Improving Quality in Mega-Casted
Products
Identification of contributors to geometrical variation

Master’s thesis in Production Engineering

Paul Adam, David Hermez

DEPARTMENT OF INDUSTRIAL AND MATERIALS SCIENCE

CHALMERS UNIVERSITY OF TECHNOLOGY
Gothenburg, Sweden 2024
www.chalmers.se

www.chalmers.se




master’s thesis 2024

Improving Quality in Mega-Casted Products

Identification of contributors to geometrical variation

Paul Adam
David Hermez

Department of Industrial and Materials Science
Production Engineering

Chalmers University of Technology
Gothenburg, Sweden 2024



Improving Quality in Mega-Casted Products
Identification of contributors to geometrical variation
Paul Adam
David Hermez

©Paul Adam, 2024, boulus@chalmers.se
©David Hermez, 2024, hermezd@chalmers.se

Industrial Supervisor: Josenfin Hansen, Geometry assurance Painted Body VCC
Academic Supervisor: Kristina Wärmefjord, Chalmers University of Technology
Examiner: Kristina Wärmefjord, Chalmers University of Technology

Master’s Thesis 2024
Department of Industrial and Material Science
Production Engineering
Chalmers University of Technology
SE-412 96 Gothenburg
Telephone +46 31 772 1000

Cover: The Volvo Mega-Casting [1]

Typeset in LATEX, template by Kyriaki Antoniadou-Plytaria
Printed by Chalmers Reproservice
Gothenburg, Sweden 2024

iii



Improving Quality in Mega-Casted Products
Identification of contributors to geometrical variation
Paul Adam
David Hermez
Department of Industrial and Material Science
Chalmers University of Technology

Abstract
When using a process to produce various parts, geometrical variation describes the
differences in deviations among those parts. These deviations arise due to different
factors, and minimizing them contributes to improved consistency in production,
increased overall product quality, and reduced costs in terms of extra rework. This
master’s thesis aimed at expanding the knowledge about high-pressure die-casting
(HPDC) and identifying contributors to geometrical variations. The scope of the
study is limited to larger defects affecting the geometrical accuracy and excludes sur-
face defects, porosities, and machining. The method used for reaching the goal was
by conducting qualitative and quantitative approaches. A total of 11 professionals
in the area of HPDC were interviewed. Production tests related to the quenching
phase were also conducted in the foundry. The main finding from the study was that
design is the root cause of geometrical variations, as the varying thicknesses induce
residual stresses which in turn influences the accuracy of the final shape. One of the
findings successfully reduced geometrical variation in megacastings but led to other
problems such as sticking to the trim-press. In addition, a DoE was suggested for
future studies. The DoE provides a systematic way of working and helps in finding
the optimal combination of influential and easily changeable parameters.

Keywords: HPDC, Megacasting, Geometrical variation, Distortions.

iv





Acknowledgements
The authors would like to acknowledge every individual for the support and guidance
in completing this thesis. We want to express our gratitude to Volvo Cars for
giving us the opportunity to be the first students to expand the knowledge about
megacasting. We are extremely grateful to our industrial supervisor Josefin Hansen
for the invaluable guidance and advice throughout the project. We also want to
thank the examiner Kristina Wärmefjord for the insights and feedback. Big thanks
go to all the interviewees for sharing their expertise, insights, and experiences. Their
contributions gave us a deeper understanding and enriched this project with practical
perspectives.

Paul Adam, David Hermez, Gothenburg, June 2024

vi





List of Acronyms

Below is the list of acronyms that have been used throughout this thesis listed in
alphabetical order:

AC Air Cooling
ANOVA Analysis of Variation
BIW Body in White
CAD Computer-Aided Design
DOE Design of Experiment
DOF Degrees of Freedom
EFL Effective Flow Length
EV Electric Vehicle
GROMCAP Geometrical Robustness of Mega Casted Aluminum Parts
HPDC High Pressure Die Casting
IHTC Interfacial Heat Transfer Coefficient
MRA Mold Release Agent
PAG Polyalkylene Glycol
PFZ Precipitate Free Zone
RDM Research Design Methodology
SQ Spray Quenching
UTS Ultimate Tensile Strength
WC Wind Cooling
WQ Water Quenching
YS Yield Stress

viii





Contents

List of Acronyms viii

List of Figures xii

1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Geometrical Variations . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Purpose and Goal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.3.1 Project Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.5 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Theory 4
2.1 Important Terminologies . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Overview of the High-Pressure Die Casting Process . . . . . . . . . . 5
2.3 The Gating Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.4 Study Findings of Contributors to Geometrical Variations . . . . . . . 11

2.4.1 Residual Stresses in HPDC . . . . . . . . . . . . . . . . . . . . 11
2.4.2 Process Parameters . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4.3 Extra-large Thin-walled HPDC Castings . . . . . . . . . . . . 14
2.4.4 Insights from MAGMA Simulation Runs . . . . . . . . . . . . 17
2.4.5 Vacuum Assisted Casting . . . . . . . . . . . . . . . . . . . . 18
2.4.6 Soldering at the Ejection Phase . . . . . . . . . . . . . . . . . 18
2.4.7 Alloy Composition . . . . . . . . . . . . . . . . . . . . . . . . 19

2.5 Quenching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.5.1 Quenching Mechanism . . . . . . . . . . . . . . . . . . . . . . 23
2.5.2 Cooling Rates and Quench Severity . . . . . . . . . . . . . . . 23
2.5.3 Quenching Media . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.5.4 Spray Quenching . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.5.5 Study on Different Quenching Types and Rates . . . . . . . . 28
2.5.6 Quenching Orientation . . . . . . . . . . . . . . . . . . . . . . 30
2.5.7 Immersion Rate . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3 Method 35
3.1 Methodological Framework . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 Screening of Literature Study . . . . . . . . . . . . . . . . . . . . . . 35
3.3 Qualitative Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

x



Contents

3.4 Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.5 Quantitative Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3.5.1 Tests in Foundry . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.6 Ethical Aspect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4 Results 40
4.1 Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.3 Summary of Experts Insights . . . . . . . . . . . . . . . . . . . . . . 45

4.3.1 Gating System . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.3.2 Shot Sleeve . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.3.3 Die Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.4 Die Temperature . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.5 Melt Temperature . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.3.6 Alloy Composition . . . . . . . . . . . . . . . . . . . . . . . . 48
4.3.7 Vacuum Assistance . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3.8 Intensification Pressure . . . . . . . . . . . . . . . . . . . . . . 49
4.3.9 Cooling Time Inside The Cavity . . . . . . . . . . . . . . . . . 50
4.3.10 Ejection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.3.11 Quenching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.3.12 Trimming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.3.13 Residual Stresses . . . . . . . . . . . . . . . . . . . . . . . . . 53

4.4 Cause and Effect Diagram for Geometrical Variation . . . . . . . . . 55

5 Discussion 56
5.1 Answering Research Questions . . . . . . . . . . . . . . . . . . . . . . 58

6 Conclusion 59

7 Future Recommendations 60

Bibliography 62

xi



List of Figures

2.1 Simple sketch of a HPDC machine. . . . . . . . . . . . . . . . . . . . 4
2.2 Representation of a casting and intended part in blue. . . . . . . . . . 5
2.3 The Process chain of HPDC [8]. . . . . . . . . . . . . . . . . . . . . . 6
2.4 V-shaped gating system [11]. . . . . . . . . . . . . . . . . . . . . . . . 7
2.5 The filling percentage and absolute velocity for the different simula-

tions[13]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.6 Teslas’ and Volvos’ mega-castings of the rear floor. . . . . . . . . . . 9
2.7 Teslas’ gating design of their mega-casting. . . . . . . . . . . . . . . . 9
2.8 A whole mega-casting of the Xiaomi rear floor. . . . . . . . . . . . . . 10
2.9 A whole mega-casting of the Zeekr rear floor. . . . . . . . . . . . . . . 10
2.10 Simple flowchart describing the differences in three stages . . . . . . . 12
2.11 a) stress lattice, b) V-shaped component [16]. . . . . . . . . . . . . . 13
2.12 Representation of the extra-large thin-walled casting [24]. . . . . . . . 15
2.13 A representation of how temperature decreases throughout different

stages of the filling process [24]. . . . . . . . . . . . . . . . . . . . . . 15
2.14 Displacement results for a rear floor mega-casting. . . . . . . . . . . . 17
2.15 Temperature results for a rear floor mega-casting. . . . . . . . . . . . 17
2.16 Cross-section that shows the soldering intermetallic layer [29]. . . . . 19
2.17 The GT models’ lid frame. Weight: 11.6 kg. Dimensions: ca. 1230 x

1250 x 390 mm [9]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.18 BMWs’ cross beam. Weight: 4.1 kg. Dimensions: ca. 810 x 450 x

210 mm [9]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.19 Audis’ rear connector. Weight: 9.9 kg. Dimensions: ca. 1454 x 375

x 552 mm [9]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.20 Cooling curves for AA6082 alumnium sheet quenched in different me-

dias [10]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.21 Immersion quenching and its cooling stages and mechanisms [35]. . . 23
2.22 Cooling curves as a function of water temperature for 1/2 in. 7075

aluminium platec [10]. . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.23 Tensile strength of 1 in. 7075 aluminium plate as a function of water

temperature [10]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.24 Comparison of water quenching and PAG 20 percent polymer quench-

ing and its impact on distortion for a plate with 1.5 mm thickness.
Upper left panel is 300×250 mm; Upper right panel is 100 mm×300
mm; lower left and right panels are 300×300 mm [34] . . . . . . . . . 27

2.25 cooling curves at each quenching condition [38]. . . . . . . . . . . . . 28

xii



List of Figures

2.26 The parameters of the investigated quenching conditions [38]. . . . . 28
2.27 calculated IHTC curves for different quenching methods [38]. . . . . . 29
2.28 Boiling curves with different spray quenching parameters [38]. . . . . 29
2.29 Comparison of the mechanical properties and residual stress of the

alloys with different quenching cooling rates [38]. . . . . . . . . . . . . 30
2.30 Water quench orientations: RE - rear face up, CC - cover face up, IN

- intake port up, FR - front face up, JF - joint face up, EX - exhaust
face up [18]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

2.31 Vapor pattern and vapor entrapment inside cylinder heads after 20
seconds of quenching [18]. . . . . . . . . . . . . . . . . . . . . . . . . 32

