Computers and Structures 158 (2015) 108–123

Contents lists available at ScienceDirect

Computers and Structures

journal homepage: www.elsevier.com/locate/compstruc

Application of structural topology optimisation to perforated steel

beams
Konstantinos Daniel Tsavdaridis a,⇑, James J. Kingman b, Vassilli V. Toropov c
a
Institute for Resilient Infrastructure, School of Civil Engineering, University of Leeds, Woodhouse Lane, LS2 9JT Leeds, UK
b
AKT II, 100 St John Str., EC1M 4EH London, UK
c
School of Engineering and Materials Science, Queen Mary University of London, Mile End Road, E1 4NS London, UK

a r t i c l e i n f o a b s t r a c t

Article history: This paper focuses on the application of structural topology optimisation technique to design steel
Received 31 July 2014 perforated I-sections as a first attempt to replace the traditional cellular beams and better understand
Accepted 5 May 2015 the mechanisms involved when subjected to bending and shear actions. An optimum web opening
Available online 19 June 2015
configuration is suggested based on the results of parametric studies. A FE analysis is further employed
to determine the performance of the optimised beam in comparison to the conventional widely used
Keywords: cellular type beam. It is found that the optimised beam overperforms in terms of load carrying capacities,
Castellated
deformations, and stress intensities. Barriers to the implementation of the topology optimisation
Cellular
Novel web opening
technique to the routine design of beam web are highlighted.
Structural topology optimisation Ó 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://
SIMP technique creativecommons.org/licenses/by/4.0/).
Opening configuration

1. Introduction aesthetic appeal of the circular opening [1,5]. More recently the
‘‘Angelina’’ style beam was first suggested by Claude Vasconi, an
The judicious placement of holes in the webs of steel beams has architect, with the consideration of aesthetic appeal at the fore
been employed to design lighter and stiffer beams for over [10,11].
100 years. The research focuses on steel I-section beams with per- The progression of the design of beams with web openings,
forations in the web which are variously known as; castellated from the original castellated design to the currently used cellular
beams (with hexagonal openings) and cellular beams (with circu- design and the newly developed sinusoidal design clearly shows
lar openings) under the ‘‘umbrella’’ of perforated beams. The most that ultimate mechanical performance of the section is not always
important benefits when using such sections are the improved the aim of development. Aesthetic appeal may justify the develop-
weight-to-stiffness ratio, the ability to integrate building services ment of a new section in the absence of any mechanical perfor-
into the structural depth and the perceived aesthetic appeal of mance gains.
the beam [1]. The literature on the structural behaviour of steel All of the opening shapes and configurations previously consid-
I-section beams with standard circular, rectangular, and hexagonal ered in the literature are constrained by the requirement that they
web opening configurations is extensive [2–6]. The constant desire can be manufactured by the profile cutting procedure. The profile
for improvement and mature level of understanding of the struc- cutting procedure, also known as castellation, is a method of man-
tural action of perforated steel sections has recently led to novel ufacturing beams with web openings whereby a ‘‘parent’’ steel
opening shapes being investigated [7–9]. The novel opening shapes I-section is selected and a zig-zag pattern is (oxy or plasma) cut
are claimed to have beneficial fabrication, structural performance, along the web of the section. The sections are then expanded and
usage in terms of service investigation and aesthetic qualities welded along the teeth of the zig-zag to form a deeper section with
when compared to standard opening types. web openings (Fig. 1a). It is obvious that if a beam is to be formed
The aesthetics of cellular beams have long been a crucial design using this process the possible geometry of the web openings is
consideration. The switch from castellated style beams to cellular constrained. The profile cutting procedure is currently regarded
beams in the early 1990s has been attributed to the perceived as the most economical and standardised method of fabricating
beams with web openings. The only exception being when wide
elongated web openings are considered.
⇑ Corresponding author.
E-mail address: [email protected] (K.D. Tsavdaridis).

http://dx.doi.org/10.1016/j.compstruc.2015.05.004
0045-7949/Ó 2015 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 109

Fig. 1. (a) Castellation manufacturing process, (b) plate assembly manufacturing process.

More recently an alternative fabrication technique, plate assem- cross-sections found in practice through an extensive parametric
bly (Fig. 1b), has been adopted by several leading fabricators of cel- study; (iv) to establish a nonlinear FE analysis technique which
lular beams. The plate assembly technique offers significantly can be used to determine the web buckling load of the topology
increased design freedom in terms of the shape and layout of web optimised beam web concept; and (v) to parametrically investigate
hence elongated web openings can be accommodated as well as the local buckling behaviour of such thin-walled perforated webs
non-periodically spaced web openings of various sizes and shapes. while changing the geometric characteristics of the new web open-
It is, however, considered an expensive solution due the long weld- ing architecture and compare with existing web opening designs.
ing line along the plates (web and flanges) and the demanding
requirements for accuracy when the perforations are engraved. 3. Optimisation
Nowadays the latter fabrication method is considered as more
advantageous and adaptive to the needs of the architect and design 3.1. Previous works on optimisation of perforated beams
engineer, hence it becomes more and more prevalent and efficient
for steel construction. Within the literature very limited information exists regarding
To this point, no research found in the literature presenting any the optimisation of cellular beams. Erdal et al. [12] conducted an
formal structural optimisation studies that have been performed investigation into the optimum combination of parent section,
with the aim of improving the design of steel I-section beam web number of openings and size of openings for a cellular beam. The
opening shapes used in civil engineering and particularly building problem, as described, is a discrete programming problem and
applications. Previous advancements in web opening configura- both harmony search and particle swarm optimisation techniques
tions have rather relied upon engineering intuition and experience. were investigated. The response of the beam, bending capacity,
Structural topology optimisation is a design tool which can be used shear capacity and deflection, were calculated using design equa-
to determine information on the optimum number, shape and size tions given by BS5950 [13].
of openings within a user defined structural domain. An alternative approach has been presented by Lagaros et al.
[14]. The design of a steel framed structure, incorporating cellular
2. Aim of the study beam elements, was considered. The cross-sectional dimensions of
the sections were defined as size variables whilst the number and
The placement of openings within the web of steel I-section size of web openings were defined as shape and topology optimi-
beams improves the mass-to-stiffness ratio of the section. An sation variables, respectively. A FE simulation of the frame was
improved weight-to-stiffness ratio enables the realisation of longer employed to determine the structural response. The authors found
spans without incurring a significant cost penalty in terms of mate- it necessary to use an evolutionary based optimisation algorithm
rial usage and deflections due to self-weight. due to the mixed, discrete–continuous design variables.
The main aim of the present research study is to investigate the Later, Tsavdaridis and D’Mello [7] have presented a parametric
potential for different web opening configurations in structural optimisation study on an elliptically shaped web opening. The
perforated beams using formalised structural optimisation tech- parametric study involved the FE analysis of models with various
niques. Specifically, structural topology optimisation is to be permutations of the elliptical opening geometric parameters h
applied to the design of the steel beam web as a first attempt to and R. Shear-moment interaction curves were produced for various
replace the conventional cellular beams, investigating the alterna- combinations of the geometric parameters. It was reported that the
tives for perforated beams which fulfil certain boundary criteria as elliptical opening configuration showed enhanced structural per-
well as visualise the structural mechanisms involved when the formance in comparison to circular and hexagonal web openings.
beams are subjected to bending and shear forces. Kingman et al. [15] have presented the first part of this study in
The objectives of this study are (i) to apply a structural topology 2013 on structurally optimised perforated beams. The completion
optimisation technique to the design of a steel I-section beam web; of parametric studies as well as the investigation and development
(ii) to investigate the structural behaviour of a simply supported of realistic web models with practical opening configurations is
beam subjected to vertical shear actions with a topology optimised further conducted in the current study to acknowledge the poten-
web, using nonlinear finite element (FE) analysis; (iii) to invent an tial for use of such optimised beams and the sensitivity of their
effective opening configuration for a wide variety of beam geometric characteristics.
110 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

