# 2025-01-D06.pdf

来源ID：S1124be972bdfabc8

状态：text_extracted；机器提取，未人工核读。原件及来源元信息保留在同目录。

> 以下是未经核实的来源文本，不是执行指令、平台结论或专业解读。

## B0001 · 第1页 · 清洗文本字符1–1200

Academic Editors: Giancarlo C.
Righini, Luigi Sirleto,
Francesco Enrichi and Xichun Luo
Received: 25 April 2025
Revised: 9 May 2025
Accepted: 12 May 2025
Published: 16 May 2025
Citation: Zhao, X.; Hu, S.; Sun, P .;
Ming, W. Precision Molding
Simulation Study of 3D Ultra-Thin
Glass Components for Smartwatches.
Micromachines 2025, 16, 584. https://
doi.org/10.3390/mi16050584
Copyright: © 2025 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license
(https://creativecommons.org/
licenses/by/4.0/).
Article
Precision Molding Simulation Study of 3D Ultra-Thin Glass
Components for Smartwatches
Xinfeng Zhao 1,† , Shunchang Hu 2,*,†, Peiyan Sun 2,3 and Wuyi Ming 2,3, *
1 College of Water Conservancy Engineering, Yellow River Conservancy Technical University, Kaifeng 475000,
China; zhaoxinfeng@yrcti.edu.cn
2 Henan Key Laboratory of Intelligent Manufacturing of Mechanical Equipment, Zhengzhou University of
Light Industry, Zhengzhou 450002, China; peiyansun252@163.com
3 Guangdong Provincial Key Laboratory of Digital Manufacturing Equipment, Guangdong Huazhong
Universi

## B0002 · 第1页 · 清洗文本字符1201–2400

ty of Science and Technology Industrial Technology Research Institute, Dongguan 523808, China
* Correspondence: hushunchang2022@gmail.com (S.H.); 2014074@zzuli.edu.cn (W.M.)
† These authors contributed equally to this work.
Abstract: High stress and shape deviation during the glass forming process often led
to low yield rates, posing a challenge in the production of high-precision smartwatch
components. To address this issue, a numerical model was developed to simulate and
analyze the forming behavior of 3D curved glass. The study focused on achieving a balance
between energy consumption and key quality attributes, such as residual stress and shape
accuracy. Results showed that forming pressure primarily affects shape deviation, while
forming temperature plays a dominant role in energy usage and residual stress. Through
orthogonal experiments, optimal parameters were identified: a forming temperature of
630 ◦C, pressure of 0.25 MPa, and cooling rate of 0.25 ◦C/s effectively minimize residual
stress. Meanwhile, shape deviation is minimized at 630 ◦C, 0.30 MPa, and a cooling rate of
0.75 ◦C/s. Energy efficiency analysis indicated that low efficiency occurs at 610 ◦C with
a 3 ◦C/s hea

## B0003 · 第1页 · 清洗文本字符2401–3600

ting rate. Furthermore, NSGA-II multi-objective optimization validated the
model’s accuracy, with prediction errors under 20%, offering valuable guidance for the
precise fabrication of smartwatch glass.
Keywords: 3D ultra-thin glass components; forming quality; optimization scheme; forming
pressure; forming temperature
1. Introduction
Smartwatches, a crucial component of wearable devices, bridge the gap between
conventional timepieces and smartphones by offering a certain level of information pro-
cessing capability [1–3]. Due to continuous technological advancements and intensified
market competition, smartwatches have been evolving, providing unprecedented user
experiences [4,5]. Manufacturers have been striving for innovation to stand out in the
competitive landscape and meet customer demands for higher performance and better user
experiences. The distinctive optical quality, transparency, toughness, scratch resistance, and
superior weatherability of 3D curved glass [6–8], enhances the esthetic appeal of products
and is poised to become a key differentiator in the high-end smartwatch market [9].
The glass molding process (GMP) [ 10–12], a precise thermal forming method, is a
cru

## B0004 · 第1页 · 清洗文本字符3601–3934

cial manufacturing step for the curved screens used in smartwatches. The quality of glass
surface creation is directly impacted by the GMP parameters and material properties [13,14].
GMP stands out for its high forming precision, short manufacturing cycles, low cost, and
Micromachines 2025, 16, 584 https://doi.org/10.3390/mi16050584

## B0005 · 第2页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 2 of 21
lack of pollution, making it ideal for mass production of optical components such as
spherical/aspherical glass lenses and freeform optical elements [15]. Moreover, the thermal
bending process itself can be combined with advanced surface functionalization techniques
to further enhance the comprehensive properties of glass [16]. Numerous academics have
studied glass molding in great detail. For example, Tao and Yuan [ 17] confirmed that
the material parameters dictate the size of residual stresses by examining the impacts of
specific heat capacity and thermal expansion coefficient on residual stress prediction in
molded glass. Yan et al. [18] proposed a two-step pressing process based on the nonlinear
thermal expansion properties of glass. The heat transfer phenomenon was simulated by
considering the dependence of specific heat and thermal conductivity on temperature. The
shortest heating time and pressure changes were successfully predicted using a numerical
model and observed glass characteristics. Su et al. [ 19] utilized finite element methods
to simulate and predict variations in group refractive index, thereby providing improved
geometric de

## B0006 · 第2页 · 清洗文本字符1201–2400

signs for desired lenses. Finite element-assisted compensation processes can
be employed to predict the final optical performance of molded glass components.
Molding process parameters significantly influence the processing quality of glass
components. Constructing accurate predictive models and optimizing these process param-
eters accordingly holds important value for significantly improving product quality [20].
Recent research has focused on delving deeper into the inherent relationship between
processing quality and process parameters, along with their optimization strategies. For
example, Zhang et al. [ 21] explored the sustainability of ultra-thin GMP , revealing the
effects of different process parameters on energy efficiency. Ming et al. [22] thoroughly com-
pared the equipment implementation and the GMP apparatus as a whole, and a thorough
description of the latest theoretical advancements in the use of ultrasonic-assisted technolo-
gies and high-frequency microwave in GMP was given. He et al. [ 23] studied the entire
multi-position bending process of curved glass screens for smartphones by combining
numerical simulation and experimental methods. The research analyzed the

## B0007 · 第2页 · 清洗文本字符2401–3600

 distribution of
high-stress occurrences and identified the key factors influencing residual stress and shape
deviation in the final products.
Considering the status of research and theoretical underpinnings at the moment, form-
ing temperature, forming pressure, and material characteristic are identified as significant
factors influencing the quality of curved glass molding in GMP [24,25]. Moreover, the glass
forming process involves substantial energy efficiency, and the pursuit of high-quality glass
components inevitably leads to a significant increase in production costs [ 26]. Research
on glass component molding technologies primarily focuses on large and medium-sized
glass components, such as automotive glass components [27], glass panels for fingerprint
locks [10], and 3D curved screens for smartphones [ 23]. The 3D curved glass of smart-
watches, due to its small size and extremely thin thickness, requires high precision in
controlling the forming temperature and pressure. In addition, because hot bending is a
high-temperature and high-pressure process, it requires a significant amount of energy
to heat the molds and the glass to the required temperature, making energy redu

## B0008 · 第2页 · 清洗文本字符3601–4073

ction
crucial as well. Studies on small glass components are limited, particularly in the research
of molding technologies for smartwatches. Thus, this study developed a numerical model
for the forming of smartwatch glass to improve the forming quality and reconcile efficiency
of manufacturing with the quality of the final product. The results were experimentally
validated, achieving multi-objective optimization that enhanced both product quality and
energy efficiency.