2.32 Cooling curves for all water quenching configurations [18]. . . . . . . 32
2.33 Maximum temperature gradients for all water quenching configura-

tions [18]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.34 Sheets are 500×500 mm×1.0 mm thick immersed in PAG with two

different immersion rates. On top, a sheet with slow immersion rate.
On the bottom, a sheet with quick immersion [34]. . . . . . . . . . . 34

3.1 A funnel-like method for literature study. . . . . . . . . . . . . . . . . 37

4.1 Inspected points on the mega-casting. . . . . . . . . . . . . . . . . . . 40
4.2 Red ellipses marking the sensitive areas and red arrows the bending

behavior. Picture adapted from [41]. . . . . . . . . . . . . . . . . . . 41
4.3 4 Red graphs represent no vacuum-assisted casting runs while 3 blue

graphs had vacuum assist. . . . . . . . . . . . . . . . . . . . . . . . . 41
4.4 Distance measurement between the front legs. Picture adapted from

[41]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.5 Different tests and their effect on distance 1 and 2. . . . . . . . . . . 43
4.6 Quench temperatures and its effect on the measured distance 1 and 2. 44
4.7 Cooling effect of quench tank on distance 1 and 2. . . . . . . . . . . . 44
4.8 Simple representation of what E3, E4, and E5 means. . . . . . . . . . 50
4.9 Final fishbone diagram for geometrical variation. . . . . . . . . . . . . 55

7.1 A simple DoE with 16 trials. . . . . . . . . . . . . . . . . . . . . . . . 61

xiii



1
Introduction

This chapter introduces the topic of mega-casting, provides a comprehensive view
of the concept, and the reason behind manufacturing companies’ shift toward mega-
casting. It also describes the high-pressure die-casting (HPDC) process chain, the
goals of the research, and its limitations.

1.1 Background
The world is evolving rapidly, new innovative products are manufactured every day
to reduce the environmental footprint, and one of these products is the electric ve-
hicle (EV). However, with new products, new challenges arise and it’s well known
that electric batteries increase the weight of the vehicle. Therefore weight reduction
becomes a new challenge in the new upcoming EVs to enhance energy efficiency [2]
and increase range to travel longer distances [3].

The Swedish car manufacturer, Volvo Cars, is investing in a new revolutionary way
of HPDC to manufacture parts for their upcoming EVs. The process includes large-
scale casting which will minimize the resources needed to assemble large floors,
an essential part of Body in White (BIW) products. The concept is called – mega-
casting. Mega-casting simplifies the assembly process by reducing welding processes,
shortening the assembly line, and decreasing the likelihood of assembly errors. This
new concept aims to replace many small parts with a single larger part [4]. It will
lead to at least 35% CO2 emission reduction and 50% of the aluminium used will
be a secondary alloy. Volvo also expects design optimization of mega-casted parts
to enable a 15% weight reduction compared to steel alternatives. Several automak-
ers, including Mercedes-Benz, are investing in large injection molding machines.
Moreover, Chinese startups like NIO and Xpeng have also placed orders for these
machines [3].

Limited experience with mega-casting poses new challenges. One of the main chal-
lenges with large-scale parts is the tendency of geometrical variation. Geometrical
variation refers to the deviation of a product’s shape and size from its nominal de-
sign [5].

A research project, Geometrical Robustness of Mega Casted Aluminium Parts, (GROM-
CAP), is conducted by Chalmers University of Technology in collaboration with
Volvo Cars and other actors to expand the knowledge about mega-castings. This

1



1. Introduction

master thesis is part of the GROMCAP project within the work package that covers
the identification of sources of geometrical deviation and variation.

1.2 Geometrical Variations
The development of modern technology has led to an increase in accuracy in manu-
factured parts, however, geometrical variations are not eliminated [6]. Geometrical
variation refers to the differences in deviations among various parts produced using
the same process [2]. In fact, these geometrical variations are process-related. In
addition, the deviation is present in all stages of product realization [6]. These de-
viations can impact the product’s shape, size, and characteristics [7]. When these
deviations are large enough the product might not function, causing a huge loss
for the manufacturer and in the best scenario extra rework. During the use phase
of a product, physical phenomena such as thermal expansion, add up the geomet-
rical variation causing a loss of product quality [6]. Twisting, bending, warping,
and shrinkage are examples of geometrical deviation. In this thesis, these terms are
commonly referred to as distortion and deformation.

1.3 Purpose and Goal
The goal of this thesis work is to establish a deeper knowledge of the HPDC pro-
cess geometry effects and establish a way of working for further studies within the
research project.

The purpose is to identify the contributors to geometrical variation to improve the
quality of mega-casted products. Distortion and shrinkage are common geometrical
challenges in casting and addressing their causes in mega-casting becomes essential.

1.3.1 Project Scope
The project scope involves researching key aspects of the HPDC process. Since the
concept of mega-casting is new, the research will be based on HPDC products in
general. small, medium, and large-sized products. The project scope consists of the
following:

• Literature study to understand and map high pressure die casting (HPDC)
process.

• Literature study to understand and map contributors to geometrical variation
in high pressure die casting.

• Perform workshop and interviews with experts within casting and casting sim-
ulation.

• Create cause and effect diagram for contributors to geometrical variation in
high pressure die casting for mega casted parts.

• Perform correlation study of measurement data and process data.

2



1. Introduction

• Create a proposal of ranking of variation sources that are of importance on
how a mega-casted part will deviate on part level.

1.4 Research Questions
Key questions that the thesis work will answer, are the following:

• What are the contributors to geometrical variation in mega-casted products/-
parts?

• What are the most crucial contributors?

1.5 Limitations
This project has a limited time frame and therefore the scope of the project also re-
quires clear limitations. The areas excluded due to limitations could be investigated
in future work. The limitations of the project are:

• Limited Resources: The mega-casting is under development and not many
studies have been published, which makes the exploration of this area confined.

• Human Factor: The mistakes made by the operator will be excluded. As-
suming that all parameter settings in the machine are correct.

• Post-Casting Processes: Machining, painting and joining will be excluded.

3



2
Theory

This chapter is a summary of the literature review that gives the reader an under-
standing of the concepts necessary to fully understand the process of HPDC.

2.1 Important Terminologies
In this section, the terminologies needed for understanding the thesis will be ex-
plained, covering both the machine level and part level. Figure 2.1 shows the basic
terminology of the HPDC machine that will be used in this project.

• Fixed Die: The stationary half of the die that remains fixed during the
casting process and the extraction of the final part.

• Moving/Mobile Die: The movable half of the die that opens to be able to
extract the final solidified part.

• Mold Cavity: The hollow space between the fixed and moving die. It’s the
space that forms the casting when the molten metal is injected.

• Ejection Pins: The extraction components that push the final solidified cast-
ing out of the die when the moving die opens.

• Shot Sleeve: A tube where the molten metal is poured in from the furnace.
It guides the metal into the mold cavity.

• Plunger: A cylindrical part that injects the molten metal into the mold cavity.

Figure 2.1: Simple sketch of a HPDC machine.

4



2. Theory

In the process of HPDC, several elements that are not intended and wanted in
the final part need to be included in the casting design. These are additional and
necessary elements of the casting to ensure the overall quality of the final part.
For example, the gating system (runner and ingate) is a crucial part of the casting
because it facilitates the flow into the mold cavity. Overflows are extra parts that
capture air and gases and the biscuit is the leftover metal that solidifies in the shot
sleeve. See figure 2.2

Figure 2.2: Representation of a casting and intended part in blue.

2.2 Overview of the High-Pressure Die Casting
Process

High-pressure die casting is a manufacturing process used in multiple industrial
fields. This type of manufacturing corresponds to approximately 50% of the world’s
production of lightweight parts [8]. Since the process of HPDC consists of several op-
erations, it is considered a complex process. See figure 2.3 for a better understanding
of the HPDC process steps.

5



2. Theory

Figure 2.3: The Process chain of HPDC [8].

In HPDC, molten metal is poured into a shot sleeve for each cycle. A piston presses
the molten metal with high pressure, resulting in high metal velocities (30 m/s - 60
m/s). When the mold is filled, the part/component will be cooled and when solidi-
fied extracted from the moving/mobile die. But before the injection of the molten
metal, the dies need to be sprayed for the dies’ safety [8].

HPDC has been used for over 100 years and was mainly used for zinc; today it has
become a common way for mass production of light-weight components [9]. In the
world of automotive, this process has become more common since there are several
categories that it could be applied to, namely:

• Powertrain
• Interiors and electronics
• Chassis and body structure

The most demanding application in terms of mechanical properties, highest strengths
and highest elongations, is the body structure since it needs to absorb large loads
in case of a crash [9].

Generally, the complexity of a part is determined by multiple factors, such as weight,
size, and geometry, but also its’ deformation capabilities, and therefore parts of the
body-structure could be classified as complex [9].

Quenching, also known as cooling, is a heat treatment step after ejection. It in-
volves rapid cooling of the casting to room temperature by immersing it in different

6



2. Theory

media such as water, or other media. There are other ways of quenching such as
spray quenching or air quenching. According to [10], alloying elements disperse
from the solution when it is slowly cooled. In addition, they start to concentrate at
grain boundaries, small gaps, unassimilated particles, dislocations, and other "im-
perfections" in the aluminium lattice. In contrast, it’s preferred to slow down the
dispersion and keep the elements in the solid solution until the age-hardening step,
to acquire corrosion resistance, ideal strength, and toughness [10]. In addition, af-
ter quenching, aging is done to create a fine distribution of elements and further
enhance the strength of the material [10].

Trimming is a subsequent step following the quenching of the mega-casting [2]. It
entails mechanical trimming to cut off unintended elements attached to the casting.
But it’s also a step where the geometrical quality could be affected due to the
spring-back effect [2].

2.3 The Gating Design
The design of runners, gating systems, and venting systems determine the quality
of HPDC parts. The most common designs of the runner systems are the vertical
runner system, V-shaped runner system, and horizontal runner system [11]. In ad-
dition, in a large and thin-walled casting, avoiding distortion becomes more critical,
due to the non-homogeneous heat transformation during mold filling [12]. This can
potentially create an unfavorable solidification pattern which in turn causes the part
to warp [11]. Furthermore, the contraction of the gating system during solidifica-
tion can drag the other sections of the casting causing distortion and hot tearing [12].

