique for archi-
tectural applications of previously unexploited spaces, providing the means
for reconfiguring and bonding new qualities to existing infrastructures. FOAM
technique is tested using different configurations based on orientation and mor-

219
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 55 Robotic coordination. Two numerical contorl mechanisms are synchronized sharing a same coordinate system
Source: Author

220
 2. Free oriented additive manufacturing
Case study discussion

phology of the base surface: planar and single-curved surfaces. The current
state of the research is focusing on an improved version of the tool that would
allow the use of reinforced thermoplastic materials.

2.3.3. Information handling, optimization and

complexity in computational design
The digital postmodernity is a revolution in science by use of information and
simulation. Through massive computational trial and error one can calculate
and optimize design, opening inquiry into the nature and aesthetics of design.

Through robotic means one can materialize with unparallel precision the vir-
tual model, resolution and complexity have become part of the design toolbox.

Contemporary form of great mechanical complexity, locally differentiated ar-

chitectural qualities and locally engineered performance conceivable through
information technologies of differentiated data computational design and ad-
ditive manufacturing requires a production generalization by algorithmic and
constructive definition.

The precision of the robotic tools and the computational power that discrete
design affords are unified to propose a novel construction system potentially
capable of producing ultra-light-weight structures of enhanced mechanical
performance to test the construction limits of large-scale membranes and com-
plex surfaces .

The emerging role of data in design is important to the technical development

of new design strategies. Data is available as another resource in the designer
toolbox and computational power translates in a growing capacity of handling
more and more complex parameters of design and information.

FOAM incorporates data management into the material and geometrical com-
position.

By testing print angles, reorientation of the end effector and robotic coordina-
tion, free oriented additive manufacturing presents a construction method for
the production of functional complex geometries whose tolerance, adaptabil-
ity, variable economy and locally differentiated representation of architectural
qualities that enable robust and flexible production systems.

The next chapter presents a data organization method for the digital configura-
tion and 3D printing of shell structures.

221
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 1 Electromechanics of AM process xxx

222

[ 15
 223

5]
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

COMPUTATIONAL SPACE: COMPUTATIONAL INSTANTATION:

GRID # (0,0,0,0,3,2) ARXEL # [0,1,4,6,0,7] i:1 j:2 k:0

code = [0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))] c.potTerr: wetAreas

temp1dCode.append(code) c.size = .35m
temp1dInd.append('i:{}, j:{}, k:{}\nCode:'.format(i,j,k)) c.volume = 1.3m3
odVisited.append(0) c.material = ABS
odCont.append([
odCont.append([None,None,None,None,None,None]) c.conections = 4
temp1dBases.append([None,None,None,None,None,None]) c.strucPerformance = 0.4
temp1dBasesCode.append([0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))])
temp1dBranchPts.append([None,None,None,None,None,None])-
temp1dBranchCode.append([0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))])

[ 12 ]

Fig. 3. 2 Discrete methods 3-dimensional grid establishes individually indexed computational spaces allowing mul-
tiple levels of information to be incorporated. (Ladrón de Guevara, et al., 2019)

224
 3. Discrete modelling computational design

3. Discrete Modelling
Computational Design

Additive manufacturing is evolving towards more sophisticated territory for

architects and designers, mainly through the increased use of scripting tools.

Slicing software and easy- to-use hardware has made it possible for seamless
connection between a digital 3D model and its physical relative. Designers do
not have to think about the intricate internal processes of either software or
hardware. Consequently, as more architects and designers embrace program-
ming, more complex alternatives for fabrication through additive manufactur-
ing are being studied, expanding design and fabrication space.

Not only are 3d printing equipment and technologies changing the way in
which we design to accommodate the many constraints of the technique, but
also requests to reflect upon the continuous nature of space, performance and
form. Digital design for Robotic manufacturing, unlike traditional manufac-
turing, must adhere to very strict rules of code.

Current advances in computational design present a novel model of the digital,

where designers conceptualize production of a fundamentally different order,
beginning to end within the particlised logic of computation and disregarding
entirely analog representation.

The digitization of the particle in discretized forms allows a digital constitu-

tion form, fundamentally different from the digital approximation of continu-
ous forms in folding, parametricism and curvilinearity, losing parametric con-
tinuity but in turn, gaining unprecedent control over the constituent element.
Discretization strategies and methods for indexing allow control over each
individual unit.

This research studies the strategy by multi-resolution to channel or distrib-

ute computational power towards areas of interest. When combined with FEA
workflows, the particle can be engineered to perform distinctively, tailoring to

225
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 3 Variable Resolution design by discrete methods of Ulna Bone

(Ladrón de Guevara, et al., 2019)

226
 3. Discrete modelling computational design

the requirements of the individual unit, but performing continuously by func-

tionally graded aggregation.

In essence, the impact of computational power on design outcome is explored,

examining in-depth the concept of resolution as a design driver.

This technique has proven to be successful in printing onto situated built sur-
faces regardless their orientation, achieving a large-scale without the need of
assembly.

The second, SFD, on delta 3D printers presents successful results for rapid
prototyping and for fully finished products at a 3D printer scale. The research
presents how indexing allows computational access the particular, offering a
way to highly customize the form and fabrication of an object, adding more
parameters to tune during the design process.

227
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 4 Discrete methods 3-dimensional grid establishes individually indexed computational spaces allowing mul-
tiple levels of information to be incorporated. Own elaboration.

228
 3. Discrete modelling computational design
Informational techniques in architecture

3.1 Informational techniques in

architecture
The infusion of digital tools to the practice is changing not only how we design
but also what we design (FREDERICK NORMAN 2001).

CAM – CAD practices merged in the 1950 as a cybernetic theorization of

design as a human-machine endeavor (Cardoso, 2012). The emergence of
ubiquitous and economically accessible computational power aligned the
computer aided design practice with the original conception of computerized
aide(Negroponte, 1969; Negroponte, 1975). Daniel Cardoso presents CAM –
CAD tools as a thinking apparatus.

“Software systems are not merely “instruments” or “tools” but

rather versatile metaphors reconfiguring conceptions of design,
materiality, work, and what it means to be creative”

Archeology of the digital, G. Lynn (Lynn, Greg, 2013b) typifies the main ar-
chitectural practice hurdles and goals of the digital pioneers through the selec-
tion of seminal projects. This research is vastly influenced by pioneer projects
such as, S. Yoh’s Prefectura Gymnasium fully digitized spatial structure (Yoh
& Lynn, 1993) and E. Ruiz-Geli and his team’s digitally informed, activist,
performative designs.

Today architecture is not only designed differently but effectively thought and
decreed fundamentally different from the modern architecture paradigm based
in the project drawing (Hauck et al., 2017). Digital design and manufacturing
practices have effectively transformed both traditional ways of representation
and traditional ways of construction with digital design and build opportunities
(Hauck et al., 2017).

