SCHOOL OF ENGINEERING

JÖNKÖPING UNIVERSITY

ANALYSIS OF CASTING PROCESS FOR

COMPLEX ELECTRONIC UNIT

Endrias Teklu
Rebal Marcos

THESIS WORK 2009

PRODUCT DEVELOPMENT AND MATERIALS

ENGINEERING
 SCHOOL OF ENGINEERING
JÖNKÖPING UNIVERSITY

ANALYSIS OF CASTING PROCESS FOR

COMPLEX ELECTRONIC UNIT

Endrias Teklu
Rebal Marcos

This thesis work is performed at Jönköping Institute of Technology within the

subject area Product Development and Materials Engineering. The work is
part of the university’s master’s degree.

The authors are responsible for the given opinions, conclusions and results.

Supervisor: Magnus Wessén

Credit points: 30 ECTS credits

Date:

Archive number:

Postal Address: Visiting Address: Telephone:

Box 1026 Gjuterigatan 5 036-10 10 00
551 11 Jönköping
 Abstract

Abstract
Most aircraft component are currently being manufactured by machining, forging, welding
and also assembling such parts. However, the possibilities of cutting cost from a single
component has brought about a growing trend towards looking into casting as a possible
option for manufacturing aircraft parts.
This thesis was done at the request of Saab Avitronics. It evaluates the possibilities of one
aircraft part, a chassis for an electronic unit that was first designed to be machined from a
blank, to be cast. The thesis goes through the multifaceted tasks of product development.
Casting process selection, cast alloy selection as well as geometry modification were some of
these tasks that were performed in this thesis. It also evaluates the performances of chosen
casting processes, the design of gating systems as well as various process parameters set, by
simulating the casting processes.
The alloy chosen was A356.0 with a T6 temper and the casting processes chosen were plaster
mold casting and rheocasting. The geometry of the original chassis, which had very thin
sections and undercuts which were complex to cast, was modified and made easier to cast
with an acceptable slight increase of mass and size. The modification done on the geometry as
well as the gating systems used had proven to be worthwhile, as the simulation of both
process showed that such a part can be casted with no crucial defects foreseen. However,
probable cavities might occur at the very tip of the chassis‟s thin-fins – that it has for carrying
away heat. Minor subsurface porosities might also be formed, which would not impair the
function of the chassis. The modified chassis was made as close to as finished piece as
possible, for the purpose of reducing machining costs. The cost of producing such a part by
casting was also seen to be much less than machining it from blank. This could be taken as
rationale for casting the chassis with thicker sections, to avoid problems that may arise in
casting, and to subsequently machine these faces later, as it would still be cheaper than
machining the chassis from a blank.

Key Words: design for casting, casting process selection, simulation of rheocasting,
simulation of plaster mold casting, casting cost estimation.

I
 Acknowledgements

Acknowledgements

We want to start with a big thanks to Saab Avitronics in Jönköping that gave us the
opportunity to carry out this thesis. We also want to thank everyone who expressed interest
and got involved with this thesis.

A special thanks goes to our supervisors Mats Johansson and Pierre Ekström at Saab
Avitronics and Magnus Wessén at Jönköping University, School of Engineering.

We would also like to pass our gratitude to Hackås Precisionsgjuteri AB, Comptech i
Skillingaryd AB and RheoMetal AB.

Finally, a special thanks goes to our families for their support during our studies.

II
 Table of Contents

Table of Contents

1 Introduction ............................................................................. 1
1.1 COMPANY DESCRIPTION ......................................................................................................... 1
1.2 BACKGROUND ....................................................................................................................... 1
1.3 THE PRODUCT ........................................................................................................................ 1
1.4 PURPOSE AND AIMS................................................................................................................ 2
1.5 DELIMITS .............................................................................................................................. 2
1.6 OUTLINE ............................................................................................................................... 2

2 General concepts ..................................................................... 4

2.1 PRODUCT DEVELOPMENT ....................................................................................................... 4
2.2 MATERIAL SELECTION ........................................................................................................... 4
2.2.1 General ....................................................................................................................... 5
2.2.2 Cast Alloys .................................................................................................................. 5
2.2.3 Designation ................................................................................................................. 5
2.2.4 Standard materials in the aviation industry .................................................................. 6
2.3 PROCESS SELECTION .............................................................................................................. 8
2.3.1 Expendable mold processes.......................................................................................... 8
2.3.2 Non-expendable mold processes ................................................................................. 11
2.3.3 Basic casting terminologies........................................................................................ 13
2.4 FUNDAMENTALS & ASSOCIATED PROBLEMS .......................................................................... 15
2.5 DEFECTS IN ALUMINUM CASTING .......................................................................................... 16
2.6 SIMULATION ........................................................................................................................ 22
2.6.1 MAGMASOFT (V 4.4)................................................................................................ 23
3 Groundwork ........................................................................... 24
3.1 PRODUCT REQUIREMENT ...................................................................................................... 24
3.2 MATERIAL SCREENING ......................................................................................................... 24
3.3 PROCESS SCREENING ............................................................................................................ 25
3.3.1 Selected processes...................................................................................................... 28
3.4 REVISION: CHOICE OF ALLOYS ............................................................................................. 30
3.5 GEOMETRY MODIFICATION .................................................................................................. 31
3.5.1 Fins ........................................................................................................................... 34
3.5.2 New box ..................................................................................................................... 35
3.5.3 Allowances ................................................................................................................ 38
3.6 GATING SYSTEMS ................................................................................................................ 42
3.6.1 Gating system for plaster casting................................................................................ 42
3.6.2 Gating system for rheocasting .................................................................................... 46
3.7 SIMULATION SETUP ............................................................................................................. 47
3.7.1 The preprocessor - Plaster casting ............................................................................. 47
3.7.2 Enmeshment parameter - Plaster casting .................................................................... 48
3.7.3 Simulation parameters - Plaster casting ..................................................................... 48
3.7.4 The preprocessor - Rheocasting ................................................................................. 49
3.7.5 Enmeshment parameter - Rheocasting ........................................................................ 49
3.7.6 Simulation parameters - Rheocasting ......................................................................... 49
3.8 SUBSEQUENT PROCESSES ..................................................................................................... 51
3.8.1 Heat Treatment .......................................................................................................... 51
3.8.2 Machining ................................................................................................................. 52
3.8.3 Coating...................................................................................................................... 54
3.8.4 Testing & Inspection .................................................................................................. 55

4 Results ..................................................................................... 56

III
 Table of Contents

4.1 MESH DENSITY.................................................................................................................... 56

4.2 RHEOCASTING RESULTS ....................................................................................................... 57
4.2.1 Filling results ............................................................................................................ 57
4.2.2 Solidification results .................................................................................................. 68
4.3 PLASTER CASTING RESULTS.................................................................................................. 74
4.3.1 Filling results ............................................................................................................ 74
4.3.2 Solidification results .................................................................................................. 79

5 Cost estimation ...................................................................... 85

5.1 TOTAL COMPONENT COST .................................................................................................... 85
5.2 MATERIAL COST .................................................................................................................. 85
5.3 LABOR COST ........................................................................................................................ 88
5.4 ENERGY COST ...................................................................................................................... 89
5.5 TOOLING COST .................................................................................................................... 90
5.6 OVERHEAD COSTS................................................................................................................ 91
5.7 ESTIMATED TOTAL COST FOR A UNIT CHASSIS ....................................................................... 91
6 Conclusion .............................................................................. 92
6.1 OVERALL ............................................................................................................................ 92
6.2 FURTHER WORK ................................................................................................................... 92

7 References .............................................................................. 93
8 Appendix................................................................................. 95
8.1 APPENDIX - COMPOSITION AND PROPERTIES OF ALLOYS ......................................................... 95
8.2 APPENDIX - ALLOY PROPERTIES IN SAND AND PERMANENT MOLDING ..................................... 96
8.3 APPENDIX - PROCESS ATTRIBUTES ........................................................................................ 97
8.4 APPENDIX - PLASTER CASTING MATERIAL PROPERTIES USED IN SIMULATION .......................... 98
8.5 APPENDIX - RHEOCASTING MATERIAL PROPERTIES USED IN SIMULATION .............................. 101
8.6 APPENDIX - SIMULATION PARAMETERS ............................................................................... 105

IV
 List of figures and tables

List of figures
FIGURE 2.1: INTERRELATIONSHIP BETWEEN PROCESS, MATERIAL, SHAPE AND FUNCTION [12] ..................... 4
FIGURE 2.2: SAND CASTING [5] ............................................................................................................. 9
FIGURE 2.3: PLASTER MOLD CASTING [5] ............................................................................................... 9
FIGURE 2.4: CERAMIC MOLD CASTING [5] ............................................................................................ 10
FIGURE 2.5: INVESTMENT CASTING [5] ................................................................................................. 10
FIGURE 2.6: LOST-FOAM CASTING [5] .................................................................................................. 10
FIGURE 2.7: GRAVITY-DIE CASTING [5] ................................................................................................ 11
FIGURE 2.8: LOW PRESSURE DIE CASTING [5] ....................................................................................... 11
FIGURE 2.9: HIGH PRESSURE DIE CASTING [5] ...................................................................................... 12
FIGURE 2.10: CENTRIFUGAL CASTING [5] ............................................................................................ 12
FIGURE 2.11: SQUEEZE CASTING [5] ................................................................................................... 12
FIGURE 2.12: RHEOCASTING [6] ......................................................................................................... 13
FIGURE 2.13: TWO-PART MOLD OF A SIMPLE GEOMETRY [7] .................................................................. 13
FIGURE 2.14: COLD CHAMBER DIE CASTING MACHINE [7]...................................................................... 14
FIGURE 2.15: BUBBLES AND BUBBLE TRAILS [9] .................................................................................... 17
FIGURE 2.16: BIFILMS (ENTRAINMENT DEFECTS ) [9] ............................................................................ 18
FIGURE 2.17: EFFECT OF INCREASING HEIGHT OF FALLING STREAM [9] .................................................. 18
FIGURE 2.18: SKETCH OF A SURFACE ENTRAINMENT EVENT [9] .............................................................. 19
FIGURE 2.19: FORMATION OF LAPS [9] ................................................................................................ 20
FIGURE 2.20: CONFLUENCE WELDS [9] ............................................................................................... 20
FIGURE 3.1: RELATIVE COST OF CASTING AN ALUMINUM ROD VERSUS THE NUMBER OF PARTS TO
BE CAST [12] ............................................................................................................................ 27
FIGURE 3.2: DIFFERENCE BETWEEN THE MICROSTRUCTURE OF A NORMAL SEMISOLID METAL AND AN
AGITATED ONE [6] ..................................................................................................................... 29
FIGURE 3.3: THE RAPID S METHOD [6] ................................................................................................ 29
FIGURE 3.4: COMPARISON OF POROSITY FORMATION IN RAPID-S’S SEMI-SOLID VS. TRADITIONAL ............... 30
FIGURE 3.5: ORIGINAL GEOMETRY ....................................................................................................... 31
FIGURE 3.6: SECTION VIEW OF ORIGINAL GEOMETRY - VERTICAL SECTION ............................................... 32
FIGURE 3.7: SECTION VIEW OF ORIGINAL GEOMETRY – HORIZONTAL SECTION .......................................... 32
FIGURE 3.8: ORIGINAL BOX AND FINS ................................................................................................... 33
FIGURE 3.9: ORIGINAL UNDRILLED BOX AND UNCUT FINS ...................................................................... 34
FIGURE 3.10: DIMENSION OF ORIGINAL FINS ........................................................................................ 34
FIGURE 3.11: DIMENSION OF NEW FINS ................................................................................................ 35
FIGURE 3.12: CHANGES MADE ON THE NEW BOX ................................................................................... 36
FIGURE 3.13: POSITION OF INCREASED DISTANCE BETWEEN SCREWS ....................................................... 37
FIGURE 3.14: TWO OPTIONS CONSIDERED FOR DRAFTING ...................................................................... 39
FIGURE 3.15: ASSEMBLY BY ONE SIDED FASTENING ............................................................................... 40
FIGURE 3.16: ASSEMBLY BY TWO SIDED FASTENING (FIRST TYPE) ............................................................ 40
FIGURE 3.17: ASSEMBLY BY TWO SIDED FASTENING (SECOND TYPE) ........................................................ 41
FIGURE 3.18: POURING METHOD IN PLASTER MOLD CASTING ................................................................. 42
FIGURE 3.19 TWO VIEWS OF A SAMPLE SLEEVE AND SPRUE ..................................................................... 43
FIGURE 3.20: A TYPICAL CERAMIC FOAM FILTER ................................................................................... 45
FIGURE 3.21: VARIOUS FILLET TYPES FOR SPRUE-RUNNER JUNCTION ...................................................... 45
FIGURE 3.22: SPRUE EXIT AREA (A2) AND ITS INCREASE AT BEND AND AFTER FILTER ................................. 45
FIGURE 3.23: GATING SYSTEM FOR RHEOCASTING ................................................................................. 47
FIGURE 3.24: AN EXAMPLE OF A PLASTER CASTING SET UP AS SEEN IN THE PREPROCESSOR........................ 48
FIGURE 3.25: AN EXAMPLE OF A RHEOCASTING SET UP AS SEEN IN THE PREPROCESSOR............................. 49
FIGURE 3.26: DATUM PLANES FOR MACHINING ..................................................................................... 54
FIGURE 4.1: COMPLICATIONS DUE TO MESHING CURVES AND TAPERS ...................................................... 56
FIGURE 4.2: V6-V1, 4MM GATE AT THE RECOMMENDED GATE VELOCITY OF 5 M/S..................................... 58
FIGURE 4.3: V9-V2, 6MM GATE SHOT AT A HIGHER GATE VELOCITY OF 5M/S (RECOMMENDED 4M/S) .......... 59
FIGURE 4.4: V9-V3, 6MM GATE SHOT AT A HIGHER GATE VELOCITY OF 8M/S (RECOMMENDED 4M/S) .......... 61
FIGURE 4.5: V9-V1, 6MM GATE SHOT AT A HIGHER GATE VELOCITY OF 15M/S (RECOMMENDED 4M/S)......... 62
FIGURE 4.6: SLURRY STAYING BEHIND AT TURN OF CORNERS AT 55% FILL (V9-V1) ................................... 63
FIGURE 4.7: V9-V1, 86% FILLED (LEFT) AND 94% FILLED (RIGHT) ......................................................... 63
FIGURE 4.8: STAYING BEHIND OF SLURRY AT HIGH MESH DENSITY AT 55% FILL (V9-V5) ........................... 64

V
 List of figures and tables

FIGURE 4.9: V9-V5, NO AIR POCKET FORMATION FOR THE HIGH DENSITY MESH AT 86% AND 94% FILL ...... 64
FIGURE 4.10: POCKETS NEAR FIN TIPS (V9-V5 HIGH MESH DENSITY) ....................................................... 64
FIGURE 4.11: V11-V1, LAYOUT TWO .................................................................................................... 65
FIGURE 4.12: V11-V2, LAYOUT THREE ................................................................................................. 66
FIGURE 4.13: FILLING OF INSIDE WALLS IN LAYOUTS TWO AND THREE ..................................................... 67
FIGURE 4.14: PROBABLE CAVITIES AT TOP OF FINS FOR RHEOCASTING .................................................... 67
FIGURE 4.15: V9-V1, SOLIDIFICATION .................................................................................................. 68
FIGURE 4.16: V11-V1 (LEFT) AND V11-V2 (RIGHT) ............................................................................... 68
FIGURE 4.17: HOTSPOT V9-V1 ............................................................................................................ 69
FIGURE 4.18: HOTSPOT, V11-V1 (LEFT) AND V11-V2 (RIGHT) ................................................................ 69
FIGURE 4.19: THE TWO TYPES OF GATE USED THIN (V6-V1) AND THICK (V9-V1, V11-V1 & V11-V2) .......... 70
FIGURE 4.20: FSTIME_69, V6-V1 (LEFT) AND V9-V1 (RIGHT) ............................................................... 70
FIGURE 4.21: FSTIME_69, THIN GATE V6-V1(LEFT) AND THICK GATE V9-V1(RIGHT) .............................. 70
FIGURE 4.22: FSTIME_69 ,FEEDING PATH FOR V6-V1 (LEFT) AND V9-V1 (RIGHT) .................................. 71
FIGURE 4.23 FSTIME_69, V6-V1 (LEFT) AND V9-V1 (RIGHT) ................................................................ 71
FIGURE 4.24: FSTIME_69, SLURRY BLOCKED FROM BEING FED V9-V1 .................................................. 72
FIGURE 4.25: FSTIME_69, V11-V1 (LEFT) AND V11-V2 (RIGHT) ........................................................... 72
FIGURE 4.26: SINGLE GATED VERSION (PLASTER_ITERATION7_V2) ........................................................ 74
FIGURE 4.27: MULTIPLE GATED VERSION (PF4) ................................................................................... 75
FIGURE 4.28: APPARENT SPRAYING AS MELT BENDS AT TOP .................................................................... 75
FIGURE 4.29: DIFFERENT TYPES OF FILTER CONFIGURATIONS ................................................................ 76
FIGURE 4.30: RESULT OF FILTER ARRANGEMENT 1 (PF3) ...................................................................... 76
FIGURE 4.31: RESULT OF FILTER ARRANGEMENT 2 (PF4) ...................................................................... 77
FIGURE 4.32: RESULT OF FILTER ARRANGEMENT 3 (P_X_1) .................................................................. 77
FIGURE 4.33: TWO TYPES OF RUNNER EXTENSIONS CONSIDERED ............................................................. 78
FIGURE 4.34: MELT ENTRANCE WITH SECOND TYPE OF RUNNER EXTENSION (P_X_1_V2).......................... 78
FIGURE 4.35: TEMPERATURE DISTRIBUTION AT 837.68 SECONDS ............................................................ 79
FIGURE 4.36: HOTSPOT RESULTS, SHOWS JUNCTION OF INSIDE WALLS AS AREAS LAST TO SOLIDIFY ............. 80
FIGURE 4.37: POROSITY RESULT .......................................................................................................... 80
FIGURE 4.38: SOLIDIFICATION STAGES AT 8.02 (LEFT) AND 897.04 (RIGHT) SECONDS............................... 81
FIGURE 4.39: SOLIDIFICATION STAGES OF NEW VERSION AT 8.02 (LEFT) AND 887.08 (RIGHT) SECONDS ..... 81
FIGURE 4.40: FEEDING THROUGH GATES SEEN ON THE FSTIME30 RESULT ............................................. 82
FIGURE 4.41:POROSITY RESULT OF NEW VERSION.................................................................................. 82
FIGURE 4.42: MELT FALLING IN SLEEVE ............................................................................................... 83
FIGURE 4.43: POURING BY USE OF A POURING CUP ............................................................................... 84

List of tables
TABLE 2.1: FOUR-DIGIT SYSTEM FOR DESIGNATING CAST ALUMINUM ALLOYS [3] ........................................ 5
TABLE 3.1: CASTING PROCESSES AND THEIR ATTRIBUTES [5] .................................................................. 26
TABLE 3.2: RECOMMENDED FAST SHOT VELOCITIES VS. GATE THICKNESS (RHEOCASTING) ......................... 50
TABLE 3.3: TYPICAL HEAT TREATMENTS FOR ALUMINUM ALLOY AND PERMANENT MOLD CASTING [14] ........ 51
TABLE 3.4: PERMISSIBLE DEVIATIONS FOR LINEAR DIMENSIONS, NON-CASTING PROCESSES [20] ................. 52
TABLE 3.5: PERMISSIBLE LINEAR DEVIATIONS FOR LINEAR DIMENSIONS FOR CASTING PROCESSES [21] ....... 53
TABLE 3.6: ROUGHNESS RANGES THAT CAN BE ACHIEVED BY SELECTED PROCESSES .................................. 53

VI
 Introduction

1 Introduction
This chapter gives brief information on the thesis’s topic, aim and its structure.

1.1 Company description

Saab Avitronics is a business unit, within the Saab Group, which develops and manufactures
advanced electronics primarily for military applications. It is one of the world's leading high
technology companies with extensive involvement in defense, aviation and space. The head
office is located in Järfälla, but the company also has offices in Kista, Linköping and
Jönköping, all in Sweden. It also has divisions in Centurion and Cape Town, South Africa.
The office in Jönköping develops and manufactures high-quality electronics, software and
mechanical parts for aircrafts, helicopters and other demanding applications. Most projects
are done for large international companies like Airbus and Boeing.

1.2 Background
From the very dawn of the aircraft industry, casted components have played an essential role
as aircraft parts. However, the use of casted components still remains limited to parts with
„non-crucial‟ role – parts having lax requirements. Nevertheless, in recent decades, the
industry has seen the potential of cutting cost from a single component, when switching its
mode of production to casting, instead of the conventional forged, machined, welded and
assembled parts. Speed of production is also another cost related attribute that merits casting
as a production process.

Other than the cost related benefits mentioned above, improvements in product predictability
as well as growing technologies in casting methods, process control, testing methods and
computer simulations, have also contributed to this growing trend towards using casting for
producing aircraft components.

Despite the growth in technology and knowledge related to the process, memories of poor
casting performance, as far as a few decades back, still haunts the industry and has slowed the
utilization of casting benefits. Moreover at that time, the use of a “casting factor”, as a safety
factor for casted components, increased the mass of a casted part, and thus has been another
impediment in choosing casting as a process. Thus, it will take decades, if ever, to see crucial
structures, having strict requirements, be produced by casting [1]. This thesis tries to study the
possibilities of one conceptual aircraft component to be cast.

1.3 The product

The product was a chassis for an electric unit that was proposed to be used on a passenger
plane. It was a concept for a business plan and did not exist on any passenger plane until the
time of finishing this thesis. This chassis was planned to be placed inside a wing of a plane, in
a somewhat harsh and open environment. The main function of the chassis would be holding
and protecting an electronic unit from a harsh outside environment as well as carrying away
heat. It also would have an additional function of conducting electricity at some surfaces. The
power outage near this chassis would be about 3 to 5𝑘𝑊.

It had an overall dimension of 430 x 275 x 120 mm with a minimum section thickness of 1
mm. The original chassis was designed to be machined from a blank, which obviously was an
expensive choice, leading to the motivation of looking into casting as another possible method
of production. As the original chassis was to be made from wrought aluminum alloy, the
choice of alloy for the cast chassis was set to be aluminum cast alloy. It should also be

1
 Introduction

mentioned that the geometry of the original chassis was difficult to cast, in that it had varying
cross-sectional thicknesses as well as having very thin sections. The chassis also had many
holes, of which most were threaded and also had undercuts not attainable with most
conventional casting processes – thus adding to the difficulty of casting it.