For high-precision and compact parts, the V-shaped gating system is the most suit-
able. However, more amount of melt will be needed [11].

Figure 2.4: V-shaped gating system [11].

A FEM approach was used to simulate three different gating designs on a HPDC

7



2. Theory

casted component [13]. Through these simulations, the authors analyzed the in-
fluence of the gating system on the quality of the castings. The different designs
of the gating system showed different amounts and types of defects e.g. defects of
simulation (A) are mainly related to porosities and entrapped gases. However, simu-
lation (B) showed deformation-related defects, due to the uneven shrinkage stresses.
Simulation (C) showed similar defects as simulation (A) but to a reduced extent,
meaning the porosity and entrapped gas defect returned. See figure 2.5.

Figure 2.5: The filling percentage and absolute velocity for the different simula-
tions[13].

8



2. Theory

A YouTube video by TMIO Tesla 1 provided a picture of the Tesla Giga press, see
figure 2.6 (a). As it appears in the pictures, the notable difference between the Volvo
mega-casting (a prototype) and the Tesla rear floor is the bar marked in red (a1)
and the legs for the backend of the rear (a2).

Figure 2.6: Teslas’ and Volvos’ mega-castings of the rear floor.

Another YouTube video by Tech Charge 2 shows the gating system of the same
casting in figure 2.6 (a) marked in red. The Tesla mega-casting has a fan-like gating
system and is connected to the backend of the casting, in between the legs. See
figure 2.7.

Figure 2.7: Teslas’ gating design of their mega-casting.
1https://www.youtube.com/watch?v=uh0rx4A6qf8&t=115s
2https://www.youtube.com/watch?v=fydlvpDZrAw&t=124s

9

https://www.youtube.com/watch?v=uh0rx4A6qf8&t=115s
https://www.youtube.com/watch?v=fydlvpDZrAw&t=124s


2. Theory

Another car manufacturer that adopted the mega-casting technology is the Chinese
Xiaomi 3. Xiaomi calls it "Hyper Casting" and their rear floor casting is quite similar
to the Tesla casting. However, the design of the Xiaomi gating system (marked in
red) is different. The Chinese producer has a fingers-like gating system which is also
located at the backend of the casting, see figure 2.8.

Figure 2.8: A whole mega-casting of the Xiaomi rear floor.

Another YouTube video 4 revealed the mega-casting of the Chinese manufacturer
Zeekrs’ rear floor. As shown in figure 2.9, the design of the gating system (marked
in red) is different from the Tesla and Xiaomi mega-castings. The gating is placed
in the middle of the casting as the biscuit is in the center of the floor. The melt
seems to be distributed to the cavity through a finger-like gating design. However,
unlike the above-mentioned rear floors, the Zeekr floor has no "free" front legs (red
rectangles marked). The legs are completely united with a complete floor. Another
important point to comprehend is that a massive gap will appear in the middle when
the gating is removed by the trimming process.

Figure 2.9: A whole mega-casting of the Zeekr rear floor.
3https://www.youtube.com/watch?v=wRNRg_PXZe0
4https://www.youtube.com/watch?v=C8lZMQOkLwY

10

https://www.youtube.com/watch?v=wRNRg_PXZe0
https://www.youtube.com/watch?v=C8lZMQOkLwY


2. Theory

2.4 Study Findings of Contributors to Geometri-
cal Variations

The following section will discuss studies that examine different process parameters
and residual stresses to highlight their impact on the distortion of HPDC castings.

2.4.1 Residual Stresses in HPDC
Residual stresses are internal stresses that are induced in HPDC without any ex-
ternal loads applied [14], [15]. These stresses are induced because the contraction
is constrained due to different cooling rates within the casting and also because the
casting is not free to move when it’s locked in the mold. Another important factor
that induces these stresses is the phase transformation during the cooling process.
Furthermore, castings with larger stress gradients and complex stress distributions
tend to vary the stress measurements [14].

Residual stresses are classified as primary and secondary stresses and the difference
between them is that the primary stresses do not depend on the geometry of the
part, whereas the secondary does. The primary stresses occur when the molten
metal starts to solidify and cools in the mold. Secondary stresses occur when dif-
ferent parts of the casted component have different temperatures. The temperature
difference between the two parts results in one side being under tensile stress and the
other under compression, resulting in distortion in the final shape. The distortion
is a result of plastic deformation when the tensile stresses exceed the elastic limit
and the yield strength, and enter the plastic strain phase [16].

Residual stresses are also affected by the quenching process due to cooling differences
between the surface area and the interior of the component. Thermal gradients will
be induced which also affect the contraction of the casting. These gradients will dis-
appear, though their impacts on the component will be perceptible in later stages.
The uneven stress distribution is also a result of quenching and these quenching
stresses are a significant contributor to component instability during the following
post-processing operations e.g. machining. This often leads to geometric variation
in the components while the component is in service [17]. The combination of resid-
ual stresses and high external stresses during the use phase often results in cracks
in sensitive areas of the part [18].

2.4.2 Process Parameters
A study [16] was conducted to examine different process parameters in the casting
process, e.g. die temperature and cooling regime, to analyze their effects on the
distortion of the final casting shape. The study highlights that the different wall
thicknesses may be the cause of residual stresses, which in turn cause distortions.
With the presence of residual stresses, the ejection forces tend to cause distortions
during the extraction of the casting from the die. Nonetheless, smaller defects, such

11



2. Theory

as hot tears, porosities, blistering, or cold shuts, have a major effect on the casta-
bility of the part and thus have a significant impact on the geometry of the final
shape [16]. Hot spots, sink marks, differential shrinkage, thermal residual stress,and
warpage are example of quality defects of the casting that could be minimized by
having an efficient cooling system that achieves a uniform temperature distribution
[19].

Another numerical study [20] was conducted to reduce residual stresses and deforma-
tions of HPDC castings by testing several process parameters and establishing their
relationship with the stresses. To obtain accurate results, the authors simulated the
component using the commercial software AnyCasting, 3D scanned the casting, and
compared the simulation results with the 3D scans. The results showed that filling,
cooling/solidification, and ejection are process steps that lead to distortion in the
final shape due to residual stresses. These stresses are a result of uneven cooling as
well as non-uniform shrinkage [20].

The authors also claim that differences in wall thickness could cause differences
in the temperature gradients and lead to stress differences, see figure 2.10 for a
clearer description. These stress differences are the result of the thermal shrinkage
differences between the thick and thin areas during the cooling phase. In such cases,
the thinner parts of the component are exposed to compression, while the thicker
parts are exposed to tension stress [20].

Figure 2.10: Simple flowchart describing the differences in three stages

Two different components - a stress lattice and a V-shaped component, see figure
2.11, were simulated in the MAGMAsoft tool, to minimize defects affecting the
components’ shape through optimizing the filling and solidification phases [16]. The
results showed that differences in wall thickness are an important factor to consider
since they affect the geometry of the part when cooled due to inhomogeneous cool-
ing. Calculations to determine global distortion were also conducted using Hooke’s
law. The measurements are based on reference points before and after the cooling
phase.

12



2. Theory

Figure 2.11: a) stress lattice, b) V-shaped component [16].

The results also showed that the quenching phase influences the geometrical accu-
racy of the stress lattice [16]. This is due to temperature differences between different
parts of the stress lattice. However, the result showed that the die temperature had
more of an effect on the V-shaped component. High die temperature combined with
shorter time in the cavity results in higher temperatures during the ejection phase,
potentially leading to plastic deformation due to lower yield strength. The authors’
main conclusion is that the global distortion of a component also depends on its
geometry and thickness [16].

Another study that investigated the distortions and residual stresses was conducted
at the University of Jönköping in Sweden [21]. However, the experiment was based
on magnesium alloy. The component tested is an engine crankcase of a chain saw.
The process parameters studied were:

• (A) First-phase injection speed
• (B) Temperature of fixed half of the die
• (C) Cooling time
• (D) Intensification pressure

The authors investigated the best combination and the relationship between these
parameters to obtain accurate and reliable results. To achieve this, a Design of
Experiment (DoE) was conducted in the DesignExpert™software.

Different part thicknesses in the components could be the cause of non-uniform
cooling, and for complex geometries, it could create plastic deformations [21]. Fur-
thermore, large residual stresses with high temperatures could lead to permanent
distortion when the stresses relax/rest. Therefore, finding an optimal set of param-
eters could provide high-quality castings and reduce the geometrical variations [21].

The results after applying Analysis of Variation (ANOVA) showed that an increase
in the intensification pressure minimized the distortions at a certain point. The tem-
perature of the fixed half of the die is also a significant parameter because it increases
the total distortion and makes the part distort towards it. This also means that the
temperature difference between the two halves of the die is an important factor. In
addition, increasing the mobile die temperature reduces the residual stresses. The

13



2. Theory

study also revealed that cooling time inside the die and first phase injection had less
effect on the distortions and residual stresses [21].

Another study was conducted to meet geometric tolerances in HPDC castings [22].
According to the authors, the integrity of a casted component is influenced by a
combination of the following parameters:

• Die temperature
• Filling phase
• Components’ design complexity
• Cooling rates in the mold

They also claim that the die temperature, in-gate velocity, and casting pressure
have a relation to the mechanical properties of the component. The intensification
pressure is also an important factor for the production of high-integrity components
as it reduces the porosities. Deformations of the casted component in combination
with the porosity defects are major drawbacks of the HPDC process. This occurs
during the cooling phase in the mold cavity due to plastic stretch, caused by natural
shrinkage in the component, since it’s locked between the dies and not free to move.
This phenomenon contributes to the inducement of residual stresses and springbacks
during the quenching phase [22]. The authors also mean that the porosity problem
is a well-studied topic, while the deformations remain.

An additional study investigated the effect of mold and pouring temperature on the
hardness and microstructure of a HPDC hyper-eutectic aluminium alloy (AlSi17Cu4)
[23]. The study concluded that in order to get good quality, the pouring temperature
should range between 720 and 750 degrees together with preheating the die between
126 and 130 degrees. It also concluded that extremely high or too low pouring tem-
perature impacts the allocation of the primary silicon and the mechanical properties
such as hardness and density.

