Quality Improvement of Polycarbonate Medical Device by Moldex3D and Taguchi DOE

Abstract

Incorporating a new design combined with a thermoplastic material without having a prior process database may cause numerous defects in the molded part. Along with the initial setup cost, the scrape due to defects can add additional costs to the production. To address such issues, the numerical analysis tool Moldex3D was used to conduct simulations, and Taguchi Design of Experiment (DOE) was performed to optimize the process parameters. For this study, a classic medical device like the safety goggle model has been chosen as a subject. An amorphous material, polycarbonate, was used because of its transparent and significant thermal properties. The most influential defects, like the warpage and short shot, needed to be addressed together. Process parameters like melt temperature, injection speed, coolant temperature, and packing pressure were considered to address those defects. The goal of this work was to improve both warpage and short-shot defects at the same time using the same combination of process parameters. By performing the simulation and statistical analysis method, 25% of warpage defects had been reduced in the molded part. A total of 2.3% of any potential short-shot defect reduction was recorded. Additionally, 300 °C melt temperature, 80 mm/s injection speed, 10 MPa packing pressure, and 100 °C coolant temperature were the best combinations for the part that had uneven thickness and narrow flow channels. The combinations to reduce potential short shots were 280 °C melt temperature, 40 mm/s injection speed, 10 MPa packing pressure, and 90 °C coolant temperature. Moreover, packing pressure for the warpage and melt temperature for the short shots were the most significant factors of all.

1. Introduction

The injection molding process is one of the most commercially used manufacturing processes for plastic parts used in the medical industry. The demand for high-quality plastic products in small to big corporations continues to rise. There is an increasing need to optimize the injection molding process to achieve enhanced manufacturing efficiency, improved product quality, and reduced production costs. In plastics, injection molding led the market, and it accounted for 98.3% of revenue share in 2021 owing to low-cost technology suitable for high-volume applications in intricate design and tolerances [1]. It has been around 150 years since injection molding was first invented. However, significant research and development in producing a defect-free molded part using low-cost and easy solutions are continuing. There are numerous companies that do not have technical experts in this field. Most of the company’s molders depend on real-life experiences and trial-and-error methods when they set up the molding machine. In the event of setting up a completely new material and new design, they often have no database. Hence, trial and error are their first choice.

A literature review was conducted in previous studies to find out what factors affect what type of defects in the thermoplastic injection-molded process. In Appendix A,

Table A1. A survey of previous research on molded parts defects.

No.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Method

Experimental

R

×

×

R

×

×

×

×

×

×

×

R

×

×

R

×

Simulation

×

×

×

×

×

×

×

×

×

×

×

×

×

×

×

×

×

×

×

Parameters

Mold Temp.

×

×

×

×

×

×

Melt temp.

×

×

×

×

×

×

×

×

×

×

×

×

Mold Design

×

×

×

×

Gate number

×

Gate Location

×

Melt viscosity

×

Mold Materials

×

×

Packing Pressure

×

×

×

×

×

×

×

×

x

Packing Time

×

×

×

×

Pressure Profile

×

Injection time

×

×

×

×

×

×

Injection Pressure

×

×

×

×

×

×

×

×

×

x

Injection Speed

×

×

×

×

×

×

Cycle time

×

Cooling time

×

×

×

×

Back pressure

×

Screw speed

×

Barrel Temp.

×

Spot heating

×

Tensile Strength

×

Temperature at Flow Front

×

Time to reach ejection temperature

×

Coolant Temperature

×

×

Molding Cycle

×

Pressure Drop

×

Defects

Shrinkage

×

×

×

×

×

Warpage

×

×

×

×

×

×

×

×

×

Weld line

×

×

×

×

×

Air trap

×

×

Surface Defect

×

Sink Marks

×

×

×

Incomplete Product

×

×

Flash

×

Short-Shot

×

×

x

Air Bubble

×

Silver lines

×

Surface Gloss

×

Surface Waviness

×

Impact Strength

×

Surface Roughness

×

×

Material Rheology

×

Void

×

Optimization

x

x

x

x

x

x

x

x

x

x

x

Light blue row = numerical order, green rows = method of experiment, blue rows = parameters, orange rows = associated defects, yellow row = optimization.

presents a survey that has been conducted on injection molding defects associated with the factors affecting it. It also represents if there was any optimization conducted on it and what type of work it was based on—experimental or simulation. A study shows that an average scrap rate detection in injection molding can be up to 10% of its total production. Among these, some of the defects can cause scrap up to 2% [2]. Thus, from the literature review, warpage and short shot defects have been chosen to optimize in this study.

