Minimizing Residual Stress in Aesthetic Automotive Components: A Study of Injection Pressure in ICM and IM Processes

2.1 Literature review

Three parameters: injection pressure, clamping force and shrinkage allowance, and the benefits of adopting ICM over IM, were selected for the initial search using Google Scholar and UniSA Open Athens. Initially search indicated limited availability of recent research publication for IM & ICM processes comparison and most of the literature is 5 years or older. The search was then refined by accessed directly through the UniSA library and Organisation details and product-specific information from their respective websites which successful found additional literatures.

The literature search results were categorised into (i) the benefits and limitations of ICM over IM, (ii) comparison of process parameters - injection pressure, clamping force and shrinkage allowance, (iii) the current research on the methodologies adopted in the ICM process and (v) identified the knowledge gaps in developing the ICM process.

2.2 Simulation approach

A multiple computer simulation on specimens to derive the relationship between ICM and conventional IM and then compare the result with a prototype tool was setup. An expert knowledge approach was commenced by an assumption (raw data), and then a quantitative analysis performed on the specimens to refine the results after an application in a case study. Only a specific grade of Polycarbonate (PC) thermoplastic material is considered to simplify the research. A research flowchart for the proposed methodology is given in Figure 2.

The simulation-based experimental analysis was used to derive the relationship between different conditions.

• Flow simulation was performed on the specimens with various combinations of with, length and thickness using Autodesk Moldflow® Ultimate 2022 software.

• ICM and IM processes were performed to analyse the injection pressure variation for each specimen.

• Injection pressure results were collected for statistical analysis.

The injection pressure of ICM and IM clamping force collected through multiple analyses on prepared specimens using Autodesk Moldflow® Ultimate software. Specimens of different wall thicknesses and sizes were modelled using CatiaV5 software. The data thus collected through simulation experiments were analysed and compared using SPSS software, and a relationship between ICM and IM on injection pressure and clamping force were derived. A prototype injection mould was used to verify this later in a selected case study.

2.2.1 Specimen preparation and analysis setup

2.2.1.1 Specimen preparation

Aesthetic automotive components generally have wall thicknesses ranging from 1.5 to 3.5 mm, so the research excludes wall thicknesses below 1 mm and above 4 mm. Similarly, the study was exclusively for the flow lengths up to 300 mm as most of the products are within this length or multiple gates are placed in series within this flow length. The flow length and wall thickness range were selected by reviewing various literature that underpins the generation of high injection pressure for high aspect ratio parts [5, 6].

3D models of these specimens were designed using Catia V5 R30 software [7]. The width of the part was set to 20 mm for consistency in comparing the injection pressure results. The feed system was designed with minimal runner length, and the gate thickness was set to the same wall thickness of the part. The gate width was set to 5 mm, and the length was set to 3 mm on all specimens. The runner diameter was set to 6 mm on all specimens for a length of 20 mm. Designing a gate thickness not less than the part wall thickness can avoid any possible pinch point that can affect the pressure. As the injection node was placed at the end of the runner, the flow length was increased by 25 mm, but the pressure drop along the cold feed system was disregarded in this study as the pressure was measured at the end of the gate, thus measuring only the cavity pressure.

Table 1 lists models to be analysed using the IM & ICM processes. In total, 42 combinations of wall thickness and flow length were analysed.

		Flow length
		50 mm	100 mm	150 mm	200 mm	250 mm	300 mm
Wall Thickness	1.0 mm	1.0 × 50	1.0 × 100	1.0 × 150	1.0 × 200	1.0 × 250	1.0 × 300
	1.5 mm	1.5 × 50	1.5 × 100	1.5 × 150	1.5 × 200	1.5 × 250	1.5 × 300
	2.0 mm	2.0 × 50	2.0 × 100	2.0 × 150	2.0 × 200	2.0 × 250	2.0 × 300
	2.5 mm	2.5 × 50	2.5 × 100	2.5 × 150	2.5 × 200	2.5 × 250	2.5 × 300
	3.0 mm	3.0 × 50	3.0 × 100	3.0 × 150	3.0 × 200	3.0 × 250	3.0 × 300
	3.5 mm	3.5 × 50	3.5 × 100	3.5 × 150	3.5 × 200	3.5 × 250	3.5 × 300
	4.0 mm	4.0 × 50	4.0 × 100	4.0 × 150	4.0 × 200	4.0 × 250	4.0 × 300

Table 1.