2.4.3 Extra-large Thin-walled HPDC Castings
A common feature in HPDC components is thin-walled castings due to economical
as well as weight savings reasons. This common feature is a contributor to distor-
tions in castings [22]

Effective Flow Length (EFL) is an approach introduced in a study [24] that asses the
castability and the quality of extra-large thin-walled HPDC castings (Aluminium
alloy). The flow length is the maximum distance that molten metal can travel in a
die. The studied and simulated component is shown in figure 2.12. Several tests on
different process parameters were conducted to study their relationship and assess
the quality of the final shape. Some of the parameters were the die temperature,
melt temperature, and cavity fill time (which corresponds to the in-gate velocity).

14



2. Theory

Figure 2.12: Representation of the extra-large thin-walled casting [24].

The in-gate system and the overflows for the studied component were optimized in
design as well as for the heating channels, vacuum system, and spraying control.
After running a test with the die temperature at 200 C, melt temperature of 720 C
and cavity fill time of 41.1 ms (which correspond to an in-gate speed of 50 m/s),
the results showed that it is more likely that the molten metal will solidify before
reaching the overflows due to a decreased temperature, see figure 2.13. Such differ-
ences in the solidification phase could affect the castability of the component since
it could affect the mechanical properties [24].

Figure 2.13: A representation of how temperature decreases throughout different
stages of the filling process [24].

15



2. Theory

Regarding the topic of extra-large thin-wall castings (in this case-study 2.8 mm
thickness), there is a combination of process parameters that affects the fluidity of
the molten metal into the mold cavity which in turn affects the castability and the
final quality of the casted part. The results were obtained from a similar study of a
DoE where multiple values of the parameters were tested.

The idea behind the EFL is to detect quality concerns by measuring the distance a
molten metal has traveled inside a die before it solidifies. If it is less than the EFL,
it indicates that the molten metal didn’t reach the overflows, and the cast part is
defective. This is because the overflows can capture dirty melt and air entrapment
and thus are important areas to consider. Therefore, one of the authors’ impor-
tant highlights is that the in-gate velocity is an essential factor for the castability
of extra-large thin-wall castings [24]. However, the study continued with another
experiment where the relative in-gate velocity was tested again, but with a higher
temperature of the molten metal to assess the castability as well as the combination
between the in-gate velocity, the temperature of the melt, and the die temperature.
The results were still almost the same, which indicates that the velocity still needs to
be high relative to the temperature of the die and the melt. To elaborate, increasing
the melt temperature and die temperature together with higher ingate velocity had
similar results.

Moreover, the authors also claimed that defects such as cracks, misruns, surface
flaws, and casting integrity are considered as important to study since castability
refers to good finishing as well as the accuracy of the geometry.

The heat loss of the molten metal during the filling process could cause premature
freezing as well as incomplete castings. Because of this phenomenon, fast accel-
erated filling becomes an important factor for thin-walled castings since it could
prevent such defects. To prevent this premature freezing, superheating the molten
metal could be a solution because it improves the fluidity. However, this approach
introduces other problems, such as gas-porosities [12].

16



2. Theory

2.4.4 Insights from MAGMA Simulation Runs
MAGMA Foundry Technologies, Inc. has uploaded a short video on Linkedin 5

that reveals a simulation result of the displacement of a mega-casting, a rear floor.
The result shows the distortion that the mega-casting undergoes during the whole
process chain of HPDC (solidification, cooling, and trimming). The result shown in
figure 2.14 is magnified to make the distortion more visible.

Figure 2.14: Displacement results for a rear floor mega-casting.

Another video that MAGMA has published is the temperature reduction the molten
metal undergoes during the filling process, see figure 2.15

Figure 2.15: Temperature results for a rear floor mega-casting.
5https://www.linkedin.com/company/magma-foundry-technologies-inc./posts/

?feedView=all

17

https://www.linkedin.com/company/magma-foundry-technologies-inc./posts/?feedView=all
https://www.linkedin.com/company/magma-foundry-technologies-inc./posts/?feedView=all


2. Theory

2.4.5 Vacuum Assisted Casting
One common defect in the process of HPDC is the air and gas entrapments in the
melt when injected into the mold cavity. Experiments were conducted to compare
the mechanical properties of aluminium castings using conventional HPDC and su-
per vacuum-assisted HPDC. The process of using vacuum-assisted HPDC involves
evacuating air from the mold cavity and the shot sleeve. This evacuation of air
seems to reduce porosities in the alloy and result in improved repeatability of the
properties. Moreover, the ductility was also improved and the fluctuations of it were
reduced [25]. The tensile strength was also enhanced. The study showed that the us-
age of super vacuum-assisted HPDC provides improved repeatability with enhanced
mechanical properties through a more uniform stress distribution. Fewer porosities
also mean that cracks will be reduced, and with a vacuum, the process will produce
high-quality die castings.

2.4.6 Soldering at the Ejection Phase
The ejection phase is the removal of the part from the die when it has solidified
and cooled down. Since the part shrinks onto the die, ejection forces in the form of
pins are used to push out the part [26]. Before locating the ejection pins, the forces
should be predicted for each pin in order to prevent casting warpage. However, it’s
still important to release the casting smoothly from the die [27].

One common defect during the ejection phase is the soldering effect. Soldering is
the sticking of the casting to the die and this defect often occurs in the hotter areas
of the die [28], [29]. This problem causes surface damage to the casting, leads to
dimensional defects [28], and increases the production downtime [29].

The general idea and assumption is that an increased interface temperature between
the mold and the casting would contribute to an intermetallic layer. Intermetallic
layers are chemical compounds formed between the die and the casting alloy and
cause soldering, see figure 2.16. This means the traditional point of view of the sol-
dering problem is through a thermodynamic and kinetic perspective [28]. However,
in industrial practices, the soldering could start much faster than the thermodynam-
ical models predict, due to the growth rates of these layers.

18



2. Theory

Figure 2.16: Cross-section that shows the soldering intermetallic layer [29].

A suggestion is to study and understand the mechanical interactions like galling,
which is believed to have a major impact on the soldering effect because it’s a wear
process caused by sticking between sliding surfaces [28]. The point is that to mitigate
the soldering effect during the ejection phase, a broader study should be conducted
to apprehend the mechanical interactions and not only the chemical and the thermal
ones.

It is also shown that aluminium alloys with silicon content tend to reduce the solder-
ing problem and the growth of these problematic layers. This means that increasing
the silicon content in the aluminium alloy decreases intermetallic layers’ growth and
thus minimizes the sticking effect [30].

Another important factor that reduces the soldering phenomena is the lubrication
of the mold before injecting the molten metal [31]. Since the lubrication acts as a
mold release agent (MRA), it creates a protective layer between the die surface and
the actual casting. In addition, the MRA improves the fluidity during the filling,
facilitates the ejection of the casting, and reduces the unbalanced die temperature
[31].

2.4.7 Alloy Composition
Although the trend of mega-casting and Giga press has been shown in recent years,
a book from 2012 [9] did discuss the idea of larger castings and specifically discusses
the HPDC process for automotive parts. A chapter also discusses the alloy compo-
sitions for such automotive applications.

The BMW Series 5 GT rear lid is shown in figure 2.17. According to the authors,
the dimensions of the casting are accurate for a 3 mm thick walled casting. One of

19



2. Theory

the requirements for this larger casting is the long-term crash properties and this is
fulfilled by an aluminium casting which also allows a great deal of design-freedom
than steel alloys.

Figure 2.17: The GT models’ lid frame. Weight: 11.6 kg. Dimensions: ca. 1230
x 1250 x 390 mm [9].

Another part that was discussed in the chapter is an integral crossbeam, also for a
BMW car, see figure 2.18. The authors’ claim is that the quenching phase would
cause a distortion of the arms [9]. This geometrical variation in the final shape
does not meet the requirements for high dimensional accuracy and would therefore
be unsuitable to use. What makes it unfit for usage is that aluminium casting is
welded to the rest of the chassis at several points [9].

Figure 2.18: BMWs’ cross beam. Weight: 4.1 kg. Dimensions: ca. 810 x 450 x
210 mm [9].

20



2. Theory

The continuation of casting larger components with complex designs makes the tra-
ditional AlMg alloys difficult to use due to poor flowability. However, by adding
silicon to the alloy (AlSiMg) the flowability would be improved [9]. This means
that the alloy composition is also an important factor to consider for producing
accurate die-castings. One of the most important elements for die-casting alloys is
silicon since it improves the castability, fluidity, and hot tear resistance [7]. With
subsequent quenching, the desired properties could be achieved. However, solution
heat treatment would be a problematic process due increased risk of distortions [9].
Silicon is also responsible for low shrinkage, and low density which is beneficial in
the weight reduction of products[32]. Increasing silicon content also increases the
ultimate tensile strength [32].

Another part mentioned in the chapter is Audis’ rear connector for the A8 model,
see figure 2.19. This component is relatively long compared to the other mentioned
castings, 1.45 m long, and weighs 10.2 kg. Reinforcements in the form of ribs need
to be strategically located to balance the weight, and this is done by mapping the
loads [9].

Figure 2.19: Audis’ rear connector. Weight: 9.9 kg. Dimensions: ca. 1454 x 375
x 552 mm [9].

2.5 Quenching
Strength, ductility, and thermal stresses are examples of material properties that
are affected by the cooling process of age-hardenable aluminium alloys. Thermal
residual stresses are closely related to the uneven temperature distribution within
the metal [18]. Reducing the cooling rate from solution heat treatment minimizes
thermal stress [10]. However, solute atmos diffuse to grain boundaries and disor-
dered regions when the quenching rate is not rapid and thus provides no enhanced

21



2. Theory

strength during age hardening [33]. On the other hand, the rapid cooling rate in-
creases the risk for distortion [10] and residual stresses[33]. Thus, optimizing quench
parameters and minimizing the unwanted deformation becomes the main challenge
in quench setup [10]. The most significant factor controlling the cooling rate is the
media temperature [34].

Generally, the hardness of the as-quenched, which is the part after being quenched,
is enhanced with higher cooling rates. The enhanced hardness is due to solute ele-
ments locked in the solution at a high cooling rate. With decreasing cooling rates
and lower homogenization temperature, the tendency to form non-hardenable pre-
cipitates or in other words, "Quench Sensitivity" increases. [10].

A metal sheet of aluminium alloy (AA6082) with dimensions of 20x20x2 mm was
quenched in different media. Figure 2.25 shows the cooling curves for different media
with different start temperatures. It also indicates the cooling rate e.g. still water
has the highest cooling rate while cooling in still air has the lowest cooling rate. [10].