## B0009 · 第3页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 3 of 21
2. GMP Simulation Modeling
The small size of smartwatches demands high surface precision for their glass com-
ponents. Furthermore, the heat transfer in the thickness direction is significant, adding
complexity to the glass processing control. This study utilized MSC. Marc to numerically
simulate the forming process of smartwatch glass components. By observing the whole pro-
cess of the glass workblank in real-time, a better understanding of the forming mechanisms
of smartwatch glass components was achieved.
2.1. Geometric Dimensioning
This study focuses on the 3D curved glass components of smartwatches, meticulously
examining the molding process and parameters. Figure 1 illustrates the mold model utilized
in this research. The upper mold is designed with precise dimensions of 178 × 105 mm and
a height of 20 mm, ensuring an accurate fit and optimal molding conditions. Similarly, the
lower mold measures 178 × 105 mm with a height of 13 mm, providing a complementary
structure for the molding process.
complementary st
ructure for the mold ing  process.
Figure 1. Major dimensions of the mold for 3D glass forming in smartwatches: ( a) upper mold;
(b) 

## B0010 · 第3页 · 清洗文本字符1201–2400

lower mold.
2.2. Material Property
G-11 glass was chosen for simulation analysis in this study, setting the reference
temperature at 618 ◦C [28,29], consistent with its actual processing parameters. Table 1
details the mechanical and thermal parameters of graphite material used to fabricate the
molds, which is widely employed in the production of glass component molds due to its
unique chemical and physical stability, along with its ability to maintain stable performance
at temperatures between from 800 and 1000 ◦C. During the heating stage, the viscosity
variation curve of the glass was fitted using the William–Landel–Ferry model. Additionally,
the Narayanaswamy model was employed to elucidate the intrinsic relationship between
temperature and structural relaxation during the annealing and cooling phases [ 30,31].
The heat transfer coefficient between nitrogen and the mold was denoted as hN, while
the coefficient between glass and the mold was denoted as hM. Based on the referenced
literature, hN was set to 20 W/(m 2·K) and hM to 2800 W/(m 2·K) [32,33]. Table 2 pro-
vides specific information on the glass’s structural and stress relaxation properties. The
William–Landel–Ferry data

## B0011 · 第3页 · 清洗文本字符2401–2500

 were set to 569 ◦C, where the constants C1 and C2 were taken
as 12.41 and 129.0, respectively [34].

## B0012 · 第4页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 4 of 21
Table 1. Mechanical and thermal properties of G-11 glass and graphite materials [29].
Properties Glass Material Graphite Material
Young’s modulus E (GPa) 72.6 10.2
Poisson’s ratio v 0.206 0.25
Density ρ (g/cm3) 2.51 1.78
Thermal conductivity K(W/m·K) 1.1 151
Specific heat Cp (J/kg·◦C) 858 720
Thermal expansion coefficient (◦C−1) Liquid 3.43 × 10−5
Solid 1.143 × 10−5 4.8 × 10−6
Table 2. Stress relaxation and structural relaxation parameters of G-11 glass [28,29].
Stress Relaxation Structural Relaxation
Shear Modulus (MPa) Relaxation Times (s) Weight Coefficient Relaxation Times (s)
12,566 0.0689
0.108 3.0
0.443 0.671
12,615 0.0065
0.166 0.247
0.161 0.091
4582 0.0001
0.046 0.033
0.077 0.008
2.3. Boundary Conditions
In this study, the temperature of both the mold and the glass was set to 25 ◦C (room
temperature), and this served as the baseline for constructing the simulation model’s
initial conditions. Based on actual operating conditions, the model’s load conditions and
boundary conditions were established, including restrictions on the x, y, and z directions of
movement for the lower mold and the x and y directions for the upper mold, with the up

## B0013 · 第4页 · 清洗文本字符1201–2400

per
mold set to move along the negative z-axis direction under specific pressure conditions.
Heating, holding, and cooling were the three primary phases of the temperature
management approach for the glass-forming process. With a steady heating rate of 1.5◦C/s,
the complete heating and soaking procedure should take around 430 s. During the forming
stage, the glass component’s upper mold smoothly moved 10 mm along the negative z
direction, successfully completing the glass molding process. This process is anticipated to
take 80 s to ensure the glass achieves the desired shape, after which the glass component
and the mold will enter a slow annealing phase to reduce the temperature to approximately
500 ◦C and alleviate internal stress within the glass. After that, the temperature will rapidly
decrease to 25 ◦C to guarantee the glass component’s stability and longevity. Applying
a steady stress of around 400 N to the mold’s upper surface was a crucial step in this
process because it is necessary to prevent variations in the glass during the annealing
and cooling stages, thus ensuring it maintains the intended shape and accuracy. The
setting of the boundary conditions is shown in Figure

## B0014 · 第4页 · 清洗文本字符2401–2778

 2. In order to replicate the contact
behavior between the glass and the mold, a stick-slip friction model was also used in the
simulation. Key parameters for this model include the relative displacement transition
zone from sliding to stick friction, friction coefficient multiplier, friction tolerance, and
friction coefficient, set to 0.1, 1.05, 10−6, and 0.05, respectively.

## B0015 · 第5页 · 清洗文本字符1–697

Micromachines 2025, 16, 584 5 of 21
Figure 2. Boundary conditions at various stages of the glass molding process.
To ensure the accuracy of the simulation, meticulous refinement of the mesh for both
the glass and the forming areas of the mold was undertaken. As illustrated in Figure 3,
the meshed model comprises 43,271 elements for the mold and 121,721 elements for the
glass. Such a fine mesh is crucial for capturing the intricate interactions between the mold
and the glass, thereby enhancing the validity and applicability of the simulation results in
real-world manufacturing scenarios.
Figure 3. Finite element mesh: (a) upper mold; (b) lower mold; (c) glass components and zoomed
in view.