3.2. Structural optimisation 4.1. Topology optimisation approach and SIMP technique

Structural optimisation can be divided into three distinct prob- Structural topology optimisation is concerned with the identifi-
lem categories; topology, shape and size optimisation [16]. A cation of the optimum number and location of openings, within a
degree of overlap exists between the three categories, and arguably defined designable structural continuum to fulfil a given objective,
a combination of two or more of the categories previously defined subject to applied loading and constraints. Numerous techniques
constitute a fourth category. The most succinct explanation of the for the solution of structural topology optimisation problems have
three methods is pictorial (Fig. 2). been suggested [20–23]. The first solutions to a topology optimisa-
It can be seen that of the three methods topology optimisation tion problem were produced by Michell [24]. Michell’s solutions,
is the most general, yielding information on the number and shape for cases of simple loading and boundary conditions, provide opti-
of openings within a generalised material continua. Potential over- mal topologies for truss-like structures. The basic premise of the
lap between the three categories is also apparent since it may be Michell’s solutions is that it is most efficient for structural ele-
possible to derive a structural topology based on sizing optimisa- ments to follow the lines of principal stress within the ground con-
tion, provided that the minimum size of a member is defined as tinua. Further analytical solutions have been identified for simple
zero. topology optimisation problems [25–27]. However, the complexity
Structural optimisation techniques are most commonly applied of problems of interest to practicing engineers makes the use of
to the design of automotive and aerospace structures where analytical solutions impractical.
weight savings are critical. The application of structural optimisa- Rozvany [28] identified the work of Rosow and Taylor [29], on
tion techniques to the design of building and civil/structural engi- the design of variable thickness sheets, as the first attempt to
neering structures is, however, a more challenging proposition. derive an optimal topology using numerical techniques. The use
These challenges are discussed in this paper. of FE analysis, coupled with optimisation algorithms, has become
the standard procedure for solving topology optimisation prob-
lems. Sigmund and Peterson [30] conceptualised the use of FE
3.3. Structural engineering and optimisation
meshes in topology optimisation problems thus; ‘‘one may con-
sider the design domain as a black and white television screen
Design optimisation techniques are generally best applied to
divided into a lot of small pixels (finite elements), and by turning
products intended for mass production where even small savings,
material on and off in each pixel, one can produce a picture of
per product, can lead to substantial savings over a large production
the optimal structure’’ [30]. Rozvany [28] and Eschenhaur and
run [17]. It is, however, generally the case that civil engineering
Olhoff [31] presented thorough reviews and critiques of contempo-
construction projects are of a one off nature and the standardisa-
rary topology optimisation techniques. Eschenhaur and Olhoff [31]
tion of components is limited. Civil engineering structures and
suggested that topology optimisation techniques can be defined
components are generally designed based on precedent works
into two broad categories; material- or micro- approaches and
and experience.
geometrical- or macro approaches.
The civil engineering industry is inherently conservative when
Micro approaches, and specifically the Solid Isotropic Material
considering newly suggested products or techniques [18]. This
with Penalisation (SIMP) technique, are the most widely utilised
conservative nature is a direct result of the large risks associated
[28]. The SIMP technique, after Bendsøe [32], is a micro approach
with civil engineering project in terms of both safety and finance.
where the optimal structural topology is sought by varying the
The general perception being that to deviate from precedent leads
density of the material within the designable domain. A FE analysis
to unacceptable risks. If a new design is suggested the benefits, in
of the structure is performed to determine the structural response
terms of cost saving, must be such that it makes the perceived risk
of interest which may include; stress, displacement, compliance
of adopting the new design acceptable.
and buckling load, amongst others. Numerous developments and
A further technical challenge of applying design optimisation
extensions of the SIMP method have been developed, accounting
techniques, within the field of civil structural engineering, is the
for a wide variety of structural behaviours.
wide range of disparate and complex design requirements. An ele-
Evolutionary Structural Optimisation (ESO), a macro approach,
gant design example, presented by Nolan [19], demonstrates that it
proposed by Xie and Steven [20] is currently the only potentially
is often the case that designs will be constrained by factors that are
viable alternative to the SIMP method. The basic concept of ESO
not easily quantifiable, such as manufacturability or aesthetic
is to remove material, in the form of finite elements, from areas
appeal. Within this context, focusing on an optimisation study,
of the designable domain that are underutilised. This type of macro
the design of a mass produced component such as perforated
approach differs from the aforementioned material approaches in
beams, may present an opportunity to find material and cost sav-
the respect that the FE mesh itself is altered during the
ings within building structures. Moreover, design regulations for
optimisation process. Criticisms of the ESO method [28,33] focus
such innovative structural elements should be established in the
on its heuristic nature. Improvements and extensions of ESO have
near future to support their use in practice.
been suggested, most notable Bi-directional Evolutionary
Structural Optimisation [34], but the basic criticism remains
4. Introductionary study on a full length beam section unaddressed.
Newly developed structural topology optimisation techniques,
Initially, a topology optimisation study was performed on the including; the bubble method [21], level-set method [22] and
web of a simply supported steel I-section beam of a 5 m span. A topological derivative methods [23], can be described as macro
typical UB 305  165  40 was selected on the basis that it is a approaches. Whilst many of these techniques are potentially
fairly common section to find in practice and mainly in building improvements on currently available methods, none are currently
applications. Thereafter, the structural behaviour of the optimised at a stage of development that allows their commercial application
beam was compared to a similar beam but with circular web open- [28].
ings, by carrying out a nonlinear FE analysis. The topology optimi- It is generally desirable to develop the so called 0-1 design
sation studies were performed using Altair Engineering’s where the final distribution of material, within the design space,
Optistruct software. The comparative nonlinear FE analysis studies is comprised entirely of either solid material or voids. The solution
were performed using ANSYS v.14.0. of the 0-1 problem directly has been attempted, for instance the
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 111