Warpage is a shape distortion that happens after part ejection from the mold and cooling process. It causes unintentional bending or twisting in the parts [3]. When a melted polymer is injected at a high speed into the mold, the chain molecular orientation of the polymer can be in different directions and shapes due to process conditions. To fill the entire cavity, the melt front changes its type abruptly and affects the molecular orientation. Moreover, the temperature difference in both side mold plates creates an asymmetric melt front [4]. Consequently, uneven cooling is observed in the molded shape. It happened due to thermally induced stress in the molecular chain structure. Both molecular orientation and thermally induced stress result in uneven shrinkage at various locations of the molded part and eventually produce a distorted part or warpage defect [5,6]. The twisted or bent surfaces of parts mainly happen due to uneven mechanisms such as uneven cooling. Recent studies show that mold design, process parameters (including melt temperature, injection speed, packing pressure, etc.), cooling time, material rheology (amorphous or semicrystalline), differential cooling rate, and others are responsible for warpage [6]. Melt temperature and packing pressure are two of the most important factors [7]. Those have various effects on amorphous and semicrystalline materials. Semicrystalline material polypropylene shows decreasing warpage when the packing pressure rises. It also showed an increase in warpage when melt temperature rises. If the melt temperature increases from 200 °C to 260 °C, the part shrinks from 1.606% to 1.615%. Changing packing pressure from 28.3 MPa to 120 MPa changed volumetric shrinkage from 11.828% to 6.295% [8]. Moreover, research shows that for semicrystalline materials, increasing injection speed reduces warpage in molded parts. It shows that 50 kg/cm² has a higher signal-to-noise ratio (SN) than at 40 kg/cm² injection speed. Higher injection speed gives a robust molded part [9]. A few research studies have been conducted on coolant temperature effects to improve warpage defects. It is proved both in simulation and experimentally that the difference in cavity coolant temperature and core coolant temperature up to 2 °C can give improved warpage up to 75.2% [10].

A mold cavity can complete a cycle even without filling it completely. When a melt starts to flow, it dissipates the heat by conducting through the mold due to low-temperature coolant flow. The flow can solidify completely before reaching the endpoint. It produces heat from the shear of flow layers. This process depends on many factors, such as mold design, mold thickness, process condition, material properties, cooling channel design, and flow rate. If the melt front solidifies completely, it will no longer fill up the cavity and create a short shot [11]. As short-shot defects can disable the functionality of a molded part, manufacturers are concerned about it. Numerous research works are available addressing the causes of short shots and their elimination methods. From the literature work, many relevant causes of the short-shot defects are identified. The solution method is also established based on the causes. Melt temperature is one of the significant process parameters affecting short shots. Low melt temperature can solidify the melt front way before the flow length. Moreover, a very high melt temperature makes melt materials less viscous. Low viscosity flow cannot fill all corners of the cavity precisely. Mehdi Moayyedian et al. showed that melt temperature is a significant factor in reducing short shots. According to them, melt temperature contributed 74.25% experimentally and 75.04% in simulation in polypropylene material. Increasing melt temperature from 200 to 280 °C reduced short shot by increasing signal-to-noise ratio [12]. Injection speed and pressure are two vital factors for short-shot minimization. Increasing pressure and speed gives better results, reducing short shots in polypropylene. Low injection speed and injection pressure result in a low flow rate. A low flow rate is unable to complete mold cavities. In polypropylene thermoplastic, 2000 bar of injection pressure and 350 cm/s of injection speed are best for minimizing short shots [13]. Besides, packing pressure impacts short shots both directly and indirectly. It is the secondary pressure that applies to melting materials to adjust the incomplete area. The incomplete area remains inside the mold cavity due to faster solidification of melt materials at the first stage of cavity filling. According to Satoshi Kitayama et al., high packing pressure at the beginning of the packing pressure results in no short shots [14]. However, there was no prior work on the coolant temperature effect.