List of specimens.

Autodesk-Moldflow® [8] software recommends limiting the injection pressure to 100 MPa as a design guideline, though depending on case to case in industry practice, this could be accepted up to 150 MPa. Research by Yao and Kim [6] identified that injection pressure for the PC part with 1 mm thickness and 50 mm flow length was approximately 125 MPa, and the pressure increases exponentially. So utmost care should be taken while considering the pressure results of thin specimens with longer flow lengths, which may give unrealistic results.

2.2.1.2 Moldflow® analysis setup

All 42 models to be analysed were imported to the Moldflow® software. Then the mesh was generated on the model containing the part and the feed system. A global edge length of 0.5 mm was selected for all models, and the minimum number of tetra layers was selected to 10 to increase the prediction accuracy. Once the meshing was done, the tetras comprising the cold runner and cold gate were assigned as cold runners and moved to a different layer for easy interpretation of analysis results. As the research was only to find the injection (cavity) pressure distribution, only Fill + Pack analysis was selected for the analysis sequence. No cooling channels were designed for the project, thus assuming uniform and constant cooling throughout the moulding phase. Material was then selected to LEXAN™ Resin LS1 for analysis. This material was thoroughly tested for the filling and packing analysis, thus bearing a ‘gold rating’ for both. This means the viscosity, Specific Heat Capacity, pvT and Thermal Conductivity of this material were tested and verified by Autodesk Moldflow®, so the accuracy of results would be the highest.

The IM & ICM process settings were different, so the analysis needs to be set up differently. The melt temperature, mould temperature and injection flow rate were set to the same value for both processes. The mould temperature was set to 95°C, and the melt temperature was set to 300°C. A pilot analysis was carried out with the above temperature settings, with the filling control set to automatic. This provided the optimum flow rate of 5.5 cm³/s required for the part to fill 100% under optimum conditions. The same flow rate was populated for all specimens for analysis. Further, the V/P switchover was set to automatic for uniformity between analyses.

For ICM analysis, the press compression direction was set to the Z direction, and the press compression was set to start at the end of packing. This ensures that the press compression has no influence on the pressure profile during the packing phase. Press speed and press compression cap were set to a higher-than-usual amount to negate any influence due to limitations. Multiple analyses were carried out to set up the right compression gap and found that a press open distance of 0.26 mm was the optimum for compressing the press back to zero when packing. Specimens were then ready to run the simulations on IM and ICM processes and to review the results in detail.

2.2.1.3 Material selection

PC LEXAN™ Resin LS1 has been selected as suitable for these products and compatible with various PVD coatings. Moldflow® software has the complete material data information available, which was tested and confirmed at Autodesk Moldflow® Plastic Labs. PC LEXAN™ Resin LS1 available in the Moldflow® database has a gold rating (a high confidence in the material quality, thus expecting the most reliable result) for all the analysis types.

2.2.1.4 Selection of mesh for analysis

The model was imported as the native CATIA Part (CATPart) file and generated Dual Domain mesh (one of the three available mesh types: Midplane, Dual Domain and tetrahedral-3D Mesh). An edge length of 0.5 mm was selected to generate at least two outer layers along the thickness for the smallest wall thickness to be analysed, which was 1 mm. Then the mesh statistics were diagnosed and confirmed that the match and reciprocal percentages were more than 99% which was suitable for analysis. Then it was re-meshed to 3D mesh with a minimum of 10 layers of tetra layers as recommended by Moldflow® for plastic parts.