Architects have started to embrace AM and automation technology in their

projects both at the prototype and built scales. In all this, architecture remains
anchored to a fundamental design process, namely, the manipulation of geom-
etry.

Since the existence of CAD software, designers have been constantly learning
new drawing skills through computational means. Accordingly, the knowledge
of the tool has been a steady feature for architects in order to draw, develop
and communicate their projects. Architects, for the last thirty years, have been Fig. 3. 5 Continuous logic of
adapting their skills to new software in order to be prepared for market exigen- design.
cies. In this computational era for designers, the tool presents a direct imprint Fig. 3. 6 Discrete logic of de-
sign. Own elaboration based
in the shape of their buildings (Carpo, 2017).
on (Medium, 2016)

229
 Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

COMPUTATIONAL SPACE:
GRID # (0,0,0,0,3,2)
SPACE: COMPUTATIONAL INSTANTATION:
) code = [0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))] ARXEL # [0,1,4,6,0,7] i:1 j:2 k:0
temp1dCode.append(code)
,('i:{}, j:{}, k:{}'.format(i,j,k))] temp1dInd.append('i:{}, j:{}, k:{}\nCode:'.format(i,j,k)) c.potTerr: wetAreas
odVisited.append(0)
end(code) c.size = .35m
odCont.append([None,None,None,None,None,None])
odCont.append([
d('i:{}, j:{}, k:{}\nCode:'.format(i,j,k)) temp1dBases.append([None,None,None,None,None,None]) c.volume = 1.3m3
(0) temp1dBasesCode.append([0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))]) c.material = ABS
None,None,None,None,None,None]) temp1dBranchPts.append([None,None,None,None,None,None])- c.conections = 4
end([None,None,None,None,None,None]) temp1dBranchCode.append([0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))]) c.strucPerformance = 0.4
e.append([0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))])
append([None,None,None,None,None,None])-
de.append([0,0,0,0,0,0,('i:{}, j:{}, k:{}'.format(i,j,k))])

[ 12 ]

[ 12 ]

Fig. 3. 7 Design of sinclastic surface by discrete methods. Detail of indexed pattern of floors, columns and rows in-
dexed at coordenates [x axis, y axis, z axis] (Ladrón de Guevara, et al., 2019).

230
 3. Discrete modelling computational design
Informational techniques in architecture

According to M. Carpo, during the second turn in architecture, architects might

start using computation in a raw format (i.e. coding directly a script rather
than using already built-in plugins or software) as ways to explore novel de-
sign methodologies. This does not only have consequences in the shape of the
design, but also and more importantly, raises new ways to think architecture,
therefore, more likely than not, new sensibilities are about to have an impact
in architecture. During the following years, we might face a coexistence of
already stablished sensibilities with new emerging insights coming from these
new methodologies.

Design, as it is becoming a virtual – material equivalent exercise, is changing

definitely the processes by which geometry is interpreted, transformed, mea-
sured or communicated. One goal of the research is to bridge virtual – material
realizations to the fabrication logics of AM. It is such geometrical manipula-
tions in conjunction with AM that are of interest in this research. Specifically,
this chapter provides a generalization of an informed alternative to current
black-box end-to-end AM pipelines, namely, a multi-resolution based method-
ology supported by discrete methods.

3.1.1. Design Scripting tools for particle

tectonics of AM
The use of programming languages in design opens up unexplored and previ-
ously unworkable territories, mainly, in conventional architectural practice. In
the 1990s, languages of continuity, smoothness and seamlessness dominated
the architectural inquiry with the CNC milling machine as its manufacturing
tool (Carpo, 2012b). Today’s computational design and fabrication technology
look at languages of synthesis of fragments or particles, with the 3D printer as
its fabrication archetype.

Fundamental to this idea is the concept of resolution —the amount of informa-

tion stored at any localized area. This paper explores a novel design methodol-
ogy that takes this concept of resolutions on discrete elements as a design driv-
er for architectural practice. This research has been tested primarily through
additive manufacturing techniques.

Additive manufacturing is evolving towards more sophisticated territories for

architects and designers, mainly through the increased use of scripting tools.
Slicing software and easy-to-use hardware has made it possible for seamless
connection between a digital 3D model and its physical relative. As a result,
the use of 3D printing in architectural practices and schools has blossomed Fig. 3. 8 Discrete components
conforming a continuous com-
over the past decade. Designers do not have to think about the intricate internal plex tessellation
Own work

231
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 9 Open membrane design

The study and development of custom instruments for design and fabrication is applied in the configuration of freeform membranes for
the large scale manufacturing of shell strucutres through discrete methods.
Own work

232
 3. Discrete modelling computational design
Informational techniques in architecture

processes of either software or hardware.

Consequently, as more architects and designers embrace programming, more

complex alternatives for fabrication through additive manufacturing are being
studied, expanding design and fabrication space.

The research presents a class of techniques for fabrication and methods of de-
sign through discrete computational models. In essence, the research explores
the impact of computational power on design outcome, examining in-depth the
concept of resolution as a design driver.

The elements of style of architecture have been traditionally related to the

apparent and superficial, however, there has been a constant growth in the
awareness of the performance of architecture (Moussavi & Kubo, 2008; Reiser
& Umemoto, 2006).

While material articulation, form, ornament and performance have been ex-
tensively studied (Rosenbauer, 1947), it was not until the advent of 3D print-
ing technologies and algorithmic design that designers started exploring the
nature of the constituent element as an exercise of design associating com-
plexity with productive function (FABRICATED); rather than only an aesthetic
and cultural purpose (Rosenbauer, 1947). This is partially due to the absolute
change in essence of fabrication, for example, 3D printing creates and poses
matter where there was none, fundamentally different from forming or sub-
tracting digital fabrication techniques.

The method of digital fabrication is, therefore, intimately related to the con-
ceptualization of form where digital relatives can become true equivalents of
their material counterparts.

Within this research, printing is how the virtual–physical bridge is mate-

rialized, and particle is the unit of construction.

This research explores two methods of printing. In study cases I and II of this
chapter, the research takes a directional 2D approach based on patterns that lie
on a surface. This means that this method prints based on layers, distributed
in such means that in order to form a volume, the orientation of these layers
should change. For instance, layers in a positive X direction crisscrossed with
layers in positive Y direction. In study case III, the approach is rather different.
We study 3D patterns that exist in each cubical volume. Instead of organizing
a series of surfaces to make a volume, this approach discretizes a given general
volume into smaller fragments.