The above-mentioned process parameters play a vital role in warpage and short-shot reduction. However, the causes of warpage and short shots have been identified in numerous studies over the years. It was also studied how it can be reduced by numerous methods. While it may be true all of such work has been conducted before, the aim of this research is to show a combined method of simulation in the user-friendly CAE software Moldex3D (2023), analysis of flow and defects, and optimization of the process parameters in Minitab—all in one study. These combinations of process parameters, defects, simulation, and Taguchi DOE optimization were never conducted together before. The scope of this study is limited to only process parameter analysis and optimization. This study will not discuss any design parameter of the part or cooling channels. This work will be significant for polycarbonate injection molding as it provides process parameter data on a particular design. Moreover, it shows how efficiently and with no prior experience in simulation, one can start building models in Moldex3D and solve some destructive defects at a low cost.

2. Materials and Methods

2.1. Medical Device Selection

The experiment was completed numerically using Moldex3D simulation software. A protective goggle, which is a class 1 medical device, was selected for this experiment [15]. It was selected because of its uneven thickness, 3D structures, holes with narrow flow channels, and symmetrical structure along with one axis. The CAD model of this product was imported from GrabCAD, shown in Figure 1 [16].

For simulation purposes, the SolidWorks 2023 CAD model was simplified to reduce some intricate corners. The length of the part is 154.50 mm, and the width is 72.50 mm. The top view of the engineering sketch is shown in Figure 2. It varies from 2.08 mm to 0.033 mm.

2.2. Mold Design in Moldex3D

Moldex3D 2023 software was used for this research [17]. In the Modlex3D software, a mold was designed to accommodate goggles shaped. After the pin gate, a cooling channel was created in both the core and cavity sides of the molded part. The cooling channels were not conforming to the cooling channels. However, it was created following geometry and within the 5–10 mm proximity of the part. The cooling channel diameter of the core side is 7.5 mm. The diameter of the cavity side cooling channel is 9.5 mm. This diameter was a default suggestion by the software itself. A square-shaped mold was created after the completion of the cooling channel. The mold is 290 mm × 290 mm for both core and cavity sides. The depth or height of both the cavity and core was 140 mm each. The total height was 280 mm. Figure 3 shows the complete details of the numerical model in Moldex3D. Figure 3a shows the selected part for our analysis along with a gate and very small runner as it is a single cavity mold. In Figure 3b, the cooling to the core was introduced. Figure 3c has the cooling channels of the cavity. Figure 3d represents a complete simulation model with a mold box including core, cavity, gate, runner and cooling channels.

2.3. Materials Selection

Polycarbonate is an amorphous thermoplastic polymer. It has been widely used in the medical industry. Due to its high thermal resistance, it is suitable for the sterilization process of medical products. Another great characteristic of polycarbonate is its transparent appearance, which can replace glass usage. It has high strength and resistance at elevated temperatures and light transmission properties [18]. A few properties of polycarbonate are listed in

Table 1. Properties of Polycarbonate [19].

Materials Properties	Value
Melting Temperature	267 °C
Density	1210 kg/m3
Melt Flow Index	6 g/10 min @(300 °C, 1.2 Kg)
Thermal Expansion	68 × 10−6/K
Elongation	95%
Specific Heat	1380 J/kg·K

.

2.4. Mesh Setup for the Model

In the Moldex3D mold structure model, each component has a different mesh type, as shown in

Table 2. Mesh type for each part by Moldex3D.

Geometry Attribute	Mesh Type
Part	5 layers of BLM
Runner	5 layers of BLM
Cooling Channel	3 layers of BLM
Mold Insert	Pure Tetra

. The goggles part and the runner/gate have five layers of Boundary Layer Meshing (BLM) with an offset ratio of 3. The cooling channel has three layers of BLM with an offset ratio of 3. Mold inserts, or mold, have a mesh type of pure tetra. These mesh types were commended by the Moldex3D company experts (Farmington, Michigan, USA). There was no part insert. The mesh size for the part was 0.9 mm. Also, the mesh densities for the part and runner are mentioned in

Table 3. Mesh Densities.