2.2.1.5 Selection of appropriate analysis sequence & boundary conditions

Moldflow® software has various analysis sequences appropriate for evaluating different results. While the packing phase is more controlled with the set packing pressure, the injection pressure is mainly driven by the factors such as injection time and flow rate to fill the part. Thickness and flow length have a more significant impact on injection pressure. The Fill + Pack sequence was selected to study the variation of injection pressure for IM & ICM processes. The Z direction was selected for the research as the compression direction. Only the part’s surface was considered for compression surface, while the feed system did not go through the compression sequence.

2.2.1.6 Selection of process settings

Optimum process settings to mould the part within the maximum shear stress and a shear rate of the material were found by multiple analyses on one of the specimens. Below are the process settings selected for the research.

Injection Moulding

Mould Temperature: 95°C

Melt Temperature: 300°C

Flow Rate: 5.5 cm³/s

V/P Switchover: 98%

Viscosity model: Standard viscosity model (from material data)

PVT model: Standard PVT model (from material data)

Solidification model: Standard Transient Temperature model (from material data)

Additional settings for Injection Compression Moulding

Press open distance: 0.5 mm.

Press Compression: starts at the end of packing.

Press force switchover: 99%.

Press compression time, Press speed cap and press compression force cap were set to the machine’s maximum limit.

2.2.1.7 Moldflow® simulation IM & ICM (data collection)

Moldflow® analysis on the IM process was first checked the predicted injection pressure. The machine injection pressure has 240 MPa limit, so considering the Moldflow® recommendation and the industry best practices, any specimen with a pressure less than 120 MPa should be considered for this research (Figure 3).

Higher injection pressure (hp) above 120 MPa was predicted for a few specimens (Figure 4) with high aspect ratios (thin and long), and short shot (ss) was predicted for extreme aspect ratios. The predicted injection pressure for all specimens was recorded (Figure 5).

The specimens with the injection pressure less than 120 MPa for the IM process were further considered only for the ICM analysis to draw a meaningful comparison. This means the specimens with high pressure or short shot predicted were neglected for further analysis. The exact process settings adopted for IM analysis used for ICM analysis, and the corresponding results are recorded (Figure 6).

3.1 Benefit and limitation of ICM over IM

IM technology, widely used to manufacture plastic products, has limitations, resulting in borrowing other technologies from various branches of science to develop high-quality products [9]. Being a versatile process with many variables involved, it is essential to study the effect of process parameters on ICM. Most of the literature aimed at deriving the benefits of ICM over IM; As Shoemaker [10] explained, the ICM process is an extension of conventional IM, which can produce dimensionally stable, relatively low residual stress parts at a lower clamping force. Further, Yilmaz et al. [11] found that the ICM technique minimised the required injection pressure compared with IM. At the same time, ICM is considered a complex process; the selection of ICM over IM is based on accuracy, the geometry of the part and cost [1, 4].

Aesthetic automotive components are critical for their transmission performances, which moulding can enhance at minimal residual stress. This can be achieved by moulding the part with very low injection and packing pressure in the ICM process, as Michaeli et al. [12] concluded in their comparison study of IM and ICM on the optical performance of moulded parts. Chen et al. [3], Young [13] and Zhu et al. [14] also highlighted in their research the higher optical performance, especially lesser birefringence when using the ICM process.

Though ICM & IECM are used in industries and studied by a few research groups, obtaining the process in detail is challenging due to the confidentiality or limitation of studies [3]. The same is underpinned by Malkin and Isayev [15] that there are few theoretical and practical studies in the literature on ICM & IECM. Further, Loaldi et al. [1] highlighted the significance of the compression gap in their study; a lower compression gap is preferred for optimal conditions. Thus, this study indirectly points out the benefit of IECM, as it is very similar to an ICM process with a minimal compression gap. This is also highlighted earlier in a patent filed by Clarke [16] on how the IECM process differs from ICM and the benefits of adopting the process in terms of lower residual stress.

Nevertheless, a comparison study on both processes is essential to determine the risk of producing poor-quality parts; Michaeli and Wielpuetz [17] highlighted that the processing window is very narrow for ICM compared to IM. Similarly, a study at IKV highlighted the importance of calculating the clamping force at the design phase [18]. Even though none of these studies provided any quantitative comparison of ICM or conventional IM, thus resulting in the inaccurate selection of moulding machine and appropriate shrinkage.