1.4 Purpose and aims

The main objective of this master thesis was to go through the product development aspects
of a product, a chassis for an electronic unit, as related to switching its manufacturing method
from machining to casting. This concept of proposing a cast chassis, encompassed going
through, geometry modification, material (alloy) selection, process selection, simulation of
selected process(es), prediction of defects, improvement of process parameters as well as an
estimation of cost.
As an extension to the objectives above, the thesis also tried to show difficulties faced while
simulating a casting process with certain material models, geometry features and process
attributes, and thus, would hopefully serve as a guideline for future simulations of casting
processes of similar shaped parts under similar conditions.
In trying to achieve the objectives mentioned above, the thesis also tried to involve casting
companies, with the hope of having their knowhow as an input, and getting their confirmation
to some assumptions as well as improvements made on this thesis.
The aim of this thesis work was to:
 Modifying the products geometry to have better castability
 Choose a casting process
 Choose an alloy material
 Have a sample gating system
 Show areas susceptible to defects
 Proposal of secondary processes and testing
 A rough cost estimation

1.5 Delimits
The scope of this thesis was limited to the simulation of casting processes, with more
emphasis given to filling results – in an attempt to evaluate and minimize potential defects.
This was achieved by changing proposed gating systems as well as varying process
parameters. The use of insulation, chills and cooling channels, to improve solidification
pattern, was not considered. This thesis did not include a physical trial production of the
chassis and thus comparison of simulation results with actual production outcome was also
out of the scope of this thesis.

1.6 Outline
This thesis consists of six main chapters.

Chapter 1 gives a general introduction to the thesis.

Chapter 2 discusses the general concepts involved with process and material selection as well
as brief descriptions of aluminum alloys, casting processes and simulation of casting
processes.

2
 Introduction

Chapter 3 lays down the main tasks performed and decisions made prior to simulation, which
includes modification of the geometry as well as choice of material (alloy) and process.

Chapter 4 presents some simulation results and discusses various changes and improvements
made on the simulation of the processes.

Chapter 5 gives a rough cost estimation of the chosen processes and the last chapter.

Chapter 6 gives the conclusion part of the whole thesis.

In addition, reference materials used for this thesis and an appendix, consisting of various
tables and figures, are provided in chapters 7 and 8 respectively.

3
 General Concepts

2 General concepts
This chapter gives introduction to the typical cast aluminum alloys as well as casting
processes considered in this thesis. It also describes the production steps involved with
casting as well as their associated problems. The common types of defects in casting
aluminum are also introduced as related to what should be observed during simulation of
casting processes (especially filling).

2.1 Product development

Product development is defined as: „Creation of products with new or different characteristics
that offer new or additional benefits to the customer. It may involve modification of an
existing product or its presentation, or formulation of an entirely new product that satisfies a
newly defined customer want or market niche‟ [2].

Material selection and process selection are two important decision steps for a product
concept. A product concept is first initiated by a new idea or a market need. This concept for a
new product, with new characteristics or modified ones, will have to perform a certain task
(function), which is associated to its geometry (shape), to the material used as well as to the
method of producing it. This interrelationship between material, process, shape, and function
is the central problem in realizing a product concept. A change or choice of one affects the
other. This dependency, of the four areas of concern in product development, can be
represented with a diagram as shown in figure (2.1).

Figure 2.1: Interrelationship between process, material, shape and function [12]

2.2 Material Selection

A material has properties, both physical and mechanical, that makes it a candidate to perform
a certain function and/or to be shaped by a certain process. These properties are density,
strength, corrosion resistance, thermal conductivity, specific heat capacity, etc. In addition to
the properties mentioned above, the cost of the material is also another factor affecting its
choice. Thus, based on the requirements from its function, constraints from the process(es) as
well as its shape, an affordable material, having the best possible properties has to be chosen.
In this thesis, a choice would be made from a list of applicable aluminum cast alloys.

4
 General Concepts

2.2.1 General

Aluminum alloys, together with titanium, magnesium and beryllium constitute the group
Light Alloys¨. The obvious primary advantage of aluminum (light alloys) in engineering
(structural) applications is the high strength per density value it has, compared to other
materials in general or iron or steel in particular. Moreover aluminum is very abundant in the
earth‟s crust. This has made aluminum a major choice as a material for different components
in the transportation industry, with aerospace and vehicles being the primary application
areas. The low density of aluminum also has an extra advantage in stiffness that gives it
resistance to buckling. However, other than the advantages associated with low density,
aluminum alloys have very attractive properties like high corrosion resistance as well as high
thermal and electrical conductivity. Aluminum alloys are part heat treat able and part not.

2.2.2 Cast Alloys

Aluminum cast alloys have the advantage of having a relatively low melting temperature
when compared with traditional cast iron. Thus, low energy input is required. Moreover, cast
aluminum alloys are less soluble to all gases, except hydrogen. Good surface finish and
fluidity, (for most) are also other properties that make cast aluminum alloys a candidate for
choice. A major disadvantage of using aluminum as a casting material is its problem with
shrinkage. Aluminum alloys have a relatively high shrinkage while solidifying and this should
be put into account while designing, so as to produce a product that is dimensionally accurate,
less susceptible to defects and a minimized residual stress.

Generally, aluminum cast alloys have inferior mechanical properties compared to wrought
aluminum alloys except in creep. However, if a sound casting can be produced, cast alloys
will come in handy in reducing cost of a product.

The relatively low melting temperature and general ease of handling of cast aluminum, has
been an aid for developing processes like rheocasting and squeeze casting – two methods of
casting that have the advantage of producing products with relatively better mechanical
properties.

2.2.3 Designation

Cast aluminum alloys have designations to identify as well as group them based on alloy
constituents. One such designation is the one developed by Aluminum Association of the
United States, which uses a four digit system and as shown in table (2.1).

Current Former
designation designation
Aluminum, 99.00% or greater 1xx.x
Aluminum alloys grouped by major alloying
elements:
Copper 2xx.x 1xx
Silicon with added copper and/or magnesium 3xx.x 3xx
Silicon 4xx.x 1 to 99
Magnesium 5xx.x 2xx
Zinc 7xx.x 6xx
Tin 8xx.x 7xx
Other element 9xx.x 7xx
Unused series 6xx.x

Table 2.1: Four-digit system for designating cast aluminum alloys [3]

5
 General Concepts

The first digit indicates the major alloying element which is the base for forming the alloy
group shown in table (2.1).

For the 1xx.x group alloys, the second and third digits indicate the minimum percentage of
aluminum for the 1xx.x group alloys- they are equal to the percentage of impurity (150.x
indicates a composition containing a minimum of 99.50% aluminum). For the alloy groups
2xx.x to 9xx.x, they have no individual significance but to serve as a number to identify the
different aluminum alloys.

In all casting alloys, the last digit after the dot indicates the product form: where 0 is for
castings, 1 for standard ingot and 2 for ingots having narrower composition limits, but still
being within the ranges of the standard ingot.

A serial letter is usually included before the numerical designation to indicate a specific
modification carried out on the original alloy, or specifying variation in impurity limits from
an original alloy. These letters are assigned starting with A in alphabetical order, leaving out
I, O, Q and X (X used for experimental alloys).

As with the case with aluminum alloys in general, which were mentioned earlier to be part
heat treatable and part not, cast aluminum alloys are also part heat treatable and part not.
Specific designation, which works for both wrought and cast alloys, exists for describing the
temper an alloy should have. This temper designation follows after a hyphen put at the end of
the alloy designation described above, and usually has two characters. The first character is a
letter for the basic temper designation and the latter a digit that indicates the temper
subdivision (in some cases two digits are also found).

Typical heat treatment types are:

 F, as cast
 O, annealed
 T4, solution treated and aged
 T5, precipitation hardened
 T6, solution heat treated, quenched, and precipitation hardened
 T7, solution heat treated, quenched and overaged

Another common designation variant, commonly used is addition of the letter “P” added to a
standard temper (e.g. T6P), indicating a producers variation of the standard heat treatment
process [3].

2.2.4 Standard materials in the aviation industry

In relation to this thesis, as per the requirements of the aviation industry, there exists a list of
materials that can be used for different aircraft applications.

Cast aluminum alloys, along with the temper requirement, that can be used for structural
applications are:
 A201.0-T7
 354.0-T6
 355.0-T6
 C355.0-T6
 356.0-T6
 A356.0-T6(T6P)
 A357.0-T6

6
 General Concepts

 D357.0-T6
 359.0-T6

General properties and applications of the alloys mentioned above are described below [4].
A more detailed information on the properties of these alloys is provided in appendix (8.1).

A201.0 – an Al-Cu-Ag alloy, heat treatable. In its T7 temper it has high strength (among the
highest for commercial cast materials) strength and hardness, moderate ductility and very
high resistance to stress corrosion cracking, can be sand cast, permanent mold cast and
investment cast, limited weldability.
Applications include aerospace housings, missile fins, and other applications where strength
at elevated temperatures, high strength with moderate elongation is required.

354.0 – an Al-Si-Mg alloy, heat treatable. Has high strength and hardness, very good fatigue
strength, processing, moderate general corrosion resistance generally processed by permanent
mold casting and have moderate weldability.
Typically used for castings requiring very high strength.

355.0 – an Al-Si-Mg alloy, heat treatable. Has good strength and hardness but very low
elongation, good castability and pressure tightness, moderate general corrosion resistance,
generally permanent mold casted and have moderate weldability.
Applications where good castability and pressure tightness is required.

C355.0 – an Al-Si-Mg alloy, heat treatable. Similar to 355.0 but with impurities controlled to
be low, resulting in relatively higher strength and hardness with improved elongation, have
moderate weldability
Applications include aircraft supercharger covers, fuel-pump bodies, liquid-cooled aircraft
engine crankcases, general applications where good castability and pressure tightness is
required.

356.0 – an Al-Si-Mg alloy, heat treatable. Has intermediate strength but high hardness and
low elongation, good corrosion resistance, very easy to handle (among the easiest) in variety
of casting processes and have good weldability.

A356.0 – an Al-Si-Mg alloy, heat treatable. Similar to 356.0 but with impurities controlled to
be low, resulting in relatively higher strength and hardness with improved elongation, high
corrosion resistance, and good weldability.
Applications include aircraft pump parts, aircraft fittings and control parts, aircraft structures
and other applications where excellent castability, good weldability, pressure tightness, and
good resistance to corrosion are required.

A357.0 – an Al-Si-Mg alloy, heat treatable. Special properties developed by controlled

casting and chilling. Has high strength and hardness, moderate elongation, generally good
weldability, very good corrosion resistance and good toughness, typically permanent mold
casted.
Applications include aircraft and missile parts requiring good weldability, strength, and
toughness, critical aerospace applications.

D357.0 – an Al-Si-Mg alloy, heat treatable. It is a modification of A357.0 alloy, requiring

narrower composition limits and a more controlled processing. Has very a relatively higher
strength and hardness, moderate elongation, with very good fatigue strength, good
weldability.
Applications include critical aerospace applications and other uses requiring high strength and
good toughness.

7
 General Concepts

359.0 – an Al-Si-Mg alloy, heat treatable. Has very high strength and hardness, moderate
elongation, high, with very good fatigue strength, generally good weldability, very good
castability, typically permanent mold casted.

All the alloys mentioned above are not restricted to the casting processes they are associated
with; instead all can either be sand casted, permanent mold casted and investment casted.
Almost all casting processes can be used to processes the alloys depending on the property
requirement of a product or other constraints like cost or section thickness, etc.

2.3 Process Selection

Product manufacturing has usually three stages, which are, shaping, joining and finishing. A
selection of process is affected by or affects the choice of material, the shape (usually
minimum section thickness), the mass of the product as well as the dimensional precision and
surface finish requirements. Usually, a final surface finish and dimensional tolerance is
attained by a secondary machining or coating process. At the end of the day, the choice of
process(es) made, should be cost conscious. It was mentioned that, the shaping process to be
used on the product, was set to be a casting process.

Casting (metals) is one of the oldest methods of shaping a product. It is a primary shaping
process and involves melting of metals (usually superheating to a proper temperature),
followed by pouring of the liquid metal into a pre-prepared cavity or mold, where it is
allowed to solidify and thus take the form of the cavity. Generally, any metal that can be
melted can be cast. Compared to other shaping processes, it has the advantage of forming
complex geometries having features like internal cavities, holes as well as complex surfaces.
It also has the advantage of being able to shape metals that are otherwise difficult to machine.
Its major disadvantages are the common defects it entails – shrinkage and porosity being
some examples.

Casting processes are most commonly categorized into two constituents, depending on the
use-frequency of mold. These are, permanent mold casting (non-expendable mold casting)
and expendable mold casting.

Expendable Mold Casting includes processes like, sand casting, shell casting, investment
casting (lost-wax casting), lost foam casting, plaster mold casting, ceramic mold casting, etc.

Permanent Mold Casting includes (high pressure-, low pressure-, gravity-) die casting,
centrifugal casting and special processes like rheocasting and squeeze casting.

2.3.1 Expendable mold processes

A. Casting processes with sand based molding

- Green-sand molding: A casting process that has green (term stands for non-cured
sand, its color is actually black) sand as a mold material. It is the most widely used
casting process. The sand is blended with clay water and additives. It is the leading,
out of the least expensive casting processes, and is a process with low tooling cost.
Depending on whether one dries the entire mold or the part contacting the casting,
dry-sand mold and skin-dried mold (respectively), are variants of green- sand mold
casting that mitigate its problems tied with moisture.
- Sodium silicate-CO2 molding: This has sand with addition of 3% to 6% sodium
silicate. The mold gains its strength from the silicate that quick hardens when exposed

8
 General Concepts

to CO2. However, it collapses relatively easily and additives might be added to

improve this. It is usually used when better accuracy, thinner sections, or deeper
draws are required than usually achieved by ordinary sand molding.
- No-bake molding: An alternate process that uses reaction between an organic and
inorganic resin binder with liquid curing agents serving as a catalyst. These are
combined with sand. It has a relatively slower mold curing time.
- Shell molding: A mixture of silica sand that has fine grains is mixed with a
thermosetting resin that is dropped, blown or shot onto a heated metal pattern (usually
some cast iron). This heat partially cures a layer (about 10 mm thick) of sand-plastic
material that forms the shell. A complete mold is made of two shell halves. The shell
is usually supported by sand in a box from the outside, in order to withstand the
stresses that arise during the process.

Figure 2.2: Sand casting [5]

B. Plaster mold casting: In this process the mold is made of plaster of paris (gypsum),
combined with additives to give it strength, permeability and formability. The process
gives excellent surface finish and dimensional accuracy (especially with metal
patterns). The plaster has low thermal conductivity and low specific heat and thus
cooling of casting is very slow. This helps molten metal to stay alive (stay hot) and
make it flow to fill thin sections and replicate fine details. Sometimes plaster molds
are preheated prior to pouring of molten metal. Unfortunately, it is usually limited to
low melting point metals (aluminum, magnesium and zinc).

Figure 2.3: Plaster mold casting [5]

C. Ceramic mold casting: It is similar to plaster casting except that the mold is made of
ceramic material. Like plaster casting this process is able to give fine details, smooth
surface and achieve thin sections. Ceramic molds are often preheated to ensure proper
filling of molten metal. Contrary to plaster casting it is able to cast metals with high
melting point. Unfortunately, the mold material is expensive (and of course, it is not
reusable).

9
 General Concepts

Figure 2.4: Ceramic mold casting [5]

D. Investment casting: It is also known as the Lost-Wax casting process. Mold material
for investment casting is ceramic material that is similar to ceramic mold casting.
This process is different in that the pattern used is disposable i.e. a wax material
around which the mold is set. The pattern is not removed prior to pouring of molten
metal (cavity less casting), instead the wax melts and is vented out, as it is being
displaced by the molten metal poured over it. The mold is then destroyed to deliver
the casting. Two variant of molding processes exit. These are, the solid mold type
(has cope and/or drag) and the ceramic shell molding type (stand alone ceramic that is
built up on pattern, no container) of processes. The latter is the frequented method of
the two. In ceramic shell molding, multiple parts, making a single pattern are
arranged in a tree-like structure and share the same gating system.

Figure 2.5: Investment casting [5]

E. Lost-foam casting: Another cavity-less casting process, similar to investment casting

but having sand as a mold material. Pattern in this case is made of polystyrene foam.
It needs the use of a flask to hold sand together and a single cavity or multiple
cavities (tree-structure) can be formed. The mold is destroyed to get the final casting.

Figure 2.6: Lost-foam casting [5]

10
 General Concepts

It should be noted that, in both Investment and lost foam casting, there is no need to remove
pattern and mold material is expendable. Thus, almost no draft is required on patterns and
features like undercuts are no trouble to shape.

2.3.2 Non-expendable mold processes

A. Die casting
- Gravity-die casting: It is also known as permanent mold casting. A die is made out
of a suitable material usually nodular cast iron, cast steel or wrought steel (for low
temperature melting metals). The mold is usually is preheated and coated prior to
filling with molten metal which fills the cavity by gravity force. Near net shape cast
can be achieved. It has the advantage of controlling cooling through good design by
introducing cooling channels. Since the molds are not permeable, vents must usually
be provided. A graphite mold makes casting of iron and steel possible, but can be
very expensive. Generally, with this process, a very high quality casting (surface,
dimension) requiring very little machining is achieved.

Figure 2.7: Gravity-die casting [5]

- Low pressure die casting: Here, the permanent molds are mounted over a sealed
furnace with a refractory tube placed between the two. By pressurizing the furnace,
molten metal is driven up and fills the cavity. Thus, the metal is exceptionally clean,
as it is extracted from melt center and fed directly to the cavity, with the tube
protecting it from reactions. The casting solidifies from top down, as pressure is kept
from bottom, till it solidifies. It has a relatively slower cooling than gravity mold
casting. It is suitable for metals with low melting points.

Figure 2.8: Low pressure die casting [5]

- High pressure die casting: Here, molten metal is injected, under high pressure, into
a metal die. The metal flows through channels and pressure from a plunger is kept
until the casting solidifies. Due to very high pressure, die halves require a clamping
mechanism. It is often used to cast metals with low melting points, typically zinc, due
to theirs low reaction with the mold, during the long and pressured contact. Very high

11
 General Concepts

surface quality castings as well as excellent dimensional accuracy can be achieved.

Two varieties of this process are common, namely cold chamber and hot chamber
pressure die castings. The former has melt poured in front of a plunger, while the
latter has the plunger already sunk in a melt container. High pressure die casting has a
very expensive tooling and equipment cost.

Figure 2.9: High pressure die casting [5]

B. Centrifugal casting: In this process, molten metal is distributed into mold cavity by
the internal (centrifugal) forces of a rotating die. Rotation could be on a vertical or
horizontal axis. Mold materials include baked sand, plaster or graphite.

Figure 2.10: Centrifugal casting [5]

C. Squeeze casting: Here, molten metal is introduced into die, with low velocity and
after the cavity is filled a high pressure is applied on the die. It is commonly used for
metals with low melting points. The process offers a very close-to-net-shape casting.

Figure 2.11: Squeeze Casting [5]

D. Rheocasting: In this process, molten metal is cooled to a semisolid state with a

constant stirring and the melt will be made to have a slurry form. The stirring action
cuts off dendrites or actually prevents dendrites from forming and results in rounded
solid particles in the slurry. This slurry, containing up to 30-50% solid, is then shot
into a die, by use of a plunger. It is commonly used for metals with low melting

12
 General Concepts

points, typically aluminum. The process offers a very close-to-net-shape casting that
is almost porosity free.

Figure 2.12: Rheocasting [6]

2.3.3 Basic casting terminologies

A two-part mold of a simple geometry is shown in figure (2.13), followed by some basic
terminologies used in the casting world.

Figure 2.13: Two-part mold of a simple geometry [7]

- Pattern: an approximate replica of the desired casting. Geometry includes an

allowance for shrinkage plus has extra features like cores (if any) as a union.
- Mold material: a material which is aggregated around the pattern, where upon its
setting and subsequent removal of the pattern material, a provision (the mold cavity)
is left. This cavity is where melt would later be poured in and left to solidify.
- Flask: a rigid material, usually from metal or wood that holds the mold material
intact.
- Cope: top half of the flask, mold, pattern or core.
- Drag: bottom half of the flask, mold, pattern or core.
Cope and drag are only used for horizontally parted mold.
- Core: a material, usually sand (or metal) left in mold cavity around which molten
metal solidifies for purpose of making holes and other hollow feature. The core
together with the mold material makes the mold cavity.

13
 General Concepts

- Core print: a recess left in mold for purpose of holding core firmly while casting.
- Riser: an extra recess made in mold for purpose of feeding shrinkage in the mold
cavity.
- Gating system: a network of channels in the mold that deliver molten metal from a
pouring vessel (that transports molten metal from furnace) to the mold cavity. It
usually comprises of, a Pouring Cup (a cavity that receives melt from the pouring
vessel), a Sprue (a vertical portion of the gating system), Runners (horizontal portion
of the gating system) and Gates (channels that controls entrance to mold cavity).
Other features like Wells (sprue well and extension wells) and Extensions (runner and
flash) are also common features that make up a gating system.
- Parting line (parting surface): an interface that separates two mold halves.
- Draft: a taper made on pattern that results in a taper on the mold cavity and then a
taper on the solidified piece (casting) to ease removal of the pattern and the final
casting.
The terminologies above can loosely be used to all type of castings but are more endemic to
non permanent, especially gravity filled, casting processes. Analogous terms used for
permanent die casting, especially to high pressure die casting, are defined below together with
the corresponding figure (2.14).

Figure 2.14: Cold chamber die casting machine [7]

- Die: a rigid block that has the desired casting shape as a cavity. It is usually made of
two halves (the ejector die and the cover die).
- Ejector die: the part of the die half that is stationary.
- Cover die: the part of the die half that moves to take out casting.
- Plunger: a piston that forces the melt into the cavity.
- Shot sleeve: a pre-cavity container where the plunger works.
- Biscuit: an excess of ladled metal remaining in the shot sleeve of a cold chamber die
casting machine. It is part of a cast shot and is removed from the die with the casting.

14
 General Concepts

- Overflow: a recess in a die connected to die cavity by a gate to assist in proper

venting.

2.4 Fundamentals & associated problems

Casting process involves carrying out a series of activities. Some literatures divide these
activates into six processing steps [8]. These six steps along with the tasks carried out during
these steps and the associated problems would be described next.

Step One: Mold cavity making

This step consist of all processes associated with making the casting cavity. Thus, the
process of pattern making and core making are also included in this step. In
expendable molds, mold material is usually aggregated around the pattern, while in
permanent molds, the shape of the pattern is machined out of an appropriate mold
material. In both cases, the pattern must have an appropriate size to account for
shrinkage and must have an adequate draft to facilitate its removal or the removal of
the final casting. Common problems associated with this step are core blows, mold
permeability, mold erosion, mold expansion and excessive residual stress. The first
three problems are typically associated with expendable molds, while the last one is
typical of permanent molds. Thus, care should be taken in selecting an appropriate
mold and core material as well as designing the mold assembly, the pattern size and
the draft on the pattern.