Figure 2.20: Cooling curves for AA6082 alumnium sheet quenched in different
medias [10].

.

22



2. Theory

2.5.1 Quenching Mechanism
Generally, when a hot casted part is immersed into a liquid quenchant, the heat
transfer undergoes three different phases/stages [35].

1. A stage: Vapor phase
2. B stage: Boiling phase
3. C stage: Convection phase

Film boiling is known as the initial and the relatively slow step because the hot sur-
face gets enclosed by a vapor blanket, which slows the cooling process[10]. During
this step heat is mainly transferred by radiation [33]. The vapor blanket is very
steady and has an effect on the creation of surface soft spots [33].

The vapor blanket breaks as the part cools [10]. Nucleate boiling is the second stage
and it starts after the breakage of the vapor coat [10]. The temperature separating
film boiling and nucleate boiling is called "Leidenfrost temperature." During nucleate
boiling the heat transfer is fastest, the heat transfer coefficient, α, for film boiling,
is 100x lower compared to α during nucleate boiling [10]
.
The last stage is the convection cooling during this stage the heat is transferred
from the quenched part to the liquid by convection [33]. It is also influenced by
factors such as the specific heat, thermal conductivity, and temperature difference
between the quenchant and quenched part. In addition, it is usually known as the
stage where most distortion arises and the stage with the least efficient stage of three
stages [33].

Figure 2.21: Immersion quenching and its cooling stages and mechanisms [35].

2.5.2 Cooling Rates and Quench Severity
The cooling rate is influenced by multiple elements [33]:

23



2. Theory

1. Type of Quenchant:
• Water
• Air
• Polymer
• Liquid nitrogen

2. Quenching Techniques:
• Immersion
• Spray
• Still air
• Air jet

3. Material Properties and workpiece characteristics :
• Thickness
• Design
• Surface quality
• Workpiece’s internal characteristics affecting heat transfer to the surface

The quench severity is dependent on several factors such as:
• Development of vapor blanket
• Boiling properties
• Velocity of the fluid
• Tempreture of fluid
• Specific heat of the fluid
• Heat of vaporization
• Conductivity
• density
• viscosity
• Wetting properties of the fluid

2.5.3 Quenching Media
Water

Water is the most commonly used fluid for quenching aluminium alloys. In addition,
it has many benefits such as inflammability, cost-effectiveness, non-hazardous, and
easy filtration. But it also has drawbacks with the residual stresses and distortion
because the rapid cooling rate is achieved at the lower temperature ranges. An-
other drawback is that it could give a long vapor blanket stage depending on part
complexity and water temperature [33] see figure 2.22. Film boiling is expected
at the first stage when quenching the aluminium solution into water. Because the
heat treatment temperature of the aluminium solution is notably higher than the
Leidenfrost temperature [10]. Stable wetting and nucleate boiling will occur at sur-
face areas where the surface temperature is less than the Leidenfrost temperature
[10]. The existence of different boiling phases on a workpiece with different heat
transfer coefficients of factor 100 creates unequal quenching. Furthermore, this will
significantly impact the wetting procedures and lead to notable thermal gradients
and amplified distortion during quenching[10]. Increasing the water temperature

24



2. Theory

stabilizes the boiling mechanism. Figure 2.22 illustrated how water temperatures
above 54◦ [10]. Several factors need to be monitored to get repeatable quench results
such as the temperature, agitation, and contamination because they affect the vapor
blanket formation and heat transfer [33].

Figure 2.22: Cooling curves as a function of water temperature for 1/2 in. 7075
aluminium platec [10].

The non-uniformity of cooling comes with consequences: [[10],[36],[37]].

• Zones where film boiling (slow cooling) endures have more precipitate for-
mation compared to zones with rapid cooling. During age-hardening, zones
with slow cooling encounter a slight increase in hardness but in contrast, have a
higher risk of intergranular corrosion than zones with a faster cooling rate [10].

• Zones where film boiling prevails during quenching have much lower yield
strength compared to zones with marginal film phases. The uneven cooling
process introduces thermal stresses that will lead to notable plastic deforma-
tion and increased distortion [10].

Quenching aluminium alloys in cold water may produce undesirable distortions as a
result of the high thermal gradients [10]. One solution for this was delay quenching,
which means quenching the aluminium alloy in boiling water and then immersing
it into cold water at a suitable time [10]. Another solution is "hot water quenching"
(60-71 ◦C) if distortion challenges arise with cold water (10-32 ◦C) [10]. Figure 2.23
shows how tensile strength varies with water temperature.

25



2. Theory

Figure 2.23: Tensile strength of 1 in. 7075 aluminium plate as a function of water
temperature [10].

Generally, the highest strength obtained from age hardening is linked with quick
quenching rates. However, a maximum quenching rate is not always the selected
option, because it’s accompanied by both residual stress and distortions [33]. Typ-
ically, the cost for optimum attainable properties is elevated residual stresses or
distortions. On the other hand, low residual stresses and distortions are attained at
the expense of properties [33].

In the majority of instances, distortion arises due to the severity of the quenching
process. However, there are other critical factors. One of them is the racking. In
order to control distortion in the furnace and quenching process a suitable racking
becomes a very critical factor. The reason for this is:

• However, it’s not only about enlargement, if parts are constrained and pre-
vented from expanding or contracting evenly this will create substantial resid-
ual stresses or distortion. In addition, fixtures and the part have different
expansion characteristics and therefore a careful plan needs to be made to
prevent the different expansion characteristics because this will result in high
levels of residual stresses and cause geometrical variation.

• Racking Orientation: the orientation of quenching the part after heat treat-
ment has a direct impact on the degree and type of warpage. Different racking
orientations such as horizontal, vertical, or angular lead to different types of
warpage. Parts that are racked in a horizontal orientation tend to bow, how-
ever, parts that are racked vertically are susceptible to twisting if the mass is
adequate.

26



2. Theory

• The configuration is the deciding factor for how fast the immersion will be.
Usually, a slower immersion is used. However, some configurations are not
dependent on how well-designed the racking is due to the existence of surfaces
preventing the quenchant from reaching all areas, in such cases a slower im-
mersion rate is applicable.

• Condition of the basket, racks, or fixture: Distorted parts result from distorted
baskets, racks, and fixtures.

Polymer Quenchants
For many years, hot water quenching has been utilized to reduce quench distortion
in aluminium, but the severe distortion was not eliminated [10]. Quenching in water
doesn’t uniformly wet the surface of aluminium. In addition, nonuniform wetting
creates different regimes with different heat transfer resulting in enough thermal
strain to distort the part [10]. Furthermore, this resulted in the adaption of a new
quenchant, aqueous poly (alkylene glycol)-PAG copolymer solution. It improves
the wetting uniformity and creates a film around the quenched part that transfers
heat. Polymer quenchants are water-based solutions, containing a polyalkylene gly-
col copolymer and additives that prevent corrosion [10]. Figure 2.24 illustrates the
benefits of using PAG quenchants compared to both cold and hot water quenching.

Figure 2.24: Comparison of water quenching and PAG 20 percent polymer quench-
ing and its impact on distortion for a plate with 1.5 mm thickness. Upper left panel
is 300×250 mm; Upper right panel is 100 mm×300 mm; lower left and right panels
are 300×300 mm [34]

2.5.4 Spray Quenching
Spray quenching (SQ) is a technique that could range from heavy to fog spray
depending on the application. It’s usually for long parts, plates, sheets, structural
shapes, and extruded parts. The advantage of using SQ is the intense quenching
because it eliminates steam pockets. But it also requires careful planning and setup
of the nozzles, their number, and location to create a uniform cooling that could
prevent unfavorable distortion problems. In addition, there are two ways to control
the quench rate in spray quenching. The first is the velocity of the water, the
second controlling factor is the volume of water per unit area per unit time of the

27



2. Theory

interference between the water and the workpiece [33].

2.5.5 Study on Different Quenching Types and Rates
A study [38] investigated the effect of different cooling rates on microstructure,
residual stresses development and mechanical properties for high vacuum die casting
AlSi10MgMn alloy. The different tests were based on SQ, air cooling (AC), wind
cooling (WC), and water quenching (WQ). The SQ had three tests with different
nozzle distances, air pressures, and water pressures see figure 2.26.

Figure 2.25: cooling curves at each quenching condition [38].

Figure 2.26: The parameters of the investigated quenching conditions [38].

28



2. Theory

Figure 2.27: calculated IHTC curves for different quenching methods [38].

Figure 2.28: Boiling curves with different spray quenching parameters [38].

According to figure 2.27, the WQ had the highest interfacial heat transfer coeffi-
cient (IHTC) compared to the other quenching methods. It can be seen in figure
2.28 that SQ30 had a film boiling phase, unlike SQ100 and SQ300 which didn’t
experience it. They had transition boiling which is a phase where vapor film is very
unstable and collapses constantly due to the different heat transfer between droplets
and contact area. SQ30 resulted in small and spread-out droplets because it had the
highest air pressure of 0.5 MPa and the lowest water pressure of 0.2 MPa. The small
droplets resulted in the formation of vapor film because there was no major decrease
in temperature when they touched the surface. In addition, the film could prevent
the other droplets from reaching the surface and extracting heat which resulted in
lower IHTC. However other factors such as the long distance between the nozzle and
workpiece (170 mm) resulted in low droplet speed and weak impact on the vapor
film. In contrast, SQ100 and SQ300 had better heat transfer and reduced surface
temperature drastically see figure 2.27. Because they had high to medium-high wa-
ter pressure (0.3-0.4 MPa) and a shorter distance between the workpiece and nozzle
(80 mm) 2.26. In addition, it resulted in no film boiling in SQ300 and SQ100 see
figure 2.28.

29



2. Theory

The high cooling rate (SQ100, SQ300, WQ400) resulted in high density and small
Beta” precipitates which also gave high tensile strength and hardness and residual
stresses. In addition, it resulted in narrower precipitate-free zones (PFZ).

The low cooling rate AC1.5, WC4.5, resulted in wider PFZ, the presence of Beta’
and Beta’ precipitates, and resulted in weaker material properties.