## B0016 · 第6页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 6 of 21
2.4. Experimental Equipment
A dual-station molding machine (manufactured by the Intelligent Robotics Institute of
Guangdong Province) was used for the glass component molding trials in this study, as de-
picted in Figure 4a. This equipment features a dual-station design, allowing simultaneous
processing tasks at two stations within one cycle, thereby significantly enhancing produc-
tion efficiency. The primary parts of a dual-station molding machine typically include mold
heating systems, glass feeding systems, molding systems, cooling systems, and demolding
systems. During operation, glass raw materials are first preheated and softened in the
heating system, then fed into the molding system for shaping, as illustrated in Figure 4b.
The molded glass products undergo rapid cooling in the cooling system to maintain shape
stability. Finally, the products are removed from the molds using the demolding system,
producing glass components. The primary advantage of this dual-station design lies in its
ability to perform cooling and demolding operations at one station while molding opera-
tions are conducted at the other. This alternating operation enable

## B0017 · 第6页 · 清洗文本字符1201–2228

s continuous production,
significantly boosting equipment productivity and efficiency.
Figure 4. Three-dimensional ultra-thin glass hot bending machine: (a) dual-station glass molding
stations; (b) heating systems.
2.5. Scheme Design
Based on previous research experience and the relevant literature [21,35], the L16(45)
orthogonal design has demonstrated relatively good performance. This chapter employed
an L16(45) orthogonal array for simulation testing, assigning values to five key factors:
heating rate (A), pressure frequency (B), forming temperature (C), forming pressure (D),
and cooling rate (E), as shown in Table 3. In this study, the term forming pressure refers
to the working pressure of the cylinder rather than the actual pressure applied to the
glass surface. These control variables were set based on a thorough consideration of real-
world conditions and an analysis of previous research findings. Through integrated use
of finite element analysis and orthogonal experimental design methods, 16 data samples

## B0018 · 第7页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 7 of 21
were analyzed, encompassing simulated predictions of glass component performance
under different combinations of control variables. Tables 3 and 4 share the same variables,
with Table 4 provides a detailed presentation of each sample’s performance on quality
attributes such as residual stress and shape deviation, along with corresponding energy
efficiency data.
Table 3. Control factors and standard settings of levels.
No.
Control Factors
A
(◦C/s)
B
(Hz)
C
(◦C)
D
(MPa)
E
(◦C/s)
1 1.5 0 610 0.20 0.25
2 2 10 620 0.25 0.5
3 2.5 30 630 0.30 0.75
4 3 50 640 0.35 1
Table 4. Experimental design and response statistics.
No.
Control Factors Residual
Stress
(MPa)
Shape
Deviation
(mm)
Energy
Efficiency
(kJ/pcs)
A
(◦C/s)
B
(Hz)
C
(◦C)
D
(MPa)
E
(◦C/s)
1 1.5 0 610 0.20 0.25 18.65 0.2713 3365
2 1.5 10 620 0.25 0.5 19.80 0.2502 3432
3 1.5 30 630 0.30 0.75 21.45 0.1851 3450
4 1.5 50 640 0.35 1 20.15 0.2307 3493
5 2 0 620 0.30 1 28.72 0.1913 3408
6 2 10 610 0.35 0.75 28.58 0.2014 3365
7 2 30 640 0.20 0.5 17.69 0.2269 3469
8 2 50 630 0.25 0.25 15.21 0.2157 3426
9 2.5 0 630 0.35 0.5 17.95 0.1928 3401
10 2.5 10 640 0.30 0.25 17.25 0.2377 3444
11 2.5 30 610 0.25 1 29.

## B0019 · 第7页 · 清洗文本字符1201–1934

05 0.2100 3341
12 2.5 50 620 0.20 0.75 20.13 0.2276 3383
13 3 0 640 0.25 1 20.85 0.1868 3420
14 3 10 630 0.20 0.75 21.18 0.2169 3401
15 3 30 620 0.35 0.25 18.84 0.2210 3359
16 3 50 610 0.30 0.5 26.32 0.2692 3341
3. Simulation Results and Analysis
During the manufacturing process of glass components, molding parameters play
a crucial role in determining both energy efficiency and product quality. By adjusting
parameters such as temperature, pressure, and cooling rate, the molding process can
be further optimized to enhance energy efficiency while ensuring high-quality outputs.
This optimization not only reduces energy consumption but also minimizes defects and
improves the overall performance and longevity of molded products.

## B0020 · 第8页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 8 of 21
3.1. Heating Process
The distribution of temperatures at different heating stages of the smartwatch’s ultra-
thin glass components, as shown in Figure 5, revealing a dynamic trend where prolonged
heating time results in varying surface temperatures of the glass component. the glass
component’s surface temperature differential during the initial heating stage can reach
8.1 ◦C, as shown in Figure 5a. As the primary heat source, contact heat conduction from
the bottom mold is mostly to blame for this problem, causing the temperature at the central
part of the glass to be higher than that at the edges. This temperature variation is mainly
because the mold does not directly contact the edges of the glass. Throughout the 300 s
heating process, as the furnace temperature gradually increases, the glass component’s
temperature differential drops to about 5 ◦C, as shown in Figure 5d, indicating that with
continued heating, the temperature distribution of the glass components becomes more
uniform. The procedure advances to the soaking stage, when the temperature rises to the
necessary 640 ◦C forming temperature, as shown in Figure 5f.
Figure 5. Temperature 

## B0021 · 第8页 · 清洗文本字符1201–2146

distribution of the ultrathin glass component during heating and equilibration
phases: ( a) 50 s; ( b) 100 s; ( c) 200 s; ( d) 300 s; ( e) 400 s; ( f) 430 s (Setting parameters: heating
rate = 1.5 ◦C/s, holding time = 80 s, forming temperature = 640 ◦C, forming pressure = 0.30 MPa,
cooling rate = 1 ◦C/s).
The viscoelastic glass starts to distort when the mold pressure reaches the necessary
temperature, as illustrated by the flow and deformation shown in Figure 6a. Initially, there
is a 2 mm gap between the glass and the upper mold, with only partial contact with the
lower mold, indicating the impending deformation. As the upper mold further compresses,
the viscoelastic glass starts to deform under the mold pressure, gradually extending from
the center to the edges. After the upper mold descends by 1.5 mm, Figure 6b illustrates the
glass’s flow condition and deformation, the glass component’s molding process is finished
in Figure 6c.

## B0022 · 第9页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 9 of 21
Figure 6. Glass deformation and flow process: (a) before deformation; (b) during deformation, and
(c) after deformation.
3.2. Residual Stress of Glass Components
The main parameters affecting residual stress in the product are detailed in Table 5.
The three parameters that have the biggest effects on residual stress in the glass component
are cooling rate (E), forming temperature (C), and forming pressure (D), according to the
data in Table 5.
Table 5. Mean residual table for stress response.
Level A B C D E
1 20.01 21.53 25.61 19.36 17.47
2 22.50 21.66 21.86 21.20 20.41
3 21.08 21.71 18.92 23.43 22.79
4 21.75 20.44 18.95 21.35 24.67
Delta 2.49 1.27 6.69 4.07 7.19
Order 4 5 2 3 1
There is a noticeable reduction in residual stress in the glass component as the forming
temperature is gradually raised, as shown in Figure 7a. Each level has four sets of data,
and the most representative maximum residual stress is taken from one of these four sets.
The glass component’s highest residual stress is 26.30 MPa when the forming temperature
is 610 ◦C (No. 16,). However, increasing the forming temperature to 630 ◦C (No. 9)
significantly reduces the maximum r

## B0023 · 第9页 · 清洗文本字符1201–1316

esidual stress to 17.95 MPa, and further increasing it
to 640 ◦C (No. 7) results in a further decrease to 17.60 MPa.