3.2 Methodologies adopted in current research

Most of the works of literature followed an experimental method to test the cause-and-effect relationships, and the outcomes are primarily quantitative. However, as indicated by Dang [19], most approaches are purely academic and thus difficult to apply in practice due to the lack of detailed study about the scope of application. As the injection moulding process involves many variables, most researchers limited their study to very few variables with relatively few experiments to get satisfactory results [20]. At the same time, few pieces of literature focused on the responses and conclusions of various studies by different researchers, giving a complete snapshot of the [21]. Simulation-based research is quite common in the field of injection moulding, which will reduce the environmental impact by reducing the wastage of physical moulding and achieving the results though a faster and cost-effective way [3, 22]. This is underpinned by Spina [23] who adopted a Moldflow® simulation-based comparison study of hot runner systems in injection moulds to mould aesthetic automotive components. On the other hand, few researchers adopted a case study-based approach, which directly correlates with application, and the results are primarily qualitative [13].

3.3 Comparison of process parameters

Flow Length (FL) to Wall Thickness (WT) ratio is important for all injection moulded parts, the primary factor contributing to injection pressure. Yao and Kim [6] investigated the effect of the rapid increase in frozen layer thickness due to cooling happening in the thin-walled parts, which increases flow resistance. Various simulation results are discussed, which indicate the rapid increase in injection pressure for thin-walled parts, but this study has limitations as they considered only two levels of mould temperature; 25°C which is the room temperature and 265°C which is the melt temperature of the selected Polycarbonate material [6]. These two temperatures are impractical for conventional moulding as the recommended mould temperature is between 70 and 100°C. However, it is evident from their research that injection pressure for the part with 1 mm thickness and 50 mm flow length is approximately 125 MPa [6]. Industry-wide practice is to limit the injection pressure to half of the maximum limit of the moulding machine; Moldflow® software recommends limiting this pressure to 100 MPa as a design guide line [8].

The criticality of the flow-length-to-wall-thickness ratio is widely discussed in many other research and by manufacturers; Hinterdorfer [5] explained about the effectiveness of ‘ENGEL coinmelt process’ which is essentially an ICM process to consistently achieve high part quality for a ratio of 1:400. Lower injection pressure, lower clamping forces and high repeatability with highly viscous material are the advantages of this process [5]. Spina [23] highlighted the effectiveness of a multi-gate feed system in filling the aesthetic automotive components to reduce the injection pressure and improve the part quality. The sprue, runner and gate design have a greater impact on the injection pressure and the quality of the parts. Their cross-section and length vary based on the applications and the design restriction, which could increase the injection pressure at the nozzle. So, to understand the actual injection pressure on the moulded parts, it is essential to consider the cavity pressure instead of the injection pressure or the pressure developed in other areas [24]. Tsai and Lan [25] explained the variation of injection pressure along a sensor point in the cavity throughout the moulding phase.

Sortino et al. [26] compared IM & ICM processes for producing optical parts. They reviewed multiple research projects that compare these processes and found out through their experimental studies that ICM is the most precise and repeatable process among the compared. Both IM and ICM have pros and cons from an application perspective; the data collected through the literature review is summarised in Table 2.

The growing trend of automotive product design with thin-walled components for weight reduction is being researched thoroughly. Thinner components have a greater environmental impact by reducing fuel consumption and better performance by being lighter [27, 28].

3.4 Simulations: Initial consideration

High residual stress within the initial moulded parts relieve can passing through subsequent processes and causing dimensional errors in later stages with excessive warpage and may fail in the field. Parts with wall thicknesses less than 1.5 mm will exhibit excessive residual stress. On the other hand, if the wall thickness is more than 3.5 mm, this will affect the effective transmission of signals. Also, increasing the wall thickness will make the part stiffer and heavier, thus restricting the assembly process’s flexibility. Thus, most aesthetic automotive components have wall thicknesses ranging from 1.5 to 3.5 mm. So, the research excludes the wall thicknesses below 1 mm and above 4 mm, considering 0.5 mm above and below the design range.