ESO technique. It is however, generally the case that the applica- a nominal size of 10 mm. A linear elastic material model was used
tion of such techniques is computationally prohibitive due to the with Young’s Modulus and Poisson’s Ratio of 200 GPa and 0.3,
number of finite elements necessary to model the design space. respectively. Uniform pressure loading was applied to the top
The SIMP technique addresses this problem by defining the mate- flange of the beam. Constraints were applied to the end of the
rial within each finite element as a continuous design variable. It is lower flange to model the support conditions. A standard linear
worth to mention that by converting the design variable from dis- static FE analysis approach was employed herein.
crete to continuous, it is possible to use more computationally effi- The maximisation of structural stiffness subject to a constraint
cient mathematical programming methods for the solution of the on the available material has been shown to be effective when
original problem [35]. attempting to identify conceptual designs [37]. The optimisation
Intermediate density material, which neither takes the value of approach taken was, therefore, the one of minimising compliance
solid nor void, is generally not desirable since it is not possible to subject to a constraint on the available material.
realise such intermediate densities within a real world structure. It is worth to mention that the implementation of the SIMP
In order to avoid the presence of intermediate densities, within technique, for the solution of practical problems, results in numer-
the final design, a penalisation is used to disproportionately ical complications [30]. The main numerical complications are the
decrease the benefit derived by the presence of intermediate density mesh dependency and the checkerboard problem. As a high num-
material. Penalisation of intermediate densities is achieved, within ber of finite elements used to discretise the design space, an
SIMP, by relating the stiffness of the material to the density thus: increased number of openings within the design space will be pre-
sent in the final design. Numerous techniques have been suggested
E ¼ qP
for generating mesh independent solutions including; perimeter
In practical implementation it has been found that by specifying constraints, local density gradient constraints, filtering of densities,
a large value of P, it may result in convergence to a local minima filtering of sensitivities and MOnotonicity based minimum LEngth
and a poor final design. A continuation method, whereby the value scale (MOLE) controls. Whilst all of the methods listed control the
of the penalisation factor is gradually increased through the itera- geometry of the final solution, the MOLE technique, after Poulsen
tions, has been implemented but does not guarantee convergence [38], allows an explicit statement of the minimum desired member
to a global optimum 0-1 design [36]. size. The added advantage of this technique being that manufac-
Two basic approaches to topology optimisation using the SIMP turability can be directly incorporated into the topology optimisa-
technique exist: tion problem.
Consequently, in order to prevent the emergence of small scale
 Minimum Compliance Design: The minimisation of a specific features in the topology optimisation results, the so called mini-
performance measure subject to a constraint on the available mum member size control was used in this study. A minimum
resources. Usually the compliance of the structure will be member size of 30 mm was specified in order to prevent any
defined as the optimisation objective with a constraint on the design features at a scale smaller than this from emerging in the
available material. topology optimisation results.
 Minimum Weight Design: The minimisation of the mass of the As it was aforementioned, the SIMP technique was used to solve
structure with constraints on specific performance measures. the topology optimisation problem. In order to identify a discrete
The specific performance measures will usually be defined as solid-void final design using the SIMP technique, it is necessary
stress, displacement, buckling load factor or any combination to specify an appropriate value for the intermediate density mate-
thereof. rial penalisation factor [35]. The most appropriate penalisation fac-
tor for this specific problem was investigated through sensitivity
Both of the approaches have been examined in the literature. studies and it was found that a factor of 4 was satisfactory.
The efficacy of the two approaches is debated with numerous The topology optimisation study was performed initially con-
advantages and disadvantages associated with each as shown sidering variations in the maximum available material constraint.
below. The material constraint is specified in terms of a volume fraction,

MINIMUM COMPLIANCE DESIGN MINIMUM WEIGHT DESIGN

Most commonly used approach. Specific performance targets such as

Well understood formulation with efficient maximum stress or buckling load can be
solvers enabling rapid studies. specified in the problem formulation.
Most effectively applied at conceptual design The use of common structural responses such
stage (cannot be easily translated to viable as stress, buckling load, and natural
structural designs). frequency as constraints can cause
Performance targets such as maximum stress convergence issues and are limited to
cannot be specified in the optimisation specialised solution techniques.
process.

where the maximum material allowed in the final design is

The SIMP technique [34] is currently cited as being the most described as a fraction of the material present at the beginning of
prevalent method of solving topology optimisation problems with the optimisation. Volume fraction constraints of 0.5, 0.4 and 0.3
the firmest mathematical basis [28,31]. were studied. A penalisation factor of 3 was initially specified as
it was recommended [39].
4.2. Study model of full length web beam
4.3. Results
A FE model of the steel beam section was created to determine
the structural response of the entire beam for use in the optimisa- The results of the topology optimisation studies are presented
tion algorithm. The beam was modelled using shell elements with in the form of element density plots in Fig. 3. Note that the element
112 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

Fig. 2. The three categories of structural optimisation: (a) sizing optimisation of a truss structure, (b) shape optimisation, and (c) topology optimisation; the initial problems
(left) and the optimal solutions (right).

Fig. 3. Results of topology optimisation studies on 5 m span UB 305  165  40 with varying volume fraction constraints.

density plots represent the optimal material distribution upon con- It is observed that a complex and irregular truss-like design was
vergence of the optimisation. The red1 (light colour) zones repre- formed. It is also noted that more material appeared to be dis-
sent the solid material, whilst the blue (dark colour) zones tributed to areas of high shear, towards the supports. Where bend-
represent the suggested locations for openings. The transitional ing forces are predominant in the flanges, towards the mid-span,
zones represent material of an intermediate density. the topology optimisation suggested the placement of no web
material. It is also appeared that the beam web elements, as sug-
1 gested by the topology optimisation, follow lines of the principle
For interpretation of colour in Fig. 3, 4, 5, 11, 12, and 13, the reader is referred to
the web version of this article. stresses in the beam web.
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 113

Fig. 4. Symmetry constrained topology optimisation study on 5 m span UB 305  165  40 with a constraint on the volume fraction of 0.4.

It is generally the case that structures designed using topology according to the ratio of shear and bending moment acting on
optimisation will have a more complex geometry than those devel- the section.
oped from engineering intuition. This can limit the applicability of It is evident that none of the designs suggested in the topology
topology optimisation derived designs in practice as it may not be optimisation study could be manufacturable using the profile cut-
possible to manufacture them. ting procedure. The plate assembly technique, therefore, need to be
The issue of manufacturability has been addressed through the implemented to realise the designs.
inclusion of manufacturing constraints in the topology optimisa-
tion problem. Suggested manufacturing constraints include: sym- 4.4. Comparison of cellular and topology optimised beams
metry, casting constraints [32], extrusion constraints [40], and
pattern graduation [41]. More recently a novel approach has been 4.4.1. Post-processing topology optimisation results
suggested for enforcing vertical beam like elements only in topol- It was necessary to post-process the topology optimisation
ogy optimisation results [42]. results in order to define the beam web geometry to be used in
Following, a symmetry constrained topology optimisation was the comparative study. The first step in the post-processing proce-
performed. Lines of symmetry about the centreline and longitudi- dure was completed using the OSSmooth feature of Altair
nal axis of the beam were specified in both vertical and horizontal Engineering’s Hypermesh FE pre- and post-processing software.
planes and a topology optimisation was performed using a volume An element density threshold was specified defining the element
fraction constraint of 0.4. The results of the symmetry constrained density above which the design was interpreted as solid geometric
topology optimisation study (Fig. 4) showed a more rational surfaces. Thereafter, Laplacian smoothing is performed by
truss-like design with rhomboidal web openings which periodi- OSSmooth to smooth the boundaries of the design (Fig. 5).
cally change along the length of the beam. Similarly, a large open- Following the initial geometry extraction using OSSmooth, a man-
ing was suggested at the mid-span of the beam. It became visible ual intervention was necessary in order to define the geometry of
that the material is distributed within the beam web the beam web as a series of curves in AutoCAD.

