Structure analysis with molding effects for
injection-molded plastic parts
Allen Y. Peng, Wen-Hsien Yang and David C. Hsu
CoreTech System Co. Ltd., HsinChu, Taiwan
Abstract: In the recent years, an increasing number of industrial products are made of engineering
plastic for its low cost and superior material properties, such as automotive parts. Since the
requirement of quality is stricter and the amounts needed is higher day by day. Hence, how to
guarantee the good quality and reduce the cycle time of development are the major concerns. The
application of CAE analysis in plastic part is also becoming popular, especially for part structure
design and molding process optimization. Traditionally, the structure analysis for injection-molded
plastic part is to perform CAE analysis based on the assumption of one or several isotropic
materials. However, the material characteristic of plastic part is extremely dependent on molding
process. The process-induced properties, such as fiber-induced anisotropic mechanical
properties, might not be favorable to the structural requirement of final products. Besides, the
mesh requirement for different CAE analysis might be not the same. In this paper, we consider the
molding properties in ABAQUS structure analysis through the data link from mold-filling analysis
to study the effects of mold design and molding process towards part structure. The results of
ABAQUS analyses illustrate the mold design and molding process will affect the characteristic of
part structure, and the better designs are found out. Furthermore, this comprehensive investigation
will help us toward both the good quality of product and reducing the cycle time of development
simultaneously.
Keywords: Structure mechanics analysis, Molding effects, Fiber orientation, Injection-molded
plastic part
1. Introduction
In the recent years, an increasing number of industrial products are made of engineering
plastic for its low cost and superior material properties, such as automotive parts. Since the
requirement of quality is stricter and the amounts needed is higher day by day. Hence, how to
guarantee the good quality and reduce the cycle time of development are the major concerns. The
application of CAE analysis in injection-molded plastic part is becoming popular in the recent
years, especially for part structure design and molding process optimization. Users usually study
the designs and manufactures through numerous individual CAE tools. Many major corporations
have adopted ABAQUS across all engineering disciplines as an integral part of design process for
injection-molded plastic products.
However, the material characteristic of plastic part is extremely dependent on molding
process. The process-induced properties, such as fiber-induced anisotropic mechanical
properties, might not be favorable to the structural requirement of final products. The traditional
structure analysis is to perform CAE analysis based on the assumption of one or several isotropic
2005 ABAQUS Users’ Conference
1
materials. But it neglects some molding effects. Sometimes the results of analysis could be
different from reality.
The injection molding of fiber-reinforced thermoplastics is a complicated process. The
reinforced composites don’t possess isotropic material properties. The thermal and mechanics
properties of the composite strongly depend on the fiber orientation pattern. The composite is
stronger in the fiber orientation direction and weaker in the transverse direction. In this paper, we
integrate structure mechanics and mold-filling analysis to enhance ABAQUS analysis for
injection-molded fiber-reinforced plastic parts.
2. Analysis Theory
2.1.
Filling:
The governing equations to simulate the 3D, transient, and non-isothermal polymer flow
with free surface are as follows,
∂ρ
+ ∇ ⋅ ρu = 0
∂t
(1)
∂
(ρu ) + ∇ ⋅ (ρuu − σ ) = ρg
∂t
(2)
(
σ = − pI + η ∇ u + ∇ u T
)
 ∂T

ρC P 
+ u ⋅ ∇T  = ∇(k∇T ) + ηγ& 2
∂
t


(3)
(4)
where u is the velocity vector, T the temperature, t the time, p the pressure, σ the total stress
tensor, ρthe density, η the viscosity, k the thermal conductivity, Cp the specific heat and γ& the
shear rate. The FVM due to its robustness and efficiency is employed in this study to solve the
transient flow field in complex three-dimensional geometry.
2.2.
Fiber orientation:
The fiber orientation state at each point in the part is represented by a 2nd-order orientation
vector A,
Aij = ∫ ( p i p j )ϕ ( p )dp
(5)
The equation of orientation change for the orientation tensor proposed by Advani and Tucker [1] is
employed for the analysis,
∂Aij
∂t
+ uk
∂Aij
∂x k
= Aik Ω kj − Ω ik Akj +
(6)
λ (Aik E kj + E ik Akj − 2 Aijkl E kl ) + 2C I γ& (δ ij − 3 Aij )
where CI is the interaction coefficient with the value ranged from 10-2 to 10-3. For the fourth-order
2
2005 ABAQUS Users’ Conference
tensor Aijkl , a closure approximation is needed.