Similarly, this study is exclusively for the flow lengths up to 300 mm as most of the products are within this length or multiple gates are placed in series for parts more than 300 mm in length. From the industry best practices, other research, and recommendations by Moldflow® software, it is anticipated that the pressure results on high aspect ratio specimens (thin and long) may give unrealistic injection pressure results. Hence, these values need to be neglected to derive meaningful conclusions.

In practice, the gates can vary in shape and size, like edge gate, cashew gate, sub gate, fan gate etc. Similarly, the runner length can range from a few millimetres and can be as long as the flow length for the part. These factors can affect the injection pressure on the part. So, this research excludes any deviation in injection pressure due to the feed system design constraints and considers only the pressure inside the part cavity, which is the cavity pressure. However, the outcome of the results needs to be carefully interpreted as it is based on the cavity pressure and excludes the higher injection pressure generated by the relatively smaller gates in some products. This is critical, especially when moulding the parts with fan gates or sub gates, as their effective opening diameter is less than the actual part wall thickness for easy de-gating. This project explicitly studies the injection pressure differences in ICM & IM process with PC Lexan LS1 material as this is the most compatible special material identified to develop aesthetic automotive components.

Injection moulding is a versatile process that involves numerous process variables. Significant process variables affecting this research are melt temperature, mould temperature, injection time, flow rate, injection speed, packing time, and packing pressure. ICM is complicated as it involves the compression phase where the compression gap, compression speed, etc., further complicates the process. For this research, a single practical value is considered for all these variables and all the specimens are analysed using the same process settings. Thus, this research has limitations in predicting the injection pressure for other possible hundreds of combinations of process variables.

There is expected to be some difference between simulation results and the actual moulding results, but this research does not bridge any gap between the accuracy of the simulation results with respect to the physical moulded parts. This research is exclusively to identify, through simulation, the relationship in injection pressure between IM & ICM for various flow lengths to the wall thickness and validate the same difference in physical moulding. This project focuses only on the study of injection pressure on ICM and IM processes of aesthetic automotive components and does not investigate the complete manufacturing process of these components, including annealing, coating, further over-moulding and compliance tests.

Injection Pressure for both IM & ICM processes are recorded and compared using different applications. Firstly, the percentage of injection pressure between both processes is calculated using Microsoft Excel and checked for consistency. Correlation and Regression analysis on injection pressures recorded is done using IBM SPSS Statistics software. From this, a regression equation is expected to be derived to calculate the ICM process’s injection pressure from a known IM process. Injection moulding trials for both IM & ICM processes on a prototype tool are carried out, and the resulting pressure obtained is analysed and compared with simulation results.

3.5 Result comparison

Injection pressure predicted from Moldflow® analysis on both processes is compared. It is found that the pressure remains almost the same for the samples when the thickness is increased by 0.5 mm, and the flow length is increased by 50 mm for specimens of lesser flow length and thickness (Figure 7).

The current assumption is that the injection pressure of ICM is 20% less than the IM process. The injection pressure predicted for each specimen for both processes is compared in Table 3.

Table 4 shows the injection pressure of the ICM process as a percentage factor with respect to the comparison of pressure difference in specimens between the IM and ICM process, revealing that the 20% injection pressure reduction of the ICM process is invalid. This varies along the flow lengths and across the wall thickness of the moulded parts. Within the acceptable range of injection pressure (120 MPa) compared among the specimens, 3 x 300 mm has the least reduction of 30.33% and 4 x 50 mm has the most reduction, 51.09%. Another trend highlighted from the comparison study is that the variation in difference increases with an increase in wall thickness and a decrease in flow length, and this indicates that the reduction of injection pressure is not consistent for all the specimens. Thus, the current assumption is not valid, which is to consider a single value (20%) as the injection pressure reduction factor (Figure 8).