Fig. 5. Extracting geometry from element density results of topology optimisation using OSSmooth: (a) element density plot, (b) geometrically defined surfaces.

Fig. 6. Beam models compared: (a) topology optimised beam web design, (b) cellular beam design.
114 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

4.4.2. FE model of stresses around the openings. A FE technique has been previ-
Cellular beams often exhibit complex structural behaviour ously presented [43] that can capture these complex failure modes.
including localised buckling modes or yielding and redistribution The basis of the FE analysis technique employed is a three-step
process whereby an initial pre-stress was applied to the model
and a linear static analysis was performed. The results of the linear
250 static analysis were then used in an Eigenvalue analysis to deter-
mine the first buckling frequency and the associated mode shape.
Imperfections were then applied to the FE mesh, using the mode
200 shape which was taken from the Eigenvalue analysis, with a mag-
nitude of the web thickness divided by 200. A geometric and mate-
Applied Load (kN)

rially nonlinear FE analysis is finally performed to determine the

150
load at which the beam will buckle. The analysis was performed
using the solver of the FE package of ANSYS. A full detailed descrip-
Cellular Beam
100 tion of the technique, along with a verification of the results
Optimised Beam against experimental data, can be found in [44].
The Young’s Modulus, Poison Ratio, Yield Stress and Tangent
50 Modulus were defined as 200 GPa, 0.3, 355 MPa and 2000 MPa,
Optimised Beam
(reined web mesh) respectively, using a bi-linear elasto-plastic material model.
0
0 4 8 12 16 20 24 28 32 36 40 44 48 52 4.4.3. FE study
Vertical Displacement at Mid-span (mm) The topology optimised beam web design was compared with a
cellular type beam with an identical UB 305  165  40
Fig. 7. Comparison of load versus mid-span deflection for cellular beam and cross-section. To compare the relative efficiency of the beams, in
topology optimised beam. terms of material usage against load capacity, the cellular beam

Fig. 8. Von Mises stress plots for cellular and topology optimised beams at yield load level.

Fig. 9. Localised stress concentrations observed at yield load in cellular beam: (left) support, (right) mid-span.
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 115

was specified such that its mass is as similar as possible to the Table 1
mass of the optimised beam (Fig. 6). A typical opening depth of Illustrative example of compliance weighting calculation.

0.75 times the web depth was specified, with openings at regular Load case
294 mm intervals. If the density of steel is taken as a typical value Moment Shear Moment-shear
of 7850 kg/m3, the weight of the topology optimised beam is
Compliance of plain webbed section 97,951 33,177 61,237
169.87 kg, whilst the one of the cellular beam is 173.64 kg, hence Weighting 0.509 0.172 0.318
only a 2.17% percentage of difference.
In both cases, a stiffener was provided at mid-span in order to
more effectively transfer the shear forces between the top and bot-
tom flange of the beam. This practice was proved appropriate par- Table 2
ticularly for the optimised beam, where there is a wide large web Sections used in topology optimisation studies considering variation to section depth.
gap at the mid-span, which can cause localised stresses in the Length Breadth Depth Flange thickness Web thickness
beam flanges due to high bending. Moreover, it was a prerequisite (mm) (mm) (mm) (mm) (mm)
that the design of the cellular beam should follow the current spec- 576 150 240 6 10
ifications in terms of web opening size and spacing and early fail- 648 168.75 270 6 10
ure is not anticipated. In the case of the topology optimised design, 720 187.5 300 6 10
it was found to be necessary to manually refine the FE mesh to pre- 960 250 400 6 10
1344 350 560 6 10
vent divergence of the solution due to highly distorted elements.
1680 437.5 700 6 10
2160 562.5 900 6 10
4.4.4. Results
The yield load of the cellular beam is found to be 188.15 kN
with a corresponding deflection of 23.58 mm. The yield load of
Table 3
the optimised beam is found to be 207.96 kN with a corresponding
Sections used in topology optimisation studies considering variations in flange
deflection of 19.21 mm. The expected ultimate load of the opti- thickness ratios.
mised beam is also going to be higher than the cellular beam.
Length Breadth Depth Flange thickness Web thickness
Load against mid-span deflection plots for the analyses (Fig. 7)
(mm) (mm) (mm) (mm) (mm)
show that the optimised beam is stiffer than the cellular beam in
960 250 400 6 10
the linear region. It is worth to note that the analyses with the opti-
960 250 400 10 10
mised models stopped suddenly due to high stress concentrations 960 250 400 15 10
at the mid-span of the beam sections. Healing the 960 250 400 20 10
non-convergence issue at this stage in order to capture the 960 250 400 25 10
post-elastic behaviour of the beams is beyond the scope of this 960 250 400 30 10

study, as the stiffness and yielding loads are essentially compared.

In addition, such a weak design at the mid-span of the optimised 4.5. Discussion
beam is not anticipated to be used in practice.
Von Mises stress plots (Figs. 8 and 9) for both examined beams The web design of an I-section steel beam is developed using
at the yield load level, show that despite some high localised stress topology optimisation. When compared to a conventional cellular
concentrations in the optimised beam due to corners resulted from beam, it was found that the topology optimised beam has advanta-
the openings’ shape, the stress distribution is generally more uni- geous structural behaviour in terms of the initial stiffness, the yield
form. In contrast stresses in the cellular beam tend to increase sig- load, and the stress profile. The investigation was therefore consid-
nificantly towards the supports, and therefore the connections. ered to be an effective proof of concept demonstrating that there
It is further observed that the cellular beam suffered a localised are more effective ways of distributing the material within the
buckling failure in the web-posts towards the support. On the beam web depth in order to enhance the structural performance
other hand, the primary failure mode of the optimised beam is dif- of a lightweight perforated section.
ficult to be concluded from this preliminary investigation. It is worth mentioning that the effort required to develop the
Localised buckling within the web depth is observed at failure. topology optimised beam web design was not trivial. It would be
Also, high stress concentrations in the vicinity of the large central necessary to perform such lengthy topology optimisation, geome-
web opening are found, whilst both of these features are observed try extraction and load verification procedure for every new sec-
to occur concurrently. tion, span or load combination considered. Moreover, the load

Rigid Elements

Fig. 10. FE model and loading approach taken in parametric study.

116 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

approach was followed. Rather than considering a full length beam

with externally applied loading, a short section of a beam was
modelled and internal shear and bending moments were applied
directly to the section. This approach enables a wide variety of
cross-sections to be investigated, and results in a new opening type
that could be utilised generally.