2.3.
Integration approach:
This approach is to link the data between mold-filling analysis and structure analysis. The
anisotropy moduli and thermal expansion coefficients of the fiber-reinforced plastic can be
obtained through the integration of fiber orientation and composite model. The molding-induced
anisotropic mechanical properties will be taken into account in structure mechanics analysis.
Besides, the concern of mesh requirement might be not the same for mold-filling analysis
and structure analysis. The mesh for structure analysis could be focused on the area of stress
concentration. However, the mesh for mold-filling analysis is stressed on the higher element
resolution across the thickness direction. This approach further develops the mapping function to
map the element properties from mold-filling-specified mesh to structure-specified mesh. It
correctly matches the elements and maps the material properties even though the mesh
characteristics are totally different, as shown in Figure 1.
3. Results and discussion
A rectangular plate of 100x50x1 mm molded with glass-fiber reinforced PET is simulated to
validate the prediction of fiber orientation. Figure2 shows the part geometry and the filling pattern.
The gate is located in the center of the plate. Figure3 displays the fiber orientation on the different
cut planes. The orientation of the lines indicates the most favorable orientation direction and the
color of each line represents the degree of orientation. In the vicinity of mold wall, we can see that
the shearing flow tends to align the fibers along the flow. In the center cut plane, the flow is shear
free and hence the fiber orientation is perpendicular to the flow direction. In Figure4, the
prediction shows the fiber alignment along the welding line. These analysis results agree well with
the experimental observation.
Two actual industrial parts are analyzed to demonstrate the capabilities of proposed
approach. A wrench made of fiber-reinforced engineering plastic is shown in Figure 5. The model
is meshed by the 6-node prismatic element and 8-node hexahedral element. The gate is located in
the end of handle. The resin is PA66 with 30% glass-fiber. Figure 6 and Figure7 show the material
properties. Figure 8(a) is the predicted melt front distribution on the cavity surface. To further
demonstrate how the cavity is filled, the iso-surfaces of melt front are plotted in Figure 8(b). The
stronger fiber orientation direction along the longitudinal direction is illustrated in Figure 9(a). In
addition, the anisotropic mechanical properties are estimated by using Halpin-Tsai composite
model with fiber orientation, fiber aspect ratio and fiber concentration considered. The modulus in
the major-axis direction is shown in Figure 9(b). The ABAQUS mesh with fiber-induced
anisotropic properties through the integrated interfacing program is shown in Figure 10(a). In
order to further demonstrate how the molding effect is, a set of boundary condition is created, as
shown in Figure 10(b). The total deflections are shown in Figure 11(a) and Figure 11(b). The
von-mises stresses are shown in Figure 12(a) and Figure 12(b). These results show the stress
distributions and deflections of the part depend heavily on injection molding process.
An automitive bumper with two runner designs is shown in Figure 13(a) and Figure 13(b).
This model is a typical thin-walled part with average wall thickness of 2.9 mm. The traditional
2.5D approach is enough to obtain good analysis results. 7,331 3-node triangular plate element are
2005 ABAQUS Users’ Conference
3
used in injection molding analysis. The resin is PET with 45% glass-fiber. Figure 14 and Figure15
show the material properties. The fiber orientation distributions of two mold designs are illustrated
in Figure 16(a) and Figure 16(b). Different mold designs will obtain different fiber orientation
distributions. In order to further carry out ABAQUS impact analysis, a mesh composed of 4-node
quadrilateral plate elements and 3-node triangular plate elements is created to increase the analysis
accurancy and reduce the computational loading, as shown in Figure 17. Through the mapping
approach, the fiber-induced anisotropic properties is mapped correctly to this ABAQUS-specified
mesh. We assume the mass of model is 800.0 kg and impact a rigid column with 4.0 km/hour rate,
such as Figure 17. The fixed constraints on bolts are also shown in Figure 18 as red points. The
period of time is 0.5 sec. And the explicit dynamic analysis is adopted. The results of random
orientation effects are shown in Figure 18. Figure 19 and Figure 20 show the results of
molding-induced orientation effects for different mold designs, including deflections and strain
energy. Besides, to further compare the displcement histroy of sensor nodes between different
mold designs, as shown in Figure 21. All these results are reasonable and show the structure
analysis of injection-molded plastic part depends heavily on molding conditions and mold design.