The AM subfield of fused deposition (FD) is presented and analyzed from a

computational perspective. We review some computational methods based on

233
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 10 Pleated sheed membrane design by discrete methods applied to an architectural element. Algorithmic design process allow
the configuration of highly complex shapes, this research focuses on the configuration of membrane structures based on forces.
Own work

234
 3. Discrete modelling computational design
Informational techniques in architecture

discrete models. We consider two main spatial printing techniques, and test
fabricating forms at differing scales which are modeled with different param-
eters. Lastly, FEM influenced specific fabrication techniques that are backed
by the multi-resolution approach upon which, by applying these techniques
and methods to infills of closed objects, we can converge to optimal fused
deposition solutions.

This research focuses on the development of scripting tectonics of form study-

ing resolution, drifting explicitly apart from parametric design tools of superfi-
cial and ornamental performance of form to particle performance.

This is possible since architects and designers are closer than ever to compu-
tational and fabrication technologies. Same as ten years ago we were explor-
ing parametric software (i.e. Grasshopper plugin for Rhinoceros, McNeel or
Dynamo, Autodesk), we are now starting to use our own code (Python GH),
finding a higher degree of freedom in design.

Parametric software started to introduce a new design thinking approach to

architects: algorithm thinking, or procedural design-based thinking. As we
started to understand how this visual scripting software works, we can better
understand the process of design by coding algorithms. As an example, we can
see how Gramazio Kohler use this workflow in their recent work (Gramazio
and Kohler, 2017).

Fig. 3. 11 Pleated sheet. Cur-

rently the construction of com-
pelx surfaces requires a labor
intensive work, although it is
highly depending on industrial-
ized materials.
Own work

235
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 12 Multi-Resolution method in Ulna bone Computational (Ladrón de Guevara, et al., 2019).

236
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2 Multi-Resolution Based Methodology

Multi-resolution studies the multiple resolutions an object might have. Ad-
justments of many resolutions across an object may offer a new design op-
portunity, since the inherent discreteness of the process provides the means to
reconfigure the object’s morphology. An object may embrace multiple configu-
rations at once. These configurations respond to degrees of resolution that cor-
respond to regions of interest. Thus, the more interesting a region, the higher
the resolution it presents. The number of different resolutions is determined by
certain design criteria.

The multi-resolution algorithm we propose has three parts: shape analysis,

creation of a 3-dimensional grid as the computational structure, and design
of discrete elements. This method contributes to the generalization of the use
of resolution as a concept for the design of architectural components and uses
discrete computational models for its development. However, the authors ac-
knowledge this is just the first of many steps in order for this method to be ap-
plicable to the construction industry. This approach has been tested on models
of human bones. This human scale simplifies barriers, allowing us to focus
only on the design of the bone, without additional considerations, such as ex-
ternal and environmental agents, which would be included at an architectural
scale.

The computational approach to serializing the solutions for successful AM

constructions are either discrete by voxelization of serialized by algorithmic
growth. Discrete computational algorithms, because of its defined boundaries
are very suitable to all AM production scales, while continuous growth algo-
rithms require a large-scale AM construction platform.

Our method examines the potential of designing an object using a discrete

elements approach in order to configure its morphology to specific conditions
through multi-resolution adjustments. The balance between higher and lower
resolutions across the object corresponds to different design criteria such as
aesthetic values, structural behavior, density, and material properties. Higher
resolutions correspond to interesting areas, and lower resolutions target less
interesting areas. That is, the algorithm behaves differently according to the de-
gree of resolution, where the subdivision level quantifies the interesting areas
and the type of information that qualifies them.

The number of areas of interest follow a design criterion, which depends on

user demand, thus, the algorithm solution can be expanded, to a certain degree,
upon this. For this particular research, the algorithm considers variable density,

237
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

The creation and

manipulation of geometry is restricted to the management of numerical values (1). For

instance, a rotation of an arxel through angle a is given by:

x’ = x cos α – y sin α (1)

y’ = y sin α + y cos α

Fig. 3. 13 The subdivision level corresponding to these areas responds to a design criterion and limitations on the
fabrication, where the algorithm is able to compute different degrees of resolution altogether Resolutions levels and
Matrix configurations. (Ladrón de Guevara, et al., 2019)

238
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

opacity levels, and printable in-budget implementations. Understanding the

fabrication end is imperative in this approach, since the affordances and limita-
tions of material and tool define the design possibilities.

This research provides a step towards a design methodology based on multi-

resolution adjustments. For simplicity, we show the potential of the method
through the study of an ulna bone both as a design example, and as a potential
application of this design and fabrication approach to other fields. Current re-
search focuses on expanding the algorithm to a larger scale for architectural
practice. The algorithms are written in Python, and for visualization purposes,
Grasshopper is used to transform numerical data into geometrical values, using
the Rhino Common library.

3.2.1. Fabrication overview

Additive manufacturing enables the fabrication of complex physical designs.
One substantial research development specific to AM is the use of spatial depo-
sition techniques, capable of continuously extruding cellular lattice members
in a spatial organization with minimal waste. Foams and spatial lattices provide
multifunctional lightweight structures, high strength and relatively low mass,
and robotic AM the precision required to manufacture. In order to manufacture
complex geometries through a continuous deposition method, fabrication and
computational methods that serialize the deposition process into a concatena-
tion of units have been developed.

Research efforts focus on adapting resolution in the discretization process

while maintaining the index concatenation required for the geometrically cor-
rect continuous deposition. We can build upon advances in AM application to
other disciplines such as bone reconstruction of graded porosity, in particular,
to study bones at an architectural level in order to establish hierarchically and
topologically different design principles. This research provides a spatial or-
ganization system of indexable units with differentiated performance, bound-
ary and neighboring conditions of manufacture through continuous deposition
maintaining geometrical continuity. With resolution-based design, such strate-
gies may be translated into functional largescale architectural components with
potential for complex design of engineered anisotropy feasible through spatial
AM.. 3D Spatial printing technique with a commercial delta 3D printer.

239
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 14 Algorithmic design based in Python plugin for Grasshopper in Rhinoceros 3D design environment. The
design workflow is centralized in the software as the bases of the projectual procedure of architectural form. The data
and information technologies are the drivers of the design and manufacture process, for which a custom program that
intersects different softwares is proposed.
Details in Annex I. Own work based on (Ladrón de Guevara, et al., 2019)

240
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.2. Data structure

This section explains how we structured our data in python, to construct the
multiresolution algorithm. Our data is organized into two main object classes.

A 3-dimensional grid establishes the multiple resolutions across the object and
indexes them, connecting their computational units —called arxels— through
a graph data structure. The arxels correspond to the vertices, whereas the edges
of the graph are defined by the type of relationship between neighbouring arx-
els.