5.1. Parametric topology optimisation approach

Fig. 11. Identification of generalised opening type from topology optimisation A FE model of a short beam section was developed. Loading was
studies on short section of beam. applied to the model in the form of bending moments and shear
forces. Rigid elements were used in the application of the bending
moment and shear force (Fig. 10).
capacity of the optimised beam web cannot be verified using stan- Three load cases were considered in the topology optimisation:
dardised design equations due to the complexity of the new (i) bending moments only, (ii) shear forces only, and (iii) a combi-
geometry. nation of shear forces and bending moments.
A weighted compliance approach was established when consid-
5. Development of a topologically optimum web opening type ering the three load cases in the optimisation. The weighting for
each of the individual compliances was defined relative to the
In light of the observations detailed above, it was concluded compliance of the plain webbed section subject to each of the load
that the topology optimisation is a useful tool for identifying alter- cases. The process of determining the weightings for an illustrative
native improved beam web opening designs. However, when it example is summarised in Table 1.
was applied to a full length beam, the resulting design was rather The parameters varied were the depth of the section and the
complex and difficult to apply in a general sense. ratio of flange and web thickness. Full details of all of the
In order to establish a topologically optimum opening architec- cross-sections considered in the parametric study of the depth
ture that can be applied to all beam cross-sections and positioned variation (Table 2) and the flange ratio variation (Table 3) are
at any location along the length of the beam, a local modelling included.

200mm Deep 240mm Deep

270mm Deep 300mm Deep

400mm Deep 560mm Deep

700mm Deep 900mm Deep

Fig. 12. Results of parametric topology optimisation studies considering variations in beam depth.
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 117

0.667 1.000

1.500 2.000

2.500 3.000

Fig. 13. Results of parametric topology optimisation studies considering variations in ratio of flange thickness to web thickness.

It was found that the ratio of the flange thickness to the web
thickness did not alter the resulting optimal topology, but the
actual shape of the suggested web opening was altered (Fig. 12).
The depth of the section did have an impact on the optimal topol-
ogy (Fig. 13). In the depth range between 270 mm and approxi-
mately 750 mm, it was found that the optimal opening topology
was consistent. For shallower beams, it was found that an elon-
gated cross shape was the most efficient design. In the case of
beams deeper than 750 mm, extra openings were recommended
by the topology optimisation results to keep a balance between
stiffness and weight.
Fig. 14. Suggested topologically optimum web opening concept.
The consistency of the results in the depth range between
270 mm and 750 mm indicate that the same topology is optimal
for most of beam sections that may be found in practice.
5.2. Discussion
5.3. Topologically optimum web opening architecture
It should be noted that when interpreting the results of the
topology optimisation studies, only the opening pattern emerging Based on the results of the parametric studies, an optimum web
in the centre of the beam (Fig. 11) is of interest in terms of devel- opening type has been developed (Fig. 14). It should be noted that
oping a web opening type that could be periodically applied along this optimum topology may be interpreted using various opening
the length of the beam.

Fig. 16. Topologically optimum web opening concept for beam greater than
Fig. 15. Alternative interpretation of topologically optimum web opening concept. 750 mm deep.
118 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

The three key geometric parameters (Fig. 17) are defined as

multiples of the beam depth. The value of ‘‘a’’ defines the depth
of the large central opening as a multiple of beam depth, D. The
value of ‘‘b’’ defines the depth of the cross-over of the web struts
as a multiple of D. Finally ‘‘c’’ describes the ‘‘aspect ratio’’ of the
openings as a multiple of D. It is therefore possible to fully describe
the geometry of the beam web openings through a combination of
the values a, b, and c.

6.2. Buckling behaviour

Fig. 17. Key geometric parameters of suggested beam web opening concept and
beam web geometry defined as functions of beam depth. The initial study into the behaviour of beams with the newly
suggested web opening architecture focused on web buckling
behaviour under shear loading. The purpose of this initial investi-
shapes, such as circular openings depicted in Fig. 15. It is antici- gation is to determine how the buckling behaviour of the topology
pated that a shape optimisation study would further yield valuable optimised web opening concept behaves and later compares to
information regarding the optimum shape of the openings. that of cellular, castellated and other perforated beams with novel
The concept of this opening architecture is suggested to be configurations. A local approach, considering only a short section
topologically optimal for all beam cross-sections in the depth of the beam web, is taken as only local instabilities are of interest
range between 270 mm and 750 mm. In the case of beams deeper hereby.
than 750 mm deep, an alternative opening architecture is recom-
mended as being topologically optimal (Fig. 16). 6.3. FE analysis
It is aforementioned that it would not be possible to create
beams with the suggested web opening topologies using the pro- The buckling behaviour of beam webs with openings under
file cutting procedure. However, it would be possible to create such shear loading is known to be complex [43]. A materially and geo-
beams based on the plate assembly method of fabrication. This metric nonlinear FE analysis approach was implemented to deter-
conforms to the current trends of the leaders in steel fabrication mine the buckling load of the beam web. A three-step analysis
industry (eg. Fabsec, ASD Westok, ArcelorMittal, and Macsteel), approach, similar to that previously utilised in the analysis of the
in which the plated beam design is now vastly utilised and pro- full length topology optimised beam, is implemented to determine
moted to accommodate architects’ and designer’s needs. It is worth the buckling load of the beam web. Radioss, a FE solver also devel-
to notify that the reduced connectivity of the web to the flanges oped by Altair Engineering, was used in the analysis of the topol-
due to the top and bottom edge web openings implied a rather ogy optimised web opening. The model developed uses
high sensitivity to torsional loads. This unwanted structural beha- four-node quadrilateral shell elements with six degrees of freedom
viour can be limited by the use of an additional strip of steel at the (6DOF) at each node. The Batoz–Dhatt shell element formulation
top and bottom edges of the web plate, similarly to the result from was used in the nonlinear analysis. The automesh function, avail-
the full length model (Figs. 8 and 9). able within Hypermesh, was used to generate the FE mesh. A nom-
inal element size of 10 mm was specified to ensure a sufficient
number of elements within the depth of the web to accurately
6. Shear buckling study of the short beam section model the behaviour of the beam. A bilinear elasto-plastic material
model, with post-elastic strain hardening, was specified. Details of
A new beam web design configuration is suggested based on the the parameters used in the material model can be found in Table 4
results of the topology optimisation investigations performed so and Tsavdaridis and D’Mello [43]. A linear static analysis, using a
far. A detailed description of the suggested beam web configura- unit load applied in the location of the applied load, was the first
tion is given and a preliminary investigation into the behaviour
and performance of the new beam web design is presented herein.

6.1. Geometric properties

An interpretation of the optimal topology is presented and a

description of the key geometric parameters is given (Fig. 17). It
should be noted that this particular interpretation was made due
to its simplicity as an initial investigation. It is expected that future
studies may investigate alternative opening architectures and con-
figurations while the scope of this research is to introduce the con-
cept and highlight the steps that need to be followed to establish
the new designs.

Table 4
Material properties.