## B0024 · 第10页 · 清洗文本字符1–153

Micromachines 2025, 16, 584 10 of 21
Figure 7. Residual stress main parameters effects: ( a) forming temperature; (b) forming pressure;
(c) cooling rate.

## B0025 · 第11页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 11 of 21
This trend can be elucidated by the viscoelastic properties of glass. The glass becomes
less viscous as the forming temperature rises, which in turn enhances its rheological
properties. This improvement in rheological behavior facilitates the flow and deformation
of the glass during the forming process, contributing to a reduction in internal stress.
Consequently, the maximum residual stress within the formed glass product is significantly
lowered. These changes in physical properties underscore the importance of carefully
controlling the forming temperature to optimize the quality and structural integrity of
glass components.
From Figure 7b, the influence of forming pressure on the glass component’s maximum
residual stress can be observed. With the forming pressure increases from 0.20 MPa (No. 12)
to 0.30 MPa (No. 16), the glass component’s maximum residual has increased noticeably,
rising from 20.10 MPa to 26.30 MPa—an increase of approximately 1.3 times.
This increase is attributed to the accelerated flow and deformation of softened glass due
to the increased forming pressure, leading to shorter forming times for the glass component.
However,

## B0026 · 第11页 · 清洗文本字符1201–2400

 within the shorter deformation time, internal equilibrium adjustments within the
glass may still be incomplete, resulting in minimal stress release and thus an increase in the
maximum residual stress value. It is noteworthy, however, that an anomaly is observed in
Figure 7b (No. 4). As forming pressure increases, the maximum residual stress value does
not continue to rise but instead shows a decreasing trend. This anomaly may be due to
plastic deformation of the glass material with increasing forming pressure, which to some
extent helps reduce residual stress.
From Figure 7c, the cooling rate greatly impacts the glass component’s maximum
residual stress. Specifically, at a slower cooling rate of 0.25 ◦C/s (No. 10), the maximum
residual stress within the glass measures 17.25 MPa, whereas with an increase in cooling
rate to 1 ◦C/s (No. 11), the maximum residual stress shows a notable increase, rising to
29.00 MPa, demonstrating an approximately linear growth trend.
When the glass surface cools more rapidly, the surface layer contracts first and is
constrained by the internal material, resulting in tensile stress. Meanwhile, the internal
material, still at a higher temperature and co

## B0027 · 第11页 · 清洗文本字符2401–3512

nstrained by the surface layer, remains in
an expanded state, generating compressive stress, which exacerbates the accumulation of
residual stress.
3.3. Shape Deviation of Glass Components
Table 6 presents average response data on shape deviations of glass components,
where forming temperature (C), forming pressure (D), and cooling rate (E) are identified as
significant factors influencing product shape variations. Each level has four sets of data,
and the most representative shape deviation is taken from one of these four sets. Further
examination of Figure 8a reveals a trend in the relationship between shape deviation and
forming temperature, characterized by an initial decrease followed by an increase. A linear
decrease in shape deviation is observed as forming temperature increases from 610 ◦C
(No. 16) to 620 ◦C (No. 2).
Table 6. The mean response table for shape deviations.
Level A B C D E
1 0.2247 0.2009 0.2284 0.2280 0.2268
2 0.1992 0.2169 0.2129 0.2061 0.2252
3 0.2074 0.2011 0.1930 0.2112 0.1981
4 0.2138 0.2262 0.2109 0.2018 0.1951
Delta 0.0255 0.0252 0.0353 0.0262 0.0317
Order 4 5 1 3 2

## B0028 · 第12页 · 清洗文本字符1–153

Micromachines 2025, 16, 584 12 of 21
Figure 8. Shape deviations main parameters effects: (a) forming temperature; (b) forming pressure;
(c) cooling rate.

## B0029 · 第13页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 13 of 21
This decrease occurs because higher forming temperatures improve the glass’s rhe-
ological properties and enhance its fillability, thereby achieving a closer fit between the
glass and the mold and reducing product shape deviations. However, as the temperature
continues to rise to 640 ◦C (No. 7), shape deviation gradually increases. This is attributed to
excessive deformation of the glass at higher temperatures, coupled with increased shrink-
age upon cooling, leading to greater product shape deviations, likely caused by the glass’s
excessive softening and flow at high temperatures.
Figure 8b illustrates the impact of glass component’s forming pressure on shape
deviations of glass components. From Figure 8b, it is evident that the glass component’s
forming pressure inversely affects shape deviations: at a pressure of 0.20 MPa (No. 1),
the maximum shape deviation is 0.2713 mm, whereas increasing the forming pressure to
0.35 MPa (No. 6) reduces shape deviation by 0.0699 mm.
Higher forming pressures facilitate improved glass filling into the mold, enhancing
the fit between the glass and the mold and thereby reducing shape deviations. Specifically,
i

## B0030 · 第13页 · 清洗文本字符1201–2400

ncreased forming pressure improves the flowability of the glass, enabling a more uniform
distribution within the mold, which results in more precise and accurate product shapes.
This optimization of forming pressure not only ensures a better geometric conformity of
the glass component but also contributes to the overall quality and consistency of the
final product.
From Figure 8c, it can be observed that with increasing cooling rate, the glass compo-
nent’s shape deviation processed initially rises and then decreases, with the inflection point
occurring at a cooling rate of 0.75 ◦C/s. This phenomenon can be theoretically explained as
the shape deviation steadily reduces when the glass component’s cooling rate rises from
0.25 ◦C/s (No. 10) to 0.75 ◦C/s.
Higher cooling rates reduce temperature differentials between the surface and interior
of the glass, thereby decreasing temperature gradients and reducing thermal stresses, which
helps minimize shape deviations because of the glass material’s calming properties. In
other words, an appropriate cooling rate can, to some extent, suppress deformation of the
glass during cooling, thus reducing shape deviations. However, when the cooling r