The study showed that a cooling rate at 30 C/s provided the wanted mechanical
properties which is 85 percent of the ultimate tensile stress (UTS) of as-cast alloy.
Because the part will be aged and more strength will be acquired. It also showed
that the residual stresses, yield stress (YS), and UTS increased with the increase
in the cooling rate. It also showed that elongation decreased with the increase in
cooling rate.

Figure 2.29: Comparison of the mechanical properties and residual stress of the
alloys with different quenching cooling rates [38].

2.5.6 Quenching Orientation
A CFD study [18] investigated the effect of orientation by simulating six different
orientations of an aluminium cylinder head quenched in water. It can be seen in fig-
ure 2.31 that joint face (JF) orientation has the highest vapor entrapment. However
this didn’t impact the cooling curves, figure 2.32 showed that the cooling curves are
almost identical. Figure 2.33 shows that orientation has little impact on the overall
cooling characteristics. However, its effect is localized in specific locations. Another
important observation is that the duration of peak temperature gradient lasts for 15
seconds and in these 15 seconds metal could exceed yielding stress and plastic defor-
mation begins, see figure 2.33. Another study [39] on AlSi 4340 steel found a strong
correlation between different quenching parameters quenching orientation, heat ex-
traction speed, type of quenchant, and geometry of the part. The highest distortion
values were related to horizontal water quenching, while the lowest distortion values
were connected to oil quenching with vertical orientation [39].

30



2. Theory

Figure 2.30: Water quench orientations: RE - rear face up, CC - cover face up, IN
- intake port up, FR - front face up, JF - joint face up, EX - exhaust face up [18].

31



2. Theory

Figure 2.31: Vapor pattern and vapor entrapment inside cylinder heads after 20
seconds of quenching [18].

Figure 2.32: Cooling curves for all water quenching configurations [18].

32



2. Theory

Figure 2.33: Maximum temperature gradients for all water quenching configura-
tions [18].

2.5.7 Immersion Rate
Another factor that plays an important role in reducing distortion of quenched alu-
minium is the immersion rate [34]. In addition, two aspects that are often confused
to be the same are the immersion rate and quench delay. However, immersion rate
is the rate that the part enters the quenchants and this aspect is often disregarded.
Quench delay refers to the time span between the end of heat treatment and the
beginning of quenching. In a brief experiment [34] the effect of immersion rate was
studied by having a quick and slow immersion rate inside 32% PAG polymer quen-
chant. In the first sample, the immersion rate was quick, and in the second sample,
the immersion rate was slow with the same quench delay for both samples. The
second sample with a slow immersion rate had severe distortion compared to the
quickly immersed part, this can be seen in figure (2.34).

33



2. Theory

Figure 2.34: Sheets are 500×500 mm×1.0 mm thick immersed in PAG with two
different immersion rates. On top, a sheet with slow immersion rate. On the bottom,
a sheet with quick immersion [34].

34



3
Method

This chapter covers the methodology used for conducting the thesis, with the main
focus on HPDC and recent trends in this field. It presents the screening method of
the literature study, data analytics, and simulation. Since the topic of Mega-Casting
is still in the evolving stage - very few studies or research have been published and
thus a pioneering approach was used in this study.

3.1 Methodological Framework
The selected design research methodology, DRM, is inspired by the DRM presented
in [40]. It consists of 4 steps. The first step is about Research Clarifications which
starts by understanding the background and problem area of the study. It’s also
about defining the aim and objective of the research. Furthermore, constructing
and aligning the research questions with the aim of the study. This step also in-
cludes setting the main foundation of the research, such as defining the scope and
delimitation of the research. These steps are accomplished through conducting a
preliminary review of the literature.

The second step is the descriptive study. Since a preliminary or even comprehensive
literature review doesn’t provide the complete picture, a descriptive study is made
to expand the understanding by collecting empirical data. The imperial data is
qualitative data gathered from the expert interviews.

The third is the prescriptive step. It involves testing the findings from the literature
review and the qualitative data gathered from interviews in order to further analyze
the selected solution.

The fourth step is descriptive study, which includes analyzing and further testing
the optimal solution. This falls outside of the scope of the study therefore a future
recommendation was suggested.

3.2 Screening of Literature Study
In the literature review stage, a screening method was used to identify the rele-
vant sources that matched the scope of the thesis. Moreover, to find the scientific
sources, only scientific search engines and databases were utilized, such as Springer,
ResearchGate, Elsevier (ScienceDirect), Chalmers Library as well as Google Scholar.

35



3. Method

This ensured the credibility, reliability, and relevance of the findings.

The following steps were used, see figure (3.1):

1. Select Specific Keywords

Since the topic is about geometrical variation in HPDC large castings, some
keywords needed to be included in the search engines for the sake of relevance
and filtration. Such keywords are:

• geometrical variation
• dimensional variation
• distortions in HPDC components/parts,
• Hyper-casting
• Giga-press
• plastic deformation
• shrinkage
• twisting
• large casting

2. Examine the Abstract and Conclusions

After the first filtration step i.e. when finding a relevant headline/title of a
source, the next step was not to analyze the whole source but to review the
relevance through the abstract and the conclusions. If it was in the scope of
the study and was related to distortions in HPDC castings, it was saved for
the next step for detailed investigation and analysis.

3. Examine the Whole Source

After the second step, the saved sources were analyzed, organized, and docu-
mented for later usage in the thesis. Furthermore, to verify the findings, some
of the primary references were also analyzed to gain a deeper and compre-
hensive understanding but also validate the findings. Hence, a jump between
steps- 2 and 3 was also a viable method applied.

4. Organize and Combine

Lastly, to increase the clarity of the literature review, the articles were further
organized into different subareas such as pre-quenching, quenching, and post-
quenching. This was also done to easily detect any contrast between the
findings.

36



3. Method

Figure 3.1: A funnel-like method for literature study.

3.3 Qualitative Study
Since mega-casting is a new topic, a large emphasis was placed on finding rele-
vant experts with extensive years of experience in the field of HPDC. In addition,
efforts were made to find interviewees outside the organization to expand the cur-
rent knowledge. By conducting interviews with technical experts and academic
researchers, the casting manager, in the qualitative phase of the thesis obtained
several perspectives and insights. The interview questions were formulated based
on the whole process chain of HPDC to ensure a thorough analysis, gather detailed
insights, and easily find contractions between interviewees. Furthermore, the in-
terviews were transcribed and analyzed by using the Affinity Mapping approach
meaning that insights from each interviewee about a specific area were written on a
post-it note and transferred to the specific area of HPDC.

The interview questions were semi-structured, meaning that the same predefined
questions were asked for all interviewees but with some level of flexibility. The
semi-structured allowed for organizing the answers in easier way and enabled, in-
depth questions such as follow-up questions. In addition, the questions had small
adjustments for the material expert and residual stress PhD student to ensure rele-
vance. However, the questions were the same and followed the whole process chain
of HPDC.

LinkedIn was the main platform to find the interviewees but some were recommended
by the supervisor and interviewees themselves. Meaning, at the end of every inter-
view, the interviewees were asked if they could recommend other experts. Most of
the interviews were conducted digitally but some were conducted in person. Almost
all interviews lasted around 1 hour.

37



3. Method

3.4 Simulation
The findings from the literature study, and interview phase, will be further analyzed
and verified in a casting simulation software. For this, a smaller part was designed to
compensate for the long simulation time and the limited time frame of this project.
Important to note that this is out of the scope of this project. The designed com-
ponent had the essential attributes found in a large and complex mega-casting.

3.5 Quantitative Study
In order to verify some of the findings in the literature study and the interview
insights, a quantitative study was also conducted. The quantitative part of the
thesis is based on measurements and 3D scans of the mega-casting and some manu-
al/rough measurement data. The data were provided by the factory engineers and
the supervisor. A small data analysis regarding the production data of the Volvo
mega-casting was carried out to extract valuable insights regarding the geometrical
accuracy of the mega-casting. The data were cleaned, due to missing values, and
plotted with Excel and MATLAB.

3.5.1 Tests in Foundry
Two sets of tests were conducted by the foundry engineers and the supervisor. The
first set of tests was conducted under normal conditions. Most of the parameters
were the same. However, the second set was conducted mainly to examine the effects
of fast quenching (low quenching time) and exclude some parts of the mega-casting
from quenching. A more comprehensive description of set 2:

• Test 1: Quenched with a very long quenching time and without changing the
original quenching orientation

• Test 2: Very short quenching time and without dipping some areas
• Test 3: Different quenching sequence and without quenching some areas
• Test 4: Same as test 3, to elaborate, same quenching sequence and with the

same areas not being quenched as test 3 but with increased die temperature
(+30 Celcius)

• Test 5: Same as test 4 but with more areas not being quenched
• Test 6: Same as test 4 but with increased time for solidification inside the die

(+30)
• Test 7: Same as test 4 but with increased water tank temperature and medium

quenching time

38



3. Method

3.6 Ethical Aspect
To protect privacy and confidentiality the interviewees were asked if they consent to
have their names published. The method used to protect their privacy was coding
for anonymity, by assigning a label to each interviewee e.g., E1, Expert 1. More-
over, adding their job titles provided context without compromising confidentiality.
In addition, due to the confidentiality agreement with the company, all sensitive
information and data will be hidden from the public. Numbers, figures, scales, and
sensitive information will remain undisclosed.

39



4
Results

This chapter contains the project results and a summary of the interview phase.
Important to note that the results were based on the real mega-casting, however,
the mega-casting pictures show a prototype.

4.1 Simulation
Performing simulations was outside the thesis scope, however, an attempt to verify
the qualitative findings on simulation software was made. Due to unknown param-
eters and the complexity of the software, the simulations turned out to be very
time-consuming. Despite these obstacles, a CAD model was designed and could be
used in future research to validate the findings through simulation runs.

4.2 Data Analysis
For the first set of tests that were conducted in the foundry, multiple mega-castings
were produced in the Volvo factory to be scanned and measured for geometrical
variations. Figure 4.1 shows the geometrical variation of different points (total of 48
points) that have been scanned on the actual mega-casting. The middle line shows
the nominal measurement and the critical points can be detected through the plot.
The crucial points, marked in red, are - p13, p14, p17, p18, p21, p27, p28, p37, p38,
and p48.

Figure 4.1: Inspected points on the mega-casting.

40



4. Results

These critical points are located on the front legs and the upper part of the wheel-
house, see Figure 4.2. The scanning results also showed that the bending of the
front legs is inwards and the wheelhouse outwards as shown in the same figure with
red arrows.