Arxel, an acronym for architectural element, is a digital structure that stores

the following data: material, local name, geometric typology, resolution type,
location and neighbours. For visualization purposes, arxels are represented by
voxels or cuboids, and it is not until the end of the process, where the voxels
are substituted by a final designed unit.

Similar to pixels that construct a global more complex image, arxels subdivide
space into information containers and may contain as much information as the
capacity of the computer that process the algorithm. Our design typology is
mainly linear, responding to the inherently segmental nature of spatial printing,
conforming to a nomenclature of numerical matrix configurations, covering a
range of possibilities, from simple to more complex shapes.

The presented pipeline has been tested with mixed printing techniques.

Robotic arm printing with FOAM, employs a robotic arm but does not need a
planar horizontal surface upon which to deposit material. This technique has
proven to be successful in printing onto situated built surfaces regardless their
orientation, achieving large scale configurations without the need of assembly.
Gantry printing with 3 axis mechanisms, tests and presents successful results
for rapid prototyping and for fully finished products at a 3D printer scale. The
last technique is a hybrid between spatial printing and layer-wise deposition,
operating at a higher level of resolution on the indices that present weak values
in the FEM analysis.

Design and fabrication processes that involve standard 3D printing do not

present major challenges for designers because of the ease of black-box tool-
pathing software. Although these accessible solutions are active contributors
to the embracement of new technology, we believe that intervening in both the
tool-path stage (that bridges design and fabrication) and the generation of an Fig. 3. 15 Algorithm based on
iterative and recursive opera-
utterly customized object, constitute an alternative and flexible pipeline that tions to configure architectural
expands the design space. form.
Own work based on (Ladrón de
Guevara, et al., 2019)

241
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

242
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

Current computational power makes possible to think of and design objects

that combine multiple resolutions. Approximating form with discrete units of
variable levels of resolution allows the designer to work at the level of the con-
stituent element, expanding the design opportunities but requiring the atomis-
tic understanding of a continuous form as an informed addition of units and ca-
pacities. The indexing, structuring and geometrical definition of 3D printable
informed units requires the designer to work with data as abstract equivalents
of geometry and process of fabrication at the same time.

This method offers a way to highly customize the form and fabrication of an
object, adding more parameters to tune during the design process. The more
common parametric design pipelines would comprise: (1) the tuning of some
certain decided parameters in a parametric design software such as Grasshop-
per or Dynamo; (2) baking a decided instance; (3) saving the file as .STL and
sending it to a tool-pathing software; (4) 3D print it. We, instead, start by add-
ing the number and type of resolutions to the object at the same moment pa-
rameters than in step (1), and one then creates a second layer of parametric
decisions between steps (2) and (4). This second layer of decisions is data
driven. We believe that understanding the object as a set of ordered containers
of parametrizable information that directly informs directly the overall shape
and later fabrication could present a potential to augment the design space,
in a manner somewhat similar to the way parametric design software such
as Grasshopper or Dynamo provided an augmentation in the decision-making
stage.

This research takes into consideration divergent design drivers that were tested
applying SFD models as well as finite element analyses (FEA) informing cus-
tom infills using layer-by-layer deposition. These are some of key elements of
this research.

• Additive manufacturing described from a computational point of

view, with a special attention to fused deposition.

• Testing SFD at different scales using specific machines.

• An analysis of the main engine of our proposed pipeline, based on

multi-resolution adjustments.

• Convergence to a computational approach based on FEA, for opti-

mized differentiated infills of thermoplastic components.

243
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 16 3D printed ulna bone applying multi-resolution algorithm in desktop 3D printer.

Own work based on (Ladrón de Guevara, et al., 2019)

244
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.2.1. Cell Structure (from lattice structure to cells)

We discretize a volume into smaller units in a so-called multi-resolution ap-
proach. The underlying 3D grid stores information such as material and geom-
etry in each of its voxels, as well as the printing order.

The purpose of this case study is to test the complexity of printing different
configured polylines that fit into a 1.18” 3D-cell size.

Testing a range of simple to complex printable polylines gives insight of the

type of design nomenclature feasible for this fabrication technique. Acute an-
gles present a lower fidelity to the digital model than obtuse angles.

Consecutive segments shorter than 0.4” start to blur the overall appearance of
the cell. Printing temperatures above 220 degrees Celsius for PLA result into
an excess of liquid that produce sagged segments. Otherwise, an excess of cool
temperature, below 185 degrees Celsius, results into over rigid segments that
produce undesirable low-fidelity outcomes.

These constraints might be overcome with the change of feed-rate flow, wait
times and the motion of the robot. It is recommendable to find a contextualized
balance between design limitations and time cost. (TABLE)

Material-scale considerations can optimize structures further. This study also

looks at the bonding between different thermoplastics together within the same
piece, as fiber orientation and bonding between materials are the main chal-
lenges that arise in 3D printed composite parts.

245
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 17 3D printed ulna bone applying multi-resolution algorithm in desktop 3D printer (Ladrón de Guevara, et al.,
2019).

246
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.2.2. Voxel - 3D Grid

Our method uses voxels to represent the 3D-printed structure. Voxels are
the 3-dimensional equivalent of pixels. So, like rasterizing a vector graphic
(shape) into a raster image (a series of pixels), a 3D model or shape can be
voxelized into a set of 3D unit cubes (voxels). The term “voxelization” is used
in this research to describe the process of transforming any given 3D shape
into a group of voxels.

We discretize the space with an underlying grid that acts as a potential con-
tainer of information. This grid is initialized at a given resolution and the size
of each of the voxels of the grid will be constrained by the fabrication tool. For
instance, in case study III, we use a grid of voxels of 30x30x30mm.

The grid is also capable of having different controllable voxel size, in similar
manners that an octree algorithm works. The algorithm stores information of
the grid of neighboring relationships with results of collision studies within a
single cell to conform printable concatenations. The information is stores in a
type of data structure of Python called dictionary.

This characteristic could failsafe fabrication issues (i.e. the height of the geom-
etry at each arxel might not reach the desired Z due to material deformations,
so a slight overlap in the Z direction would benefit arxels at the upper level, as
they would be anchored to the top part of below floors).

Arxel

An Arxel is an informed digital architectural unit. It is a voxel

with an overlapping agenda of material properties, economy,
geometry and performative data such as structural behavior,
opacity and density variations. Arxels are understood in depen-
dency with their surrounding arxels. As a data unit, an arxel
is an information container rather than a geometrical element,
and it is structured as a set of dictionaries.

Its boundary is defined by eight coordinates corresponding to its

vertices. The boundary, however, is only for visualization pur-
poses; it does not possess any geometrical value.