From these results the better mold design would be found out.
4. Conclusions
In this paper, we propose an approach to study the structure analysis with molding effects
for injection-molded plastic parts. Through the data link between mold-filling analysis and
ABAQUS, the molding-induced anisotropic characteristics are taken into account in structure
analysis. The results from ABAQUS static stress analysis and dynamic impact analysis show the
mechanics behaviors of fiber-reinforced plastic parts depend heavily on molding conditions and
mold design. This comprehensive investigation will help part/mold designers to study the effects
of mold design and molding process towards part structure in ABAQUS. Moreover, the correlation
between part design and mold design will be cleared up. It will be a cost-effect approach toward
both the good quality of product and reducing the cycle time of development simultaneously.
5. Reference
1.
2.
3.
4
S.G. Advani and C.L. Tucker, J. Rheol., 31, 751 (1987).
W.H.Yang, David C. Hsu, Venny Yang and R.Y.Chang, “Computer simulation of 3D short
fiber orientation in injection molding”, 470, ANTEC 2003, Nashville(2003)
Allen Peng, Yorker Chang, Anthony Yang, Venny Yang and F.C.Chuang, ”3D fiber
orientation and warpage analysis of injection-molded throttle valve”, 3rd Automotive
Composite Conference, Detroit (2003)
2005 ABAQUS Users’ Conference
6.
Figures
Figure 1. Map element property between two different meshes.
(a) Melt front
(b) Fiber orientation
Figure 2. Filling results of center-gate plate.
(a) On the wall
(b) In the center cut plane
Figure 3. Fiber orientations distribution of center-gate plate.
2005 ABAQUS Users’ Conference
5
(a) Melt front
(b) Fiber orientation
Figure 4. Filling results of side-gate plate.
Figure 5. Wrench model is meshed by 8,477 solid elements.
Figure 6. Material viscosity property.
6
2005 ABAQUS Users’ Conference
Figure 7. Material mechanical properties.
(a) On cavity surface
(b) Iso-surface
Figure 8. Melt front of wrench.
(a) Fiber orientation
(b) Major modulus
Figure 9. Fiber-induced properties of wrench.
2005 ABAQUS Users’ Conference
7
(a) Model is shown on ABAQUS/CAE
(b) Boundary conditions
Figure 10. ABAQUS analysis conditions.
(a) With orientation effect
(b) Without orientation effect
Figure 11. Deflection (scale: 5).
8
2005 ABAQUS Users’ Conference
(a) With orientation effect
(b) Without orientation effect
Figure 12. Von-Mises stress distribution.
(a) Mold design 1
(b) Mold design 2
Figure 13. Bumper models with two mold designs.
Figure 14. Material viscosity property.
2005 ABAQUS Users’ Conference
9
Figure 15. Material mechanical properties.
(a) Mold design 1
(b) Mold design 2
Figure 16. Average fiber orientation distributions.
Figure 17. Models and constraints for impact analysis.
10
2005 ABAQUS Users’ Conference
(a) Deflection (0.025 sec)
(b) Deflection (0.100 sec)
(c) Deflection (0.200 sec)
(d) Deflection (0.375 sec)
(e) All strain energy for whole model
Figure 18. Results of random orientation effects.
2005 ABAQUS Users’ Conference
11
(a) Deflection (0.025 sec)
(b) Deflection (0.100 sec)
(c) Deflection (0.200 sec)
(d) Deflection (0.375 sec)
(e) All strain energy for whole model
Figure 19. Results of mold design 1.
12
2005 ABAQUS Users’ Conference
(a) Deflection (0.025 sec)
(b) Deflection (0.100 sec)
(c) Deflection (0.200 sec)
(d) Deflection (0.375 sec)
(e) All strain energy for whole model
Figure 20. Results of mold design 2.
2005 ABAQUS Users’ Conference
13
(a) Locations of sensor nodes
(b) Results of random orientation effect
(c) Results of mold design 1
(d) Results of mold design 2
Figure 21. Displacement histories of sensor nodes (node 1: red line, node 2: light blue line).
7.
Acknowledgement
The authors would like to thank APIC Corp. for the partially supporting this research. The
results of ABAQUS impact analysis for bumper case reported in this paper are finished under
APIC Corp. assistances.
14
2005 ABAQUS Users’ Conference