Figure 4.2: Red ellipses marking the sensitive areas and red arrows the bending
behavior. Picture adapted from [41].

Some graphs from figure 4.1 were filtered for analyzing the vacuum effect on the
variations of the points. The red graphs in figure 4.3 represent casting runs with no
vacuum and in the blue ones, the vacuum was activated.

Figure 4.3: 4 Red graphs represent no vacuum-assisted casting runs while 3 blue
graphs had vacuum assist.

41



4. Results

Figure 4.3 shows that the vacuum-assisted runs are producing more variations in
the critical points of the mega-casting (e.g. p18, p27, p28, p37, and p38). However,
apprehending this observation regarding the runs of the mega-casting, a conclusion
cannot be drawn. This is because other more significant parameters are unknown,
such as the die temperature and cooling time inside the mold cavity. Furthermore,
for robust analysis, the cycle runs are crucial because the runs with no vacuum as-
sistance were conducted in later stages, and thus tool conditions were unknown.

Since the front legs appeared to be a sensitive area, the distance between the legs
was measured in two places – distance 1 and distance 2, see figure 4.4. A plot was
generated to show the distance variation between the front legs. Figure 4.5 shows
that tests 6 and 7 had the closest value to the nominal distances 1 & 2.

Figure 4.4: Distance measurement between the front legs. Picture adapted from
[41].

42



4. Results

Figure 4.5: Different tests and their effect on distance 1 and 2.

The effect of the water temperature on the distortion was analyzed in figure 4.6 where
each circle represents a measured instance. The x-axis of the plot represents the
water temperature and the y-axis is the distance between the legs. The plot shows
that the water temperature has an impact on the distortion. In addition, as the water
temperature increased, the distances started to converge near the nominal distances,
presented by the blue and red straight lines. Distance 2 had better results compared
to distance 1, and it can be seen that distance 2 reached the nominal distance in the
highest water temperature measured (in the top right corner). However, increasing
the water temperature and reducing the cooling rate led to another problem during
the trimming process. Since the part was hotter after being quenched, it got stuck
to the trim-press.

43



4. Results

Figure 4.6: Quench temperatures and its effect on the measured distance 1 and 2.

Another test of the quench tank’s temperature was conducted to see the effect of
colder water on the distortion, see figure 4.7. The X-axis represents the temperature
drop of the quench tank. The plot shows that, as the water cools, the more the
distances diverge from their nominal value. The plot also showed a stabilization
after a certain level of temperature.

Figure 4.7: Cooling effect of quench tank on distance 1 and 2.

44



4. Results

4.3 Summary of Experts Insights
A total of 10 interviews were conducted with knowledgeable people in the field of
HPDC and die-casting. Additional 2 interviews were also conducted with a Ph.D.
Student and a Materials experts to expand the knowledge of residual stresses. The
interviewees are referred to as E(Number) (E for Expert) and their current work is
also mentioned below.

• E1: Casting expert and CEO of a consulting business in HPDC
• E2: Ph.D, Assistant Professor Materials and Manufacturing - Casting
• E3: Professor in materials and manufacturing - casting
• E4: Associate Professor in Materials and Manufacturing - Casting
• E5: Site and Sales Manager of a casting company that focuses on die castings
• E6: Researcher in the field of HPDC
• E7: Technical Specialist casting
• E8: Casting expert and CEO of a casting company
• E9: Technical casting expert
• E10: Die Casting Tool Engineer
• E11: Ph.D student in residual stresses
• E12: Materials Expert

4.3.1 Gating System
All the interviewees agree that the gating system of mega-casting is a very important
factor to consider when analyzing distortions. However, they highlighted different
aspects of their claims.

According to E1, the gating system is typically way thicker than the actual casting,
and thus a proper design is important to minimize the geometrical variation, espe-
cially when undergoing quenching.

E2, E3, E4, and E8, do agree with E1 and claim that a well-designed gating system
is crucial for the:

• Flowability of the melt into the mold
• The temperature loss of the melt when injected
• Accuracy of the mega-casting
• Turbulence avoidance

E9 stated that the gating is a massive mass that is needed to make the mega-casting,
and since it doesn’t become the part, its weight should be minimized. However, both
E9 and E4 explained that a good flow of the injected melt should be an even flow
and for such large structural automotive parts, the gating system may be thicker
than intended. The aspect of flowability was also discussed by E5, which explained
that the main purpose of the gating system is to transfer the melt in a good way
and thus, the turbulence should be reduced to achieve high-quality castings. E4

45



4. Results

also highlighted the turbulence issue which means that it’s an inevitable problem
that will cause the inducements of small oxides in the melt. This also means that
the gating design will affect how these oxides will emerge in the melt and could
cause casting defects. E5 claims that, for larger castings, this issue could affect the
bending behavior after the casting has solidified.

E3 stated that the wider the gating system is, the wavier the flow will be, which
also means more defects. The professor (E3) continued by explaining two different
gating designs and their differences:

• Fingers-like gating (similar to Xiaomi): made for a less wavy fill front, but
fusion lines will occur.

• Fan-like gate (similar to the Tesla): Fusion line issue could be controlled, but
a wavy fill front would occur.

However, he also suggested that, for the finger design, the fusion line issue could
be controlled through proper pressure distribution in certain places. E3 means that
the placement of the fingers could be optimized to restrict the twisting of the part.

Another important factor that was discussed by E3, E4 and E9 is the thickness of
the gating. Since the gating also undergoes the solidification process, it will contain
a lot of heat and thus will affect the temperature distribution of the mega-casting.
For this reason, some of the experts mean that minimizing the weight of the gating
system could benefit the accuracy of the mega-casting since the whole casting part
would have a similar solidification rate. This phenomenon would result in less ten-
dency or less total energy to cause warping of the casting, means E3. However, for
better flowability of the melt, the thickness of the gating might be larger.

E2 and E3 claims that the geometrical accuracy of a large casting could be influenced
by the flow of the melt and thus understanding the behavior of the bending effect
in the nearby areas of the part is also important.

4.3.2 Shot Sleeve
The experts don’t directly mention any connection between the shot sleeve velocity
and the geometrical accuracy of the casting, instead, they claim that it’s a matter
of the locations of the air entrapments in the casting. E8 claims that the shot sleeve
velocity needs to be "good enough" to prevent too much splashing into the mold.

Otherwise, without exception, all experts agreed that the shot sleeve velocity (second
phase) has minimum effect on the geometrical variations or any type of distortion
compared to the other parameters, such as the die temperature.

46



4. Results

4.3.3 Die Design
Important to note that the die design also refers to the placement of the cooling
channels.

One important point that E1 mentioned is the importance of draft angles. Since the
part shrinks and sticks to the die, the ejection forces may be increased to be able
to eject the part, which could result in deformations. To prevent this phenomenon,
the design could be changed and draft angles could be introduced.

Another innovative way that was discussed by E1 and E10 is to design the part
inaccurately, and later compensate for the distortions through the solidification and
quenching phase. However, E10 means that this concept is very limited to certain
designs and that the manufacturer should be very certain of the distortion behavior
from the nominal shape. This approach is known as design for manufacturing or
could be coined “design for distortion.”

According to E1, E2, and E8, in the future, 3D printing could be used to optimize
the design of cooling channels. E1 also mentioned jet cooling and explained that this
method can be timed and easily programmed (active control). He also claims that
jet cooling facilitates the cooling of different wall thicknesses equally. However, the
most important aspect is to have a continuous solidification that goes back into the
shot sleeve. E2s hypothesis is that a die cannot have "the best" cooling system that
compensates for different wall thicknesses. However, he believes that 3D printing
may make it possible.

E8 also claims that die design is very important since it’s mainly about solidifica-
tion and cooling. He highlighted that the different solidification rates would cause
tensions in the casting. In addition, the hot spots area would have more sticking to
the die and result in larger tension when ejected.

4.3.4 Die Temperature
E1 claims that it’s desirable to have uneven temperature distribution for HPDC
castings, especially mega-casting. The expert reasons that the temperature at the
end of the mold should be hotter than the gating area because the molten metal so-
lidifies before reaching the end. However, E1 also states that when the mold is filled
an even temperature is preferred. E2 agreed that preheating the end part of the
mold would be beneficial but would be complex to implement. E8 also highlighted
the problem of melt starting to solidify before fully filling the mold. The expert also
mentioned that such a problem could cause cracks. However, E1 also stated that
when the mold is filled, an even temperature of the die is preferred.

47



4. Results

Thermal History of the Die

E1 mentioned that the type of spraying, conventional spray or micro spraying com-
bined with how cold the die is, will have a major impact on the lifetime of the tool
because the temperature difference between the liquid metal and the surface of the
dies determines the fatigue load. In addition, having a cooler tool means quadru-
pling the fatigue load. Therefore having a hot tool is beneficial for the casting, E1
means.

4.3.5 Melt Temperature
According to E1, the melt temperature affects the mechanical properties and the
cold shuts more than it affects the geometrical accuracy. He also highlighted that
the sticking of the overflows and the gating system to the fixed die could cause ge-
ometrical variation.

According to E8, even if the filling time is in milliseconds (∼500 milliseconds), the
melt temperature will change and start to solidify. On one hand, he points out that
having a high temperature means a larger solidification range, thereby creating a
hot and cool metal with a larger temperature difference in the same tool, leading to
more tension. E5 also stated that the melt temperature does affect the shrinkage
and a higher melt temperature means a higher risk for distortion of the part. On the
other hand, E8 explains that too low temperature would cause the melt to solidify
during injection, leading to difficulties in reaching the thinner parts far away from
the gating system. Therefore the temperature of the metal is critical and thereby
the properties of the metal according to E8. E4 claimed that, depending on the
part design and cooling rates, the melt temperature could increase/decrease the
distortion effects. Moreover, he mentioned that It’s a complex combination of the
parameters to optimize the casting and its quality.

4.3.6 Alloy Composition
E1 claimed that pure aluminium has a shrinkage rate of approximately 6.6% and
adding silicon to the alloy composition would benefit the casting since silicon ex-
pands during solidification, leading to compensation for the shrinkage. E2 and E4
agreed with E1 and claimed that silicon in the alloy would be useful since it also im-
proves the fluidity of the melt during the injection and thus improves the castability.
E7 also mentioned that silicon is important for improving fluidity. E8 also noted
the improvement of the fluidity and added that silicon content affects the melting
temperature.