Step Two: Melting process

In this step the alloy material is heated and turned into liquid. The melt is usually
heated above its melting temperature. Problems associated with this step are
premature solidification and cleanliness of melt. To avoid these problems, the melt
should be superheated to an adequate temperature. Moreover, proper care should be
taken to avoid foreign inclusions from entering the melt as well as preventing
unwanted reactions of the melt with its environment. Other typical problems
associated with cast aluminum alloys are high hydrogen solubility and oxide
formation. As much as possible, lowering the super heat of metal to be cast, is desired
since it lessens these reactions that form oxides and gas solubility. A low super heat
also saves the energy required for melting an alloy.

Step Three: Pouring process

This step involves the transfer of molten metal from one container to another as well
as the filling of the mold cavity. Thus this pouring happens at various stages of the
casting process. First, the melt can be transferred from the melting furnace to a
holding furnace. Next, the molten metal can be transferred from the holding furnace
to the casting cavity or a shot chamber depending on the type of casting process.
Finally, the subsequent filling of the casting cavity by the molten metal is also part of
a casting process considered here. This step is the area where most casting defects are
generated (see section 2.5 for details). Fluidity is a term given to the ability with
which molten metal can flow. In most cases fluidity of molten metal is improved by
adding specific alloys, in a process called alloy modification. Generally, care should
be taken in designing gating systems as well choosing associated parameters like
velocity and filling time of casting processes.

Step Four: Solidification Process

This steps involves the solidification of molten metal. During solidification, shrinkage
of casting should be fed by a proper method so that internal voids and porosities
would be avoided. Hotspots, are areas that solidify last in a casting and usually takes
as an indication for areas susceptible for porosity formation. Another problem

15
 General Concepts

associated with solidification is the warping of components, which usually happens to

dished-type components. The cooling rate during solidification is an important
process parameter deciding the final property of the casting. Generally faster cooling
rate gives better mechanical properties. However, formation of a desired
microstructure can be assisted by using grain refinement and controlling the
solidification direction.

Step Five: Removal of casting from mold

In this step the casting, which has solidified, is removed from the mold. Care should
be taken so as not to chip the cast or mold (if it is to be reusable). Also non adequate
drafts might make removal of the cast part from the mold difficult. Non-Reusable
molds can be crushed away from the cast part. However, the applied crushing force
must be just enough to take away the mold and not disturb the cast part.

Step Six: Various cleaning, finishing and inspection processes

Subsequent processes like quenching and heat treatment should be studied if they
attain a desired micro structure. Appropriate surface treatment process should be
chosen to prolong the components life. Other secondary processes like machining and
joining need a proper consideration so as not to have too little or too much surface to
machine. Moreover, cast parts need a clear-cut designation of datum lines as a
reference for providing dimensions.

2.5 Defects in aluminum casting

Defects in casting originate from three areas [17]. These are the casting design, the technique
of casting and the application of the technique (workmanship). A defect may be traced back to
a single cause or be from a combination of factors which might not be easy to pin down. The
different steps in casting, and the care required in these steps to avoid defects in castings, have
been briefly touched above. It was also mentioned that, out of the six casting steps, the
pouring step accounts for the majority of defects that occur in castings (aluminum casting in
particular). Thus, here concentration would be given to entrainment defects, to highlight what
must be studied, in a flow simulation and to give emphasis on such defects as they account to
a vast majority of the defects and failures occurring during casting a product and while the
cast product is in-service.

Entrainment defects can be divided as [9]:

 Exogenous inclusions
 Bubbles
 Bifilms

Before explaining about the defect types, it is worth mentioning two characteristics of
aluminum.

The first is that aluminum has hydrogen, as the only gas appreciably soluble in it. The
solubility of hydrogen increases with temperature, especially rapidly above the liquidus
temperature. This dissolved hydrogen might eventually result in hydrogen porosity.
Depending on the distribution of these pores, this type of „defect‟ might even be desirable (if
finely distributed). Unfortunately, the technology with which to produce a product with the
above desired distribution of pores is not yet refined and hydrogen porosity is generally
thought to be undesired.

16
 General Concepts

The second characteristic of aluminum is that it readily oxidizes, either in solid or liquid form.
Oxides can be formed by reaction with air, vapor or other oxides that come in contact with
aluminum. This oxide forms on any surface that is exposed to air.

Coming back to the first type of defect, exogenous inclusions are grains of foreign matter that
find their way into the melt and subsequently ending up in the final casting. The source for
these inclusions might be some materials being dislodged from a furnace‟s, a holding
container‟s or a ladle‟s wall.

The second type, bubbles, are hollow globules of gas entrained in the melt. Eventually these
bubbles might end up as a porosity defect in the casting. Even if these bubbles escape to the
surface due to buoyancy, they result in a more critical defect termed as a bubble trail. As can
be understood from the term, bubble trail is a trail left by a bubble as it travels through the
melt. This trail looks something like the tail of a comet and can be seen in figure (2.15). These
bubbles and bubble trail defects account for the majority of defects in castings and are much
attributed to poor pouring processes, while core blow could also be another cause for their
formation.

Figure 2.15: Bubbles and bubble trails [9]

Central to the above mentioned types of entrainment defects is the third type of entrainment
defect- which is the concept of a bifilm. It was mentioned earlier that aluminum readily
oxides and that this oxide form a coat around the melt. If this outer surface folds and meets,
dry side to dry side, the resulting formation is termed as a bifilm. Naturally these biflms will
form inwards and might again find their way into the casting due to poor pouring and flow
conditions. A figure showing these bifilms and how they relate with the first two defects can
be seen in figure (2.16).

17
 General Concepts

Figure 2.16: Bifilms (Entrainment defects) [9]

In the figure (2.16,a), a new bifilm can be seen, where (2.16,b) shows how bubbles are formed
integrally to a bifilm. In (2.16,c,d,e), it can be seen how bifilms can trap liquid, a debris
entrained in a bifilm as well as sand particles trapped in a bifilm. The last part (2.16, f) shows
an old film that has a debris as an inclusion, itself trapped in a new film. Thus, that was why
the concept of bifilms were claimed to be central to entrainment defects.
As mentioned earlier, poor pouring and flow patterns result in the formation of bubbles and
bifilms as well as for them to find way into the casting. These entrainments can occur in both
turbulent and laminar type flows and also when the fall height of a melt is large.
As an example, the effect of increasing the height of drop on the formation of bubbles and
bifilms can be seen in figure (2.17).

Figure 2.17: Effect of increasing height of falling stream [9]

18
 General Concepts

In the figure (2.17, a), the molten stream is seen to fall and join the melt with the outer
boundary staying intact and thus not forming bifilms. A slight increase in height (2.17, b)
shows the stream falling and joining the melt to form a dross ring that is still intact but forms
bifilms that stay on the surface of the melt. Still another increase in falling in height (2.17, c)
shows how bifilms being formed and being carried deep into the body of the melt. Thus from
(2.17) it can be seen how increase in falling height increases the probability of bifilms to
form.

Another example of bifilm formation due to a turbulent flow is when a melt is having a high
velocity coupled with a non-proportional increase in channeling area. This might create
splashes as well as disintegration of the melt front into small forerunning melt volumes that
ultimately form bifilms upon the subsequent meeting with the tailing melt. This splashing of
molten metal and how bifilm might be formed is shown in figure (2.18).

Figure 2.18: Sketch of a surface entrainment event [9]

As hinted earlier, having a laminar flow is not a sufficient condition for ensuring bifilms are
not formed. What is wanted from a flow is to have the melt expanding, like a balloon being
inflated, at all front and thus filling the cavity by having the oxide skin end up as the casting‟s
surface. Failure to do so, even with a laminar flow (when melt not expanding at all fronts),
might also result in bifilms. As an example, in figure (2.19) shows such a case, where bifilms
form when melt laps due to an intermittent start-and-stop kind of filling.

19
 General Concepts

Figure 2.19: Formation of laps [9]

Another common phenomenon that results in the formation of bifilms is the separation of
melts, upon passing through different channels, and then meeting again to form what is
termed as a confluence weld. Confluence weld also forms when melts separate when hindered
by a core material placed in the cavity for creating holes. Confluence welds can be seen in
figure (2.20).

Figure 2.20: Confluence welds [9]

Other modes of bifilm formations exist, but the above descriptions give enough insight into
what bifilms are in particular and what entrainment defects are in general.

20
 General Concepts

These bubble formations, and entrainments cannot be explicitly evaluated by any simulation.
Nevertheless there exists a criterion for a critical velocity that must not be exceeded during
filling that might serve as a limit beyond which formation of such defects is very probable.
This critical velocity is approximately 0.5 m/s for all metals [10]. This value is almost always
exceeded, especially for pressure die casting process. For casting processes other than
pressure die casting this criteria is desirable. If this velocity can be achieved it is assumed that
the filling would give lesser defects. If not, values of up to 1 m/s particularly for thin walled
castings have also been said to give satisfactory castings, by researchers. Unfortunately,
aluminum needs to drop only 12.5 millimeters in order to attain this critical velocity! Thus, if
exceeding this criterion cannot be helped, the preferred practice is to try to contain the melt
close-bound by the channeling and cavity walls, in order to avoid splashing and sprouting.

Accordingly, during simulation, the flow velocities, whether too high or too slow, can be
taken as one indication of whether the flow has a potential for resulting in entrainment
defects. Other than the magnitude of flow velocity, splashing and sprouting occurrences in the
flow can also be observed, and tried to be minimized. Flow circulations should also be
watched out, as vortex actions might generate bubbles.

As mentioned earlier, flow velocities in pressure die casting (rheocasting included), always
exceeds the critical velocity. Thus, what should be used as a criterion for flow evaluation is
the latter method of watching out splashes and sprouts in melt.

As a conclusion what should be stressed here is that entrainment defects account for the
majority of casting defects and they are related to the pouring (filling) step of the six casting
steps described earlier. Thus observing the filling action of the melt is invaluable. It aids in
improving the filling pattern by varying the process parameters or by changing the geometry
or layout of the gating system. Ensuring that these defects are less probable to develop while
pouring would also mean that associated defects like porosity and cracks that develop by
unfurling of bifilms during solidification are also less probable to occur.

Other defects:
Other defects that could also be looked at during simulation are mis-runs and cold-shuts in
flow simulations while shrinkage cavities/porosities, surface sinks, contraction defects and hot
tears, etc. are observed in solidification and stress simulations.

21
 General Concepts

2.6 Simulation
The need to shorten the time and reduce the cost of product development has accelerated the
use of computer simulations. Simulation is a method for understanding processes and a tool
for optimizing them. To this end, the technology of simulation of casting processes has grown
extremely fast in recent decades and has been proved to provide results comparable to
physical processes.

Simulation of casting processes usually involves three areas. These are calculating the melt
pouring process-a fluid flow problem, calculating the solidification process-a heat transfer
problem and calculating stresses-a solid mechanics problem. These three areas of calculation
are the basis of any casting simulation software that is used to study process behaviors or
optimize a process.

From simulation of a mould filling process, one can study the filling pattern of the melt to see,
if air is trapped while pouring, if there will be areas in the cavity which might not fill
completely, if there is a flow circulation, if there is a folding of melt, etc. Velocities at
different areas of the melt might also be evaluated to see if a mold (non-permanent mold) is
susceptible to erosion due to high melt velocity. As mentioned earlier a very high velocity
might result in a turbulent flow as well as a non-intact melt front. Moreover, too low a
velocity might lead to laminar flow problems as well as holding back of the melt front, which
might result in cold-shuts – if temperature gets too low at these areas. Observation of the final
temperature after pouring also shows if the casting can be fed sufficiently during
solidification.

From simulation of a solidification process, one can observe the solidification pattern to study
areas which are prone to solidification shrinkage as well as hotspots. Moreover, observing the
temperature as well as time and duration of the solidification simulation helps to evaluate if a
feeding is sufficient or conversely, to design a better feeder. Study of cooling curves at areas
of interest in the casting helps to predict the resulting microstructure from the process.

From calculation of stresses, one can study the residual stresses that develop in the casting.
This aids in improving the geometry of the casting or varying the process parameters in order
to decrease residual stresses in castings. This simulation area, though equally important, is not
as frequently carried out as the first two. This is because casting industries still do not
consider studying residual stresses is very important and also because results from such
calculations are still doubted, with regards to approximating the actual physical stresses that
develop.

The advantage of simulating a process, over production trial runs or physical experiments,
was mentioned to be the saving of time and cost. Simulation also has the ability to get and
visualize different outputs at the same time (velocity, temperature, pressure, stress).
Nevertheless, since some simulations might take weeks to run and optimization an iterative
process, a project might take months to finish. Thus, a wise decision has to be made on how
long a time one is willing to sacrifice. Typically, not so many industries use simulation as a
tool for optimizing their processes. Instead, each industry has its own invaluable knowledge
reserve, from years of experience that it uses in designing a process. Unfortunately, because
of business competition between industries, a company may not be keen on sharing it‟s know-
how to an outsider. However, what one cannot shy away from is that the result of integrating
experience with simulation has proven to be a very powerful as well as rewarding method of
optimizing processes, thus triggering a growing trend, among casting companies, towards
adopting it as a practice.

22
 General Concepts

2.6.1 MAGMASOFT (V 4.4)

MAGMASOFT (V4.4) is the simulation software used in this thesis work. It is capable of
simulating melt flow as well as solidification and stress.

It is a complete system that provides a:

 Preprocessor: this is used to build or import geometries that are used to simulate a
process. The geometries normally include: the casting (the desired final shape), the
gating system (the channel that delivers melt into cavity) and the mold.
 Enmeshment: this is used to mesh the geometries.
 Simulation setup: here the various process parameters and specific materials as well
as their properties as set.
 Numerical calculation: based on chosen materials and their properties as well as
process parameters set, the software calculates for outputs the user desires (like stress
velocity, temperature, etc.).
 Postprocessor: here it gives a 3D graphical presentation of the results.

MAGMAsoft has various modules that enable it to simulate almost all types of casting
processes. Thus, in relation with this thesis, the normal shape casting(batch production)
module was used to simulate the plaster casting process while the high pressure die casting
module (MAGMAhpdc) with activation of the thixo-mode, was used for simulating the
rheocasting process.

MAGMAthixo
Rheocasting was mentioned to be a thixoforming process and thus, as a precondition, the
material used has to show two properties. These are pseudoplasticity and thixotropy. With
psuedoplasticity, the material shows a decrease is viscosity with increase in shear rate, while
with thixotropy, the stress in the material does not only depend on its current state of
deformation (motion), but also its recent deformation history [11].

A rheological model governs the way such a non-Newtonian fluid flows or deforms. This is
typically performed by making use of apparent viscosity models. MAGMAsoft has two such
models. These two models are the Carreau-Yasuda model and the Ostwald-de-Waele model,
of which, the latter was available and used for this project.

Other commercial software also exist, Starcast, QuikCAST, ProCAST, InteCAST, NovaCast
are examples of such software.

23
 Groundwork

3 Groundwork
This chapter discusses the various decisions made in choosing a cast alloy, casting processes
as well as the geometry modification carried out on the chassis. It also describes the steps
and considerations taken while designing the gating systems. The setup parameters used in
simulations as well as some aspects associated with secondary processes are also discussed
briefly.

3.1 Product requirement

As briefly mentioned in the first chapter, the original chassis was first designed with a plan to
machine it from a blank. Thus, the product attributes of the original chassis were taken as a
benchmark for setting the new chassis‟s requirement. This requirement was then used to
select suitable casting processes and the corresponding cast alloys for the new chassis.
The product attributes for the original chassis is given below:

Original Chassis:

 Material – Wrought Aluminum 6082 - T6

 Overall dimension ~430 x 275 x 120 mm
 Mass~ 4 kg
 Minimum section thickness~1 mm
 Tolerance - as per SS-ISO 2768-1
 Surface roughness – Ra 3.2 μm
 Surface treatment – Colourless chromating ACC SS-ISO 10546 Class 3
 Thread tolerance 6h-6H

3.2 Material screening

In chapter two, it was discussed that the aviation industry had a list of cast aluminum alloys
that can be used as an engineering material. These aluminum alloys were mentioned to be:

 A201.0-T7
 354.0-T6
 355.0-T6
 C355.0-T6
 356.0-T6
 A356.0-T6(T6P)
 A357.0-T6
 D357.0-T6
 359.0-T6

The physical and mechanical properties of these nine alloys as well as their alloy
compositions were evaluated against that of the wrought alloy. This served as the first step
towards narrowing down on a specific alloy.

24
 Groundwork

Accordingly, the two main functions of the chassis were taken to be:

 Protecting the electronic units that would later be placed in it. Hence, it has to guard
the units from harsh environment.
 Transferring heat away from these electronic units.

These two functions of the chassis needed having a cast alloy with corrosion resistance,
thermal conductivity and specific heat capacity values similar to or better than the wrought
alloy. In addition, it was assumed that the chassis will not be subjected to excessive
mechanical loads. Thus, strength and elongation requirements were given a lesser priority
than the thermal and corrosion properties (see appendix 8.1 & 8.2).

Thus, with respect to corrosion resistance, alloys 356.0, A356.0, A357.0, D357.0 and 359.0
come first, with better resistance to corrosion than C355.0, 355.0 and 354.0. The corrosion
resistance was better for both permanent and sand mold casting processes.

Looking at the thermal conductivity values, the alloys with large values were desired so as to
carry away heat fast. Out of the nine alloys, the seven, 355.0-T6, C355.0-T6, 356.0-T6,
A356.0-T6(T6P), A357.0-T6, D357.0-T6, 359.0-T6 had equivalent thermal property values.
This made them a better candidate over A201.0-T7 and 354.0-T6.

Thus, after comparing the alloys with respect to corrosion and thermal requirements the five
alloys, 356.0-T6, A356.0-T6(T6P), A357.0-T6, D357.0-T6, 359.0-T6 were regarded as cast
alloys that can replace the wrought one. This choice is later revised in section 3.4.

3.3 Process screening

In selecting the method of casting, the Cambridge Engineering Selector (CES) software
package was used to narrow down on few candidate processes.

Initial screening of the process was done by using the values shown below:

 Function: Chassis

 Constraint: Shaping Process: Casting (Primary)

Material: Cast Aluminum Alloy (The five candidates)
Shape: Solid 3D
Mass: 4 kg
Range of section thickness: 9 mm maximum
Minimum Section: 1 mm
Roughness: Ra 3.2 μm

 Rank By: Range of section thickness

 Objective: Minimize Cost

 Free Variable: Choice of Process(es)

Process Chain Option

Select forming processes – All stages

Some attributes of the evaluated processes is shown in table (3.1).

25
 Groundwork

Range of
section Economic
thickness Roughness Capital cost batch size
Name (mm) (µm) (kSEK) (units)
High pressure die
casting 1-8 0,8 - 1,6 1200- 5980 5e3 - 1e6
Centrifugally-aided
casting 1 - 20 12,5 - 25 12 - 120 10 - 1e4
Investment casting,
automated 1,5 - 75 1,6 - 3,2 120 - 598 500 - 5e3

Ceramic mold casting 1,5 - 999 1,6 - 3,2 5,98 – 59,8 50 - 1e3

Plaster mold casting 2 - 999 1,6 - 3,2 12 - 59,8 50 - 5e3

Shell casting 2 - 50 6,3 - 12,5 12 – 120 1e3 - 1e5
Evaporative pattern
casting, manual 3,5 - 999 9,4 - 12,5 3,59 - 23,9 1 - 1e3
Evaporative pattern
casting, automated 3,5 - 999 9,4 - 12,5 59,8 – 359 1e3 - 1e5
Low pressure die
casting 4 - 10 1,8 - 6,3 120 – 1200 2e3 - 1e5

Cosworth casting 4 - 999 3,2 - 6,3 120 -598 1e3 - 1e6

Green sand casting,
manual 5 - 999 12,5 - 25 5,98 – 35,9 1 - 1e3

Gravity die casting 5 - 45 3,4 - 6,3 120 – 478 1e3 - 1e5

Green sand casting,
automated 5 - 999 12,5 - 25 120 – 598 1e3 - 1e5

Squeeze casting 6 - 200 0,8 - 3,2 1200 - 5980 1e3 - 1e5

Table 3.1: Casting processes and their attributes [5]

It should be noted that values in table (3.1), like the minimum section thickness, are an
average, for a range of possible cast alloys. This means, compared to iron, aluminum can be
casted to even thinner sections and tighter tolerances than the values shown in table (3.1) [8].
Nevertheless, comparison of the processes was performed all the same, since the relative
capabilities of the processes were assumed to be the same.

The first criterion used for comparison was the minimum section thickness that the processes
can achieve (see table 3.1). Accordingly, the first five processes seen in table (3.1) were
selected and taken to the next level. Comparison between these five screened processes was
then done based on the surface roughness achievable. As a result, centrifugal aided casting
was eliminated from being a choice. This was done to avoid the extra machining process that
it demanded to attain the desired surface roughness value of 3.2 μm. Extra machining meant
extra cost. Thus, the chassis was desired to be casted as close to finished piece as possible.
Thus far, four processes were left; which were:

- high pressure die casting, investment casting (automated), ceramic casting and plaster
mold casting.

26
 Groundwork

The next criterion used was the economic batch size. This value indicates the number of units
that must be manufactured by a process for it to be considered economical. In other words, the
economic batch size indicates the number of products that must be produced by a process to
make it competitive with contending processes [5]. This concept of an economic batch size
could be best understood from the figure (3.1).

Figure 3.1: Relative cost of casting an aluminum rod versus the number of parts to
be cast [12]
Figure (3.1) shows the relative cost of producing an aluminum part with various casting
methods. High pressure die casting process, which has a high tooling cost but a fast
production rate, would be a cost competitive process with increasing batch size. On the other
hand, sand casting, which has a low tooling cost but a slow production rate, would tend to be
an expensive choice with increasing batch size.
The number of chassis that was planned to be produced was mentioned to be 5000 pieces.
Accordingly, table (3.1) had high pressure die casting, automated investment casting and
plaster casting as economical production methods for this planned batch size. On the other
hand, ceramic mold casting demanded a lesser batch size, for it to be a competitive in cost.
Thus, it was eliminated from the list of processes leaving three candidate processes.
Accordingly, the three processes that were remaining could be seen to be of two types of
processes.

The first type, which contained plaster casting and automated investment casting, fall under
processes that should have a low series production to be a cost competitive process. From
table (3.1), it can be seen that these two processes had the chassis‟s planned batch size of
5000 pieces as a maximum value, beyond which the processes starts to be less competitive.

The second type had only high pressure die casting, which falls under processes that needed a
high series production to be a cost competitive process. As seen from table (3.1), high
pressure die casting, had the chassis‟s planned batch size of 5000 pieces as a minimum batch
size, beyond which the process starts to be more competitive.

27
 Groundwork

One process was taken from each type. From the first group, plaster casting was the process
of choice as it had a better tool life and better tooling cost than automated investment casting
(appendix 8.3).

The second type only had high pressure die casting. However, since high pressure die casting
is naturally associated with turbulent flow that might result in porosity problems, another
process, rheocasting, was taken in its place. Rheocasting was considered since it had benefits
over high pressure die casting in particular and other processes in general, which will be
discussed in the next subsection. With regards to equipment cost, rheocasting uses the same
high pressure die casting equipment, with an additional slurry forming machine. In the next
subsection, the cost for equipment and tooling required for rheocasting, will also be argued to
have a cost comparable with a high pressure die casting process.