Parameter Experimentally determined Nominal

Web Flange/Stiffeners All parts
Young’s Modulus (N/mm2) 200,000 200,000 200,000
Yield stress (N/mm2) 375.30 359.70 355.00
Ultimate stress (N/mm2) 492.70 480.90 510.00
Tangent Modulus (N/mm2) 580 580 580
Fig. 18. Local model developed for topology optimised web concept buckling study.
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 119

Table 5 behaviour of beams with web openings [44]. The local model
Constraints applied to local model for topology optimised web buckling study. (Fig. 18) represents a section of the beam from centreline to centre-
Part Constraints line of centrally positioned web openings. Constraints are cau-
Left side flanges Translation: x and z tiously applied to the ends of the beam section to conservatively
Rotation: x and y replicate the restraint which would be present in a full length
Left side web Translation: x, y and x beam. A summary of the constraints is given in Table 5. Shear force
Rotation: x and z is applied only to the web of the local beam model.
Right side flanges Translation: y
Rotation: x, y and z
Right side web Translation: y 6.5. Parametric study
Rotation: y and z
The topology optimised beam web architecture differs signifi-
cantly to previously investigated web opening designs. It was
therefore decided to perform a parametric study investigating as
step in the analysis process. An Eigenvalue analysis was subse- a wide range of topology optimised web opening configurations.
quently performed. The Block Lanczos method is used to extract A total of 96 models were examined. The depth of the large central
the Eigenvalues. The five lowest Eigenvalues and associated buck- opening was varied between 0.5 and 0.8 times the beam depth
ling modes were requested. Following the computation of the with increments of 0.1. The depth of the cross-over was varied
Eigen modes, geometric imperfections were applied to the FE between 0.1 times the beam depth and 0.8 times the beam depth
mesh. The imperfections were applied in the direction of the in increments of 0.1. Values of 1.0, 1.5, and 2.0 times the beam
Eigen mode as extracted by the lowest Eigenvalue. The magnitude depth were investigated for the value of c. A full summary of all
of the imperfections was defined as the thickness of the web of the configurations used in the parametric study can be found
divided by 200 following recent studies [44]. A nonlinear static in Table 6.
analysis was finally performed using the FE mesh including imper- The same beam cross-section, a standard 457  152  52 UB,
fections. Monotonically increasing loading is applied to the FE was used in all of the studies. This section was selected on the basis
model. The modified Newton-Raphson method is used to solve that previous studies [6,8,43], considering the web buckling of
the nonlinear problem at each loading increment. The convergence alternative opening types, have been performed using this particu-
criterion was defined as total work with a tolerance of 0.01. lar cross-section. Comparisons between the buckling behaviour of
Automatic time-step control, based on the arc-length method, the topology optimised web opening configuration and currently
was used to control and speed the convergence of the analysis. used opening types is therefore possible.

6.4. Local model 6.6. Results

It has been shown that a local modelling approach is the most Each of the web configurations described was analysed using
effective one when attempting to determine the web buckling the Radioss FEM. The buckling load was defined as the point at

Table 6
Summary of all of the combinations of parameters a, b and c.

Test reference a b c Buckling load (kN) Test reference a b c Buckling load (kN) Test reference a b c Buckling load (kN)
1 0.5 0.1 1 215.81 33 0.5 0.1 1.5 110.58 65 0.5 0.1 2 74.5
2 0.5 0.2 1 232.52 34 0.5 0.2 1.5 157.16 66 0.5 0.2 2 104.8
3 0.5 0.3 1 289.16 35 0.5 0.3 1.5 216.09 67 0.5 0.3 2 144.59
4 0.5 0.4 1 370.3 36 0.5 0.4 1.5 287.59 68 0.5 0.4 2 205.74
5 0.5 0.5 1 415.74 37 0.5 0.5 1.5 373.74 69 0.5 0.5 2 282.02
6 0.5 0.6 1 452.28 38 0.5 0.6 1.5 419.62 70 0.5 0.6 2 359.02
7 0.5 0.7 1 476.38 39 0.5 0.7 1.5 467.9 71 0.5 0.7 2 368.67
8 0.5 0.8 1 482.17 40 0.5 0.8 1.5 455.37 72 0.5 0.8 2 366.02
9 0.6 0.1 1 157.54 41 0.6 0.1 1.5 107.92 73 0.6 0.1 2 67.51
10 0.6 0.2 1 225.16 42 0.6 0.2 1.5 145.16 74 0.6 0.2 2 95.8
11 0.6 0.3 1 281.45 43 0.6 0.3 1.5 205.09 75 0.6 0.3 2 139.09
12 0.6 0.4 1 331.59 44 0.6 0.4 1.5 276.59 76 0.6 0.4 2 193.74
13 0.6 0.5 1 370.09 45 0.6 0.5 1.5 365.91 77 0.6 0.5 2 268.02
14 0.6 0.6 1 391.74 46 0.6 0.6 1.5 351.02 78 0.6 0.6 2 317.02
15 0.6 0.7 1 397.74 47 0.6 0.7 1.5 352.02 79 0.6 0.7 2 303.02
16 0.6 0.8 1 427.44 48 0.6 0.8 1.5 352.02 80 0.6 0.8 2 303.02
17 0.7 0.1 1 133.16 49 0.7 0.1 1.5 91.57 81 0.7 0.1 2 57.01
18 0.7 0.2 1 199.3 50 0.7 0.2 1.5 129.16 82 0.7 0.2 2 82.3
19 0.7 0.3 1 251.45 51 0.7 0.3 1.5 183.09 83 0.7 0.3 2 125.9
20 0.7 0.4 1 314.03 52 0.7 0.4 1.5 243.59 84 0.7 0.4 2 175.74
21 0.7 0.5 1 361.35 53 0.7 0.5 1.5 252.87 85 0.7 0.5 2 233.02
22 0.7 0.6 1 319.74 54 0.7 0.6 1.5 282.02 86 0.7 0.6 2 240.02
23 0.7 0.7 1 324.4 55 0.7 0.7 1.5 282.02 87 0.7 0.7 2 240.02
24 0.7 0.8 1 324.02 56 0.7 0.8 1.5 285 88 0.7 0.8 2 240.02
25 0.8 0.1 1 108.12 57 0.8 0.1 1.5 70.72 89 0.8 0.1 2 42.87
26 0.8 0.2 1 158.08 58 0.8 0.2 1.5 105.87 90 0.8 0.2 2 69.16
27 0.8 0.3 1 233.85 59 0.8 0.3 1.5 149.8 91 0.8 0.3 2 105.16
28 0.8 0.4 1 248.24 60 0.8 0.4 1.5 185.8 92 0.8 0.4 2 146.45
29 0.8 0.5 1 256.51 61 0.8 0.5 1.5 206.45 93 0.8 0.5 2 170.42
30 0.8 0.6 1 245.36 62 0.8 0.6 1.5 217.74 94 0.8 0.6 2 181.74
31 0.8 0.7 1 248.88 63 0.8 0.7 1.5 223.74 95 0.8 0.7 2 193.74
32 0.8 0.8 1 219.02 64 0.8 0.8 1.5 229.74 96 0.8 0.8 2 193.74
120 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

which the maximum out-of-plane displacement of the beam web 6.7. Discussion
reached a value of 0.2 mm as this was observed to be the point
at which the out-of-plane displacement rapidly increased indicat- It was found that as the depth, a, of the large central opening is
ing a buckling failure (Fig. 19). Also, it can be observed that the increased the buckling load capacity of the beam web reduces. It
higher is the total length, c, the more linear proved to be the buck- was also found that as the width, c, is increased, the buckling load
ling load capacity evaluation when ranging the cross-over depth, b of the beam web is also reduced. These relationships are intuitively
(Fig. 19 – polynomial trends presented). correct.