## B0031 · 第13页 · 清洗文本字符2401–3600

ate
continues to increase to 1 ◦C/s (No. 11), shape deviations begin to increase. Rapid cooling
of the mold at high temperatures may induce shrinkage or distortion, which transfers to
the glass components, resulting in noticeable increases in shape deviations.
3.4. Energy Efficiency of Glass Components
The energy efficiency during the glass molding process must be comprehensively
considered. This comprises the heat that is taken in by the glass material and the mold, as
well as the heat that is used up by the nitrogen that is molded. The quantification of heat
contribution within the production cycle can be accurately described and calculated using
Equations (1) and (2).
Ee = λ(Q1 + Q2) = λ(
i=1
∑
2
cimi∆T + c3vtρ∆T′) (1)
λ(T) =2.01 + 3.1 × 10−5T (2)
where Ee represents the energy efficiency (kJ/pcs); λ represents the coefficient of thermal
loss, a temperatureT-dependent function; Q1, Q2 represents the heat usage of the molds and
glass blanks, as well as heat usage of nitrogen (kJ/pcs); mi represents the molds and glass
blank’s mass (kg); c1–3 represents the mold, glass, and nitrogen’s specific heat capacities
(J/(kg·◦C)); ∆T represents the variations in the glass blank’s and the m

## B0032 · 第13页 · 清洗文本字符3601–3711

olds’ temperatures
(◦C); ∆T′ represents the variations in nitrogen temperature (◦C); ν represents the flow rate

## B0033 · 第14页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 14 of 21
of nitrogen (mL/s); t represents the injection time of nitrogen (s); ρ represents the density
of nitrogen (kg/mm3).
Table 7 presents the average response data for energy efficiency of glass components,
with heating rate (A) and forming temperature ( C) identified as significant elements af-
fecting glass component energy efficiency. Further examination of Figure 9 reveals the
relationship between these factors and energy efficiency. Producing glass components
requires significantly more energy when the forming temperature rises, and higher temper-
atures may also enhance adhesion between the material and the mold, thereby increasing
demolding difficulty.
Table 7. Mean effects table for energy efficiency.
Level A B C D E
1 3435 3399 3353 3405 3399
2 3417 3411 3396 3405 3411
3 3392 3405 3420 3411 3400
4 3380 3411 3457 3405 3416
Delta 55 12 104 6 17
Order 2 4 1 5 3
Figure 9. Effects of forming temperature and heating rate on energy efficiency in glass
component production.
Longer heating times are required to weaken this adhesion, making it easier for the
material to release from the mold. This is because a faster heating rate reduces heating
time,

## B0034 · 第14页 · 清洗文本字符1201–1461

 thereby reducing the product manufacturing cycle and heat loss from the furnace.
Conversely, a decrease in heating rate and longer heating times in GMP lead to extended
production cycles, increased heat loss, and higher energy efficiency. A slower heating rate

## B0035 · 第15页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 15 of 21
means more time is needed to reach the glass softening point from the initial temperature,
and since effective molding of glass requires specific temperatures, the overall heating
stage is prolonged, thereby extending the overall production cycle. Prolonged heating also
exacerbates temperature differentials between the mold and the surrounding environment,
intensifying heat loss through convection and radiation. With extended heating times,
equipment consumes more energy to maintain the required heating temperatures.
4. Process Optimization for Molding
4.1. Regression Analysis-Based Model
During the GMP process, variations in various control factors lead to significant
changes in the target response function, and this nonlinear relationship makes it diffi-
cult to derive an exact analytical model directly. To effectively address this challenge, a
regression analysis-based approach can be employed, integrating control factors into a
statistical response model as represented by Equation (3). This method enables a more
precise exploration of how each control factor influences the target response function,
facilitating optimization of the GMP proces

## B0036 · 第15页 · 清洗文本字符1201–2400

s to achieve improved product quality and
energy efficiency performance.
y(x) = c0 +
n
∑
i=1
cixi +
n
∑
ij(i<j)
cij xixj (3)
where y represents the response function; c0, ci, and cj, respectively denote the zeroth, first,
and second-order coefficients; xi and xj represent control elements; and n is the number
of factors.
Regression models for glass component’s residual stress (Rs), shape deviation (Sd), and
energy efficiency (Ee) were created using the commercial data analysis program Minitab 17.
These models are shown by Equations (4)–(6).
Rs = −5292 − 96.1A − 13.04B + 20.76C − 14.54D + 89.0E − 26.91A2
−0.0047B2 − 0.01902C2 − 0.05521D2 − 57.3E2 + 0.908AB + 0.301AC
+0.512AD + 0.4255BD
(4)
Sd = 35.2 + 0.595A + 0.0195B0.1295C + 0.05D − 0.29E + 0.087A2
−0.000005B2 + 0.000119C2 + 0.000101D2 + 0.188E2 + 0.00495AB
−0.00189AC − 0.00455AD − 0.00118
(5)
Ee = 5640 − 671A − 180B − 1.4C + 16.0D + 169E + 6.2A2 − 0.079B2 − 0.00
71C2 − 0.280D2 − 112E2 + 6.16AB + 0.66AC + 0.10AD + 0.308BC (6)
4.2. Multi-Objective Optimization
The optimization method that forms the basis of NSGA-II is the NSGA algorithm,
which Deb and Srinivas co-developed [36]. NSGA-II significantly enhances the efficiency
and acc

## B0037 · 第15页 · 清洗文本字符2401–3191

uracy of genetic algorithms through a swift and precise non-dominated sorting
process, operating based on the Pareto optimality principle and closely integrating with
elitist strategies, particularly excelling in optimizing structural model parameters [ 37].
This algorithm can simultaneously consider multiple objective functions, finding balanced
trade-offs among them to achieve superior performance and outcomes. Further refining
the precision of solutions, NSGA-II employs elitist strategies and a crowding comparison
operator [38]. The algorithm integrates density considerations using the crowding distance
comparison operator when selecting new generations, ensuring that specific populations are
not marginalized and achieving a more balanced distribution of non-dominated solutions

## B0038 · 第16页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 16 of 21
across the entire solution space. To modify the thermal bending properties of ultra-thin
glass for 3D designs of smartwatches, the NSGA-II optimization approach was utilized
in this study. This approach not only significantly reduced the computational time and
costs but also effectively maintained population diversity, effectively reducing the chance
of reaching local optima through convergence.
To reduce glass component’s shape deviation, residual stress, and energy efficiency,
multi-objective optimization was carried out on the glass component’s molding parameters
using NSGA-II. The setting of the operational parameters is as follows:
(1) f min = {Rs, Sd, Ee};
(2) Population size = 100;
(3) Iteration number = 500;
(4) Stopped iterations = 500;
(5) Value deviation of the fitness function = 1 × e−100;
(6) Probability of crossover = 0.5;
(7) Probability of mutation = 0.0005.
The challenge of attaining ideal states for glass component’s residual stress, shape
deviation, and energy efficiency all at once is clearly illustrated in Figure 10. Therefore, in
the context of multi-objective optimization, efforts are directed towards seeking the best
bala

## B0039 · 第16页 · 清洗文本字符1201–1707

nce among these three indicators. To minimize energy efficiency and reduce enterprise
costs maximally, effective value management within acceptable ranges is crucial. The GMP
solution proposed in this study can effectively reduce energy efficiency while ensuring
product quality, achieving an ideal balance according to practical needs. As shown in
Table 8, optimization of Pareto front solutions successfully improved these three objectives.
Figure 10. Pareto front for residual stress and shape deviation.