Fig. 3. 18 3D Printable Ge-

ometries within discerete spa-
tial unit. (Ladrón de Guevara,
2018)

247
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

PRINTABLE TESTS PATTERN C

[ 51 ]

[4] Global information [5] List of arxels INDEX: floor:0,row:1,col:6

CONDITION: OUT
Element: CurvedWall
EMPTY ARXEL
INDEX: floor:0,row:1,col:7
iD: #0001
INDEX: floor:0,row:0,col:0
CONDITION: OUT
///////// CONDITION: OUT
EMPTY ARXEL
fabrication Date: 04/29/2018 EMPTY ARXEL INDEX: floor:0,row:2,col:0
INDEX: floor:0,row:0,col:1
CONDITION: OUT
fabrication time: 64.0 h
CONDITION: OUT
EMPTY ARXEL
filament type: PLA 2.85mm EMPTY ARXEL INDEX: floor:0,row:2,col:1
INDEX: floor:0,row:0,col:2
CONDITION: OUT
materials quantity: 4
CONDITION: OUT
EMPTY ARXEL
number of used patterns: 3 INDEX: floor:0,row:2,col:2
EMPTY ARXEL
INDEX: floor:0,row:0,col:3
number of sizes: 2, 30mm and 60mm arxel size CONDITION: FULL
CONDITION: OUT
MATERIAL: PLA A
/////////
EMPTY ARXEL
INDEX: floor:0,row:0,col:4 MATERIAL LENGTH: 90.99mm
total Volume: 12.69.m3
CONDITION: OUT PATTERN: Type 3
INDEX: floor:0,row:2,col:3
total Area: 14.91m2
EMPTY ARXEL
CONDITION: FULL
INDEX: floor:0,row:0,col:5
total Length: 1907.55m
MATERIAL: PLA A
CONDITION: OUT
total Weight: 15.09kg
MATERIAL LENGTH: 90.99mm
EMPTY ARXEL
INDEX: floor:0,row:0,col:6
number of voxels in grid: 5200 PATTERN: Type 3
CONDITION: OUT INDEX: floor:0,row:2,col:4
number of scaledUp size: 45
EMPTY ARXEL CONDITION: FULL

number of nonValid: 87 INDEX: floor:0,row:0,col:7

MATERIAL: PLA A
CONDITION: OUT
number of arxels: 2130 MATERIAL LENGTH: 90.99mm
EMPTY ARXEL
PATTERN: Type 3
number of material A: 1536 INDEX: floor:0,row:1,col:0
INDEX: floor:0,row:2,col:5
CONDITION: OUT
number of material B: 594 CONDITION: OUT
EMPTY ARXEL
number of external pieces: 38 INDEX: floor:0,row:1,col:1 EMPTY ARXEL
INDEX: floor:0,row:2,col:6

number of voids: 49 CONDITION: OUT

CONDITION: OUT
EMPTY ARXEL
number of boundary Arxels: 414 INDEX: floor:0,row:1,col:2 EMPTY ARXEL
INDEX: floor:0,row:2,col:7
number of interior Arxels: 1768 CONDITION: BOUNDARY
CONDITION: OUT
MATERIAL: PLA B
number of out boundary Arxels: 3018 EMPTY ARXEL

[ 50 ]
INDEX: floor:0,row:3,col:0
MATERIAL LENGTH: 90.99mm

PATTERN: Type 1 CONDITION: OUT

INDEX: floor:0,row:1,col:3
EMPTY ARXEL
CONDITION: FULL INDEX: floor:0,row:3,col:1

MATERIAL: PLA A CONDITION: BOUNDARY

MATERIAL LENGTH: 90.99mm MATERIAL: PLA B

MATERIAL LENGTH: 90.99mm

PATTERN: Type 3
INDEX: floor:0,row:1,col:4
PATTERN: Type 1
INDEX: floor:0,row:3,col:2
CONDITION: FULL
CONDITION: FULL
MATERIAL: PLA A

MATERIAL LENGTH: 90.99mm MATERIAL: PLA A

PATTERN: Type 3 MATERIAL LENGTH: 90.99mm

INDEX: floor:0,row:1,col:5
PATTERN: Type 3
CONDITION: OUT INDEX: floor:0,row:3,col:3

CONDITION: FULL
EMPTY ARXEL

Fig. 3. 19 Each of the units that conform the complex shape include a series of information accesible by index tag.
Own work based on (Ladrón de Guevara, et al., 2019)

248
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.3. Continuity by concatenation of individual

units with distinct properties
The level of fidelity required by the designer will be determined by the action
of discretization of form into arrays of indexed particles and the potentially
infinite amount of information each of these particles is determined to host.

This is particularly important to the exercise of continuous deposition because

virtually there is no way to produce a successful large-scale print but to break
it down into segments that approximate form with a determined level of reso-
lution. What this research acknowledges is that precise level of resolution and
exploits it to pre-scaled level so that particles are determined based on both
their properties and their eligibility to be printed. The continuous form is there-
fore approximated by particle concatenation.

This does not mean that the smaller the unit the higher the resolution is. The
particle determines the index address in the neighborhood of information box-
es, as a void or empty list potential to be filled with an array of information.

The resolution is therefore the levels and complexities of information, which

means the physical-virtual relativity is not a superficial one, approximating
only levels of geometric interest, but one that emulates in the addition of par-
ticles and its properties at a determined level of fidelity.

Fidelity is not superficial interest. Resolution at the level of structured particles

can emulate the internal and external characteristics of the virtual model.

Each cell voxel contains a certain amount of information.The robot finds and
emits low voltage signals information to activate any secondary electrome-
chanical signals. In essence, geometrical and physical computing instructions
interweave at every intersection.

This technique has proven to be successful in printing onto situated built sur-
faces regardless of their orientation as (REFERENCE PREVIOUS CHAP-
TER) with the purpose of implementing a large-scale physical organization of
particles. This research presents how indexing particles allows computational
access of the individual unit to hyper tailor form and fabrication of an object.

Fig. 3. 20 Study of scalable

space filling polyhedral arrays
Discrete units can adapt to
scale if the 3D printing process
does not create self intersec-
tions in the derived cell resolu-
tion subdivision.
Own work based on (Ladrón de
Guevara, et al., 2019)

249
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 21 A network of 42 trasformations of 3 types of fibers are oriented following information about the resulting
neighborhood grouping of fibers, in this case to reduce density to create opening-like arrangement. Own work based
on (Ladrón de Guevara, et al., 2019)

250
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.3.1. Pixel continuity

Information graduality and the articulation of properties in a continuous fash-
ion is achieved in a pixel-like distribution. In a similar way that pixels that
conform an image can draw and depict a clear set of information, my the-
sis explores this concept extended to Additive Manufacturing or architectural
properties. An array of organized architectural properties, primary geometri-
cal, secondary of manufacturability and tertiary material, draw the optimized
distribution of construction of form.