Geometry Attribute	No. of 3D Solid Mesh/Volume in mm3
Part	10.00
Runner	48.51

.

Table 3 shows the mesh densities in this model.

For mesh convergences, the h-refinements process was followed where the seeding size decreased and the mesh number increased. Figure 4 shows that at around 1.4 million solid mesh numbers, the model was converging.

2.5. Process Parameter Selection

Process parameters are the input or design variables of this simulation work. Melt temperature, packing pressure, injection velocity, and coolant temperature are the four factors of the DOE. Since all factors are quantitative, three levels were decided for this study. The selected level for polycarbonate is given in

Table 4. Variable process parameters selected at different levels.

Levels	Melt Temperature, (°C)	Packing Pressure (MPa)	Coolant Temperature (°C)	Injection Speed (mm/s)
1	280	10	90	40
2	300	40	100	60
3	320	70	110	80

and

Table 5. Other fixed process parameters.

Parameters	Type/Values
Mold Temperature (°C)	100
Mold Materials	Aluminum 6061
Injection Pressure (MPa)	150
Eject Temperature (°C)	132
Coolant Material	Water
Coolant Flowrate (cm3/s)	120

, respectively. The software has limits on each process parameter. A separate study was conducted on different levels of each parameter to observe any effects on the part quality. The levels that had significant effects on the defects were selected for this experiment. Other than these process parameters, the default suggested settings were followed.

Besides the above-mentioned parameters, all other process parameters for polycarbonate are given in Figure 4.

2.6. Mathematical Assumptions

Moldex3D software has many advanced features to conduct CAE analysis of an injection molding flow. The software is made of mathematical models and assumptions. A few related models and assumptions are mentioned in this section. For standard injection molding, the plastic melt is assumed to be an incompressible and generalized Newtonian fluid. This flow considers the conservation of mass, conservation of momentum, and conservation of energy equations.

Conversation of mass equation: The mass of fluid is conserved.

$${\displaystyle \frac{\partial v}{\partial x}}+{\displaystyle \frac{\partial v}{\partial y}}+{\displaystyle \frac{\partial v}{\partial z}}=0$$

(1)

Conversation of momentum equation:

Sum of forces = Rate of change in momentum

$$\rho {\displaystyle \frac{Dv}{Dt}}=-\nabla p-\nabla ·\tau +\rho g$$

(2)

Conversion of energy:

Rate of change in energy = Rate of Heat added + Rate of work conducted

$$\rho C_p\left({\displaystyle \frac{\partial T}{\partial t}}+v·\nabla T\right)=\nabla ·\left(k\nabla T\right)+\eta \gamma$$

(3)

where v is the velocity vector, $\rho$ is the density, $C_p$ is the specific heat, $k$ is the thermal conductivity, T is the temperature, t is the time, $p$ is the pressure, $\eta$ is the viscosity, $\gamma$ is the shear rate, $\tau$ is the stress tensor, and g is the gravity.

In the post-processing section, a warpage measurement has been conducted in this research (Figure 5). A displacement of the initial dimension coordinate was measured to find out the warpage in the part. The variation u represents the displacement along the X direction in the coordinate system. The linear shrinkage is defined by the maximum displacement divided by the part’s original dimension L_x.

Maximum displacement in X direction is calculated by u_max − u_min.

Linear shrinkage in x direction is calculated by ${\displaystyle \frac{u_{max}-u_{min}}{L_x}}$.

2.7. Optimization Techniques

The Taguchi DOE was used to find out the optimum level for warpage and short shot reduction. The Taguchi method has been widely used by leading companies to improve the fundamental function of the product or process over the years. Reduction of variability is a robust strategy that is being followed in Taguchi DOE [20]. In this current study, the four different process parameters are the control factors or input. The responses are warpage, total displacement, and injected part weight. In the Taguchi method, a fractional factorial experiment was conducted instead of a full factorial. The Taguchi method provides options for collecting necessary data to determine which factors most affect product quality with a minimum amount of experimentation, saving time and resources [21].