Finally, based on the steps described above plaster mold casting and rheocasting were chosen
as two processes that are capable of economically producing the chassis, as closed to finished
piece as possible.

3.3.1 Selected processes

Plaster Casting: As introduced in the second chapter, plaster casting (plaster mold casting) is
an expendable-mold casting process that can attain a near-net shaped casting. As the name
indicates, the process has a plaster (gypsum) material for a mold. It is usually used for non-
ferrous casting alloys and is specially suited to alloys with low melting points.

The process starts with making the pattern, which will then be placed inside a frame and
plaster aggregated around it. Subsequent to setting of the plaster, the pattern is removed; the
mold parts and core (if any) assembled and baked. This baking helps to remove moisture as
well as improve the permeability of the mold. After the baking of the mold melt can be
poured into the cavity and be left to solidify.

Its main advantages are associated with casting quality, like surface roughness, dimensional
tolerance and excellence in filling of thin sections. All advantages are attributed to the slow
cooling of the mold material. This slow cooling property of the mold also aids in minimizing
warpage and promoting a uniform mechanical property and structure in castings. In plaster
casting, control of this cooling process can be simply achieved by the use of chills. This
control of the cooling process is done to provide a desired mechanical property. Another
advantage of plaster casting is its ability to shape products with undercuts.

Main disadvantages for plaster casting include, a time consuming mold making and cooling
procedures. The molding (including pattern making) process, which usually consists of a
series of processes and equipments, take a long time. In the previous paragraph, the slow
cooling of the mold was mentioned to deliver the qualities achievable in plaster casting.
However, these time consuming molding and cooling procedures are disadvantageous as they
will increase the cost of production. This increase in cost is specifically significant as the
mold material is not reusable. Another problem associated with plaster casting is that the
mold material, with the exception of foamed plaster, has a low permeability to gas. Thus, to
avoid problems associated with permeability, venting should somehow be assisted (by
pressure or vacuum).

Rheocasting: As introduced in the second chapter, rheocasting is a special casting process

that has the metal in a semi-solid (slurry) form. The central process to this method is forming
the slurry. This is usually achieved by agitating the melt. Molten metal is cooled while
constantly being agitated to have the slurry. This agitation cuts off dendrites and results in
rounded solid particles in the slurry, which easily shears to flow and fill cavities. This

28
 Groundwork

difference between the microstructure of a normal semi solid and an agitated one can be seen
in figure (3.2).

Figure 3.2: Difference between the microstructure of a normal semisolid metal and an
agitated one [6]

This agitated slurry is thixotropic; i.e. its viscosity decreases with increasing shear rate. One
such slurry forming process is a process by Rheometal (a company working with rheocasting
technology). The process is called Rapid S, which stands for rapid slurry process. This
process uses the method of controlling the enthalpy to control the solid fraction of the slurry.
Rheometal has also developed a machine that uses this technique of attaining the thixotropic
slurry. Steps used for such a slurry forming process can be seen in figure (3.3).

Figure 3.3: The Rapid S method [6]

After the slurry is formed, it is typically transferred to a pressure die casting machine, pressed
into the mold cavity and left to solidify. It is usually used for low melting point alloys.
Advantages of the process include being able to achieve a near-net shaped casting, with
almost no porosity (see figure 3.4).

29
 Groundwork

Figure 3.4: Comparison of porosity formation in Rapid-S’s semi-solid vs. traditional

casting [6]

On the left had side of figure (3.4), a traditional semi solid metal can be seen to result in
porosity, while on the right, no porosity can be seen for a rheocasted part. In rheocasting, the
slurry will relatively be free of entrainment defects compared to a conventional melt. This
happens since the melt front as well as the melt itself stays relatively intact. Rheocasting has
other advantages too. Some of these advantages are: having lower cycle time, longer die life,
lesser need for lubrication, lesser pressure requirement, lesser die soldering problems and
producing a heat treatable product. All advantages are closely linked to having a lower
pouring temperature as well as a more laminar flow. A disadvantage of the process is that it is
not suitable to thin-sectioned products. This is due to the viscosity of the slurry and the lower
pouring temperature.

As mentioned above, equipments used for high pressure die casting process can be used with
one additional machine for forming the slurry. The extra tooling and equipment costs for the
slurry forming machine was assumed to be compensated by the benefits of rheocasting. These
benefits of having a longer die life and lesser cycle time reduce the tooling and production
costs that would have been incurred for a conventional high pressure die casting process.

3.4 Revision: Choice of alloys

In section 3.2, the five alloys 356.0-T6, A356.0-T6(T6P), A357.0-T6, D357.0-T6, 359.0-T6
were screened to be good candidates for the cast chassis. In addition, in section 3.3, two
processes rheocasting and plaster-mold casting were chosen as two candidate processes. After
selection of the processes the choice of alloys was revised. Considering rheocasting, rapid
slurry forming is said to be possible for all alloy ranges of Al-Si-Mg class. However, alloys
with silicon content of between 9-12% are termed „less functional‟ for rheocasting. A356.0 &
A357 are typical rheocasting alloys, with the former being the most widely used and
researched-on alloy. Moreover, the material data available from MAGMAsoft, that models a
thxotropic alloy, was only for the A356 alloy. This limited the choice of alloy to be A356.0
(thixo-model). Looking at the alloys for plaster casting, there was no restriction of a material
model by the software. However, one company that was contacted disclosed that it uses
A356.0 for plaster casting similar products. Thus, A356.0 was used in all plaster casting
simulations, while a thixotropic model of the same alloy was used for rheocasting
simulations.

30
 Groundwork

3.5 Geometry Modification

The original geometry of the chassis was believed to be too complex to be produced by
casting due to the following features:

 Holes – it had many small sized holes, lateral and vertical, with most being threaded,
see figure (3.5).
 Undercuts – it had undercuts on various walls to reduce the weight of the chassis, see
figure (3.5).

Holes

Undercuts

Figure 3.5: Original geometry

 Very thin sections – it had sections as thin as 1 mm, at fin tip and grooves, see figure
(3.6).
 Thin and long sections – due to the undercuts, it had very thin and long wall sections,
having dimensions of 1.5 mm with a 60 mm depth, see figure (3.6).

31
 Groundwork

Thin and
Thin section long section

Figure 3.6: Section view of original geometry - vertical section

 Non-uniform walls – it had various non uniform wall sections, where sections vary
from 1.5 mm up to 6 mm see figure (3.7).

Non-uniform
walls

Figure 3.7: Section view of original geometry – horizontal section

The features discussed above were difficulties for casting the chassis and were minimized by
modifying the geometry. Moreover, as per requirements of the chosen casting processes,
provisions for draft angles, fillet radii and allowances were taken into consideration while
modifying the chassis. The allowances are usually added for reasons like: ease of
pattern/casting removal from mold, for subsequent machining processes as well as for
shrinkage. If available, company specific process capabilities and recommendations are
crucial while providing these allowances or changing some dimensions on the chassis – since
tolerances, draft and fillet requirements vary from one company to another.

32
 Groundwork

Changes that are to be made on the original chassis should have the following four points as
constraints:

1) Mass should be preferably equal to or less than the original. If not, the increase in
mass should be of an acceptable value.
2) Dimensions of the bottom areas of the three compartments should preferably stay
constant but not decrease. This is for the purpose of accommodating the electronic
units that are to be placed inside the chassis.
3) Minimum distance between an inside of any wall and a part to be placed in the
compartment should preferably stay constant but not decrease. This is for ease of
assembling/disassembling the electronic units; i.e. for provision of space that the wire
connecting and screw fastening procedures require.
4) Number of screws should preferably not decrease and/or a maximum distance of 45-
50 mm between screws should be kept.

To modify the geometry, the original chassis, weighing a total of 3.87 kg (after finishing),
was considered in two parts: These were the box and the fins.

Box Fins

Figure 3.8: Original box and fins

The box as shown in figure (3.8), with its holes and undercuts, weighs 2.911 kg while the fins,
as shown above (with cut out sections, at box holes), weighs 0.959 kg. In all cases regarding
geometry modification, the density of aluminum was assumed to be 2710 kg/m3.

The holes in the box and the cutout parts of the fins, seen in figure (3.8), were agreed to be
machined after casting. This resulted in an undrilled box and uncut fin as shown in figure
(3.9). The mass of this undrilled original box and uncut fins was 2.964 kg and 0.999 kg
respectively, summing up to total of 3.963 kg. Thus, comparison of the new model‟s mass
was done with these values of a non-cut fin and an undrilled box (as in figure 3.9). However,
the undercuts on the box still remain. Thus, the new model should have a mass comparable to
these masses of an undrilled box and an uncut fin. Upon subsequent drilling of the holes and
cutting out of the fins of the new chassis, its mass should be comparable to the original
chassis‟s mass (as in figure 3.8).

33
 Groundwork

Undrilled Box Uncut Fins

Figure 3.9: Original undrilled box and uncut fins

As mentioned above the modification was done in parts, with fins being modified first and the
box modified next. The considerations taken while modifying the chassis (fins and box) will
be discussed next.

3.5.1 Fins

The function of the fins is to carry away heat from the box and the original fins had the
dimensions shown in figure (3.10).

Figure 3.10: Dimension of original fins

It was decided that the height and length of the fins as well as the number of fins should be
kept unchanged.

Based on recommendations (preferred values) from two companies, one for rheocasting and
another for plaster casting, the profile shown in figure (3.10) was modified.

a) For the case of rheocasting, a minimum section thickness of 2 mm at top and a draft
angle of 1.5 degrees on the ejector die side of the chassis (assumed to be the fin side)
and 2 degrees on the cover die side were recommended. In addition, a full radius fillet
was also recommended at the top of the fin. These two changes gave a fin having the
following cross-section.

34
 Groundwork

Figure 3.11: Dimension of new fins

Using these recommended values while keeping the length height and number of the
original fins, increased the mass of the new fins to 1.339 kg. This was an increase of 0.34
kg from the original fins.

b) For the case of plaster casting, a minimum section thickness of 1.8 mm at top and 4
mm at the bottom of the fins were recommended. This would have resulted in a taper
angle of about 1.4 degrees and thus gave the fins a more or less similar cross-section
to the fin shown in figure (3.11). It was therefore decided to use the same fin profile
for both processes.

For ease of any future changes that may be required, the new fin was modeled
parametrically, depending on two variable dimensions – the thickness at top and the taper
angle. The number of fins might also be varied on the parametric model.

3.5.2 New box

Giving considerations to the constraints put for changing the chassis‟s geometry described
earlier and the difficult features like non uniform thickness and thin sections, etc, a new box
was modeled. The new box was believed to have a better castability than the original box and
can be seen in figure (3.12). The major changes made, its compliance to the constraints, and
its features that were believed to make it better to cast are also discussed next.

35
 Groundwork

3
Figure 3.12: Changes made on the new box

As could be seen from figure (3.12) the major changes made on the new box were:

 The new chassis had no undercuts. This benefits both rheocasting and plaster casting.
Undercuts were not possible to shape by rheocasting and thus the new model can be
made with no difficulty. Even if the undercuts were planned to be machined after
rheocasting, the new model still had the advantage of reducing machining costs.
As for plaster casting, it being an expendable mold process, undercuts were not such
difficult features to shape. Nevertheless, the new model had the benefit of reducing
the cost by decreasing the mold making time as well as the number of mold parts
required for an undercut chassis. The new model thus needed fewer mold parts as no
core material would be used.

 It had one new big recess on the outside wall, as pointed by number 1 in figure (3.12)
as well as eight smaller recesses on the inside walls – to reduce an equal amount of
mass as made possible by the undercuts of the original chassis.

 The recess pointed by number 2 in figure (3.12) demanded a change in position and
dimension of the wire channel seen next to it.

 The drain ways pointed by number 3 in figure (3.12) were pushed back and a wall
material provided.

36
 Groundwork

 It had a relatively constant wall cross-section of dimension 3 mm almost all-round

the box; as opposed to the varying section thicknesses of the original chassis (see
figure 3.7).

Considering the constraints

 The mass of the new box was 2.908 kg, which is less than the original by 0.056 kg.
(this value will increase upon draft provision as will be discussed in the next sub-
section).

 The dimensions of the compartment‟s bottom areas as well as the distances between
the compartment‟s walls and components to be placed inside were kept to the
previous minimum. This meant that no inconvenience would be introduced in the
assembly or disassembly procedures of the new chassis.

 The number of screws was one less than the original one, with the minimum distance
between two screws being 60 mm at one place. This was due to the provision of a
through-height recess on both sides of the inside wall, as shown in figure (3.13).

Figure 3.13: Position of increased distance between screws

Again, for future changes that may be required, the new box was also modeled parametrically,
using two variables. One variable was the draft angles on the cover die side of the chassis
(assumed to be the fin side). The other variable was the draft angle on the ejector die side of
the chassis (non-fin side). The details of these draft provisions will be discussed in the next
subsection.

After the described changes were made on the box and fin, the new chassis comprising the
modified box weighing 2.908 kg (0.056 kg less than the original) and modified fin having

37
 Groundwork

1.339 kg (0.34 kg more than the original), weighed a total of 4.247 kg. This value was heavier
from the original chassis by 0.284 kg. On the other hand, the new chassis was believed to be
relatively easy to cast and had the benefits of having no undercuts. The new chassis also had
the advantage of having a thicker minimum wall section as compared with the original one. It
was also believed to be better with regards to defects as well as cheaper with regards to
production.

This new box was next prepared for simulation. This preparation, brought about changes in
mass and dimensions, due to the various allowances that had to be considered, as discussed
next.

3.5.3 Allowances

A geometry that is desired to be cast has to have some allowances. This are provided to have
a final size and shape of casting that is accurate and easy to produce.

These major allowances are:

 Shrinkage allowance
 Machining allowance
 Draft allowance

Shrinkage allowance: This is added to compensate for liquid shrinkage, solidification

shrinkage as well as the subsequent solid shrinkage during cooling to room temperature. It is
usually set by experienced pattern makers. For this thesis the shrinkage allowance was
neglected.

Machining allowance: This is added to have an adequate surface to machine. Machining is

usually needed to attain required surface roughness and dimensional tolerance. Machining
might also be needed to compensate for distortions that may arise during casting. This
allowance was also not considered for reasons discussed on section (3.8).

Draft allowance: This is added for ease of removal of pattern from mold or removal of cast
from mold. These allowances were considered and values recommended from two companies,
for the two respective processes, were taken.

With respect to the draft allowance, each process has its own range of recommended values,
which varies from one company to another based on their capabilities.

a) For rheocasting: A draft angle value of 1.5 degrees on the ejector die side and 2
degrees on the cover die side was recommended by the company.

b) For plaster casting: A draft angle of 1.5 degrees was recommended on both sides.

However, for comparing both processes, the same draft of 1.5 degrees on the fin side and 2
degrees on the other, was taken for both processes.

While considering draft allowance, two options were considered, as shown on a simple
representative model in figure (3.14).

38
 Groundwork

Figure 3.14: Two options considered for drafting

The first option was to draft the faces as shown on the top part of figure (3.14). In this option,
the constraint of not decreasing the compartments floor areas, described on previous sections,
was not met. It demanded machining of the faces that decreased the floor areas. It also did not
meet the constraint of not increasing mass. All walls are seen to have added mass due to the
drafts. On the other hand, the size of the chassis was not affected as the overall dimension was
not changed.

In the second option, the faces were drafted as shown on the bottom part of figure (3.14). On
the middle wall, it had the same draft as with option one but had a different draft on the outer
walls. On the outer walls, material that is added by a draft on one side would almost be
removed by the draft on the other side. This is done while keeping the areas of the
compartments floors unchanged. This resulted in an increase of the overall dimension of the
chassis as seen in figure (3.14). This option required no machining on the walls, as compared
to option one, as the mass is relatively unchanged (except of course on the middle walls
which was the same with option one). The floor areas were not decreased. Here the outer
walls still had a dimension of about 3 mm – similar to the non drafted chassis.

After comparing the two options, the second type of draft was chosen for the purpose of
avoiding (or reducing) machining cost. On top of that, the difficulty of filling such a thin
walled component was believed to be evident when using the second option as opposed to the
first option, which had thicker walls, than the modified chassis (in figure 3.12). It has to be
remembered that the draft magnitude varies on the side of the chassis (ejector side or cover
side), as discussed earlier.

In addition, the influence of such a draft on the final assembly of the electronic units was also
considered. All holes were planned to be drilled and threaded after the chassis is casted. Upon
subsequent assembling, these threaded holes are used for mounting the electronic units. The
holes should have their axes perpendicular to the faces of the non drafted walls, similar to the
original chassis. Thus, for the new drafted chassis, the holes would not have their axes
perpendicular to the drafted face but at an angle of 1.5 or 2 degrees from the perpendicular.

With regards to the assembling procedures affected by the draft provision, the procedures
could be categorized into two, based on the type of fastening sides used. These were given the
term „one sided fastening‟ and „two sided fastening‟.

39
 Groundwork

One sided fastening, as can be guessed from term, involved fastening a part by a screw to one
side of a threaded vertical wall. As the axes of the holes were designed to be perpendicular to
the undrafted face, it would mean that a screw head would rest against a wall that is drafted
by 2º. One such fastening can be seen on the 2D sketch and 3D representation in figure (3.15).

Figure 3.15: Assembly by one sided fastening

This was assumed to cause no problem on assembly as the deviation, 2º from the undrafted
wall, was assumed to be very small and moreover the washers used and materials to be
fastened can also deform and accommodate such a variation.

The two sided fastening were again of two types. The first is fastening at the outside walls,
with five such fastening cases. The second fastening on the inside walls, with only two
fastening cases.

 The first type involves resting one component on one side, on a 2º drafted wall, and
securing it from the other side, with a threaded nut resting on a 1.5º drafted wall.
There is a variation of 0.5 degrees between the sides of a wall. One such fastening can
be seen on the 2D sketch and 3D representation in figure (3.16).

Figure 3.16: Assembly by two sided fastening (first type)

Again this was assumed to be of no significant problem as the change is so small that
the wall could be thought of having vertical sides. Moreover, since most components
assembled are of elastic material and have flexible washers, they will deform to
accommodate the variation in angles.

40
 Groundwork

 The second type involves resting one component on one side, 1.5º drafted wall, and
securing it from the other side to a threaded nut resting on another side to a 1.5º
drafted wall. Here material is added on both sides of the wall. One such fastening can
be seen on the 2D sketch and 3D representation in figure (3.17).

Figure 3.17: Assembly by two sided fastening (second type)

This would mean now there is a total of 3 degrees between the two sides of the wall. This
might necessitate the use of accommodating washer elements or machining of these selected
walls to remove the drafts.

In the end, other than the draft features put on the model, appropriate fillets of varying values
was also put at the numerous sharp edges on the model. This was done to promote smooth
melt flow as well as avoiding chipping and casting difficulties at sharp corners.

As hinted above, the draft angles increased the overall dimension of the casting to ~438 x 286
x 120 mm from ~430 x 275 x 120 mm of the new chassis. Moreover the mass of the final
chassis increased to 4.457 kg. The mass of the non drafted new chassis was 4.24 kg and that
of the original chassis was 3.96 kg. Thus there was a 0.494 kg increase from the original
chassis (figure 3.9), of which 70% was attributed to the increase in mass of the fins.

41
 Groundwork

3.6 Gating systems

In order to simulate the casting processes on the new geometry, gating systems had to be
designed first. The assumptions, calculations and recommendations [10] taken in designing
the gating systems are discussed next.

3.6.1 Gating system for plaster casting

Pouring basin:
As usually practiced with plaster mold casting, the company contacted for plaster mold
casting the component, does not make use of a pouring basin for pouring in melt. Instead, it
makes use of a sleeve to guide in the melt. The melt is poured from a hand held ladle as
shown in the figure (3.18).

Figure 3.18: Pouring method in plaster mold casting

The sleeve used by the company had an inside diameter of 50 mm and a height of 200 mm
and had no taper.

Sprue:
To calculate the theoretical dimensions of the down sprue, the well known method of
simultaneously equating the relations of a freely falling stream (from zero velocity) and of
conservation of matter was used. Thus from a free falling melt:

1/2
𝑣 = 2𝑔𝑕 (3.1)

And from conservation of matter:

𝑄 = 𝑣1 𝐴1 = 𝑣2 𝐴2 = ⋯ (3.2)

The above relations are equated at two heights of the sprue. This heights are at entrance to
sprue and at its exit, indicated by 1 and 2 respectively (see figure 3.19). Thus the two relations
(3.2 and 3.1), gave velocity values and areas at the respective heights.

42
 Groundwork

Sleeve

Sprue

2
Figure 3.19 Two views of a sample sleeve and sprue

Assuming that pouring from ladle starts just at the top of the sleeve (height 0); the melt drops
a height of 200 mm to get to the top of the sprue (height 1). It should be recalled that the
critical velocity that alloys should not exceed, so as not to have a turbulent flow is believed to
be 0.5 m/s and this is attained after a 12.5 mm drop for aluminum. Since such a large fall was
unavoidable, what must be tried was to contain the melt close-bound by the sprue walls. This
was done to avoid splashing and keeping the melt intact. To that end, the areas calculated
from the above two relations resulted in channel that pressurizes the melt by reducing the area
of flow. The two areas, at height 1 and 2, should be connected by a profile that would assume
the shape of a freely falling stream (as could be seen in figure 2.18). This profile is believed to
have hyperbolic shape. For obvious complications of shaping such a hyperbolic profiled
sprue, the usual practice is to connect the top and bottom areas straight. This results in a
tapered cone or prism, depending on the profile of the two areas.

Getting back to the calculation of the areas, the top of the sprue (height 1), should be higher
than the highest point of the component to be cast. This was done to have a head pressure for
pushing melt to mold extremities. As per the chosen layout, the height of the chassis was 120
mm. So an extra 20 mm was provided for creating a head pressure. This gave a drop of 140
mm in the sprue. This, added to the earlier 200 mm drop through the sleeve, gave a total
height of 340 mm. Using these two drop heights, 200 mm to get to height 1 and another 140
mm to get to height 2, the corresponding velocities were:

1/2 1/2
𝑣1 = 2𝑔𝑕1 = 2 ∗ 9.81 ∗ 0.2 𝑚 ~1.98 m/s (3.3)
And
1/2 1/2
𝑣2 = 2𝑔𝑕2 = 2 ∗ 9.81 ∗ 0.34 𝑚 ~2.58 m/s (3.4)

Thus 1.98 m/s and 2.58 m/s were the assumed velocities at entrance and exit of sprue, i.e. at
height 1 and height 2 respectively.

Before continuing to the calculation of the two sprue areas, the assumptions made for pouring
time/fill rate has to be discussed.

The pouring time for a component to be casted, especially a new product, is highly arbitrary.
Generally speaking both fast and slow poring rates create problems. It is difficult to ascertain
whether the problems witnessed are due to poor gating design or just the pouring time alone.