For a discretized membrane to be stable, the parts that compose it must be

continuous. FDM additive manufacturing consists of a linear, continuous, set
of commands.

This research studies continuous surfaces. The mechanical continuity in dis-

cretely conformed lattices is required. Other property qualitative and quantita-
tive continuities also apply.

The research, inspired by novel developments in AM in disciplines such as op-

timized medical bone reconstruction lattices, in essence, explores the capacity
of optimization of voxelized cellular forms in architectural form.

The study of inner formations by optimized aggregation is not an exercise in

engineering. Influenced by natural systems, the study focuses in presenting a
robust method for aggregation of architectural qualities extending the scope to
a review of architectural affect.

The main properties of pixelated graduality are:

• Dictionaries store neighborhood qualities and manufacturability for

computation efficiency.

• Geometrical patterns perform according to design requirements so

properties are induced by engineering groups of cells, reviewed in this
chapter Fig. 3. 22 Moebius strip inval-
id geometry.
• Mapping properties from simulation, studied in the next chapter. Several transformations of the
grid incur in invalid self inter-
• A Solution Space is generated section or invalid manifold
Own work based on (Ladrón de
Guevara, et al., 2019)

Fig. 3. 23 Base orthohedric

array displays a valid result
Own work based on (Ladrón de
Guevara, et al., 2019)

251
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 56 Adaptive resolution grid configuration, digitization of an ulna bone. Own work based on (Ladrón de Gue-
vara, et al., 2019)

252
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.3.2. Multi-Resolution
The resolutions of a given volume are constructed by the scrutiny of inter-
esting areas. A volume can have one or more different resolutions. A default
resolution is defined at the beginning of the process, normally informed by the
fabrication tool, and new resolutions are given by the interesting areas. In our
implementation, these are selected directly in the Rhino view-space and linked
to Grasshopper.

The level of subdivision that corresponds to these areas responds to a design

criterion and limitations on the fabrication, where the algorithm can compute
different degrees of resolution altogether. Ray casting is used to retrieve the
Boolean value at each arxel, yielding true, if its centroid falls inside the inter-
esting areas, and false, otherwise.

This information is passed in the form of a dictionary, which is then read by

a function that subdivides the arxels based on the subdivision level. Level 1
subdivides the arxel into 8 smaller units, level 2 subdivides it into 64 units,
and so on.

An algorithm similar to an octree approach is developed to form a multi-reso-

lution grid. The algorithm restricts each arxel with a false value to level 0, the
default level of resolution.

To bridge between a conventional IDE and the Rhino-3D environment and ge-
ometry created in GH-Python, the grid information is exported as a Javascript
json file.

This section is composed by two parts. The first describes a general approach
to design discreteness without considering limitations of fabrication whereas
the second implements specific features such as material budget responsive-
ness and consideration of constraints on the 3D spatial printing technique for
manufacturing purposes, such as printing order and collision checks.

Fig. 3. 24 Adaptive Voxel Grid

Refinement.
Work in collaboration with
Manuel Ladrón de Guevara

253
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 2. 57 Variable resolution grid translated in geometrical arrays. Interface of geometrical visualization. Own work
based on (Ladrón de Guevara, et al., 2019).

254
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

3.2.3.3. General Configuration

This process takes as input information relating to the resolutions of the object
from a prior exported data file. This is informed by the fabrication process,
and natural to the process is the creation of a linear language where the geom-
etry is directly used as toolpath [9], [27]. However, we insert different XYZ-
coordinates that are connected with lines only for purposes of visualization.
The creation and manipulation of geometry is restricted to the management
of numerical values (1). For instance, a rotation of an arxel through angle a is
given by:

x’ = x cos α – y sin α (1)

y’ = y sin α + y cos α

No direct manipulation of geometrical models is made. We create a design

grammar —or language— composed by a set of six numerical matrix con-
figurations, ranging from less to more complexity in shape to ensure disparity
between the options. For instance, an area within the given volume that may
require a higher structural reinforcement could be achieved by placing denser
geometries to the detriment of simpler ones.

Similarly, areas with higher degrees of opacity might be better determined by

denser geometries. Each geometrical configuration is placed within the bound-
ary of an arxel, selected upon design specifications. In order to increase options
for design and fabrication, four 90° rotations along the Z axis are also added
to the set, augmenting each initial grammar by four, having a set of 24 shapes.
Matrix configurations

3.2.3.4. Matrix Configuration

Fabrication is based on 3D spatial printing techniques: we create a design
language based on linear segments, constrained by material and tool proper-
ties. For instance, a delta type 3D printer, which extrudes 1.75mm filaments of
polylactic acid (PLA) or acrylonitrile butadiene styrene (ABS), will constrain
certain design decisions.

The nozzle of the printer defines the angle between the linear segments of a
polyline and the height of an arxel. The printing order informs the position
Fig. 3. 25 Multi-resolution
of the numerical configurations at each arxel to avoid collisions with already discrete design of locally differ-
printed geometries. entiated spatial lattice. Mockup
shape of 1 cubic meter of shell
The printing order follows the logical positioning of the arxels within the grid, structure is subdivided into 10
cm hexahedral units to allocate
in the order of columns (Y axis), rows (X axis) and floors (Z axis). That is the varying materials, density, ge-
algorithm starts positioning the first numerical matrix at the 0th column, 0th ometries. (Ladrón de Guevara,
2018)

255
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 26 Backtracking algorithm is implemented to find a valid solution following the pseudo-code. (Ladrón de Gue-
vara, et al., 2019)

256
 3. Discrete modelling computational design
Multi-Resolution Based Methodology

row and 0th floor indices, the next matrix is placed at the 1st column, 0th row,
and 0th floor. At this point it is necessary to check whether the chosen matrix
may collide with an already printed neighboring arxels. For instance, at the
position col=1, row=0, floor=0, only the arxel at position (col-1, row) needs
to be checked. At the position, say, col=3, row=5, floor=0, the list of arxels
[(col-1,row-1), (col-1,row), (col+1,row), (col,row-1)] should be checked.
A backtracking algorithm is implemented to find a valid solution following the
pseudo-code:

3.2.3.5. Collision check

To check whether an arxel might collide with previously printed arxels, we
compare each of its coordinates against the already printed coordinates, such
that zP > z, -xP < x < xP and -yP < y < yP, where xP,yP,zP already printed
coordinates. The domains {-xP, +xP} and {-yP, +yP} correspond to the nozzle
dimensions. The complexity of the algorithm increases at the subdivided vox-
els, since these arxels are formed by matrixes of matrixes equivalent to the
level of resolution.