Taguchi considers the difference between the target value of the performance characteristic of a process, $r$ and the measured value y, as a loss function shown below,

$$l\left(y\right)={K_c(y-r)}^2$$

(4)

The constant, $K_c$, in the loss function can be determined by considering the acceptable interval, delta, Δ.

When the goal is to minimize the characteristic value, the loss function is defined as follows:

$$l\left(y\right)={K_{c}y}^2$$

(5)

where $r$ = 0.

When the goal is to maximize the characteristic value, the loss function is defined as follows:

$$l\left(y\right)={\displaystyle \frac{K_c}{y^2}}$$

(6)

The current study followed L9 orthogonal arrays. The Taguchi Design DOE is based on the design matrix created by Genichi Taguchi.

Table 6. Single Design Matrix [22].

Factor Levels	2	3	4	5
No. of Runs
4	L4 (23)
8	L8 (27)
9		L9 (34)
12	L12 (211)
16	L16 (215)		L16 (45)
25				L25 (58)
27		L27 (313) L27 (322)
32	L32 (231)

shows the single design matrix associated with levels and factors. In this current study, all factors had three levels. This was a single matrix case. If there were mixed-up levels, a mixed design matrix would be followed, as shown in Table 6 [22].

To determine the effect of each variable on the output, the signal-to-noise (SN) ratio needs to be determined. To minimize the output, the SN ratio is calculated as shown below.

$$S{N}_i=-10\log \left(\sum _{u=1}^{N_i}{\displaystyle \frac{y_u^2}{N_i}}\right)$$

(7)

To maximize the output,

$$S{N}_i=-10\log \left({\displaystyle \frac{1}{N_i}}\sum _{u=1}^{N_i}{\displaystyle \frac{1}{y_u^2}}\right)$$

(8)

where i is the experiment number, u is the trial number, N_i is the number of trials for the experiment, and y_i is the value of the performance characteristic for a given experiment [21]. After finding the SN ratio and minimizing or maximizing the output, an ANOVA was performed to determine the most influential factors among the four process parameters. Minitab 22 software was used to complete the whole statistical analysis, including creating the DOE array shown in

Table 7. Design for Experiment Runs for Polycarbonate.

Case No.	Melt Temperature	Injection Speed	Packing Pressure	Coolant Temperature	Pattern
1	1	1	1	1	----
2	1	2	2	2	-000
3	1	3	3	3	-+++
4	2	1	2	3	0-0+
5	2	2	3	1	00+-
6	2	3	1	2	0+-0
7	3	1	3	2	+-+0
8	3	2	1	3	+0-+
9	3	3	2	1	++0-

, SN ratio calculation, and ANOVA analysis. The pattern column represents the combination levels of each case. The negative sign (-) represents level 1, zero (0) represents level 2 and positive sign (+) represents level 3.

3. Results

All cases listed in Table 7 were run in Moldex3D, and the warpage displacement and part weight were recorded. Warpage displacement results were discussed first, followed by the short shot results.

3.1. Warpage Displacement Measurement

The warpage displacement for all DOE cases of polycarbonate materials is listed in

Table 8. DOE for Warpage Displacement of Polycarbonate.

Case No.	Melt Temperature (°C)	Injection Speed (mm/s)	Packing Pressure (MPa)	Coolant Temperature (°C)	Warpage, mm	SN for Warpage
1	280	40	10	90	0.658	3.6355
2	280	60	40	100	0.649	3.7551
3	280	80	70	110	0.664	3.5566
4	300	40	40	110	0.646	3.7953
5	300	60	70	90	0.749	2.5104
6	300	80	10	100	0.559	5.0518
7	320	40	70	100	0.741	2.6036
8	320	60	10	110	0.630	4.0130
9	320	80	40	90	0.649	3.7550

. In the warpage, the goal was to minimize it. Thus, Equation (7) was followed, and the signal-to-noise (SN) ratio was calculated to remove noise and minimize variation. Eventually, the case that has less variation in warpage displacement will have the highest SN ratio. In Table 8, case 6 has an SN ratio of 5.05, which is the highest. It means 300 °C melt temperature, 80 mm/s injection speed, 10 MPa packing pressure, and 100 °C coolant temperature are the optimum settings for the selected goggles part for polycarbonate to reduce warpage displacement. On the other hand, Table 8 also indicates the lowest SN ratio, which is case 5. Case 5 has settings of 300 °C melt temperature, 60 mm/s injection speed, 70 MPa packing pressure, and coolant temperature of 90 °C. If both cases of warpage displacement are compared, it improves 25% of warpage defects. The Taguchi method reduces time and cost and improves warpage defects significantly.