43
 Groundwork

Nonetheless, a filling rate of 1 kg/s was assumed to be a good value. As the products mass
was 4.457 kg, this filling rate gives a poring time of about 4.5 seconds. It should be noted
that 4.457 kg doesn‟t include the mass of the gating system. However this filling rate of 1 kg/s
was not directly used as pouring rate is naturally fast at the beginning of pour and eventually
drops to zero when the cavity fills. To take this increased rate at the beginning of flow into
account, a pouring rate of 1.5 kg/s was assumed for calculating sprue dimensions.

To calculate sprue entrance and exit areas, the volume flow rate of the process needs to be
calculated, which is given by:

𝑄 = 𝑓𝑖𝑙𝑙𝑟𝑎𝑡𝑒/𝜌 (3.5)

Where ρ is density of the metal and its value was taken as 2500 kg/m3 (since it would be at
high temperature).

This gave an initial volume flow rate of :

1.5
𝑄= = 0.6 ∗ 10−3 𝑚 3 /𝑠 (3.6)
2500

Resulting in areas of ;
𝑄 0.6
𝐴1 = 𝑣 = 1.98 ~0.303 ∗ 10−3 𝑚 2 = 303 𝑚𝑚 2 at sprue entrance (3.7)
1

and the exit area,

𝑄 0.6
𝐴2 = 𝑣 = 2.58 ~0.232 ∗ 10−3 𝑚 2 = 232 𝑚𝑚2 (3.8)
2

These two areas could be circular or a polygon with equal number of sides. In this thesis, for
the purpose of making an easy sprue-runner junction, a rectangular area was chosen. Thus, the
sprue was set to have the dimensions 8 mm x 38 mm for the entrance area and 8 mm x 29 mm
for the exit. The thickness was selected because an 8 mm sprue was believed to pressurize the
melt and contain it from sprouting.

Well:
The use of wells at the exit of a sprue has been a long tradition in casting industries but a one
that is dying, as it is argued that its benefit of slowing down melt is a mine for bubble
formation. Thus, its use was avoided and the melt was directly bent 90 degrees, to join a
horizontal runner.

Filter:
The company contacted for plaster casting, uses filters for some products that have high
quality requirements. The prime importance of using a filter is slowing down the velocity that
exits the sprue. Its effect in holding back dross (or bi-films) is believed to be ineffective
(especially in aluminum). Thus, for its benefit of slowing down the melt velocity as well as
stream-lining the melt, the use of a filter was decided for. The filter was placed just after the
bend from the sprue, with different layouts being tried. The results of the various layouts
would be discussed in chapter 4. A typical filter used for casting aluminum could be seen in
figure (3.20).

44
 Groundwork

Figure 3.20: A typical ceramic foam filter

Runners:
The melt that was dropping vertically through the sprue was bent 90 degrees to join the
runner. This bend was assumed to reduce the velocity of the melt by 20 % and thus
necessitating an increase of cross-sectional area by an appropriate amount (20 %). This
increase of area, retains the velocity of the melt at the reduced level (see figure 3.22). The
area at sprue exit is increased by 1.2 times that of the runner. This was achieved by increasing
the thickness of the runner to 9.6 mm with its width staying the same as the sprue exit. This
thickenss was also assumed to pressurize the melt and contain it from sprouting. Instead of
having sharp corners at both sides of the sprue – runner junction, provision of a fillet was
made having a radius equal to the sprue thickness (8 mm). The profile of this junction could
be seen at the most right in the figure (3.21).

Not good Better Best

Figure 3.21: Various fillet types for sprue-runner junction

Moreover, the use of the filter described earlier, is believed to reduce the velocity of melt by
25 % and thus necessitate an increase in area by the same amount. However, in order to
pressurize the melt a bit, the cross-sectional area of the runner after the filter is increased by 4
times the area of sprue exit (see figure 3.22). It should be noted that the area after filter is four
times that of sprue exit not the area of the runner. This was again done to pressurize the melt.

A2

1.2xA2 4xA2

Figure 3.22: Sprue exit area (A2) and its increase at bend and after filter

45
 Groundwork

Runner extensions:
Runner extensions are usually placed to divert and contain the first melt from entering the
cavity. Runner extensions of different layouts were also investigated to study their effect. The
results of the different layouts would be discussed in chapter 4.

Ingate:
The cross-sectional area of the ingate was set to be equal to the area of the runner after the
filter. Thus, resulting in a sprue exit-to runner-to ingate area ratio of 1:4:4. The latter two
areas are equal and are four times the first area. Assuming the filter reduces velocity at sprue
exit by a factor of 4 and the bend reducing it by 20%, the reduction in velocity that is
theoretically expected is:

2.58 0.8
𝑣𝑖 = 4
∗ 𝑠
= 0.58 m/s (3.9)

Where 𝑣𝑖 , is the velocity at the ingate as the melt enters the casting. This value does not
exceed the critical velocity by much, and in fact velocities of up to 1 m/s are claimed to be
used successfully for thin walled castings as the walls leave no room for the melt to sprout or
disintegrate. However the magnitude given by equation (3.9) is expected to increase as the
initial velocity at the entrance of the sleeve is non zero as assumed in the derivation. Instead
there would be a small falling of melt before it gets to the sleeve entrance. Moreover, other
reasons like back pressure from impact with runners also increase velocities at entrance to
casting. This would also be discussed on chapter 4.

The ingate area, which was set 4 times that of the sprue exit, could be of a single gate or could
be divided to multiple gates. The effect of having a single gate as opposed to multiple gates
was also studied and would be discussed on chapter 4.

Layout:
Different directions and positioning of the ingates as well as various layouts have been
studied some of which are discussed chapter 4. When considering different type of layouts for
the gating system, the removal of pattern material as well as removal of final casting was
carefully considered. If a complex gating system or an unrealistic gating system was
designed, it was believed that it will complicate the manufacturing of the pattern and the mold
cavity as well as increasing the number of mold parts – which in turn would increase the cost
of production.

3.6.2 Gating system for rheocasting

The designing of gating systems for rheocasting is still an area that has not been researched a
lot. The typical practice is to design a similar gate as in a high pressure die casting process
and try to optimize the geometry as well as flow parameters, based on trial runs. It was hinted
earlier that casting companies have a knowledge base that serves as an aid in designing an
“optimal” gate with a relatively short time period. Unfortunately, this is usually kept as a
company secret. Thus, designing of a gating system for the rheocasting was done by adopting
an ingate from a similar high pressure die casted product. The ingate area, the ingate length as
well as the volume of the gating system was varied, while keeping the plunger diameter
constant.

Inlet:
As the diameter of the inlet has to be equal to the diameter of the plunger, the diameter of the
inlet was made 80 mm with a height of 20 mm. This plunger diameter was selected after
choosing a machine from the contacted company. The height of 20 mm was taken because the
software demanded an inlet to be at least three elements thick.

46
 Groundwork

Biscuit:
The biscuit, representing the melt left over in the shot sleeve, was obviously designed to be of
the same diameter as the plunger, while the height was set to be 50 mm.

Gating and Ingate:

As mentioned above the gate and ingate, making up the gating system were taken from
another similar product and adopted, in hopes of getting an improved flow. It was of a fan
type gate, typical to high pressure die casting process. Its geometry can be seen below (figure
3.23), while the results of the variant dimensions of the gate tried would be shown in chapter
4.

Gating Biscuit

Ingate
Inlet

Figure 3.23: Gating system for rheocasting

Overflows:
Extra volumes for containing overflow of melt has been added on some simulations, as
customary with high pressure die casting processes. Some of the simulations had over flows.
This would not vary expected results as such.

Layout:
Different directions of positioning the ingates as well as the whole set up have been studied
and some would be shown in chapter 4. Here too, the removal of casting was carefully
considered, when trying out various layouts and designs of the gating system.

3.7 Simulation Setup

As introduced in the second chapter, the simulation software used for simulating the casting
process was MAGMASOFT. The set up for simulating the two processes is briefly discussed
next.

3.7.1 The preprocessor - Plaster casting

The variant gating systems, along with the chassis were modeled in Catia V5 and imported to
the MAGMAsoft preprocessor. One example of a gating system, chassis and mold can be
seen in figure (3.24). Some control points for tracing flow as well as studying solidification
were set at appropriate points.

47
 Groundwork

Figure 3.24: An example of a plaster casting set up as seen in the preprocessor

3.7.2 Enmeshment parameter - Plaster casting

Meshing of the geometry was done with the automatic mode. As with other related software,
MAGMAsoft uses the finite difference method and makes use of hexahedral mesh elements.
The component, being of very thin section, required the use of a large number of cells – as the
software needs at least two cells across a thickness, for accurate calculation. Thus, the number
of total cells used ranged from 14,000,000 to 15,000,000. This gave metal cells of 850,000 to
980,000. The disadvantage of increasing the number of cells, which provides more accurate
results, is an increase in computation time. The different problems that may arise from not
having enough number of cells will be discussed in chapter 4.

3.7.3 Simulation parameters - Plaster casting

While setting up the simulation parameters, first the calculation for filling and solidification
was requested and then the different materials were selected. The inlet, pouring basin (sleeve),
gating (runners and sprue), and ingates had the same material – AlSi7Mg (A356.0) while the
mold had another material- „gypsum-investment‟. Both these materials were selected from the
MAGMAsoft material data base, with no change in any properties, i.e. the default setting was
used. Some properties (thermal conductivity, density and specific heat capacity) of these
materials can be found in appendix (8.4).

Initial temperature- the initial temperature for aluminum was taken as 700ºC, while for the
mold, a value of 190ºC was chosen. These values were recommended by a company
specialized with plaster casting. In addition, a heat transfer coefficient of 1000 𝑊/𝑚2 𝐾 was
set between the cast mold interface. A pouring time, which was calculated from the filling
rate of 1 kg/s, dictated the pouring time.

48
 Groundwork

3.7.4 The preprocessor - Rheocasting

The tasks performed here are similar with the other process, with the obvious change in the
gating system. Layout for one example can be seen in figure (3.25).

Figure 3.25: An example of a rheocasting set up as seen in the preprocessor

3.7.5 Enmeshment parameter - Rheocasting

The tasks performed here are similar with the other process, with the obvious change in the
gating system. The total number of cells used is between 10,000,000 and 12,000,000. The
difficulties faced for each process will be discussed in chapter 4.

3.7.6 Simulation parameters - Rheocasting

The first part of setting up the simulation was similar to the previous process. The calculation
for filling and solidification was requested and then the different materials were selected. The
inlet, gating, and ingate were assigned the same alloy – AlSi7Mg_thixo while the mold had
another material- steel. Both these materials were selected from the MAGMAsoft material
database, with no change in any properties, i.e. the default setting was used. Some properties
(thermal conductivity, density and specific heat capacity) of these materials can be found in
the appendix (8.5).

As for the preheat temperature, the initial temperature for aluminum was taken as 590ºC while
for the mold a value of 250ºC was chosen. These values were recommended by a company
specialized with rheocasting.

In addition, a heat transfer coefficient of 7000 𝑊/𝑚 2 𝐾 was set between the ejector and cover
die interface, while a temperature dependant coefficient was set between the cast material and
molds. Values of this temperature dependant coefficient can be seen in the appendix (8.5).

49
 Groundwork

The second part of the simulation setup, unique to high pressure die casting, was next
performed. The magma Demo_700 machine was selected here, with its default settings. The
diameter of the plunger was kept to be 80 mm, while changing the active length of shot sleeve
- so as to get a shot sleeve feeling of about 50%, a recommended value for rheocasting. This
was followed by setting the slow shot velocity and the fast shot velocity. The slow velocity
was kept around 0.1 m/s while the fast was varied so as to get the recommended ingate
velocity.

These values which were provided by Rheometal, a consulting company experienced with the
process, can be seen in table (3.2).

Maximum
Gate thickness speed at gate
2 𝑚𝑚 9 𝑚/𝑠
4 𝑚𝑚 5 𝑚/𝑠
6 𝑚𝑚 4 𝑚/𝑠
8 𝑚𝑚 3.5 𝑚/𝑠
10 𝑚𝑚 3 𝑚/𝑠
> 10 𝑚𝑚 2 𝑚/𝑠
Table 3.2: Recommended fast shot velocities vs. gate thickness (rheocasting)

The above values were studied, while simulating the various simulations performed by
varying ingate dimension. However, the ingate velocity was not restricted to these relations
with gate thickness. Values of double, and even more than three times these velocities were
tried; to study the effects of exceeding these values. Switching over of the slow velocity to the
fast one was generally set to be just after the filling of the runners. Die closing parameter was
set to be time dependant at 20 seconds and the core opening definition set to 15 seconds- for
both ejector and cover dies. Finally a feeding efficiency of 70%, which is customary to
rheocasting, was also set.

50
 Groundwork

3.8 Subsequent Processes

This subsection is a very general overview on expected subsequent processes.

3.8.1 Heat Treatment

Generally defined, heat treatment is the process of cooling or heating a component to give it a
desired attribute. It is mostly done to change physical and mechanical properties,
metallurgical structure as well as residual stress distribution of the component.

A typical heat treatment procedure that is practiced in industries, consisting of three steps,
was proposed here. It involves solution heat treatment, followed by quenching and a final
aging process.

Solution heat treatment – involves heating alloy to a temperature above the solvus
temperature, where all solutes are dissolved to attain a single-phased structure and holding it
for a finite time.

Quenching – involves rapidly cooling the above heated component, having a single phase, to
prevent precipitates forming

Aging – involves reheating the alloy to a suitable temperature below the solvus temperature
for a finite time that the desired precipitate needs to form.

Depending on the type of alloy used and process used to shape it, various standards as well as
company specific treatment parameters exist. Based on the modification carried out on an
alloy as well as the geometry complexity, these treatment parameters, vary for a specific
alloy. One such recommendation, for A356 alloy can be seen in table (3.3).

Alloy Temper Type of Solution heat treatment(b) Aging treatment

casting(a) (c)
Temperature Time, Temperature(c) Time, h
h
°C °F °C °F
A356.0 T6 S 540 1000 12 155 310 3-5
T61 P 540 1000 6-12 Room 8(minimum)
temperature
155 310 6-12
(a) S, sand; P, permanent mold.
(b) Unless otherwise indicated, solution treating is followed by quenching in water at 65-100 °C (150-212 °F).
(c) Except where ranges are given, listed temperatures are ±6 °C or ±10 °F.

Table 3.3: Typical heat treatments for aluminum alloy and permanent mold casting [14]

As mentioned earlier, table (3.3) was put here as an example. However, one thing that must be
stressed here is to avoid the use of water as a quenchant, which is often customary in
industries. Quenching with water develops a high residual stress that might lead to failure of
the component. Instead quenching with air or polymer can be used as alternate procedure. The
polymer quenchant is an expensive choice but faster while using air is cheap but slower. The
slight decrease in mechanical properties, about 5-10%, that might be achieved by these
alternate methods, as compared to quenching in water was believed to be acceptable and
beneficial - as the life of non-water quenched component is usually longer [10].

51
 Groundwork

3.8.2 Machining

Typically, a casting needs to be machined for three reasons [17]. These are:

a- when the dimensional tolerance achievable on the as-cast component is considered

inadequate and believed it would impair the function of the component.

b- when the surface roughness achievable on the as-cast component is considered

inadequate and believed it would impair the function of the component.

c- when some design features are believed to be too intricate for production, in the as-
cast form.

The above conditions do not take into account the compulsory machining that may be
required to separate the desired cast from the gating system.

Accordingly, while choosing the two casting processes and running the simulations, the
process capabilities were compared with that of the requirement set for the original chassis
that was to be produced by machining.

With regards to the dimensional tolerance (point a), the standard that was to be referred for
the machined chassis was ISO2768-1. This was checked against the related ISO 8062
standard for casting processes.

A table of extracted values from ISO 2768-1, concerned with permissible deviations for linear
dimensions (except broken angles), can be seen in table (3.4).

Tolerance class Permissible deviations for basic size range [mm]

Designation Description from over over over over over over over
0.5* 3 6 30 120 400 1000 2000
up to up to up to up to up to up to up to up to
3 6 30 120 400 1000 2000 4000
f fine ±0.05 ±0.05 ±0.1 ±0.15 ±0.2 ±0.3 ±0.5
m medium ±0.1 ±0.1 ±0.2 ±0.3 ±0.5 ±0.8 ±1.2 ±2
c coarse ±0.2 ±0.3 ±0.5 ±0.8 ±1.2 ±2 ±3 ±4
v very coarse ±0.5 ±1 ±1.5 ±2.5 ±4 ±6 ±8
* For nominal size below 0.5mm, the deviation shall be indicated adjacent to the relevant
nominal size(s).
Table 3.4: Permissible deviations for linear dimensions, non-casting processes [20]

Assuming a medium tolerance class was to be kept, to be fair to casting processes, a

dimensional deviation of 0.1 mm up to 0.8 mm, in accordance with the dimension ranges, was
acceptable for the machined chassis. Consulting ISO 8062 for the casting processes, shown in
table (3.5), the tolerance levels are highlighted for the chassis, which had a raw basic
dimension of 430 mm.

52
 Groundwork

Cast piece Total tolerance width

base measure Tolerance grade CT
(mm)
from to 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
10 0.18 0.26 0.36 0.52 0.74 1.0 1.5 2.0 2.8 4.2
10 16 0.20 0.28 0.38 0.54 0.78 1.1 1.6 2.2 3.0 4.4
16 25 0.22 0.30 0.42 0.58 0.82 1.2 1.7 2.4 3.2 4.6 6 8 10 12
25 40 0.24 0.32 0.46 0.64 0.90 1.3 1.8 2.6 3.6 5.0 7 9 11 14
40 63 0.26 0.36 0.50 0.70 1.0 1.4 2.0 2.8 4.0 5.6 8 10 12 16
63 100 0.28 0.40 0.56 0.78 1.1 1.6 2.2 3.2 4.4 6 9 11 14 18
100 160 0.30 0.44 0.62 0.88 1.2 1.8 2.5 3.6 5.0 7 10 12 16 20
160 250 0.34 0.50 0.70 1.0 1.4 2.0 2.8 4.0 5.6 8 11 14 18 22
250 400 0.40 0.56 0.78 1.1 1.6 2.2 3.2 4.4 6.2 9 12 16 20 25
400 630 0.64 0.90 1.2 1.8 2.6 3.6 5 7 10 14 18 22 28
630 1000 1.0 1.4 2.0 2.8 4.0 6 8 11 16 20 25 32
1000 1600 1.6 2.2 3.2 4.6 7 9 13 18 23 29 37
1600 2500 2.6 3.8 5.4 8 10 15 21 26 33 42
2500 4000 4.4 6.2 9 12 17 24 30 38 49
4000 6300 7.0 10 14 20 28 35 44 56
6300 10000 11 16 23 32 40 50 64
Table 3.5: Permissible linear deviations for linear dimensions for casting processes [21]

Which tolerance grade is achievable varies from one company to another. However, the
standard recommends a tolerance grade of CT 5-7 for pressure die casting of light alloys,
which was taken as a process comparable with rheocasting. On the other hand, for investment
casting light alloys, which was taken as a process comparable with plaster casting, a tolerance
grade of CT 4-6, was recommended.

Assuming that a tolerance grade of 5 (CT5) can be achieved by both processes; this meant
there would be a maximum deviation of 0.9 mm on all dimensions. This value was assumed
to be acceptable, as it was believed it would not impair the function of the chassis.

With regards to the surface tolerance (point b), again the requirement from the machined
chassis was taken as a bench mark, which was stated to be Ra 3.2 µm. In addition, from the
table (3.1) showing list of processes and their process capabilities extracted here in table (3.6),
it can be seen that both process are capable of achieving the required surface roughness of Ra
3.2 µm.

Process Roughness (µ𝐦)

High pressure die casting 0,8 – 1,6
Plaster mold casting 1,6 – 3,2

Table 3.6: Roughness ranges that can be achieved by selected processes

It should again be noted that rheocasting was assumed to be comparable to pressure die
casting.

Finally, with regards to the intricate features (point c); there were the holes in the box as well
as cuts in fins that were assumed to be best machined after casting. Some holes were also

53
 Groundwork

mentioned to have threads. Machining of these features was assumed to be best, if done after
performing the heat treatment process.

The positions of these holes and materials that needed to be cut out were located on a drawing
and the specific dimensions provided. However, this was not included in this thesis, but the
location of the three datum lines used for locating features of the chassis can be seen in figure
(3.26).

Figure 3.26: Datum planes for machining

The two datum lines shown to the right of figure (3.26) are given at the position where the
inside walls meet the respective floors of the compartments. Provision of datum line at such
mid planes was necessary to half any errors in dimension on both sides. However, the datum
line on the let was decided best to be at the flat top of the flange.

3.8.3 Coating

Surface coating is naturally done to prolong the life of components in service or storage. The
type of coating chosen for the original (wrought) chassis was chemical conversion coating.
This process can also be used on the cast alloys and thus was also chosen for the new chassis.
Thus, no new treatment was proposed.

The process involves producing an insoluble aluminum compound on the surface of a part by
a chemical reaction occurring with an aqueous solution containing chrome.

The process usually consists of three steps. These are:

 Preparation- here the surface of part to be coated is prepared so that it would accept
the subsequent reaction
 Coating- here the part is either immersed, sprayed or brushed for inducing reaction
 Washing and drying- here the part is cleaned

This process gives parts a good corrosion resistance, especially in humid conditions and could
also serve as a base for any subsequent painting.

54
 Groundwork

3.8.4 Testing & Inspection

Testing of casted components can be done at different production steps as well as using
different methods. These procedures should not be redundant as well as be excessive with
regards to the additional cost they would entail.

Basically the methods can be grouped into three categories. These are:

 Methods for determining surface quality

 Methods for detecting internal discontinuities
 Methods for dimensional inspection

Generally speaking, the tasks related to inspection and testing can be done on a sample basis
or be exhaustive. It could also be a destructive or nondestructive test.

In relation to the new chassis what was proposed is a preliminary visual inspection of all
castings in the as-cast state before chopping off the gating system. These would help in
avoiding unnecessary subsequent processes like machining. The chassis passing this
inspection would have their gating system chopped off and heat treated. Next, all heat treated
parts can be checked for dimensional accuracy, by making use of a template geometry to
roughly check if the compartments have the necessary dimensions to accommodate the
electronic units that would be assembled subsequently. The chassis that pass this stage would
be machined to create the various holes on the chassis. After that, a liquid penetrant test can
be performed on all machined parts, to check for surface and internal discontinuities.

55
 Results

4 Results
This section discusses the results of the simulations of the selected processes.

4.1 Mesh Density

Before discussing the various results attained from the simulations of the two casting
processes, the difficulties faced with the mesh density and its influence on the results attained,
should first be discussed.