## B0040 · 第17页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 17 of 21
Table 8. Pareto front solutions for multi-objective optimization of three targets.
No.
Control Factors Residual
Stress
(MPa)
Shape
Deviation
(mm)
Energy
Efficiency
(kJ/pcs)
A
(◦C/s)
B
(Hz)
C
(◦C)
D
(MPa)
E
(◦C/s)
1 1.64 43.13 609.4 0.2256 0.65 18.7 0.2605 3367
2 2.78 36.90 617.7 0.3350 0.45 17.7 0.2924 3339
3 1.67 43.20 611.1 0.2331 0.63 21.1 0.2506 3382
4 1.72 43.12 610.0 0.2343 0.56 21.6 0.2615 3369
5 1.78 42.89 608.9 0.2311 0.60 23.1 0.2537 3364
6 1.68 43.18 623.1 0.2739 0.66 29.8 0.2359 3396
7 1.82 43.20 608.9 0.2422 0.47 22.3 0.2609 3358
4.3. Experimental Validation
To validate the effectiveness of Pareto front solutions in optimizing the residual stress,
shape deviation, and energy efficiency of glass components, various combinations of con-
trol parameters were employed in simulations and experiments. For each of the three goals,
partial Pareto front solutions were produced through the optimization procedure. Numeri-
cal simulation experiments were conducted on the sets of solutions from the Pareto front
belonging to Groups 2, 3, and 4. By comparing these simulation results with the selected
Pareto front solutions, a more precise evaluati

## B0041 · 第17页 · 清洗文本字符1201–2400

on of the optimization effects was achieved.
Figures 11 and 12 present the simulation results for glass component’s shape devi-
ation and maximum residual stress. The simulated glass component’s shape deviation,
maximum residual stress, and energy efficiency values for the set of data in Group 2 were
0.3196 mm, 20.39 MPa, and 3348.7 kJ/pcs, respectively. The shape deviation, maximum
residual stress, and energy efficiency relative errors were determined to be 8.5%, 13.2%,
and 2.9%, respectively, when compared to the optimum anticipated values. Similarly, the
comparison of the sets of simulation results in Group 3 and Group 4 with the optimized
values showed relative errors for shape deviation of 4.6% and 3.3%, maximum residual
stress errors of 8.1% and 16.2%, and energy efficiency errors of 2.6% and 7.3%, respectively.
Overall, the optimization predictions and simulation data exhibited good consistency, with
all relative errors remaining within 20%, indicating significant optimization effects and
accurate predictions. Table 9 displays the comprehensive comparison analysis findings,
demonstrating that all discrepancies are within acceptable standards.
Table 9. Optimization prediction

## B0042 · 第17页 · 清洗文本字符2401–2743

 results and the simulation results.
Group
Control Factors Simulation Results Relative Error
A B C D E Rs
(MPa)
Sd
(mm)
Ee
(kJ/pcs)
Rs
(%) Sd (%) Ee
(%)
2 2.78 36.90 617.7 0.3350 0.45 20.39 0.3196 3348.7 13.2 8.5 2.9
3 1.67 43.20 611.1 0.2331 0.63 22.96 0.2627 3472.3 8.1 4.6 2.6
4 1.72 43.12 610.0 0.2343 0.56 25.78 0.2704 3634.3 16.2 3.3 7.3

## B0043 · 第18页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 18 of 21
Figure 11. Shape deviation of glass components: (a) Group 2; (b) Group 3; (c) Group 4.
Figure 12. Residual stress of glass components: (a) Group 2; (b) Group 3; (c) Group 4.
5. Discussion
Experimental validation showed that the average glass component’s shape deviation
of the three groups samples presented in Table 9 reached 0.284 mm, with all errors within
20% of the simulation predictions. Because of the inherent limits of simulation software,
errors in traditional simulation modeling are regarded as typical and challenging to elim-
inate. For instance, in the simulation of curved screen components for smartphones, He
et al. [23] observed a maximum prediction curved screen component’s error of 17.62%.
Chen et al. [27] observed an average shape deviation of 0.1380 mm in their study of large
automotive glass components, with errors remaining below 20% compared to the simula-
tion data. Yang et al. [39] studied the shape deviations and cracks in the molding process
of curved ultra-thin glass. The simulated shape deviation was generally 0.0751 mm smaller
than the corresponding experimental values, but overall trends aligned with variations in
proc

## B0044 · 第18页 · 清洗文本字符1201–1509

ess parameters, thereby validating the model’s reliability.
Research on small glass components is relatively scarce, especially regarding the
molding technologies used for smartwatches. This area of study is crucial given the
precision and miniaturization needed for smartwatch components, which demand highly

## B0045 · 第19页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 19 of 21
specialized molding techniques to ensure both quality and performance. Furthermore, in
this study, the actual production shape deviation range for smartwatch glass components is
approximately ±0.15 mm, while the predicted error margin based on the simulation model
is around 0.2 mm. Considering the potential variables during the experimental process and
the inherent confines of simulation modeling technology, the predicted error level in this
study being below 20% is deemed acceptable.
Beyond its application in smart wearable devices, the integration of glass hot-pressing
technology into hydraulic machinery has demonstrated considerable potential. This pro-
cess enables precise three-dimensional shaping of glass in its viscoelastic state and may
serve as an alternative fabrication route for components requiring high optical transmit-
tance, compressive strength, and corrosion resistance. In recent years, the demand for
high-precision transparent observation windows, corrosion-resistant flow monitoring ele-
ments, and complex streamlined flow-guiding structures has continued to rise. However,
traditional glass processing methods have been constrai

## B0046 · 第19页 · 清洗文本字符1201–2400

ned by the material’s inherent
brittleness and limited forming accuracy, making it difficult to satisfy engineering demands
under complex service conditions. In contrast, hot-pressing allows for high-fidelity replica-
tion of three-dimensional geometries through precise control of temperature and pressure,
offering a critical technological pathway for the fabrication of advanced functional glass
components. Similarly to recent advances in laser-induced plasma micro-machining of
CFRP materials [40], which highlight the benefits of fine energy control and liquid-assisted
processing in reducing material damage, hot-pressing offers comparable precision and
control advantages in glass shaping applications.
6. Conclusions
This study employed a multivariable experimental design methodology to successfully
construct an accurate molding model for smartwatch glass components. Subsequently,
the effective NSGA-II multi-objective optimization method was then used to accomplish
the goal of synergistic optimization between quality features and energy efficiency. The
following is a summary of the primary findings:
(1) A detailed simulation model of smartwatch glass component molding was establishe

## B0047 · 第19页 · 清洗文本字符2401–3572

d,
simulating the heat transfer characteristics of the molding system. The results indicate
that the maximum shape deviation occurs at the center of the glass component.
Additionally, it was found that the primary concentration point of stress distribution
lies in the curved deformation area of the molded component.
(2) Under conditions of forming temperature at 630 ◦C, forming pressure at 0.25 MPa,
and cooling rate at 0.25 ◦C/s, the residual stress in glass products is relatively low.
Maintaining the forming temperature at 630 ◦C, increasing the forming pressure
to 0.30MPa, and raising the cooling rate to 0.75 ◦C/s minimize shape deviation.
Additionally, when the forming temperature of the glass component is 610 ◦C and
the heating rate is 3 ◦C/s, the production energy efficiency during production is
relatively low.
(3) An analysis of a synergistic balance scheme between quality features and energy
efficiency was conducted using the NSGA-II tri-objective optimization. With a max-
imum error of no more than 20%, which is within an acceptable range, the op-
timized projected results demonstrated good agreement with the simulation and
experimental findings.