3.6 Correlation and regression analysis

The project’s main aim is to find the linear relationship between the injection pressure of IM & ICM. Correlation analysis measures the linear relationship between two variables. It has a value between −1 and + 1, where the result closer to these extremes indicates a strong correlation. A weak or no correlation exists if the value is closer to 0. IBM SPSS Statistics software is used to analyse the correlation between the injection pressures of IM & ICM. A strong positive correlation value of 0.999 is identified between IM & ICM.

Regression analysis is performed on the injection pressure data collected from both processes. Injection pressure recorded for the IM process for the samples is considered Predictors (Constant), and ICM is considered a dependent variable. The P-value identified is less than 0.001, which means the data is statistically significant.

A scattered graph (Figure 9) shows the relationship between the injection pressures on IM & ICM.

The regression equation derived for the positive correlation is Y = a + bx = −0.88 + 0.66x

when simplified considering the acceptable deviation of ±5 MPa,

This equation is cross verified by calculating the injection pressure of ICM using the known pressure of IM for the simulated samples and compared with the simulated pressure of ICM for the respective samples as given in Table 5. The maximum difference observed between simulated results and pressure calculated using the regression equation is less than 1.12%, and this is generally acceptable for predicting injection pressure for ICM, which is expected to be within ±5%.

3.7 Result validation

To ensure an error-free prediction of the injection pressure of ICM from a known injection pressure of IM, it is essential to validate the relationship derived through the regression analysis. Current Moldflow® analyses use simple specimens of uniform wall thickness and simplified gate and runner systems. It is essential to test this in a live project to ensure the underlying objectives of the research are achieved. A prototype tool is used for validating the results.

The part to be moulded is of 2.5 mm wall thickness and a total flow length of 295 mm (Figure 10). While performing the simulation studies, the injection pressure generated for the IM process for the 2.5 x 300 mm part was 125 MPa. This specimen has not been considered for statistical analysis as the injection pressure exceeds the recommended acceptable range of 120 MPa. Process settings used for the simulation are adopted for the moulding trials. The prototype tool is manufactured so that IM, ICM & IECM processes can be trialled using the movable core feature, which is mechanically equipped within the mould.

To maintain the confidentiality with the customer project, excluding the customer emblem and key identification features, a model similar to the one used for manufacturing the prototype mould is first analysed for the conventional IM process using Autodesk Moldflow® software, applying the same process settings as the research specimens. Unlike the specimens analysed, this part does not have a uniform wall thickness all over the sections. So, a difference in injection pressure results compared to the specimens is expected. The pressure result obtained after Moldflow® analysis is recorded, and the maximum injection pressure observed is 137 MPa (Figure 11).

The prototype tool is prepared for the moulding trials of both IM and ICM processes. PC LEXAN™ Resin LS1 material is preheated per the manufacturer’s recommendations, and the process settings are entered in the machine interface the same as the simulation. The conventional Injection Moulding process is first tried. Once the mould is well heated and after initial trial shots to stabilise the process, a few continuous shots are taken, and the resultant injection pressure is noted.

Observed average pressure = 1408 bar = 140.8 MPa.

The simulation result for the same prototype lens is 137 MPa, a difference of 10 MPa is observed between the simulation results and the actual. Though the trial results are expected to be closer to the simulation results, some variation is expected due to different factors, such as the mould and machine conditions.

The regression Eq. 1 is applied to the injection pressure results obtained through simulation and moulding trials.

Using simulation results, ICM _(pressure) = 0.7 x 137–0.9 = 95 MPa

Using moulding trial results, ICM _(pressure) = 0.7 x 141–0.9 = 98 MPa

The machine setting is now changed to ICM, and the compression gap is set to 0.5 mm. Mould & melt temperatures, flow rate etc., are the same value as the IM process. After a couple of trial shots with ICM to stabilise the moulding process, a few continuous shots are taken, and the pressure result is noted.

Observed average pressure = 969 bar = 96.9 MPa

The injection pressure observed from the actual moulding is 97 MPa which is very close to the calculated injection pressure for the ICM process using the derived regression equation.