As briefly outlined earlier, a total number of mesh elements ranging between 10 to 12 million
was used for rheocasting while around 15 million elements was used for plaster casting. This
gave a total metal cell (cells that make up the casting – which included the gating system) of
about 900,000 on average. MAGMAsoft, has three mesh elements as a preferred minimum
across a section. This is especially true for inlets, as was mentioned earlier, while for other
components two elements across a section can be acceptable. Of course, large number of
elements come at the expense of computation time or may not even be computable, depending
on the capacity of the computer. The computers used had an average of 2 GB RAM and 3.4
GHz of processer speed. Simulations of the rheocasting process took a duration of 10 to 15
days. The duration depended on the number of mesh elements as well as the velocity at the
ingate. Simulations of versions having faster ingate velocity took shorter time to compute. On
the other hand, simulations of the plaster casting process, took an average of three days,
irrespective of velocity at inlet, for about the same number of metal cells. In both cases,
processes simulations were mostly done to filling and solidification sub-processes.

The number of mesh element does not only influence computation time but the accuracy of
the results too, as the geometry would not be entirely represented by the mesh elements (see
figure 4.1). This was especially the case for curved features of the geometry. In addition,
other features like fillet radius and draft angels complicate the full representation of the
geometry. It should be remembered that the mesh elements are hexahedral in shape.

Figure 4.1: Complications due to meshing curves and tapers

56
 Results

Figure (4.1) shows that, due to the taper on the walls (especially the outer walls) as well as
fillet radii at the four corners, attaining a minimum of two elements across a wall was hard to
attain. Moreover, attaining two elements across a section did not necessarily ensure that the
melt always had, a two-element-thick channel guiding its flow. The melt was seen to be
constrained at curve and taper transitions. This restriction can be witnessed from the figure
(4.1), which shows how a flow that had three or two element thick channel can be constrained
at different steps of its passage. Thus, to fully comprehend the simulation results, this
complications faced by the meshing should be taken into account.

4.2 Rheocasting results

In chapter two, description of entrainment defects was provided and emphasis given to the
importance of achieving a smooth filling- as this was assumed to improve the probability of
these defects not developing during pouring. Smooth filling will in turn avoid defects
associated with solidification from forming.

In chapter three, the various changes made on the chassis‟ geometry as well as decisions made
to design the gating systems of both processes were outlined. There too, all decisions were
made with a smooth filling in mind as well as some solidification defects that might occur.

In the next two subsections, simulation results for the filling and solidification of the
rheocasting process is discussed. The results for plaster casting will then be discussed.

4.2.1 Filling results

The layout first decided on had the arrangement shown in figure (4.2). It was assumed that the
slurry would enter at the flange of the chassis, and then go down the side wall and turn to fill
the fins one by one – as the slurry expands to reach the other end. The slurry was desired to
advance equally across the chassis, as it moved from one end to the other. Moreover, a fin
was desired to fill to its extremity as soon as the slurry gets access to it, as opposed to filling
under pressure that occurs after most of the slurry passes the fins to fill the box. This was not
assumed to be easy as the slurry was naturally expected to flow through a passage where it
finds least resistance. In other words, the fins, which were shown to have very thin tips, would
exert more resistance to the slurry and thus filling them the desired way would be a hard task.

Having the layout shown in figure (4.2), different ingate areas were first tried, to see how the
fillings vary with flow rate. The difference was decided to be best achieved by varying the
thickness of the ingate, as this would lead the way to evaluating the recommended ingate
velocities shown in table (3.2).

An ingate thickness of 4 mm (version V6-v1) and 6 mm (version V9-v2) were first tried.
Beyond its effects on filling and solidification, which will be discussed next, a thick gate
demands a greater cutting power for its subsequent separation from the final casting. Both
versions were simulated with an ingate velocity of ~5m/s. Comparing this 5m/s velocity at
ingate with the recommended values given in table (3.2), it could be observed that the gate
with 4 mm thickness had an ingate velocity equal to the recommended value of 5 m/s. On the
other hand, the thicker gate, having a thickness of 6 mm and a recommended velocity at
ingate of 4 m/s, was filled with a velocity 1.5 times that of the recommended value. For a
more detailed comparison, some parameters of these two versions can be seen in appendix
(8.6).

57
 Results

The difference in the filling pattern of these two versions can be seen from the respective
screen shots of filling velocity, sampled in figures (4.2) and (4.3). The pictures are of eight
shots for each version, taken at 35%, 55%, 75% and 95% of cavity filling. One version has
four shots of the fins and four of the box at the same percentages of filling.

Figure 4.2: V6-v1, 4mm gate at the recommended gate velocity of 5 m/s

58
 Results

Figure 4.3: V9-v2, 6mm gate shot at a higher gate velocity of 5m/s (recommended 4m/s)

59
 Results

The percentage filled is calculated by adding the volume of the chassis as well as the gating
system. Thus, strictly speaking, the filling percentages of one version cannot be compare with
that of the other, as the volume of the gates vary slightly. However, comparison was done all
the same, neglecting this small distinction.

As can be witnessed from figures (4.2) and (4.3), both fillings were not good, as they had
created air pockets which, upon subsequent filling of the slurry, would definitely create
bifilms (as a form of confluence welds). For more clarity, these areas are circled in the figures
for the two versions. The areas were thought to be potential defect areas for such a cast
because the process has a relatively low temperature and the slurry at this pocket areas had a
longer contact time with the air that they enclosed – which had no place to escape. Thus, the
subsequently forming confluence welds might not have their full strength. What was more
important was that these confluence welds were formed on the outer walls of the chassis. The
chassis needed to be leak proof and such weld areas on the outer walls, might be possible
areas for forming discontinuities that create leakage problems. Thus, attempt was made to
avoid this air enclosing and would be discussed in here. On the other hand, the joining of
slurry at the inner walls, were assumed to be tolerable because, if it even forms a discontinuity
there, it would not result in leakage problems. Nevertheless, attempt was made to prevent this
as well.

It was also mentioned that the fins were desired to fill up to their extremities as early as the
slurry got access to them. However, from figures (4.2) and (4.3) slurry was seen to pass the
fins and fill the box readily, instead of completely filling a fin that it entered. The slurry, did
of course fill all extremities finally, due to pressure from the plunger, but this was desired to
happen sooner. This was not considered as critical as the enclosing of air mentioned earlier,
since it would result in no major defects, other than the probable chipping at fin tips. This
probable chipping at fin tips will be discussed towards the end of this subsection.

Nevertheless, when comparing the results of the two versions, it can be seen that the version
with the thick gate (V9-v2) had a velocity a bit larger than the recommended value by
Rheometal (table 3.2), but its filling pattern was slightly better, especially when looking at the
air pockets described above and the depth of fins filled at various steps. This might be hard to
see from the figures (4.2) and (4.3), but was still evident.

Of course for both cases, the filling would have been better if a large mesh density had been
used, but the relative filling patterns of the two versions was assumed to be similar and further
simulation was carried on the thicker type gate.

This thicker-gated version was next studied by varying the ingate velocities. It was believed
that the recommended velocity values at ingate, shown in table (3.2), can be increased to
avoid the air pockets. This increase of velocity was assumed to have no downside, as the
chassis was too thin and the slurry too viscous for it to disintegrate. Thus, two new variants of
simulations were carried out on the thicker-gated version. These two versions had ingate
velocities of 8 m/s (V9-v3) and 15 m/s (V9-v1) respectively. It should be remembered that the
two new versions had a 6 mm thick ingate and a recommended velocity of 4 m/s. This meant
that the versions had a velocity of double and 3.75 times that of the recommended value. The
varying parameters between these versions can also be compared from the table in appendix
(8.6). The results of these two new versions are shown in figures (4.4) and (4.5).

60
 Results

Figure 4.4: V9-v3, 6mm gate shot at a higher gate velocity of 8m/s (recommended 4m/s)

61
 Results

Figure 4.5: V9-v1, 6mm gate shot at a higher gate velocity of 15m/s (recommended 4m/s)

62
 Results

It can be witnessed from the above images that the version with the highest velocity (V9-v1)
had a better filling pattern as compared with the others. This was the case with regards to both
the air pockets and filling of fins. Moreover, since the slurry being filled was semi-solid, it
showed no tendency to separate or splash even at such high velocities. Splashes would surely
have occurred if a conventional melt was high pressure die casted to shape the chassis.
Unfortunately, in all versions the slurry was seen to stay behind, especially when it turned the
corners, at the lower part of the side walls. This staying behind of slurry can be seen in figure
(4.6).

Figure 4.6: Slurry staying behind at turn of corners at 55% fill (V9-v1)

This was believed to be due to the mesh problem discussed earlier (see figure 4.1), that
showed how slurry can face constrictions at curved features. In addition, two other pictures
shown in figure (4.7), point out two incidents where air enclosing were noticed, at 86 and 94
percentage of filling. Both pockets were found at the outer walls, which was mentioned to be
an unsought filling characteristic. This was assumed to be a direct consequence of the staying
back of slurry, which in turn was assumed to be due to a low mesh density.

Figure 4.7: V9-v1, 86% filled (left) and 94% filled (right)

It was thus believed that increasing the mesh density would address these mishaps of slurry
staying behind and air pockets shown in figure (4.6) and (4.7) respectively. Accordingly, the
mesh density of (V9-v1) was increased to 12,000,000 (which resulted in 1,200,000 metal
cells) and the difference in result studied. This version with a higher mesh density, V9-v5,
showed the slurry to fill the chassis smoothly and advancing equally across the chassis, with
the slurry in the side flange exceeding the slurry in the side wall by a small margin. Moreover,
the pockets showed in figure (4.7) were not observed. Comparison of the slurry staying
behind of figure (4.6) can be made with the new version‟s shown at the same 55% of fill, in
figure (4.8). The formation of pockets in figure (4.7) can also be compared with that of the
new versions‟ shown at the same 86% and 94% percentages of fill, in figure (4.9).

63
 Results

Figure 4.8: Staying behind of slurry at high mesh density at 55% fill (V9-v5)

Figure 4.9: V9-v5, No air pocket formation for the high density mesh at 86% and 94% fill

From the above two figures it can be seen that increasing the mesh density mitigated the two
mishaps, slurry staying behind and air pocket formation, witnessed on the previous version.
The slurry is still seen to advance quickly in the top flange (see figure 4.8). However, this was
a better filling as melt was seen to advance relatively equally across the chassis. Moreover, no
air pockets were seen to form at the walls. However, some air pockets were noticed near the
tip of the fins and are shown in figure (4.10). This was relatively less critical as it is less likely
to cause leakage problems.

Figure 4.10: Pockets near fin tips (V9-v5 high mesh density)

64
 Results

This chosen value of high velocity (~15 m/s) as well as the same gate (6 mm thick gate) were
also used to evaluate its results on a two other different layouts of the gating system. These
two new layouts can be seen in figures (4.11) and (4.12). The mesh size used for the new
layouts was the same as V9-v1 (10,000,000 elements). The parameters of this version can also
be compared from the table shown in appendix (8.6).

Eight filling velocity pictures are sampled for the new layouts, termed Layout two (V11-v1)
and Layout three (V11-v2). The results can be seen in figures (4.11) and (4.12).

Figure 4.11: V11-v1, layout two

65
 Results

Figure 4.12: V11-v2, Layout Three

In figures (4.11) and (4.12), both new layouts, were seen to fill the chassis with the slurry
advancing equally across the chassis. The fins were also seen to more or less fill to their tip,
as soon as the slurry got access to them. Both layouts did not show any enclosing of air at the
outer walls of the chassis. However, layout three (V11-v2) was noticed to have a confluence
weld forming at the inside wall at 40 percentage of filling. This might be a potential area for a
discontinuity of slurry, but even if such a discontinuity occurs it would be less critical as it
was formed at the inside wall. On the contrary V11-v1 did not have such a confluence weld.

66
 Results

This could be attributed to the reason that the slurry had a longer way to travel to reach the
junction of the outside wall with inside wall, and had lost energy due to friction and
developed enough back pressure that the slurry filled the inside wall from the floor of the
chassis. These incidents can be seen for both versions in figure (4.13).

V11-v1 V11-v2

Figure 4.13: Filling of inside walls in Layouts two and three

In all the last three simulations (figures 4.5, 4.11 and 4.12), it was observed that such a casting
can be filled smoothly and be done so at higher than recommended velocity values. Moreover,
the slurry was not seen to disintegrate when being filled at such high velocities. This was
possible since the chassis was thin walled and the slurry viscous for making a high velocity
shot possible without splashes or sprouting. On the other hand, the simulations having slower
or equal velocity values with the recommended ones did not have the energy to advance
across the chassis equally and the slurry was seen to stop at various positions in the cavity.
This retarding or stopping of flow would result in areas of slurry with slightly lower
temperature and thus was not desired for the subsequent joining of slurry – as it would be an
area susceptible to defect. With regards to the undesired incidents in filling, like the enclosing
of air and confluence welds discussed earlier, version V9-v1 was seen to have two areas
where this happens, although a finer mesh, version V9-v5, showed this wouldn‟t occur.
Version V11-v1 was seen to fill the chassis without such happenings while V11-v2 had a
confluence weld at an inside wall which was considered to be not critical as this was just a
possible but not certain area for a discontinuity to form. Plus, even if a discontinuity forms,
leakage would not be a problem, as it would form in an inside wall. As for the filling of the
fins, all versions somewhat showed the slurry passing to fill the box before completely
reaching a fin‟s tip that it started to fill. These fin tips were observed to fill completely at the
end of fill, however, it might result in some exposed cavities, of various sizes, isolated or
grouped at the tip of the fins. What these cavities might look like can be seen in the figure
(4.14). Comparing the filling of the three layouts, layout one (V9-v5) showed the best filling.
However, a simulation with a finer mesh was not done for layout two and layout three.

Figure 4.14: Probable cavities at top of fins for rheocasting

A test production was assumed necessary to ascertain if these cavities would form. Again, the
problem with the mesh density, on the results of the simulations, cannot be stressed enough.

67
 Results

4.2.2 Solidification results

All filling simulations discussed in the previous section had their solidification process also
simulated. In all the simulations, at the end of filling the chassis had a maximum temperature
of 590ºC and a minimum of around 550ºC, with solid fraction value in between 50% and
56%. This was welcomed, as having a slurry at a higher solid fraction, before filling was
complete, might result in cold-shut or mis-run formations. The direction of solidification of
all simulations that had the first layout, were more or less similar. The thinnest parts of the
chassis like the fins and the side walls solidified fast while the T-junction where the two inner
walls meet was last to solidify. One version (V9-v1) is shown in figure (4.15) as an example
to see this solidification pattern. The pictures were at 25%, 50%, 75% and 100%
solidification.

Figure 4.15: V9-v1, solidification

The other layouts, Layout two (V11-v1) and Layout three (V11-v2), had a more or less
similar solidification direction, except with expected variations that resulted from the position
of the gate (hot slurry). Both versions are shown in figure (4.16) at 100% solidification.

Figure 4.16: V11-v1 (left) and V11-v2 (right)

68
 Results

As for porosity formation in the chassis, MAGMAthixo was not able to calculate porosities
and showed zero porosity for rheocasting. This does not mean that the process would not have
any porosity. The customary way to identify areas where shrinkage porosities might form is to
either look at the hotspot results or the feeding criteria results (FSTIME_69) which will be
discussed next.

The hotspot result is one way of identifying areas where shrinkage porosities might likely
form. The results for the three versions, (again V9-v1, V11-v1, V11-v2), is shown in figures
(4.17) and (4.18).

Figure 4.17: Hotspot V9-v1

Figure 4.18: Hotspot, V11-v1 (left) and V11-v2 (right)

From the above figures it was seen that all the three versions had more or less similar areas of
hotspots, which was expected. These pictures could be one indication for possible porosity
areas. On the other hand, the feeding criteria results (FSTIME_69), that would be discussed
later, could also be taken as another indication for such areas.

The two gates having 4 mm (V6-v1) and 6 mm (V9-v2), that were compared for filling results
on the previous section were also compared for their feeding performance during
solidification. The other thing to be mentioned about the gates is that the thicker one had a
shorter ingate length while the thinner one had a longer one, as can be seen in figure (4.19).

69
 Results

6mm
4mm

Figure 4.19: The two types of gate used thin (V6-v1) and thick (V9-v1, V11-v1 & V11-v2)

It should be remembered that, when discussing the setup of the rheocasting simulations
(chapter 3), it was mentioned that a feeding efficiency of 70% was set. This means shrinkage
was assumed to be fed macroscopically until the fraction of solid in the casting was 70%.
Accordingly, the difference with regards to feeding was compared for the two types of gates,
by looking at the areas of the slurry that did not attain this local 70% fraction of solid criteria.

The result shown in figure (4.20) is the FSTIME_69, which indicated the time left, in
seconds, that the slurry had before it attained the solid fraction criteria of 70% - after which
feeding through an element will be blocked. A section view of both type of gates is seen in
figure (4.20).

Figure 4.20: FSTIME_69, V6-v1 (left) and V9-v1 (right)

It was seen from both pictures (figure 4.20) that the spectrum was continuous from entrance at
gate to its exit, with the hot part in the middle. This implied that, as long as there is a contact
between the slurry that had not reached the 70% criteria in the chassis and the hot slurry in the
gate, then feeding will continue. Figure 4.21 shows how this feeding was achieved.

Figure 4.21: FSTIME_69, thin gate V6-v1(left) and thick gate V9-v1(right)

70
 Results

In both versions (figure 4.21) it was noticed that feeding cannot be done through the outer
walls, as the walls had already passed the 70% solid fraction criteria. However, a section view
of both versions can show that, feeding can occur through the gate‟s contact with the inner
walls. Such a section view along planes parallel to both the XZ and YZ planes is shown in
figure (4.22).

Figure 4.22: FSTIME_69 ,feeding path for V6-v1 (left) and V9-v1 (right)

Further increasing the solidification time, both versions were compared at a time when the
volume of slurry that has not passed the 70% criteria was almost equal. At this time the
thicker version was seen to still have contact with the hot slurry in the chassis and the gate
while the thinner version lost its contact. This is shown in figure (4.23).

Can't feed Can feed

Figure 4.23 FSTIME_69, V6-v1 (left) and V9-v1 (right)

Thus, the thicker version of the gate was taken to be a relatively better choice. However, both
versions had an almost equal volume of slurry blocked from being fed by hot slurry in the
gate. This is shown in figure (4.24) and was mentioned to be another way of identifying
potential areas for porosity formation.

71
 Results

V9-v1

Figure 4.24: FSTIME_69, Slurry blocked from being fed V9-v1

Similarly the FSTIME_69 result of the other two layouts, layout two and Layout three, can be
seen in figure (4.25).

Figure 4.25: FSTIME_69, V11-v1 (left) and V11-v2 (right)

When comparing the feeding of these last two layouts V11-v1 and V11-v2 with V9-v1, it was
observed that V11-v1 had a slightly larger volume of slurry blocked from being fed while
V11-v2 had a slightly lesser volume as compared to V9-v1 (shown in figure 4.24). This could
be explained by looking at the position of the gate in each case and their position of the last
slurry needing to be fed. In V11-v1, since the gate was far from the inside walls, the hot slurry
in the gate lost its path to the last solidifying slurry quickly, while for V11-v2 the gate had a
closer contact with an inner wall through which feeding was done for a longer time. Not only
did the gate of V11-v2 had access to an inner wall, but also this inner wall was at the center of
the gate as opposed to an offset position for V9-v1, which made it better at feeding.

It was mentioned that looking at the slurry remaining from being fed, could serve as an
indication for possible areas for porosity formation. Accordingly, figure (4.24), which showed
the last slurry that needed feeding for version V9-v1, could be taken as one possible
indication for assuming areas where porosity might form. This method was actually assumed
a better indication than the hotspot results discussed earlier.

72
 Results

Rheocasting was described to be advantageous as the slurry is already part solid even at start
of pouring and that this solid fraction of the slurry would continue to increase during pouring.
Thus, the subsequent shrinkage is without any liquid shrinkage, and thus has a lesser
solidification range than conventional casting methods, ridding it from most problems
associated with liquid or solidification shrinkage. This is one reason, that with regards to
porosity, rheocasting was said to be a “no-porosity” process. However, as the software was
not able to calculate porosity results, the results from the feeding criteria or hotspots, were
taken as possible indication for areas where porosities might form. From both these indicating
results, no significant porosity can be expected at the outer walls. This means that there would
be no defect related to porosity that can have the potential to cause leakage problems.

73
 Results

4.3 Plaster casting results

In the next two subsections the filling and solidification simulation results for plaster mold
casting will be discussed.

4.3.1 Filling results

In order to have a smooth filling of the casting cavity, various layouts and gating system
geometries were simulated. The guide for calculating the gating systems, to preferably have
the melt below the critical velocity value at entrance to chassis, was briefly outlined in chapter
three. With regards to the number of ingates, two type of gating systems were used. These
were mentioned to be a single-gated version and a multiple-gated version.

A single gate means that melt will not separate at an early stage of filling, to subsequently join
at a later stage, and thus has the advantage of not having defects related to confluence
welding. However, it requires a gate that has a wide area which might be susceptible to
hotspot related defects. On the other hand a multiple gate has numerous gates that have small
areas. These gates will not have problems associated with hotspots as they take short time to
solidify.

The respective layouts of these versions are shown in figure (4.26) and (4.27). The steps and
guidelines used to design the gating system were outlined in subsection (3.6.1). However, the
single-gated version can be seen to have a pouring cup receiving the melt while the multiple-
gated version had the sleeve. It was assumed that this variation in receiving the melt will not
affect the reason one version was merited over the other.

Two pictures from each versions is given in figures (4.26) and (4.27), sampled at different
percentage of fill.

Figure 4.26: Single gated version (Plaster_Iteration7_V2)

74
 Results

Figure 4.27: Multiple gated version (PF4)

From the two figures, it was observed that the single-gated version had the melt not equally
rising up all-round the chassis. Instead, the melt was seen to rise rapidly through the side wall
near the gate, to later curve at the top and drop down the adjacent side walls to meet the
portion of melt rising up the side. As the fall height from the top of the chassis through the
side walls was much greater than necessary to accelerate the melt beyond the critical velocity
of 0.5m/s, the melt was seen to disintegrate and spray while dropping to meet the melt at the
sides. This was an undesired incident, as it would be a mine for generating bifilms as well as
being a potential area for discontinuities to occur in the chassis. On the other hand, the
multiple gated version was observed to have the melt rise relatively equally all-round the
chassis. The melt arriving through the different gates met at the flange. This meeting of melts
will most likely form bifilms which will be entrained and end up in the chassis. As compared
to the single gated version, this was assumed to be less detrimental. Thus, most simulations
performed on this method of casting were of the multiple-gated layout.

Another observation made on almost all plaster casting simulations was that the melt front
appears to form a minor spraying as it rises the walls and curve to join the plane above (see
figure 4.26).

Figure 4.28: Apparent spraying as melt bends at top

75
 Results

This was assumed to be the result of having a low mesh density, as the melt requires a smooth
radius to take a turn. However, as described in section (4.1) of this chapter, fillet radii were
not fully represented by the created mesh elements (see figure 4.1).