Similar to the backtracking strategy explained above, an algorithm that finds

a valid solution responding to material length limit is computed. To adjust
the overall piece within a material constraint, we calculate the distance of the
coordinates at each voxel in order to know the total printed length of the piece.
If the total length is more than the given material budget value, the algorithm
selects pseudo-randomly non-subdivided arxels, since these correspond to less
interesting areas.

Their matrix is changed and tested procedurally, passing tests if the overall
length is less than the previous state and still above the material budget. If so,
another arxel is selected, changed and tested until finding a solution that fits in
the budget. If no solution is found, earlier design decisions such as the size or
quantity of interesting areas, or the base resolution should be modified.

257
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 27 3D printed ulna bone applying with different resolutions. Areas of interest can be configured to accomodate
more information. (Ladrón de Guevara, et al., 2019)

258
 3. Discrete modelling computational design
Ulna bone case study

3.3 Ulna bone case study

We test superficial geometric complexity high resolution surface geometry
[12]. Commercial Printers WASP 2040 with a 20cm diameter print bed is used
to produce discrete assembly of complex formworks. Hyper-tailored engi-
neered concrete proves to be significantly lighter and functional [13].

Deposition of engineered reinforcement matrices through Stress Line Additive

Manufacturing [13] is also tested. Industrial 3d Printer WASP 3MT and robotic
AM with ABB IRB 6640 are evaluated.

3.3.1. A new digital design pipeline based on multi-

resolution discrete design method
This research proposes a new pipeline based on multi-resolutions within a
DM approach. The pipeline is composed by of three main layers: fabrication
techniques, tools, and a set of computational design objectives. Each layer is
potentially expandable. We have explored several workflows that result from
combinations of these nodes..

An ulna bone is used as the case study since its complex morphology requires
different levels of resolutions. Furthermore, applying our design approach at
a human scale first will inform bigger and more complex architectures. We
apply a higher degree of resolution to the ulna’s radial and semicircular notch
through the definition of a localized sphere, and also, we reinforce the ulna’s
coronoid process via definition of a spline. We maintain lower resolutions in
the rest of the bone.

The case study begins with an analysis of the ulna, extracting its dimensions
and decreasing the scale by two thirds for fabrication in a conventional 3D
delta printer. Initial tests with different default resolution are made to assess the
degree of information that satisfies the purposes of this research. For the initial
analysis, we use the arxels’ boundary representation for visualization purposes.

Upon determining the interesting areas of the ulna, we compute the levels of
subdivision based on the amount of information at the notch and at the inter-
nal stress line as. The design process continues by automated selection of the
type of geometry that best suits the specific location at the bone. That is, since
the grid stores information on the piece such as the number of components or
parts, areas of interest or structural analysis, we define denser areas where a
higher structural capacity is required, and the algorithm selects those gram-
mars within the set that fit the needs.

Once the selection of the resolutions and type of geometry of the discrete ele-

259
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 3. 28 Collision check in unit pairs is coupled with collision violations in neighbourhoods by backtracking algo-
rithm to review the constructability of a general tessellation design. Own work based on (Ladrón de Guevara, et al.,
2019)

260
 3. Discrete modelling computational design
Ulna bone case study

ments that form the bone satisfies the design conditions, we enact backtracking
algorithm for fabrication where a fabrication solution is found without altering
the situated conditions across the piece.

For simplicity, no multi-material is implemented since we use a conventional

nozzle in the 3D printer. Instead, PLA is used for early prototyping .

3.3.2. Case study discussion

This paper contributes to a design methodology based on multi-resolution ad-
justments as areas of opportunity. This is only now feasible as the design meth-
odology represents a change in design thinking spurred by a computational
revolution and the advent of 3D printers. It is now when we can start talking
about resolution in architecture.

The design method presents numerous limitations that should be overcome

with more research in both the computational part and physical tool develop-
ment.

This research uses a recursive strategy for subdivision; this is computationally

expensive and relatively complex.

On the other hand, subdivision follows a recursive logic that might be quicker
to apply than other alternatives. Data structures such as dictionaries or sets are
used, as their hashable nature optimizes the operation time. Also, designing
within a frame of spatial printing presents even more constraints.

The fidelity between the digital model and the physical model fabricated in
a 3D delta printer is significantly limited by the vitrification process. As with
other 3D printing techniques, slow processes tend to better fidelity. The seg-
ment-like nature of the geometry also limits the design scope.

Combined strategies between spatial printing and conventional 3D printing are

under research. The ulna bone was chosen as the design study, since the degree
of complexity of a bone is convenient to assess the potential benefits and limi-
tations of this design approach.

Moreover, understanding it at a human scale informs possible design consider-

ations at an architectural scale. We acknowledge that this is only a small step
towards a larger design process. However, this design approach can be under- Fig. 3. 29 Collision check in
unit pair.
stood at a higher level, by considering new computational and fabrication tools Geometrical violations are
available to explore alternative design methods. based in toolpath movements
of the nozzle intersecions with
We have used a multi-resolution approach that is limited to different geometri- previously printed geometries
or impossible addition of struts.
cal and scale outputs that define structural behavior and different density lev- Own work based on (Ladrón de
Guevara, et al., 2019)

261
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

[through
11 different
] resolutions
voxelization of the space digital unit ‘arxel’ mapped arxel is the overlap of information,
material and geometry

Fig. 3. 30 Centroids of discretized complex shape form a continuous fibrous structure. (Ladrón de Guevara, 2018)

262
 3. Discrete modelling computational design
Ulna bone case study

els. However, this approach can be expanded in multi-materiality and multi-

fabrication techniques, and also the exploration of novel aesthetic meanings.

Similar to the evolution of the meaning of the value of the ornament in archi-
tecture, from cultural to performative, (Moussavi & Kubo, 2008; Rosenbauer,
1947), new aesthetic values might arise as the result of design alternatives.
Although this chapter attempts to generalize a discrete approach, we under-
stand that a relatively high level of computation is needed in order to follow
this design process. Also, a sophisticated understanding of the material and the
machine is required if the design seeks to leave the digital environment.

The physical pieces presented in this paper require some level of knowledge
about hacking commercial 3D printers. The design space is conducive to cre-
ative exploration, and the understanding of resolution through other fabrica-
tion techniques remains open.

3.3.3. Discrete design

Discrete design enables control over the individual element with significant
methodological and aesthetical consequences. The application of discrete
computational models of design to architecture and its potential augment in
productivity by construction automation is studied.

A novel class of techniques for fabrication and methods of design through

discrete computational models of recursive subdivision based in resolution is
presented.

The discretization of continuous form by octree recursive subdivision is ap-

plied to partition a three-dimensional form to enable ordered value allocation.
Approximating form with discrete units of recursive levels of resolution allows
the designer to work at the level of the constituent element, expanding the de-
sign opportunities but requiring the atomistic understanding of a continuous
form as an informed addition of units and capacities.