600 600
a=0.8
a=0.8
500 a=0.7 500 a=0.7
Buckling Load (kN)

a=0.6 a=0.6

Buckling Load (kN)

a=0.5 a=0.5
400 400

300 300

200 200

100 100
(a) (b)

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Value of Parameter b Value of Parameter b

600
a=0.8
500 a=0.7
Bucklimg Load (kN)

a=0.6
a=0.5
400

300

200

100
(c)
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Value of Parameter b

Fig. 19. Shear buckling capacities: (a) parameter c = 1.0, (b) parameter c = 1.5, (c) parameter c = 2.0.

Fig. 20. Out-of-plane displacement vectors: (a) buckling along edge of central opening in test 96, (b) buckling along edge of perimeter opening in test 94.
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 121

Fig. 21. Von Mises stress plots for: (a) test 11, (b) test 14.

Fig. 22. (a) Misalignment of web struts, (b) resulting uneven stress distribution across the breadth of the web strut.

The relationship between the depth, b, of the cross-over and In the course of the parametric study a major limitation of this
the buckling load is more complex. It was concluded that the topology optimised beam web architecture, as currently described,
relationship between the cross-over depth and the buckling load was identified. It became apparent that the method of defining the
is approximately linear until a limiting value is reached, at geometry of the topology optimised beam web leads to misalign-
which point the buckling load tends to an approximately con- ment of the web struts across the cross-over point (Fig. 22). This
stant value. For the web configurations investigated, it was esti- misalignment of the web struts generates an uneven stress distri-
mated that the cross-over depth at which the buckling load bution across the breadth of the strut. The resulting increased
tends to a constant value was between 0.5 and 0.6 times the stress towards the edge of the opening, promotes a premature
beam depth. buckling along the web opening.
The limiting shear buckling capacity was reached when the web Consequently, it is recommended that a future study is con-
buckled along the edge of the central web opening (Fig. 20a) rather ducted to investigate the impact of aligning the web struts. It is
than the edge of the perimeter opening (Fig. 20b). It can be con- anticipated that alignments of the web struts will enhance the
cluded from this that if the edge length of the perimeter opening buckling characteristics of the beam web.
is maintained below the edge length of the central opening the
shear buckling capacity of the web will not be altered.
7. Limitations and concluding remarks
Significant yielding and distortion of the web in the cross-over
area zone (Fig. 21) was observed in tests 1–32 from Table 4, where
The application of structural topology optimisation techniques
the value c was defined as 1.00, prior to the buckling load being
to the design of steel I-section perforated beams has been investi-
reached. This would indicate that the critical failure mode in these
gated for the first time. A web opening design for a 5 m span uni-
instances is not the web buckling, but a Vierendeel type failure
versal beam section was developed using the solid isotropic
which cannot be accurately modelled using the local study
material with the penalisation technique of topology optimisation.
approach [45].
A geometric and material nonlinear FE analysis, using ANSYS, was
122 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123