Figure 6 shows the main effect plot of all four parameters. It shows the effect of different levels and their relationship to the defects. Injection speed and packing pressure levels have a great effect on reducing warpage defects in this part. On the other hand, melt temperature and coolant temperature levels have a low effect on reducing warpage in polycarbonate molded parts.

Table 9. Analysis of variance of warpage displacement SN ratio of polycarbonate.

Source	DF	Adj SS	Adj MS	F-Value	p-Value
Regression	4	4.02310	1.00578	7.79	0.036
Melt temp	1	0.05516	0.05516	0.43	0.549
Injection Speed	1	0.90407	0.90407	7.01	0.057
Packing Pressure	1	2.70654	2.70654	20.97	0.010
Coolant temp	1	0.35733	0.35733	2.77	0.171
Error	4	0.51620	0.12905
Total	8	4.53931

shows the analysis of variance (ANOVA) of the warpage displacement SN ratio, where DF is the degree of freedom, SS is the sum squares, and MS is the mean squares. The purpose of performing an ANOVA is to find out the variance between individual factors and find out the significant factor. It calculated the F-value and p-value. The closer the F-value is to 1, the less effect the factor has on warpage reduction. Also, if the p-value is less than 0.05, that factor has a significant effect on reducing the defect. Packing pressure has a p-value of 0.010, which is less than 0.05. It can be said that, among all other factors, packing pressure is the most statistically significant factor.

Figure 7 shows the warpage defects distribution on the product. The displacement varies at different locations based on a number of factors, such as its structure, inconsistent melt temperature, cooling channel distribution, injection speed, and packing pressure.

3.2. Short Shot Reduction

Table 10. Design of experiment for part weight of polycarbonate.

Case No.	Melt Temperature (°C)	Injection Speed (mm/s)	Packing Pressure (MPa)	Coolant Temperature (°C)	Part Weight, gm	SN Ratio
1	280	40	10	90	37.558	31.494
2	280	60	40	100	37.412	31.460
3	280	80	70	110	37.330	31.441
4	300	40	40	110	37.261	31.425
5	300	60	70	90	37.205	31.412
6	300	80	10	100	37.114	31.391
7	320	40	70	100	37.301	31.434
8	320	60	10	110	36.701	31.294
9	320	80	40	90	36.736	31.302

shows the part weights of nine cases of DOE. The column signal-to-noise ratio listed the SN ratio value using equation (8), which maximizes the output and reduces noise. Case 1 shows the highest SN ratio of 31.494 with a part weight of 37.558 gm. It has optimum settings of 280 °C melt temperature, 40 mm/s, 10 MPa packing pressure, and 90 °C coolant temperature. On the other side, case 8 has the lowest SN ratio of 31.294 with a part weight of 36.701 gm. The improvement of part weight or short shot defect compared between case 1 and case 9 is 2.3%.

Figure 8 shows the main effect plots for the part weight SN ratio. It shows how those factors affect different levels of output. It looks like the melt temperature has a great effect on part weight. Also, injection speed looks like a promising factor for part weight optimization. However, packing pressure and coolant temperature do not have much effect on part weight.

Table 11. Analysis of variance of part weight SN ratio of polycarbonate.

Source	DF	Adj SS	Adj MS	F-Value	p-Value
Regression	4	0.032706	0.008177	7.87	0.035
Melt temp	1	0.022280	0.022280	21.45	0.010
Injection Speed	1	0.008052	0.008052	7.75	0.050
Packing Pressure	1	0.001989	0.001989	1.91	0.239
Coolant temp	1	0.000386	0.000386	0.37	0.575
Error	4	0.004156	0.001039
Total	8	0.036862

shows the analysis of the variance of the SN ratio for the part weight of polycarbonate. Considering the F-value and p-value, melt temperature, which has a p-value of 0.010, is the significant factor for part weight for polycarbonate among all other factors.