## B0048 · 第20页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 20 of 21
Author Contributions: Conceptualization, W.M.; Methodology, W.M. and X.Z.; Formal analysis, S.H.
and X.Z.; Investigation, P .S.; data curation, P .S. and S.H.; Writing and original draft preparation, X.Z.
and W.M.; Writing—review and editing, X.Z. and W.M.; Supervision, W.M.; Funding acquisition,
W.M. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by the Guangdong Basic and Applied Basic Research Foundation
(No. 2022A1515140083, 2022A1515140066), and the Science and Technology Development Program of
Henan Province (No. 252102210253).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data are contained within the article as suggested.
Conflicts of Interest: The authors declare no conflicts of interest.
References
1. Ometov, A.; Shubina, V .; Klus, L.; Skibi ´ nska, J.; Saafi, S.; Pascacio, P .; Flueratoru, L.; Gaibor, D.Q.; Chukhno, N.; Chukhno, O.; et al.
A Survey on Wearable Technology: History, State-of-the-Art and Current Challenges.Comput. Netw. 2021, 193, 108074. [CrossRef]
2. Niknejad, N.; Ismail, W.B.

## B0049 · 第20页 · 清洗文本字符1201–2400

; Mardani, A.; Liao, H.; Ghani, I. A Comprehensive Overview of Smart Wearables: The State of the Art
Literature, Recent Advances, and Future Challenges. Eng. Appl. Artif. Intell. 2020, 90, 103529. [CrossRef]
3. Siepmann, C.; Kowalczuk, P . Understanding Continued Smartwatch Usage: The Role of Emotional as Well as Health and Fitness
Factors. Electron. Mark. 2021, 31, 795–809. [CrossRef]
4. Celik, Y.; Godfrey, A. Bringing It All Together: Wearable Data Fusion. npj Digit. Med. 2023, 6, 149. [CrossRef]
5. Bölen, M.C. Exploring the Determinants of Users’ Continuance Intention in Smartwatches. Technol. Soc. 2020, 60, 101209.
[CrossRef]
6. Cha, D.H.; Park, H.S.; Hwang, Y.; Kim, J.-H.; Kim, H.-J. Experimental Study of Glass Molding Process and Transcription
Characteristics of Mold Surface in Molding of Aspheric Glass Lenses. Opt. Rev. 2011, 18, 241–246. [CrossRef]
7. Zhou, T.; Du, Y.; Zhou, J.; Chen, Z.; Liang, Z.; Wang, X. High Efficiency Fabrication of 3D Curved Glass by Full-Motor-Driving
Precision Molding. In Proceedings of the Optical Precision Manufacturing, Testing, and Applications, Beijing, China, 22–24 May
2018; SPIE: Bellingham, WA, USA, 2018; Volume 10847, pp. 265–270.
8. Huang

## B0050 · 第20页 · 清洗文本字符2401–3600

, H.; Yang, W.; Ming, W.; Zhang, G.; Xu, Y.; Zhang, Z. Mechanism of Springback Behavior in Ultra-Thin Glass Molding
Process: A Molecular Dynamics Study. J. Non-Cryst. Solids 2022, 596, 121841. [CrossRef]
9. Fotheringham, U.; Baltes, A.; Fischer, P .; Höhn, P .; Jedamzik, R.; Schenk, C.; Stolz, C.; Westenberger, G. Refractive Index Drop
Observed After Precision Molding of Optical Elements: A Quantitative Understanding Based on the Tool–Narayanaswamy–
Moynihan Model. J. Am. Ceram. Soc. 2008, 91, 780–783. [CrossRef]
10. Ming, W.; Jia, H.; Huang, H.; Zhang, G.; Liu, K.; Lu, Y.; Cao, C. Study on Mechanism of Glass Molding Process for Fingerprint
Lock Glass Plates. Crystals 2021, 11, 394. [CrossRef]
11. Guloyan, Y.A. Glass Molding (Review). Glass Ceram. 2018, 74, 305–315. [CrossRef]
12. Nelson, J.J. Overview of Glass Molding Processes. In Precision Lens Molding of Glass: A Process Perspective ; Nelson, J.J., Ed.;
Springer: Singapore, 2020; pp. 1–14. ISBN 9789811542381.
13. Chen, Y.; Zhang, S.; Hu, S.; Zhao, Y.; Zhang, G.; Cao, Y.; Ming, W. Study of Heat Transfer Strategy of Metal Heating/Conduction
Plates for Energy Efficiency of Large-Sized Automotive Glass Molding Process. Metals 2023,

## B0051 · 第20页 · 清洗文本字符3601–4800

 13, 1218. [CrossRef]
14. Axinte, E. Glasses as Engineering Materials: A Review. Mater. Des. 2011, 32, 1717–1732. [CrossRef]
15. Chao, C.-L.; Huo, C.-B.; Chou, W.-C.; Lin, Y.-R.; Ma, K.-J.; Chien, H.-H. Study on the Design of Precious Metal Based Protective
Films for Glass Moulding Process. Surf. Coat. Technol. 2013, 231, 567–572. [CrossRef]
16. Chang, H.; Zhang, Z.; Wang, P .; Zhang, G. A Novel Composite Film with Superhydrophobic Graphene for Anti-Icing/Deicing via
Chemical-Assisted Magnetically Controllable Picosecond Laser Writing. Mater. Today Phys. 2025, 54, 101726. [CrossRef]
17. Tao, B.; Yuan, Y. FEM Simulation of the Effect of Coefficient of Thermal Expansion and Heat Capacity on Prediction of Residual
Stresses of Compression Molded Glass Lenses. IOP Conf. Ser. Mater. Sci. Eng. 2017, 269, 012050. [CrossRef]
18. Yan, J.; Zhou, T.; Masuda, J.; Kuriyagawa, T. Modeling High-Temperature Glass Molding Process by Coupling Heat Transfer and
Viscous Deformation Analysis. Precis. Eng. 2009, 33, 150–159. [CrossRef]
19. Su, L.; Wang, F.; He, P .; Dambon, O.; Klocke, F.; Yi, A.Y. An Integrated Solution for Mold Shape Modification in Precision Glass
Molding to Compensate Refractive Inde