In trying to have a smooth flow of the chosen gating system (multiple gates), attempt was
made to reduce splashes and sprouts that were observed in the first trial (figure 4.27). One
such defect generating area was at the entrance and exit of the filters. The use of filters was
mentioned to be very important for the purpose of reducing velocity. Thus, by trying different
arrangements of filter position, attempt was made to minimize the turbulence created before
and after the filter. These different positions of filters can be seen in figure (4.29).

Filter arrangement-1 (PF3) Filter arrangement- 2 (PF4) Filter arrangement- 3(P_X_1)

Filter Filter Filter

Figure 4.29: different types of filter configurations

The results of these three positions are shown in the next three figures (4.30), (4.31) and
(4.32).

Figure 4.30: Result of filter arrangement 1 (PF3)

76
 Results

Figure 4.31: Result of filter arrangement 2 (PF4)

Figure 4.32: Result of filter arrangement 3 (P_X_1)

It was easily seen that the last layout had a smooth flow of melt before and after the filter
while the first two were seen to form lots of splashes. These splashes can surely cause lots of
bifilms that can easily find their way to the final casting. Other than the splashes, in the first
two models (4.30 and 4.31), the filters‟ purpose of reducing velocity was not fully achieved.
The last filter, however, was seen to have a flow that was less likely to generate bifilms with
the additional advantage of reducing the melt velocity as soon as the melt exits from it. When
the melt gets to the sprue exit, it had dropped 320 mm which gives it a theoretical velocity of
2.5m/s. However the melt was seen to exit the third type of filter with a velocity near the
critical velocity of 0.5m/s. Thus, the third filter arrangement was chosen (see figure 4.32).

77
 Results

Another comparison that was made was between different types of runner extensions that
were included to trap the first melt from entering the chassis. This first melt is best to avoid,
as it might have various defect forming entrainments. To this end, two types of runner
extensions were tried and can be seen in figure (4.33). The effectiveness of these extensions,
with regard to trapping first melt, can be compared from the Enter Timestep result shown in
figure (4.33). The result shows at what time step the respective particles entered at the
pouring point (entrance to sleeve).

Figure 4.33: Two types of runner extensions considered

In both cases, the runner extensions trapped the first melt to some extent. This particles
designating the first melt has a light blue color in figure (4.33). However, few particles were
noticed to circulate and find their way back into the cavity. This was not that much critical, as
the large percentage of the first melt stayed trapped.

The function of these runner extensions was not only to trap first melt but also to lessen the
effect of the melt‟s impact with the end of the runner channel. This impact would create a jerk
that would force the melt into the cavity with a higher velocity and thus create sprouting of
melt at entrance. With regards to this incident, the impacts of the two extensions shown above
were also compared.

Figure 4.34: Melt entrance with second type of runner extension (P_X_1_v2)

78
 Results

It was observed that both types of runner extensions had a more or less similar effect on the
impact of melt. One example is shown in figure (4.34) to see the entrance of melt. However,
the preferred type of runner extension was the one shown in figure (4.34), since it was
assumed that the rising melt in the vertical extension, creates a gradual increase in back
pressure and thus gradually decrease the velocity of the melt entering the chassis. This
gradual increase in head pressure enabled the melt to enter the cavity more smoothly.

In addition the influence of tilting the cavity was also investigated. Tilting is done to
decelerate the melt in the runner as it flows against gravity. The other advantage of tilting is to
avoid defects that may occur due to circulation of melt at the top of the chassis. Circulation is
common in products that have a wide and flat horizontal plane. A tilting of about 1 degree
above the horizontal gave a slightly better result (lesser circulation patterns at top), as
compared to a non-tilted casting.

4.3.2 Solidification results

The solidification results were more or less the same for all simulations. Thus, discussion will
be restricted to one version of the simulation.

All fillings were completed with no substantial decrease of temperature. As the melt was fully
liquid at the end of fill, no potential for having a mis-run or cold-run was observed. The
direction of solidification can be seen in figure (4.35). The side walls solidified fast while the
junction where the inside walls meet solidified last.

Figure 4.35: Temperature distribution at 837.68 seconds

Related to the solidification, the hotspot areas and porosity results can be seen in figures
(4.36) and (4.37).

79
 Results

Figure 4.36: Hotspot results, shows junction of inside walls as areas last to solidify
The location of the hotspots in figure (4.36) was expected as the inside walls, which were
relatively thicker, had a T-junction there and would thus, solidify last. This could be taken as
an indication of a possible area for shrinkage porosities to form.

The porosity result is shown in figure (4.37). It mainly shows the tip of the fins as areas that
need to be fed as cavities were formed forms due to liquid shrinkage. This cavity at tip can
also be noticed in figure (4.35).

Figure 4.37: Porosity result

Usually such cavities that form in castings due to shrinkage are fed by provision of feeders.
However, putting such a reservoir of hot melt at the top of the chassis was not an option as the
geometry of the fins complicate the location of a feeder. Such a feeder was assumed to have a
negative effect as it might lead to warpage and deformation of the fins as well as the chassis
as a whole. This clearly complicated locating a thick reservoir of melt to feed the porosities.
In addition, feeding of such thin walled casting is not advised and is naturally believed to be
counterproductive [10].

80
 Results

It was thought better to feed the casting through the gates. Melt from the gating system was
feeding the chassis through the gates. This can be seen in figure (4.38).

Figure 4.38: Solidification stages at 8.02 (left) and 897.04 (right) seconds

Figure 4.38 shows the level of melt at two stages during the solidification of the chassis. The
one on the left is an image at the beginning of solidification. The melt filled the chassis up to
the top of the sprue and the runner extensions. The one on the right shows melt at about the
end of solidification. In this image the melt in the sprue as well as the runner extension is seen
to have a decreased height. This was due to the feeding of shrinkage through the gates.
However, the feeding was not adequate as the top of the fins are also seen to have cavities due
to shrinkage (see the right image in figure 4.38).

It should be remembered that the feeding criterion was set to be 30% for plaster casting. This
means that feeding can take place through the dendrites from hot melt in the gates that had not
passed this feeding criterion. After the melt has passed this criterion feeding through an
element will be blocked.

The cavities noticed at the fin tips were assumed to be due to insufficient hot melt in the
sleeve. In the simulation of the process filling was be done up to the top of the sprue.
However, in the physical process the melt will also fill the sleeve to a certain height. This was
assumed to be sufficient for feeding the chassis and was thus simulated. The result of this new
simulation (P_X_1_v4) can be seen in figure (4.39).

Figure 4.39: Solidification stages of new version at 8.02 (left) and 887.08 (right) seconds

Figure 4.39 shows a 50 mm height of melt in the sleeve. This created enough head pressure to
feed the cavities that would have formed due to liquid shrinkage. As was assumed there were
no cavities seen at the tip of the fins.

81
 Results

This feeding through gates occurs upto the feeding criterion mentioned earlier. After that,
for such thin castings, shrinkage is assmed to be fed by solid contraction [10]. This means
that, outside walls might move inwards resulting in an offset of 0.5 percent per face. For a
3 mm thick walled part like the chassis, this offset can be translated to a worst of 15 µm
surface sink at a face. This small offset was neglected and believed that it would not
impair the function of the chassis.

However, one thing was noticed when observing the FSTIME_30 result. This result can be
seen in figure (4.40).

Figure 4.40: Feeding through gates seen on the FSTIME30 result

Figure 4.40 shows that longer feeding through gates will happen through the middle gates
found on both sides. It was thus assumed that the location of these middle gates can be offset
to have the inner walls at their middle and thus get a slightly prolonged feeding.

The porosity for this new version is seen in figure (4.41). The porosity areas seen in this new
version are not at the tip of the fins like the previous versions‟ (see figure 4.37). Instead the
porosity areas were seen to decrease and were at the junction of the fins with the plane at top.

Figure 4.41:Porosity result of new version

82
 Results

Another feature considered is the way melt is received at pouring. As emphasized earlier, the
greatest care that should be taken to ensure a defect free casting is to try to minimize
formation of entrainment defects during pouring. If too many bifilms, bubbles or bubble trails
are generated at pour and find their way into the casting, they will be suitable areas for surface
discontinuities and cracks to form. To this end, the effect of using a pouring cup or a sleeve
was investigated.

The simulations that had a sleeve receiving the melt had a less turbulent flow after the exit of
the sleeve. However, the fall of melt in the sleeve was seen to be a potential area for
generating defect. As the fall in the sleeve was large (about 200 mm), it would accelerate the
melt to a velocity of 1.98m/s. This is about four times the critical velocity and thus
disintegration of the melt was eminent. This was especially significant since the sleeve had a
wide cross-sectional area and was unable to contain the melt close-bound. This disintegration
of melt will create entrainment defects. This fall through the sleeve and the disintegration of
melt can be seen in figure (4.42).

Figure 4.42: Melt falling in sleeve

Accordingly, the use of a pouring cup, integral to the mold (or not) was believed to give a
better result as its use would reduce the velocity of the melt at an intermediate fall height and
then directing flow to the sprue, where it is easily contained by the walls. Thus, melt would be
kept from disintegrating and the flow in the sprue will be full (no air falling alongside melt).
Melt was assumed to fill the cavity by pushing out the air in front. Such a layout was modeled
(this changed the dimensions in the gating system) and filling simulated. The considerations
taken to design this gating system was similar to the steps discussed earlier with change in fall
heights as the drop height of the pouring cup is lesser. This change in height changed the
cross-section of the sprue, runner and gates as per the relations discussed in sub-section
(3.6.1).

83
 Results

Figure 4.43: Pouring by use of a pouring cup

From the figure above (4.43), it could be seen that the melt rests at an intermediate height
(less than the sleeve height) and was directed into the sprue – resulting in a full melt flow in
sprue, with no air pockets, and thus avoiding entrainment defects. This was thus assumed to
be better than the direct pouring to sleeve that was discussed to be frequented in plaster mold
casting.

Another thing that was discussed earlier was that plaster molds are less permeable to air. This
might make the filling of the fins defect prone, as the air in the cavity would have no way to
escape when melt is poured in. It was also mentioned that one way to mitigate this in plaster
castings is by vacuum assisting the filling process. This is done by drilling a hole to the cavity
and sucking air out along channels that are provided to the mold while filling is done.

84
 Cost estimation

5 Cost estimation
This chapter gives a rough cost estimation of the two processes.

In this chapter weight and mass are used interchangeably and both terms are used to
designate the mass of a body in kg.

5.1 Total component cost

As the motivation for looking into casting as a possible method of producing the chassis was
the reduction of production cost. A rough cost estimation of the process must be provided.
Total casting cost can be calculated by the equation [19].

𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 = 𝐶𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 + 𝐶𝑙𝑎𝑏𝑜𝑢𝑟 + 𝐶𝑒𝑛𝑒𝑟𝑔𝑦 + 𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 + 𝐶𝑜𝑣𝑒𝑟 𝑕𝑒𝑎𝑑𝑠 (5.1)

Where the above variables mean costs for casting, material, labor, energy, tooling and
overheads, as per their respective subscript designation.

Equation (5.1) is used for calculating manufacturing costs only, while other aspects like
logistics, fixed cost as well costs for secondary processes like machining, coating and heat
treatment, are not looked at in this thesis. These five cost components would be discussed
next.

5.2 Material cost

The material cost is divided into two parts. These are the cost for direct and indirect materials.
Direct material cost is the cost incurred for buying the cast alloy while indirect materials cost
is for other essential materials that are required for production but are not meant to be
included in the final product. The latter includes costs for mold (for expendable type molds),
dispensable cores, insulating sleeves chills etc.

Thus,
𝐶𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 = 𝐶𝑑𝑖𝑟𝑒𝑐𝑡 + 𝐶𝑖𝑛𝑑𝑖𝑟𝑒𝑐𝑡 (5.2)

Where the direct cost, which must account for various losses that occur during production, is
given by equation (5.3).

𝐶𝑑𝑖𝑟𝑒𝑐𝑡 = 𝑐𝑢𝑛𝑖𝑡 _𝑎𝑙𝑙𝑜𝑦 ∗ 𝑤𝑐𝑎𝑠𝑡 ∗ 𝑓𝑚 ∗ 𝑓𝑝 ∗ 𝑓𝑓 (5.3)

Where the terms on the right side of the equation (5.2) stand for: cost per weight of an alloy,
weight of casting, factor for metal loss during melting (𝑓𝑚 ), factor for metal loss during
pouring (𝑓𝑝 ), factor for metal loss in fettling (𝑓𝑚 ), respectively. Fettling cost accounts for loss
when mold is separated from casting (especially for expendable molds) and when cutting
system is cut away.

 𝑓𝑚 – depending on the type of furnace used this factor usually rages between 1.05 –
1.12
 𝑓𝑝 – this factor usually ranges between 1.01 – 1.07
 𝑓𝑓 – this factor also usually ranges between 1.01 – 1.07

85
 Cost estimation

𝑤𝑐𝑎𝑠𝑡 is the weight of the casting given by the relation

𝑤𝑐𝑎𝑠𝑡 = 𝜌𝑐 ∗ 𝑉𝑐𝑎𝑠𝑡 (5.4)

Where the terms on the right in equation (5.4), stand for density of cast alloy (𝜌𝑐 ) and volume
of casting (𝑉𝑐𝑎𝑠𝑡 ) respectively.

Accordingly the direct material cost was assumed to be the same for both processes. So the
mass of the new chassis before machining, which was mentioned to be 4.5 kg, was taken as
the value for 𝑤𝑐𝑎𝑠𝑡 . In addition, values of 20 SEK/kg, 1.12, 1.07 and 1.07 were assumed
for 𝑐𝑢𝑛𝑖𝑡 _𝑎𝑙𝑙𝑜𝑦 , 𝑓𝑚 , 𝑓𝑝 and 𝑓𝑓 respectively. It should be noted that, although material loss might
be a bit larger in plaster casting, the highest values were taken for the three factors. This
resulted in a direct material cost of :

𝐶𝑑𝑖𝑟𝑒 𝑐𝑡 =4.5𝑘𝑔 ∗ 20 𝑆𝐸𝐾/𝑘𝑔 ∗ 1.12 ∗ 1.07 ∗ 1.07

= 115 𝑆𝐸𝐾 (𝑓𝑜𝑟 𝑏𝑜𝑡𝑕 𝑝𝑟𝑜𝑐𝑒𝑠𝑠𝑒𝑠)

While the cost for indirect materials is given by:

𝐶𝑖𝑛𝑑𝑖𝑟𝑒𝑐𝑡 = 𝐶𝑚𝑜𝑢𝑙𝑑 _𝑠𝑎𝑛𝑑 + 𝐶𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 + 𝐶𝑚𝑖𝑠𝑐𝑒𝑙𝑙𝑎𝑛𝑒𝑜𝑢𝑠 ….. (5.5)

Here the terms on the right of the equality sign stand form cost for the molding material (sand
or plaster), cost for core sand and other costs. The first two terms contribute to the major
share of indirect costs and as described are for expendable mold type casting processes. Thus
this indirect cost would only be used for plaster casting – as the miscellaneous cost was
neglected in this thesis.

The cost for sand is given by the relation

𝐶𝑚𝑜𝑢𝑙 𝑑 𝑠𝑎𝑛𝑑 = 𝑐𝑢𝑛𝑖 𝑡 𝑚𝑜𝑢𝑙 𝑑 ∗ 𝑓𝑟𝑒𝑐𝑦𝑐𝑙𝑒 ∗ 𝑓𝑟 ∗ 𝑓𝑚𝑜𝑢𝑙 𝑑 𝑟𝑒𝑗 ∗

𝑠𝑎𝑛𝑑

𝜌 𝑐 ∗𝑉𝑚 1
𝑟 𝑚𝑒𝑡𝑎𝑙 _𝑠𝑎𝑛𝑑
− 𝜌𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 ∗ 𝑉𝑐𝑜𝑟𝑒 ∗ 𝑛 … (5.6)
𝑐

Where the terms on the right stand for:

 𝑐𝑢𝑛𝑖𝑡 _𝑚𝑜𝑢𝑙𝑑 _𝑠𝑎𝑛𝑑 = cost of unit mold sand

 𝑓𝑟𝑒𝑐𝑦𝑐𝑙𝑒 = factor of recycled sand = 0.1 – 1.0
 𝑓𝑟 =factor of casting rejection= 1.00 – 1.12
 𝑓𝑚𝑜𝑢𝑙𝑑 _𝑟𝑒𝑗 = factor of mould rejection
 𝑉𝑚 = metal volume per mould
 𝑟𝑚𝑒𝑡𝑎𝑙 _𝑠𝑎𝑛𝑑 =metal to sand ratio
 𝜌𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 = density of core sand
 𝑉𝑐𝑜𝑟𝑒 =core volume
 𝑛𝑐 = number of cavities per mould

𝑉𝑚 is given by,

𝑉𝑚 = 𝑛𝑐 ∗ 𝑉𝑐𝑎𝑠𝑡 + 𝑉𝑔 + 𝑉𝑓 (5.7)

86
 Cost estimation

 Where 𝑉𝑔 and 𝑉𝑓 stand for volume of the entire gating system and volume of all
feeders per mould, respectively.

The cost for core sand, which was not used for this thesis (since no core material is required),
is given by the equation (5.8).

𝐶𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 = 𝑐𝑢𝑛𝑖𝑡 _𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 ∗ 𝑓𝑟 ∗ 𝑓𝑐𝑜𝑟𝑒 _𝑟𝑒𝑗 ∗ 𝜌𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 ∗ 𝑉𝑐𝑜𝑟𝑒 (5.8)
Where:
 𝑐𝑢𝑛𝑖𝑡 _𝑐𝑜𝑟𝑒 _𝑠𝑎𝑛𝑑 = cost of unit core sand
 𝑓𝑐𝑜𝑟𝑒 _𝑟𝑒𝑗 = factor of core rejection

With regard to 𝐶𝑖𝑛𝑑𝑖𝑟𝑒𝑐𝑡 - neglecting cost for miscellaneous, and also having no cores
necessary in this thesis, the indirect cost would only comprise of the cost for mold material
which is only applicable for plaster casting (for rheocasting mold material is considered as a
tooling cost).

Thus, for plaster casting process taking a unit mold cost of 15SEK/kg and values of 1, 1.12
and 1.1 for 𝑓𝑟𝑒𝑐𝑦𝑐𝑙𝑒 , 𝑓𝑟 and 𝑓𝑚𝑜𝑢 𝑙𝑑 _𝑟𝑒𝑗 the remaining terms in equation (5.6) can be simplified
to be the mass of mold material used. The mass of the mold was assumed to be around 70 kg
where the gypsum powder was assumed to account for 40% of this mass- the rest being water.
Thus, assuming a gypsum material of 70 kg for casting one chassis, the indirect material cost
was calculated to be:

𝐶𝑚𝑜𝑢𝑙𝑑 _𝑠𝑎𝑛𝑑 = 15𝑆𝐸𝐾/𝑘𝑔 ∗ 1 ∗ 1.12 ∗ 1.1 ∗ 70 ∗ 0.4

𝐶𝑚𝑜𝑢𝑙𝑑 _𝑠𝑎𝑛𝑑 ≅ 517.5𝑆𝐸𝐾

Obviously the above would significantly increase the cost of producing a unit chassis,
specifically since mold material was assumed to be not recycled.

87
 Cost estimation

5.3 Labor cost

Labor cost accounts for the costs incurred by labor on the various activities carried out in
production. It also account for the duration of these activities. It is given by the relation

𝑛
𝐶𝑙𝑎𝑏𝑜𝑢𝑟 = 𝑓𝑟 ∗ 𝑎𝑐𝑡 =1 𝑓𝑟𝑒𝑗 _𝑎𝑐𝑡 ∗ 𝑐𝑢𝑛𝑖𝑡 _𝑙𝑎𝑏𝑜𝑢𝑟 ∗ 𝑙𝑎𝑐𝑡 ∗ 𝑡𝑎𝑐𝑡 (5.9)

Where:
 𝑐𝑢𝑛𝑖𝑡 _𝑙𝑎𝑏𝑜𝑢𝑟 = cost of unit labour
 𝑙𝑎𝑐𝑡 = number of workers involved in activity 𝑖
 𝑡𝑎𝑐𝑡 = time for activity 𝑖 per component
 𝑛 = number of activities
 𝑓𝑟𝑒𝑗 _𝑎𝑐𝑡 = rejection factor for activity 𝑖

=𝑓𝑐𝑜𝑟𝑒 _𝑟𝑒𝑗 for core making activity = 1.00 – 1.20

= 𝑓𝑚𝑜𝑢𝑙𝑑 _𝑟𝑒𝑗 for mould making activity = 1.00 – 1.10
= 1 for other activities

Since the values used to calculate the labor cost were hard to attain, especially since no
company was keen to divulge such information, an average value of 1500 SEK/hr was
assumed for rheocasting and a value of 800 SEK/hr for plaster casting. This value was
assumed to account for labor, overhead as well as machine cost – thus is assumed to be equal
to production cost.

Thus for rheocasting assuming an average of 40 shots/hr the above gave a labor cost of:

𝑆𝐸𝐾 1
𝐶𝑙𝑎𝑏𝑜𝑟 = 1500 ∗
𝑕𝑟 40 𝑠𝑕𝑜𝑡𝑠
𝑕𝑟
𝐶𝑙𝑎𝑏𝑜𝑢𝑟 = 37.5𝑆𝐾

For plaster casting, since it is a very slow process a production rate of 1𝑢𝑛𝑖𝑡/𝑕𝑟 was
assumed, this gave a labor cost of:

𝑆𝐸𝐾
𝐶𝑙𝑎𝑏𝑜𝑢𝑟 = 800 ∗ 1𝑢𝑛𝑖𝑡/𝑕𝑟
𝑕𝑟
𝐶𝑙𝑎𝑏𝑜𝑢𝑟 = 800𝑆𝐸𝐾

88
 Cost estimation

5.4 Energy cost

The cost of energy has two parts. The first, termed as cost for melting energy, accounts for the
majority of cost incurred for energy. The second is termed as cost for other energy. The
relation for total cost of energy is given in equation (5.10).

𝐶𝑒𝑛𝑒𝑟 𝑔𝑦 = 𝐶𝑚𝑒𝑙𝑡𝑖𝑛𝑔 + 𝐶𝑜𝑡 𝑕𝑒𝑟 _𝑒𝑛𝑒𝑟𝑔𝑦 (5.10)

The cost for the melting energy is given in equation (5.11).