The indexing, structuring and geometrical definition of 3D printable informed

units requires the designer to work with data as abstract equivalents of geom-
etry and process of fabrication at the same time.

This research contributes to the generalization of the use of resolution as a

concept for the design of architectural components and uses discrete compu-
Fig. 3. 31 Geometrical stud-
tational models for its development. However, we acknowledge this is just the ies of geometries with capicity
first of many steps for this method to be applicable to the construction industry. to accommodate periodically
subvidided 3D printable linear
Further research is needed to tackle the main issues of FD models such as struc- segments
Own work based on (Ladrón de
tural feasibility or scalability. Additionally, alternative permutations, modifica- Guevara, et al., 2019)

263
 Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces
[ 56 ]

Fig. 3. 32 Large scale 3D printing geometrical arrays with different resolutions. (Ladrón de Guevara, et al., 2019)

264
 3. Discrete modelling computational design
Ulna bone case study

tions and combinations of the techniques presented in this paper might result in
research that contribute further to the advancement in additive manufacturing.

We believe that inserting the data manipulation section between the tool and
output layers augments the flexibility in the design output, leveraging informed
designs and offering different platforms of inquiry.

3.3.3.1. Form and material, towards discrete

characterization of structural lattices.
Mechanical work by shape is the result of arrangement of material, linking the
strength of the whole and the composition of its parts. This also implies the
union between the materials, that is why the form and the material are linked
aspects and the possibility of executing them refers to the possibility of being
built.

Discrete methods of design and fabrication create ordered data structures that
extend the design realm to the particular.

The 21st century search is for free forms, giving new character to spaces by be-
ing more fluid and complex. For this, the evolution of construction systems and
the study of materials and is essential. From nature this symbiosis is expressed
since natural forms are a consequence of the physical-mechanical laws that
compose it. They involve forms, systematization, materialization and every-
thing concludes in an element of balance that continuously responds to intrin-
sic and environmental changes.

The following section expands the concepts of discrete design to create biolog-
ically inspired, mechanically continuous, complex surfaces through discrete
methods.

Fig. 3. 33 Ulna bone adap-

tive grid. Own work based on
(Ladrón de Guevara, et al.,
2019)

265
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

hexahedral matrix double dodecah

10cm 17cm

10cm3

1 hexaedral matrix at segment of

print boundary condition and
continuity of fabrication

dodecaedral matrix for uniform

2 distribution of stress towards boundary
through isostatic vector matrix
80c
design texture through stocastic
3 distribution of open polygons

ø 6mm

80cm 50cm 30cm

80cm3 sample of freeform shell

Fig. 4. 6 Multi-resolution discrete design of locally differentiated spatial lattice. Mockup shape of 1 cubic meter of shell structure is
subdivided into 10 cm hexahedral units to allocate varying materials, density, geometries.
Own work.

266
 Mechanical grading for AM of cellular structures
Abstract

hedron matrix open polygon with stocastic distribution

10cm3 10cm3

cm

267
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Digital configuration of a continuous surface

Digital 3D Printing. Custom tools, code and materials for spatial printing lattic-
Materials and instruments developed es based on (MESH MOULD, WIREPRINT and Borunda)

Serialization of the print process based on (Retsin, LIU,

Borunda)

Concatenation of discrete units with variable resolution

information processing with index creation and orthoedric
matrix organization based on (Borunda)

Digital 3D Printing Cellular membrane structures linear and non-linear static

Process applied to ceullular mem- simulation model
branes configuration. Methods of
discrete arcthiectural design and
computation for additive manufactur- Membrane digitization and information processing
ing of shell structures.

Geometric studies in space filling cellular forms that ap-

proximate hierarchical biomechanics of biological struc-
tures

Hierarchical fibrous structure computation

Characterization of the ceullular membrane

Fiber alignment, characterization and deposition to form

complex cellular structures based in endogenic forces

Prestress cellullar membranes. Characterization and depo-

sition to form complex cellular structures based in exogenic
forces

Principles for 3D printing continuous Goal: Continuous and discontinuous robotic manufactur-
membranes ing of cellular membranes based in forces in equilibrium

Result: Hierarchical fibrous membrane engineer-

ing with fiber characterization based in endogenic
and exogenic forces in equilibrium for increased
mechanical strenght

Fig. 4. 7 Research framework. The goal of this research is to identify and sablish principles of digital 3D printing membranes in
equilibirum, along the additive process local particle deposition and in global monocoque, shaped by a system of forces in space. To
reach this goal, this chapter presentes convering independent studies that configure continuous complex surfaces based in in cellular
structures, spatial 3D printing, biomimicry studies of trabecular and cortical bone and fiber mechanical characterization within a
centralized program of design and projectual basis.
Own work

268
 4. Geometrical Characterization and Mechanical Grading

4. Geometrical
Characterization and
Mechanical Grading

Recent advances in computation allow designers to work at the material level,

designing not only the geometry but also its precise mechanical properties
thanks to numerical control methods. This chapter explores the configuration
of space filling geometries based on a force field.

Hierarchical structures are predicted to have superior mechanical properties,

enhanced strength and very low weight to strength ratio. Cellular structures
prove to be functional and efficient alternative lightweight structures, coupled
with current robotic manufacturing advances, an area of research in additive
manufacturing of spatial lattices is emerging.

Digital is synonym of number based, today’s computers offer automatic calcu-

lations, current computational tools allow designers to compress unprecedent
levels of information and digital fabrication renders this transformation physi-
cal. To provide insight into the additive manufacturing of architectural shells,
I study the mechanical and geometrical properties of highly informed spatial
lattices across meso to large scales. This research investigates the digital Con-
strue, Tensional Characterization and Additive Manufacturing of space filling
tessellations.

Geometric arrays were specially studied in membrane applications and chal-

lenges for creating 3D infills assessed.

Spatial 3D printing of cellular structures allows an efficient increment of

strength allowing the lattice to create isotropic or anisotropic spatial struc-
tures.

This research presents developments in design and manufacturing systems

of functional, light, and high structural capacity digital parts. It is based on
the biomechanics of formation of trabecular or spongy bodies and cortical or
compact bodies. Taking advantage of the precision that robotic manufacturing
offers, we optimize the use of the material, hierarchizing the mechanical be-
havior and varying the density of the deposition in function of tension systems. Fig. 4. 8 Fibrous structures
The computational modeling and fabrication of trabecular and spongy bodies, based on space filling polyhe-
dra
and cortical and dense bodies in porous cellular structures is engineered for Own work

269
Impresión digital 3d, diseño y fabricación digital de superficies continuas
Digital 3d printing, design and fabrication of continuous surfaces

Fig. 4. 9 Hierarchical hypar design by discrete methods.

Own work

270