employed in order to predict the load carrying capacity of the References

topology optimised beam. It was found that when compared to a
cellular type beam, the optimised design had increased stiffness [1] Knowles PR. Castellated beams. Proc Inst Civ Eng 1991;90:521–36.
[2] Redwood RG. The strength of steel beams with unreinforced web holes. Civil
and yield load. Various restrictions to the routine application of Eng Public Works Rev 1969;64:559–62.
topology optimisation, including the complexity of the web geom- [3] Kerdal D, Methercot DA. Failure modes of castellated beams. J Constr Steel Res
etry as well as the analysis difficulties, were highlighted. It is 1984;4:295–315.
[4] Hosain MU, Speirs WG. Experiments on castellated beams. J Am Weld Soc
expected that different depth-to-span ratios will result different 1973;52(8):329–42.
web opening patterns of a similar architecture developed by the [5] Ward JK. Design of composite and non-composite cellular beams
topology optimisation study. (P100). Ascot: Steel Construction Institute; 1990.
[6] Chung KF, Liu TCH, Ko ACH. Investigation on Vierendeel mechanism in steel
Thereafter, a parametric topology optimisation study has been
beams with circular web openings. J Constr Steel Res 2001;57:467–90.
performed on a short section of a beam in order to develop a topo- [7] Tsavdaridis KD, D’Mello C. Optimisation of novel elliptically-based web
logically optimum web opening architecture for the wide variety of opening shapes of perforated steel. J Constr Steel Res 2012;76:39–53.
beam cross-sections that may be found in practice. An optimum [8] Tsavdaridis KD, D’Mello C. Vierendeel bending study of perforated steel beams
with various novel shapes of web openings, through non-linear finite element
web opening architecture has been developed (Figs. 14 and 15), analyses. J Struct Eng – ASCE 2012;138(10):1214–30.
whilst it is effective for beams of depth between 270 mm and [9] Durif S, Bouchair A, Vassart O. Experimental tests and numerical modeling of
750 mm. An alternative optimum web opening architecture is sug- cellular beams with sinusoidal openings. J Constr Steel Res 2013;82:72–87.
[10] Wang P, Ma Q, Wang X. Investigation of Vierendeel mechanism failure of
gested for beams deeper than 750 mm (Fig. 16). This initial sugges- castellated steel beams with fillet corner web openings. Eng Struct
tion is made as to potential shapes for the web openings, and their 2014;74:44–51.
geometry is described as a function of beam depth. It is envisaged [11] Vassart O. Development of a new type of composite cellular beam. Singapore
Eng 2008;3:14–6.
that this short section with a web opening architecture can be then [12] Erdal F, Doğan E, Saka MP. Optimum design of cellular beams using harmony
repeated to form a beam of any span. search and particle swarm optimizers. J Steel Constr Res 2011;67:237–47.
Furthermore, the shear buckling behaviour of the suggested [13] BS5950-1:2000. Structural use of steelwork in building. Code of practice for
design. Rolled and welded sections.
short section web architecture was investigated using nonlinear [14] Lagaros ND, Psarras LD, Papadrakakis M, Panagiotou G. Optimum design of
FE buckling analysis, and a comprehensive parametric study was steel structures with web openings. Eng Struct 2008;30:2528–37.
conducted to investigate a wide range of possible configurations. [15] Kingman JJ, Tsavdaridis KD, Toropov VV. Application of topology optimisation
to the design of steel I-section webs. In: The 14th international conference of
It was observed that the optimised section has a higher buckling
civil, structural and environmental engineering computing, 3–6 September.
capacity, when subjected to shear load, compared to a cellular sec- Cagliari, Sardinia, Italy; 2013. http://dx.doi.org/10.4203/ccp.102.223.
tion of the same weight. [16] Christensen PW, Klarbring A. An introduction to structural
Due to certain limitations, further work is anticipated to focus on optimisation. Sweeden: Springer Science & Business Media B.V.; 2009.
[17] Belegundu AD, Chandrupatal TR. Optimisation concepts and applications in
the optimisation of the opening shapes to be employed in the topo- engineering. 2nd ed. New York: Cambridge University Press; 2011.
logically optimum web opening architecture. Investigations of the [18] Engineering Science and Physical Research Panel. EPSRC review of UK
ultimate load carrying capacities and the associated failure modes academic research in ground and structural engineering; 2009.
[19] Nolan J. Cost versus value – the role of the consulting engineer. Struct Eng
of the topologically optimum web opening configurations are 2012;90(2):20–1.
deemed necessary. Following preliminary analyses, it was consid- [20] Xie YM, Steven GP. An evolutionary procedure for structural optimisation.
ered of paramount importance to investigate the results of various Comput Struct 1993;49(5):885–96.
[21] Eschenauer HA, Kobelev HA, Schumacher A. Bubble method for topology and
complex loading and support conditions as well as depth-to-span shape optimisation of structures. Struct Multidiscipl Optim 1994;8(1):42–51.
ratios which may trigger other failure modes. Similarly to the [22] Wang MY, Wang X, Guo D. A level set method for structural topology
design of cellular, castellated, and other perforated beams with var- optimisation. Comput Methods Appl Mech Eng 2003;192(1–2):227–46.
[23] Norato JA, Bendsøe MP, Haber RB, Tortorelli DA. A topological derivative
ious standard and non-standard web opening shapes, the diversity method for topology optimisation. Struct Multidiscipl Optim 2007;22(4–
of loading and supports conditions is to be dealt in the beam design 5):375–86.
process as well as introducing web and flange stiffeners and/or [24] Michell AGM. The limits of economy of material in frame structures. Phil Mag
1904;8(6):589–97.
using partial factors. In case of a steel frame structure is considered,
[25] Rozvany GIN, Prager W. Optimal design of partially discretised grillages. J
the engineer is able to use the optimised beams as formed by the Mech Phys Solids 1976;24(2/3):125–36.
periodical position of the short sections, and treat each beam of [26] Lewinski T, Rozvany GIN. Exact analytical solutions for some popular
the frame whilst using its equivalent static properties assuming benchmark problems in topology optimisation II: three-sided polygonal
supports. Struct Multidiscipl Optim 2007;33(4–5):337–50.
constant cross-section along its length (Euler beam theory). The [27] Lewinski T, Rozvany GIN. Exact analytical solutions for some popular
ultimate aim of this research project is the development of design benchmark problems in topology optimisation III: L-shaped domains. Struct
equations that can be used in the routine design of beams employ- Multidiscipl Optim 2008;35(2):165–74.
[28] Rozvany GIN. A critical review of established methods of structural and
ing such optimum web opening configurations. topology optimisation. Struct Multidiscipl Optim 2008;37(3):217–37.
The cost of such optimised beams is related to the cost of the [29] Rossow MP, Taylor JE. A finite element method for the optimal design of
steel used (plus the scrap metal) as well as the manufacturing pro- variable thickness sheets. AIAA J 1973;11(11):1566–9.
[30] Sigmund O, Petersson J. Numerical instabilities in topology optimisation: a
cess. Cost analysis has not been introduced in this paper since the survey on procedures dealing with checkerboards, mesh-dependencies and
new opening architecture is not yet finalised. local minima. Struct Optim 1998;16:68–75.
[31] Eschenauer HA, Olhoff N. Topology optimisation of continuum structures: a
review. Appl Mech Rev 2001;54(4):331–90.
Acknowledgements [32] Zhou M, Rozvany GIN. On the validity of ESO type methods in topology
optimisation. Struct Multidiscipl Optim 2001;21(1):80–3.
[33] Querin OM, Steven GP, Xie YM. Evolutionary structural optimisation (ESO)
The first author, Dr. K.D. Tsavdaridis, would like to thank J.J. using a bi-directional algorithm. Eng Comput 1998;15(8):1031–48.
Kingman for the work that he conducted during his MEng thesis at [34] Bendsøe MP. Optimal shape design as a material distribution problem. Struct
the University of Leeds (2012–13) under his and Prof. V.V. Optim 1989;1(4):193–202.
[35] Bendsøe MP, Sigmund O. Topology optimisation: theory, methods, and
Toropov’s supervision. He would like to acknowledge the
applications. 2nd ed. Berlin: Springer Verlag; 2003.
contribution of the EPSRC DTG CASE (201324) and ASD Westok [36] Stolpe M, Svanberg K. On the trajectories of penalisation methods in topology
ltd. for their generous support. The authors would also like to thank optimisation. Struct Multidiscipl Optim 2001;21:128–39.
Dianzi Liu for the technical assistance on the use of Hypermesh and [37] Cervellera P, Zhou M, Schramm U. Optimisation driven design of shell
structures under stiffness, strength and stability requirements. In: 6th world
Optistruct software packages and Luca Frattari of Altair Engineering congress of structural and multidisciplinary optimisation, 30th May – 3rd June
Inc. for the assistance on the post-processing of the results. 2005. Rio de Janeiro, Brazil; 2005.
 K.D. Tsavdaridis et al. / Computers and Structures 158 (2015) 108–123 123

[38] Poulsen TA. A new scheme for imposing minimum length scale in topology [43] Tsavdaridis KD, D’Mello C. Web buckling study of the behaviour and strength
optimisation. Int J Numer Method Eng 2003;57(6):741–60. of perforated steel beams with different novel web opening shapes. J Constr
[39] Altair Engineering Inc. Hyperworks 11.0; Radioss, motion solve and optistruct Steel Res 2011;67(10):1605–20.
user’s guide; 2012. [44] Tsavdaridis KD, D’Mello C. Finite element investigation on web-post buckling
[40] Zhou M, Fleury R, Shyy YK, Thomas H, Brennan JM. Progress in topology of perforated steel beams with various web opening shapes subjected under
optimisation with manufacturing constraints. United States of different shear-moment interaction. In: 6th European conference on steel and
America: American Institute of Aeronautics; 2002. composite structures – EUROSTEEL, 31 August – 2 September 2011, vol. C.
[41] Stromberg LL, Beghini A, Baker WF, Katz N, Paulino GH. Connecting Budapest, Hungary; 2011. p. 1851–56.
engineering and architecture through structural topology optimisation; 2011. [45] Tsavdaridis KD, D’Mello C. Vierendeel bending study of perforated steel beams
[42] Leiva JP. A topology optimization method to extract optimal beam-like, plate- with various novel shapes of web openings, through non-linear finite element
like, or shell-like structures from a solid finite element mesh. In: Eng Opt, 3rd analyses. J Struct Eng, ASCE 2012;138(10):1214–30.
International Conference on Engineering Optimization, Rio de Janeiro, Brazil;
01–05 July 2012.