𝐶𝑚𝑒𝑙𝑡𝑖𝑛𝑔 = 𝑐𝑢𝑛𝑖𝑡 _𝑒𝑛𝑒𝑟𝑔𝑦 ∗ 𝑓𝜂 ∗ 𝑤𝑐𝑎𝑠𝑡 ∗ 𝑓𝑦 ∗ 𝑓𝑟 ∗ 𝑓𝑚 ∗ 𝑓𝑝 ∗ 𝑓𝑓 ∗ 𝑐𝑝𝑠 ∗ 𝑡𝑚𝑒𝑙𝑡 − 𝑡𝑟𝑜𝑜𝑚 + 𝐿 +

𝑐𝑝𝑙 ∗ (𝑡𝑡𝑎𝑝 − 𝑡𝑚𝑒𝑙𝑡 ) (5.11)

Where:
 𝑐𝑢𝑛𝑖𝑡 _𝑒𝑛𝑒𝑟𝑔𝑦 = cost for unit energy
 𝑓𝜂 = factor for efficiency of furnace 3 – 3.5
 𝑓𝑦 = factor for overall yielding (gating and fettling)
 𝐿 = latent heat
 𝐶𝑝𝑠 = specific heat of alloy at solid phase
 𝐶𝑝𝑙 = specific heat of alloy at liquid phase
 𝑡𝑚𝑒𝑙𝑡 = pouring temperature of melt
 𝑡𝑟𝑜𝑜𝑚 = room temperature,
 𝑡𝑡𝑎𝑝 = tapping temperature – temperature of the molten metal when
removed from furnace

Here a cost for unit energy of 1 SEK/kWh was assumed and values of 3.5 and 2 were used for
𝑓𝜂 and 𝑓𝑦 , respectively. As for the remaining values: values of 1.1 kJ/kgK and 1.15 kJ/kgK
were used as 𝐶𝑝𝑠 and 𝐶𝑝𝑙 values, while 321kJ/kgK was used for the latent heat. Temperature
values of 700ºC, 20ºC and 800ºC were used as 𝑡𝑚𝑒𝑙𝑡 , 𝑡𝑟𝑜𝑜𝑚 and 𝑡𝑡𝑎𝑝 values respectively.

Also using 𝑓𝑟 , 𝑓𝑚 , 𝑓𝑝 and 𝑓𝑓 values as the previously taken ones the total cost was calculated
to be:

1𝑆𝐸𝐾 1.1 700 −25 +321+1.15 800 −700 𝑘𝐽

𝐶𝑚𝑒𝑙𝑡𝑖𝑛𝑔 = 𝑘𝑊𝑕 ∗ 3.5 ∗ 4.5 ∗ 2 ∗ 1.12 ∗ 1.12 ∗ 1.07 ∗ 1.07 ∗ 𝑘𝑔
∗
2.27 ∗ 10−4
𝐶𝑚𝑒𝑙𝑡𝑖𝑛𝑔 = 12𝑆𝐸𝐾

89
 Cost estimation

5.5 Tooling cost

Tooling cost account for the various cost of tools used for enabling production like patterns
and pattern making tools for sand casting and cost metal moulds in die casting as well as the
different machining required for pattern and mold making. The relation is given by equation
(5.12 and 5.13).

𝐶𝑟𝑒𝑙 _𝑡𝑜𝑜𝑙 _𝑐𝑜𝑠𝑡 = 𝑒𝑥𝑝 0.629 ∗ 𝑉𝑐𝑎𝑠𝑡 + 0.048 ∗ 𝐶𝑎𝑐 + 0.023 ∗ 𝐶𝑠 + 0.739 (5.12)

𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 = 𝑐𝑖𝑛𝑑𝑒𝑥 ∗ 𝑐𝑟𝑒𝑙 _𝑡𝑜𝑜𝑙 _𝑐𝑜𝑠𝑡 /𝑄 (5.13)

Where:
 𝐶𝑟𝑒𝑙 _𝑡𝑜𝑜𝑙 _𝑐𝑜𝑠𝑡 = relative tooling for type of tooling used
 𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 = amortized for tooling
 𝑐𝑖𝑛𝑑𝑒𝑥 = tooling cost index that varies with manufacturer, currency and time
 𝑉𝑐𝑎𝑠𝑡 = casting volume in 𝑚 3
 𝐶𝑎𝑐 = accuracy index 1-100 scale
 𝑄= order quantity
 𝐶𝑠 = casting shape complexity

Where 𝐶𝑠 :

𝐶𝑠 = 0.3 ∗ 𝐶𝑎 + 𝐶𝑐 − 14 (5.14)

and:
 𝐶𝑠 = Shape complexity
 𝐶𝑎 =Area ratio=
100 ∗ 1 − 𝑠𝑢𝑟𝑓𝑎𝑐𝑒 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑐𝑢𝑏𝑒 𝑜𝑓 𝑒𝑞𝑢𝑎𝑙 𝑣𝑜𝑙𝑢𝑚𝑒 𝑠𝑢𝑟𝑓𝑎𝑐𝑒 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑝𝑎𝑟𝑡
 𝐶𝑐 = Core complexity factor = 100 ∗ 1 − 1 1 + 𝑛 , where n is the number of
cores

Calculating tooling cost, using the above relation was found difficult as some values were
hard to determine. Instead the tooling cost values from CES were taken as an assumption for
evaluating this cost.

Taking the tooling cost shown on appendix (8.3), the maximum tooling cost for high pressure
die casting was assumed to be the tooling cost for rheocasting. This gave a total tooling cost
of 800,000SEK. From the same table, it can also be seen that such a tool has the capacity of
producing 2,000 up 1,000,000 units. The tool life estimated for this thesis was around 150,000
units. However, this tool life was not used as the batch size for the chassis was 5,000 and
production of extra chassis was not planned at the time. This gave a tool cost per unit of:

𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 = 800,000𝑆𝐸𝐾/5000 𝑢𝑛𝑖𝑡𝑠

𝐶𝑡𝑜𝑜𝑙 𝑖𝑛𝑔 = 160𝑆𝐸𝐾(for rheocasting)

For plaster casting, consulting the same table on appendix (8.3), a total tooling cost of
12,000SEK was assumed. From the same table, it can be seen that the tool has the capacity of
producing 500 to 5000 units. The tool life estimated for this thesis was around 2,500 units.

𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 = 12,000𝑆𝐸𝐾/2,500 𝑢𝑛𝑖𝑡𝑠

𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 ≅ 5𝑆𝐸𝐾
𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔 = 5𝑆𝐸𝐾(for plaster casting)

90
 Cost estimation

5.6 Overhead costs

This cost accounts for administrative overhead cost as well as depreciation cost. Both
components making up this cost are on a weight basis and are usually calculated by
calculating how much cost is incurred, for each, over a long period of years and dividing it
with the total weight of castings produced in those years. Their relation is given in equation
(5.15-17).

𝐶𝑜𝑣𝑒𝑟 𝑕𝑒𝑎𝑑𝑠 = 𝐶𝑎𝑑𝑚𝑖𝑛𝑠𝑡𝑟𝑎𝑡𝑖𝑜𝑛 + 𝐶𝑑𝑒𝑝𝑟𝑖𝑐𝑖𝑎𝑡𝑖𝑜𝑛 (5.15)

𝐶𝑎𝑑𝑚𝑖𝑛𝑠𝑡𝑟𝑎𝑡𝑖𝑜𝑛 = 𝑤𝑐𝑎𝑠𝑡 ∗ 𝐶𝑎𝑑𝑚𝑖𝑛𝑠𝑡𝑟𝑎𝑡𝑖𝑜𝑛 _𝑟𝑎𝑡𝑒 (5.16)

𝐶𝑑𝑒𝑝𝑟𝑖𝑐𝑖𝑎𝑡𝑖𝑜𝑛 = 𝑤𝑐𝑎𝑠𝑡 ∗ 𝐶𝑑𝑒𝑝𝑟𝑖𝑐𝑖𝑎𝑡𝑖𝑜𝑛 _𝑟𝑎𝑡𝑒 (5.17)

Where:
 𝐶𝑎𝑑𝑚𝑖𝑛𝑠𝑡𝑟𝑎𝑡𝑖𝑜𝑛 = cost for administration per component
 𝐶𝑎𝑑𝑚𝑖𝑛𝑠𝑡𝑟𝑎𝑡𝑖𝑜𝑛 _𝑟𝑎𝑡𝑒 = administration cost rate per weight of alloy
 𝐶𝑑𝑒𝑝𝑟𝑖𝑐𝑖𝑎𝑡𝑖𝑜𝑛 = cost for deprecation per component
 𝐶𝑑𝑒𝑝𝑟𝑖𝑐𝑖𝑎𝑡𝑖𝑜𝑛 _𝑟𝑎𝑡𝑒 = depreciation cost per weight of alloy

However, in this thesis, it was mentioned that the overhead costs were lumped with the labor
cost, as detailed values that should have been used in the respective equations were hard to
attain.

5.7 Estimated total cost for a unit chassis

In the end, using equation (5.1), the estimated cost for rheocasting a unit chassis was:

𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 = 𝐶𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 + (𝐶𝑙𝑎𝑏𝑜𝑢𝑟 + 𝐶𝑜𝑣𝑒𝑟 𝑕𝑒𝑎𝑑𝑠 ) + 𝐶𝑒𝑛𝑒𝑟𝑔𝑦 + 𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔

𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 = 115𝑆𝐸𝐾 + 37.5𝑆𝐸𝐾 + 12𝑆𝐸𝐾 + 160𝑆𝐸𝐾
𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 ≅ 325𝑆𝐸𝐾
𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 = 325𝑆𝐸𝐾

Considering that the tool life is significantly higher than the planned 5000 units, it should be
emphasized that the total cost per unit can be significantly reduced if the number of products
needed is increased.

On the other hand, for plaster casting, the estimated cost of producing a unit chassis was:

𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 = 𝐶𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 + (𝐶𝑙𝑎𝑏𝑜𝑢𝑟 + 𝐶𝑜𝑣𝑒𝑟 𝑕𝑒𝑎𝑑𝑠 ) + 𝐶𝑒𝑛𝑒𝑟𝑔𝑦 + 𝐶𝑡𝑜𝑜𝑙𝑖𝑛𝑔

𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 = (115𝑆𝐸𝐾 + 517.5𝑆𝐸𝐾) + 800𝑆𝐸𝐾 + 12𝑆𝐸𝐾 + 5𝑆𝐸𝐾
𝐶𝑐𝑎𝑠𝑡𝑖𝑛𝑔 ≅ 1450𝑆𝐸𝐾

The price of the plaster casting was much higher than that of rheocasting. This was attributed
to the slow production rate of the process as well the mold being non reusable. It should be
noted that no plaster material was assumed to be recycled here and a very slow production
rate were assumed.

In both cases, the calculation of cost did not include the cost for machining and coating as
well as heat treatment that would be done on the chassis.

91
 Conclusion

6 Conclusion

6.1 Overall
The choices made in modifying the chassis as well as designing and adopting the gating
systems had made casting of such a „complex‟ chassis achievable. The simulations of both
selected casting processes have shown that it is possible to have a sound casting of the
modified chassis.
The filling results in rheocasting, showed the slurry having a smooth flow and not seen to
disintegrate at high velocities. Recommended velocities at ingate which were given as a
function of gate thickness were also studied. It was shown that having a much larger velocity,
of up to three times the recommended one, did not result in any slurry disintegration.
However, the filling of fin tips was not done as soon as slurry gets access to a fin. This may
result in surface cavities at fin tips. During solidification slurry was seen to be fed
sufficiently. Rheocasting is said to be a „no porosity‟ process, however, the casting may have
some small subsurface porosities dispersed near the junction of the inner walls.
The filling results in plaster casting showed a smooth filling of the chassis as well as the fin
tips. The use of filters was shown to be valuable in decreasing melt velocity as well as
streaming flow. Moreover, the advantage of having a pouring cup instead of using a sleeve to
receive poured melt was made evident by the relatively intact melt flowing down the sprue.
Solidification results showed melt being fed through the gates. The fins had cavities at their
tip and needed to be fed. This was assumed to be due to liquid shrinkage that occurs quickly.
Increasing the volume of extra melt in the sleeve reduced this greatly, as it would result in a
greater head pressure to push melts to fin tips. The chassis might also have subsurface
porosities dispersed near the junction of inside walls and the junction of fins with the box.
In this thesis, preference to one type of casting process was avoided for the purpose of being
impartial and thus judgment was left for the reader.
Comparing the cost of casting such a piece, it was believed that both processes would be
much cheaper than the machining planned for the original chassis.

6.2 Further work

It is a known fact that there is a difference between results of process simulations and their
physical outcome. This difference might be magnified as thin castings have difficulties with
meshing. Thus a trial production is necessary and should be carried out to see if the processes
are able to produce a sound casting. It can also be useful to compare if the defects predicted
would really form.
On the other hand, other process features could be included to the simulations carried out.
One very important thing to include would be cooling channels which are usually put to
control solidification. In both processes, control of the solidification direction could
physically be achieved by such mechanisms and thus their consideration in simulations are
invaluable.
The modified chassis can also be varied slightly, as it was designed to be as closed to as
finished piece as possible. All walls were assumed to require no machining. Instead, adding
thicknesses to some sections to subsequently machine them can be an option that could be
considered. This might be done after a trial production.

92
 References

7 References

[1] http://www.thefreelibrary.com/Aircraft%2FAerospace.+(Casting+Market+Focus)-
a087742837
(Acc. 2009-04-10)

[2] http://www.businessdictionary.com/definition/product-development.html
(Acc. 2009-02-12)

[3] I. J. Polmear (2006), Light Alloys 3. ed.

Butterworth Heinemann, ISBN 0-7506-6371-5

[4] John Gilbert Kaufman, Elwin L. Rooy (2004), Aluminum alloy castings
ASM International, ISBN 0-87170-803-5

[5] CES EduPack software 2008

[6] http://www.rheometal.com/rapidp4.php
(Acc. 2009-02-05)

[7] http://www.custompartnet.com
(Acc. 2009-03-25)

[8] J T. Black, Ronald A. Kohser (2008), DeGarmo’s materials and processes in

manufacturing
John Wiley & Sons, ISBN 978-0-470-05512-0

[9] John Campbell (2003), Castings 2.ed.

Butterworth Heinemann, ISBN 0-7506-4790-6

[10] John Campbell (2004), Castings practice

Butterworth Heinemann, ISBN 0-7506-4791-5

[11] MAGMAsoft 4.4 Manual

[12] Michael F. Ashby (2005), Materials Selection in Mechanical Design

Butterworth Heinemann, ISBN 0-7506-6168-2

[13] ASM Handbook (1988), Casting Volume.15

ASM International, ISBN 0-87170-007-7

[14] ASM Handbook (1991), Heat Treating Volume.4

ASM International, ISBN 0-87170-379-3

[15] ASM Handbook (2000), Mechanical Testing and Evaluation Volume.8

ASM International, ISBN 0-87170-389-0

[16] ASM Handbook (2000), Nondestructive Evaluation and Quality Control Volume.17
ASM International, ISBN 0-87170-007-7

[17] Peter Beeley (2001), Foundry technology 2. ed.

Butterworth Heinemann, ISBN 0-7506-4567-9

93
 References

[18] Geoffrey Boothroyd, Peter Dewhurst, Winston A. Knight (2002), Product design for
manufacture and assembly 2. ed.
Dekker, ISBN 0-8247-0584-8

[19] http://www.me.iitb.ac.in/~bravi/lab/paper/2005CIM_CastingCostEstimation.pdf
(Acc. 2009-06-02)

[20] International Standard ISO 2768-1:1989

[21] International Standard ISO 8062:1994

94
 Appendix

8 Appendix

8.1 Appendix - composition and properties of alloys

Alloy composition
Al(%) Be(%) Cu(%) Fe(%) Mg(%) Mn(%) Si(%) Ti(%) Zn(%)
6082(T6) 97 - - 0,9 0,7 1 - -
0,15- 0,2- 0,15-
A201.0(T7) 94-95 - 4--5 0,1 0,35 0,4 0,05 0,35 -
354.0(T6) 87-89 - 1,6-2 0,2 0,4-0,6 0,1 8,6-9,4 0,2 0,1
355.0(T6) 90-92 - 1-1,5 0,6 0,4-0,6 0,5 4,5-5,5 0,25 0,35
C355.0(T6) 92-94 - 1-1,5 0,2 0,4-0,6 0,1 4,5-5,5 0,2 0,1
0,2-
356.0(T6) 90-92 - 0,25 0,6 0,45 0,35 6,5-7,5 0,25 0,35
0,25-
A356.0(T6) 91-92 - 0,2 0,2 0,45 0,1 6,5-7,5 0,2 0,1
0,04-
A357.0(T6) 91-93 0,07 0,2 0,2 0,4-0,7 0,1 6,5-7,5 0,04-0,2 0,1
0,04- 0,55-
D357.0(T6) 91-93 0,07 - 0,2 0,6 0,1 6,5-7,5 0,1-0,2 -
359.0(T6) 89-90 - 0,2 0,2 0,5-0,7 0,1 8,5-9,5 0,2 0,2

Physical properties
K Y.S. Elogn Hard Cp Fatig Rho Fracture Elec.res.
(w/mk) (Mpa) (%) (HV) (J/kgK) (Mpa) (kg/m3) (Mpam1/2) (ohm)
6082 169- 240- 5-11 95- 882- 90- 2,67- 33-35 3,8-4
(T6) 175 290 105 918 100 2,73
A201.0 121- 331- 1,5-5 148- 921- 97-99 2,78- 18-24 4-6
(T7) 131 366 152 959 2,81
354.0 126- 228- 2-3 110- 963- 117- 2,78- 18-26 4-6
(T6) 137 290 120 1000 175 2,81
355.0 152- 117- 0,4- 80-85 963- 60-64 2,7- 22-26 4,7-4,9
(T6) 165 129 0,465 1000 2,73
C355.0 152- 193- 1-3 90-95 963- 62-97 2,7- 18-23 4,7-4,8
(T6) 165 276 1000 2,73
356.0 152- 103- 0,7-1 120- 963- 57-73 2,67- 18-22 5,8-6
(T6) 165 114 121 1000 2,7
A356.0 152- 152- 2-5 77- 963- 60-90 2,67- 18-24 4,1-4,3
(T6) 165 234 103 1000 2,7
A357.0 152- 193- 3-5 95- 963- 62- 2,67- 18-24 4-6
(T6) 165 276 114 1000 103 2,7
D357.0 152- 248- 2-3 103- 963- 124- 2,67- 18-24 4-6
(T6) 165 283 107 1000 169 2,7
359.0 152- 207- 3-4 110- 963- 110- 2,67- 18-24 4-6
(T6) 165 262 120 1000 162 2,7

95
 Appendix

8.2 Appendix - alloy properties in sand and permanent

molding

Sand
Resistance
to Hot Pressure Fluidity Shrinkage Corrosion Machinability Weldability
Alloy Cracking Tightness Tendency Resistance
A201.0
201.0 4 3 3 4 4 1 2
354.0 1 1 1 1 3 3 2
355.0 1 1 1 1 3 3 2
C355.0
356.0
A356.0 1 1 1 1 2 3 2
357 1 1 1 1 2 3 2
A357.0
D357.0
359.0 1 1 1 1 2 3 1

Permanent
Resistance
to Hot Pressure Shrinkage Corrosion
Alloy Cracking Tightness Fluidity Tendency Resistance Machinability Weldability
A201.0
201.0 4 3 3 4 4 1 2
354.0 1 1 1 1 3 3 2
355.0 1 1 1 2 3 3 2
C355.0 1 1 1 2 3 3 2
356.0 1 1 1 1 2 3 2
A356.0 1 1 1 1 2 3 2
357 1 1 1 1 2 3 2
A357.0 1 1 1 1 2 3 2
D357.0
359.0 1 1 1 1 2 3 1

1 stands for excellent and 4 for poor

96
 Appendix

8.3 Appendix - process attributes

Tooling Cost Tool Life

Name Mass range (kg) (SEK) (Units)
High pressure die casting 0,05 - 15 5,38e4 - 7,77e5 2e4-1e6
Centrifugally-aided casting 0,05 - 20 598-5,98e3 50-200
Investment casting,
automated 0,05 - 20 1,2e4-1,79e5 500-5000

Ceramic mold casting 0,2 - 50 1,2e3-1,2e4 50-500

Plaster mold casting 0,05 - 10 2,39e3-1,2e4 500 – 5000

Shell casting 0,3 - 10 1,2e4-5,98e4 1e3-1e4

Evaporative pattern casting,
manual 0,5 - 25 598 - 2,39e3 1 - 10
Evaporative pattern casting,
automated 0,5 - 25 5,98e3-5,98e4 5e3-1e4

Low pressure die casting 5 - 25 5,98e4-1,79e5 3e4-1,79e5

Cosworth casting 0,2 - 50 5,98e3-3,59e4 1e4-1e6

Green sand casting, manual 0,5 - 1,2e4 598-5,98e3 100-1e3

Gravity die casting 0,5 - 50 3,59e4-1,43e5 1e4-1e5

Green sand casting,
automated 0,5 - 1,2e4 5,98e3-3,59e4 1e4-1e5

Squeeze casting 0,5 - 12 9,56e4-2,39e5 1e4-5e4

Tooling Tooling
Tooling Tool life min max
Name Cost min Cost max For 5000 For 5000
Plaster mold casting 2390 12000 500 23900 120000
High pressure die casting 53800 777000 20000 13450 194250
Ceramic mold casting 1200 12000 50 120000 1200000
Investment casting,
automated 12000 179000 500 120000 1790000

97
 Appendix

8.4 Appendix - plaster casting material properties used

in simulation

Thermal conductivity– A356.0

Density– A356.0

98
 Appendix

… Appendix 8.4 continued

Specific heat capacity – A356.0

Thermal conductivity – gypsum (plaster mold)

99
 Appendix

… Appendix 8.4 continued

Density– gypsum (plaster mold)

Specific heat capacity – gypsum (plaster mold)

100
 Appendix

8.5 Appendix - rheocasting material properties used in

simulation

Thermal conductivity – A356.0 (thixotropic model)

Density – A356.0 (thixotropic model)

101
 Appendix

… Appendix 8.5 continued

Specific heat capacity – A356.0 (thixotropic model)

Thermal conductivity – Steel (die)

102
 Appendix

… Appendix 8.5 continued

Density – Steel (die)

Specific heat capacity – Steel (die)

103
 Appendix

… Appendix 8.5 continued

Thermal dependant heat transfer coefficient used between casting and die in rheocasting

104
 Appendix

8.6 Appendix - simulation parameters

Simulation parameters used for some versions in rheocasting

V6-v1 V9-v1 V9-v2 V9-v3 V9-v5 V11-v1 V11-v2

Rheo_v6_ Rheo_v9_ Rheo_v9_ Rheo_v9_ Rheo_v9_G Rheo_v11_ Rheo_v11_
GS_1_Fast GS_1_v01 GS_1_v02 GS_1_v03 S_1_v05 GS_2_v01 GS_2_v02
Mesh
10000000 10000000 10000000 10000000 12000000 10000000 10000000
elements
Ingate
area 6.402 9.537 9.537 9.537 9.537 10.293 10.293
2
[cm ]
Gate
thickness 4 6 6 6 6 6 6
[mm]
Cavity
filling time 205.688 109.339 328.018 218.679 109.339 110.200 110.200
[ms]
Velocity at
the ingate 12.562 15.812 5.271 7.906 15.812 14.650 14.650
[m/s]

105