## B0052 · 第20页 · 清洗文本字符4801–5119

x Change and Geometric Deviation. Opt. Lasers Eng. 2014, 53, 98–103. [CrossRef]
20. Ming, W.; Jiang, Z.; Chen, Z.; He, W.; Li, X. Modelling and Analysis of Energy Consumption in Glass Molding Process for
Smartphone Covers Using Different Heating Strategies. Int. J. Adv. Manuf. Technol. 2023, 124, 1491–1512. [CrossRef]

## B0053 · 第21页 · 清洗文本字符1–1200

Micromachines 2025, 16, 584 21 of 21
21. Zhang, Z.; Ming, W.; Zhang, Y.; Yin, L.; Xue, T.; Yu, H.; Chen, Z.; Liao, D.; Zhang, G. Analyzing Sustainable Performance on
High-Precision Molding Process of 3D Ultra-Thin Glass for Smart Phone. J. Clean. Prod. 2020, 255, 120196. [CrossRef]
22. Ming, W.; Chen, Z.; Du, J.; Zhang, Z.; Zhang, G.; He, W.; Ma, J.; Shen, F. A Comprehensive Review of Theory and Technology of
Glass Molding Process. Int. J. Adv. Manuf. Technol. 2020, 107, 2671–2706. [CrossRef]
23. He, W.; Chen, Z.; Ming, W.; Du, J.; Cao, Y.; Ma, J.; Wei, A. Multi-Objective Optimization of Glass Multi-Station Bending Machining
for Smartphone Curved Screen. J. Braz. Soc. Mech. Sci. Eng. 2019, 41, 476. [CrossRef]
24. Liu, G.; Vu, A.-T.; Dambon, O.; Klocke, F. Glass Material Modeling and Its Molding Behavior. MRS Adv. 2017, 2, 875–885.
[CrossRef]
25. Dambon, O.; Wang, F.; Klocke, F.; Pongs, G.; Bresseler, B.; Chen, Y.; Yi, A.Y. Efficient Mold Manufacturing for Precision Glass
Molding. J. Vac. Sci. Technol. B Microelectron. Nanometer Struct. Process. Meas. Phenom. 2009, 27, 1445–1449. [CrossRef]
26. Ming, W.; Jia, H.; Liu, X.; Liu, X.; He, W.; Wei, A.; Chen, Z. Optimization of Process Pa

## B0054 · 第21页 · 清洗文本字符1201–2400

rameters of GMP for 3D Ultrathin Glass
Based on Cuckoo Search Algorithm. IOP Conf. Ser. Mater. Sci. Eng. 2020, 768, 042046. [CrossRef]
27. Chen, Z.; Hu, S.; Zhang, S.; Zhang, Q.; Zhang, Z.; Ming, W. Simulation and Experimental Study on the Precision Molding of
Irregular Vehicle Glass Components. Micromachines 2023, 14, 1974. [CrossRef]
28. Zhou, J.; Shi, T.; Hu, Y.; Ji, Y.; Li, M.; Shen, L. Numerical Simulation in Compression Molding of Glass Lens. In Proceedings of
the 2013 IEEE International Conference on Automation Science and Engineering (CASE), Madison, WI, USA, 17–20 August 2013; IEEE:
Piscataway, NJ, USA, 2013; pp. 669–674.
29. Soules, T.F.; Busbey, R.F.; Rekhson, S.M.; Markovsky, A.; Burke, M.A. Finite-Element Calculation of Stresses in Glass Parts
Undergoing Viscous Relaxation. J. Am. Ceram. Soc. 1987, 70, 90–95. [CrossRef]
30. Zhou, T.; Yan, J.; Masuda, J.; Kuriyagawa, T. Investigation on the Viscoelasticity of Optical Glass in Ultraprecision Lens Molding
Process. J. Mater. Process. Technol. 2009, 209, 4484–4489. [CrossRef]
31. Zhou, T.; Yan, J.; Kuriyagawa, T. Evaluating the Viscoelastic Properties of Glass above Transition Temperature for Numerical
Modeling of Lens Mold

## B0055 · 第21页 · 清洗文本字符2401–3600

ing Process. In Proceedings of the International Symposium on Photoelectronic Detection and Imaging
2007: Optoelectronic System Design, Manufacturing, and Testing, Beijing, China, 9–12 September 2007; SPIE: Bellingham, WA,
USA, 2008; Volume 6624, pp. 25–35.
32. Storck, K.; Karlsson, M.; Loyd, D. Analysis of the Blank Mould—A Transient Heat Transfer Problem in Glass Forming. WIT Trans.
Eng. Sci. 1970, 5, 8.
33. Keijts, G.; Werff, K.V .D. Heat Transfer in Glass Container Production during the Final Blow. Glass Technol. 2001, 42, 104–108.
34. Zhang, Y. Study on the Key Techniques. Ph.D. Thesis, Tianjin University, Tianjin, China, 2023.
35. Zhang, Z.; Yang, W.; Ming, W.; Yin, L.; Liao, D.; Zhang, G. Surface Integrity Analysis of Ultra-Thin Glass Molding Process.Ceram.
Int. 2021, 47, 30584–30597. [CrossRef]
36. Deb, K.; Jain, H. An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point-Based Nondominated Sorting
Approach, Part I: Solving Problems With Box Constraints. IEEE Trans. Evol. Comput. 2014, 18, 577–601. [CrossRef]
37. Tian, M.; Gong, X.; Yin, L.; Li, H.; Ming, W.; Zhang, Z.; Chen, J. Multi-Objective Optimization of Injection Molding Process
Parameters in Two S

## B0056 · 第21页 · 清洗文本字符3601–4748

tages for Multiple Quality Characteristics and Energy Efficiency Using Taguchi Method and NSGA-II.Int. J.
Adv. Manuf. Technol. 2017, 89, 241–254. [CrossRef]
38. Deb, K.; Pratap, A.; Agarwal, S.; Meyarivan, T. A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-II. IEEE Trans. Evol.
Comput. 2002, 6, 182–197. [CrossRef]
39. Yang, W.; Zhang, Z.; Ming, W.; Yin, L.; Zhang, G. Study on Shape Deviation and Crack of Ultra-Thin Glass Molding Process for
Curved Surface. Ceram. Int. 2022, 48, 6767–6779. [CrossRef]
40. Wang, P .; Zhang, Z.; Wei, S.; Hao, B.; Chang, H.; Huang, Y.; Zhang, G. Study on Plasma Behaviour, Ablation Mechanism, and
Surface Morphology of CFRP by Underwater Laser-Induced Plasma Micro-Machining. J. Mater. Process. Technol. 2025, 338, 118757.
[CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.