Figure 9 represents the distribution of viscosity in the mold cavity. The higher the viscosity, the higher the part weight will be in a particular process parameter setting. Case 1 set has the best part weight. On the other hand, case 8 has the least part weight of all. However, the difference between the two extreme cases is not significantly high. However, there was no visible short shot on the part. It may be considered that the most part weight will reduce any chances of obtaining a short shot on the part.

4. Discussion

This study aimed to find an easy and low-cost solution for an injection molding setup for a particular design and material for molders with no prior simulation or optimization experiences. Two severe defects and their affecting process parameters were selected to conduct the study. The research has been conducted on Moldex3D software, which is very user-friendly. The optimization was conducted in Minitab. A goggle was selected as the product to conduct the entire simulation. It was limited to four process parameters only. The structure of the part or design of the cooling channel was not the subject of study. This study did not explore the product structure or cooling channel design. Moreover, warpage and short shots can be caused by other factors as well, which are not studied in this research.

Warpage was one of the severe defects. Several combinations of process parameters showed different improvement patterns for warpage. The best combination of process parameters to reduce warpage was observed as 300 °C melt temperature, 80 mm/s injection speed, 10 MPa packing pressure, and 100 °C coolant temperature. This study also showed that using the Taguchi DOE can improve warpage defects by 25% compared to the case that showed the most warpage in the part. In addition to that, packing pressure was selected as the most significant factor among all other factors to reduce warpage. Prior studies showed melt temperature and packing pressure were very influential in reducing warpage. This research is aligned with prior findings.

On the other hand, another defect, the short shot, was not significantly visible on the part. This work focused on maximizing the part weight. The more the part weighs, the less short the shot might occur in the molded part. Thus, the combination to reduce potential short shots in selected parts was 280 °C melt temperature, 40 mm/s injection speed, 10 MPa packing pressure, and 90 °C coolant temperature. The improvement of the potential short-shot defect was 2.3% by using Taguchi DOE. Furthermore, the most significant factor for short-shot reduction was found to be melt temperature. Hence, this finding aligns with the previous studies where it showed melt temperature was a significant process parameter to reduce any potential short shot or improve total part weight. Finally, to reduce both defects in the product, combinedly, 10 MPa packing pressure and 280 °C melt temperature can be selected with other factors mentioned above.

This study showed a combined improvement strategy with Moldex3D and statistical analysis for beginners. Moldex3D is user-friendly software with the necessary features, and it is easy to navigate as well. This research shows that this entire model and associated optimization work can be used to identify any other injection molded defects. It shows how the defects can be analyzed, and the Taguchi DOE method can be utilized quickly to find the best combinations that improve selected defects significantly. This work also provides a process database on polycarbonate material for the goggle part. Particularly a part that has an uneven thickness, which is responsible for warpage, and narrow flow channels, which are responsible for the short shot defects.

Further research activities can be performed to obtain some groundbreaking results. This research was focused on the single cavity. In industry, companies prefer multicavity injection molding. Multicavity simulation can be performed in the next work. This research work was completely focused on the process parameters. Besides process parameters, mold design, cooling channel design, mold materials, barrel screw speed and temperature, and other factors can be considered for the next simulation model. Moldex3D offers features to detect other injection molding defects as well. While observing warpage and short shots, weld line defects can be observed too. Statistical analysis can include weld lines to find optimum levels for process parameters. Further studies will greatly advance plastic injection molding manufacturing, including the medical industry and all other industries.

5. Conclusions

This study investigated a simple and cost-efficient approach for plastic injection molding defect reduction. It shows the model setup in Moldex3D, defects analysis, and factor selection, followed by optimization in Minitab as a beginner guideline. The results for warpage showed a 25% improvement, which indicates the selected process parameters were very crucial, especially packing pressure. Although there was no visible short shot in the part, the chance of getting a potential short shot was explored by improving the part weight. The improvement plan showed 2.3% better results. These findings provide significant guidelines for local companies that do not have molding experts with simulation and optimization knowledge. This study was limited to simulation, and further experiments with a few improvements of the model are needed. Future studies could explore multicavity models, change cooling channel design and optimization, use different thermoplastic materials, and explore different process parameters.