However, E4 stated that silicon generates more heat during the solidification process
and the transformation phase could affect the geometrical accuracy of the casting.
A significant amount of heat is also generated when different areas of the mega-
casting solidify at different rates. This heat can cause the already-solidified areas
to warm up again. He also highlighted that adding silicon to the alloy composition
can enhance the strength of the casting. E9 claims that “from a theoretical point of

48



4. Results

view”, silicon could affect the final dimensions of the mega-casting, but the impact
wouldn’t be huge.

Regardless, E4 highlighted that it cannot be concluded that residual stresses in-
crease due to the higher temperature associated with the silicon content, and cause
distortions. Instead, E4 claims that the distortion phenomenon is related to the
cooling rates and the geometrical variations of the different sections in the casting.
Nevertheless, E4 also noted that if the geometry is “well designed” to mitigate the
stresses, then it’s still “good to go” even with high silicon content. Consequently, a
greater amount of heat needs to be managed and extracted from the mold.

4.3.7 Vacuum Assistance
E2 stated that HPDC with vacuum assistance reduces the air inside the mold before
the filling process, contributing to fewer defects in the material. E4 noted that
vacuum assistance makes it easier to fill the cavity and is also good for avoiding
trapped air. E8 pointed out that vacuum is important because it affects the porosity
and indirectly the geometrical variations. In addition, according to E3, the vacuum
assistance affects the cooling conditions of the die and does not have a big impact
on the final geometry of the casting. E9 agreed with E3 and claimed that vacuum
has minimum impact on the casting in terms of geometrical variations.

4.3.8 Intensification Pressure
The effect of intensification pressure depends on the size and shape of the part ac-
cording to E2, E3, E4, and E5. However, E9 claims that the intensification pressure
does not have a big impact on the geometrical accuracy of the casting. Areas far
away from the ingate, in a complex mega casted part, might not have the effect of
the intensification pressure like areas close to the ingate according to E2. In other
words, E2, believes that it wouldn’t affect the whole component, only areas close to
the ingate. E1 highlighted that intensification pressure is very important to com-
pensate for the shrinkage porosity during the solidification and prevent degradation
of mechanical properties. In addition, E1 explained that during solidification, the
molten metal starts to form dendritic shapes, and if there are a lot of dendrites, a
microporosity would emerge and impact the properties. Therefore the setup should
be correct from the beginning according to E1.

E3, E4, and E5 means that the intensification pressure could be a contributing factor
if the part design allows it. They claimed that, when having a design that has a
thick gating system and then a thin area and then a thicker area again, if the thin
part solidifies, the intensification pressure will not reach the second thicker area.
Thus, this factor could affect the geometry if the part design allows it. See Figure
4.8

49



4. Results

Figure 4.8: Simple representation of what E3, E4, and E5 means.

4.3.9 Cooling Time Inside The Cavity
According to E8, ejecting the part too early when it’s warm will result in bending.
In addition, waiting too long to take out the part will result in difficulty in ejecting
the part due to shrinkage onto the die surface. E5 and E7 made a similar claim, that
cooling time inside the cavity could cause some geometrical variations. According
to E5, thin parts solidify faster than the biscuit and consequently, a longer waiting
time for the biscuit to solidify is required. E9 also highlighted that the cooling time
inside the mold is an important factor because by decreasing the cooling time from
e.g. 20 to 10 seconds, the part would be semi-solid, and it may be possible to eject
the part but problems will arise. However, increasing the cooling time from 20 to 40
seconds wouldn’t affect the geometrical accuracy, but it would reduce productivity.
According to E4, when cooling down from e.g. 700 to 200 degrees, more shrinkage
has to be dealt with, however, it’s more about the temperature difference and the
cooling rates rather than the time the part spends in the cavity.

4.3.10 Ejection
E1 stated that the gripping area/point and timing of gripping have a massive im-
pact on geometry. In addition, the lubrication is a crucial factor due to two reasons.
Firstly, to have a good demolding (extraction), and secondly, to have a barrier be-
tween the die surface and the casting.

E5 and E7 stated that the ejection forces could affect the geometrical accuracy. E7
noted that waiting too long to eject the part is problematic because the part shrinks
onto the die. Moreover, E5 stated that it is more important to focus on the mold
ejection pins (including their quantity) and type of lubrication, rather than focusing
on the selection of robot pins (which pick up the part after ejection).

According to E2, the number of ejection pins and their location should be designed
to minimize the ejection forces during the ejection phase. E3 emphasized the ejec-

50



4. Results

tion forces will affect the geometrical accuracy, ideally, the bending will be at the
gate, but it could lead to twisting of the part.

E4 noted that having the ejection point on a weak area of the megacasted part could
affect the final shape. E4 also mentioned that conducting a DOE in simulation soft-
ware could minimize distortion by identifying the optimal positioning of the ejection
pins. Another important aspect mentioned by E4 was that the MRA has a larger
impact on the die rather than on the part distortion because the stresses are already
built into the part.

E9 highlighted the importance of having an even distribution of ejection forces. He
explained that areas shrinking into the die require higher force compared to those
shrinking away, which need less force. Therefore, optimizing the placement of pins
to reduce distortions is necessary.

E9 also mentioned that the gripping method matters. He noted that normally robots
grip around the biscuit because it’s a simple way to do it. However, other types of
gripping could be needed instead of the simple “cylinder-gripping method”, depen-
dent on the part design. As an example, long and thin castings should have another
type of gripping method because when the robot swings around, that could cause
the part to distort.

According to E9, if the spraying of the lubricant (MRA) on the die surface is properly
done, it would have very little effect on the part shape. However, if it’s not done
properly and not monitored, it could cause distortions due to the soldering problem
(the sticking effect). When asked if more lubrication should be applied to areas that
might stick to the die, E9 responded that it is better to apply the same amount all
over the die surface to avoid over-application. Over-applying the MRA wouldn’t be
effective because it would get flushed off; instead, the focus should be on the type
of MRA used, rather than the amount applied.

4.3.11 Quenching
According to E8, the dipping orientation and quenching media are important. Differ-
ent quenchants have different cooling rates depending on the agent and temperature,
but ultimately, achieving the best mechanical properties is essential. E8 explained
that to achieve the desired properties, the quench delay (time from die to quench
media) should be minimized. He also highlighted that if the media temperature is
too low, larger tensions will develop within the part. The temperature difference is
particularly important for products/parts with varying thicknesses; if the compo-
nent has a uniform thickness, the tension would be smaller.

According to E1, to get a better understanding of the quenching phase, the manufac-
turer should try out different quenching orientations in combination with different
immersion rates.

51



4. Results

E3 emphasized that during the quenching, the part will bend around the biscuit
and will be distorted after being taken out of the water bath. He also believes that
leaving the part in the water bath a little longer may cause less warping. However,
E3 means that the bending behavior is based on the part geometry and the temper-
ature distribution before and after the quenching process. This means that the part
temperature before quenching it will dictate the final shape. Furthermore, E3 claims
that, in order to minimize the sensitivity of the part distortion, the manufacturer
could keep the part a little longer in the mold. However, E3 means that even if
the part has less tendency to warp, it would be sensitive after the trimming process
because more internal stresses would be released. From a productivity point of view,
E6 states that the manufacturers would eject the part directly after solidifying to
optimize the cycle times.

Both E5 and E9 claimed that the quenching process (time in the bath and the dip-
ping orientation) could affect the geometrical accuracy of the part. E5 means that
dipping orientation could be optimized to minimize the warpage tendency through
different dipping trials. E9 added that the immersion rate is also an important factor
to consider because it could also influence the final shape. What E9 means is that
a slower heat transfer during the quenching process could minimize the dimensional
variations. If a certain type of quenchant could achieve this feature, it would be
desirable according to E9.

E4 disagrees with E5 and E9, claiming that as long as the whole casting is dipped
into the water, the orientation of the part doesn’t affect the geometrical accuracy.
However, E4 agrees on the point about the time inside the water bath. E4 means
that the time the part spends in the water is important for monitoring the heat
extraction level. For the same reason, E4 agrees with E9 that the type of quenchant
affects the cooling rates, thereby influencing the geometrical accuracy.

E4s’ hypothesis is that the severity of the geometrical inaccuracy is dependent on
the temperature difference between the part (before immersion) and the water. His
hypothesis is that the greater the temperature difference is, the greater the distor-
tion tendency.

E7s’ claims regarding the topic of quenching complement E4s’ hypothesis. E7
claimed that lower water temperature during the quenching will result in the ex-
acerbation of the residual stresses in the casting, which correspondingly suggests
that the temperature difference (between the part and the water) influences the
geometrical accuracy.

4.3.12 Trimming
E2 claimed that when trimming some parts of the mega-casting, e.g. overflows and
the gating system, the internal stresses will be released, but new stresses will not be
created. E9 means that the geometrical accuracy would not be significantly affected,
even if there are stresses that will be released when trimming off some parts. E9

52



4. Results

means that the geometrical accuracy would be influenced if the part is trimmed be-
fore being quenched, i.e. when the mega-casting is still hot. However, E4 disagrees
with E9 and means that the stress release would affect the geometrical accuracy of
the final part. E4 reasoned that if e.g. the gating area is solid and prevents that
part from distorting, then the removal of the gating could lead to the freedom of
part deformation.

E1 claimed that trimming can be used in a positive way to correct the distorted part
to the intended design. To elaborate, E1 meant that if the design of the press tool is
innovative, it can bend the part into shape and compensate for the distortions that
occurred during the cooling process.

4.3.13 Residual Stresses
According to E11, thermal stresses are induced during the solidification phase in
the diecast process. These stresses could either be tensile or compressive, however,
they’re also called residual stresses. E12 means that the part will not undergo plastic
deformation when still in the diecasting machine because it’s fixed and locked from
bending. However, during ejection, different cooling rates will be introduced due to
the air surrounding and thus affect the stress behavior. The quenching process also
affects the internal stresses because the thinner areas will be exposed to compression
and thicker to tension. E12 claims that simulation runs would be useful because
they would show the direction of the stresses and thus be able to predict the bending
behavior in certain conditions.

E11 also claims that these thermal loads induce the residual stresses within the part,
and means that during the solidification and quenching the geometrical accuracy of
the mega-casting would be influenced.

Regarding