1 Introduction

The aesthetic impression of a car interior is dependent on the combination of many different plastic components. The appearance of each component is mainly controlled by the colour, gloss, and texture. Surface textures are deliberately imposed to achieve an exclusive appearance and improve the scratch resistance. The surface texture is a replication of the texture on the walls in the mould cavity, and it is therefore important to achieve a good replication to obtain uniformity in texture, colour, and gloss [1].

The injection moulding replication is controlled by three main factors: the driving force, the material deformability, and the surface topography [2]. The driving force is controlled by the cavity pressure that arises due to the cavity filling and later the holding pressure, while the material deformability is controlled by material properties such as viscosity and elasticity. Material deformability is also affected by the thickness of the frozen layer which is formed during mould filling when the plastic melt is in contact with the surfaces of sprue, runner, gate, and mould surface [2]. Finally, regarding surface topography, this parameter generally affects the replication in such a way that smaller structures with higher aspect ratios are more difficult to replicate. These smaller structures are known to affect the surface gloss [3, 4], and controlling these structures is critical for manufacturing injection-moulded interior components.

Earlier investigations regarding surface topography in injection moulding have focused on the replication of specific micro-structured surfaces with lines, pins, or pyramid textures [5,6,7,8,9], and few investigations have also focused on replication of grain textures generally used in the automotive industry [1, 10, 11]. Grain texture refers to the deliberate addition of a specific pattern on the surface of the mould and, consequently, the moulded part, in this case to emulate a leather finish. Research results from [5, 10] suggest that replication of the mould cavity texture can be better achieved by using a low melt viscosity. Low melt viscosity can for instance be obtained by low molecular weight, high melt temperature, and/or high injection speed. A high tool temperature has also been noticed to be beneficial for good replication of the mould cavity texture [2, 6], since a high tool temperature facilitates filling of the texture and prolongs the time for build-up of a frozen layer. In an investigation on replication quality of micro-structures carried out by Meister et al. [6], it was noticed that a high tool temperature, higher than the glass transition temperature for amorphous plastic materials and higher than the crystallisation temperature for semi-crystalline plastic materials, gives a good mould cavity texture replication. Furthermore, the cooling conditions can also have a large effect on the replication of the mould cavity texture [2, 6, 12, 13], especially for semi-crystalline plastic materials, such as polypropylene (PP), since the cooling conditions affect the degree of crystallinity and thereby the volumetric shrinkage.

The mould temperature had the largest effect in the investigations on the influence of process variables on replicating high aspect ratio micro-structured surfaces of polystyrene where the average height of micro cavities was considered to evaluate the replication capability [14]. Other controllable process variables, injection speed and holding pressure, corresponded with significant increases of 50% and 16% respectively in average cavity height. Notably, the micro-structure quality along the cavity is uneven, and in micro injection moulding, large surfaces with high aspect ratio are better replicated at low injection speed and high holding pressure. Similar studies conducted by Masato et al. [15] investigated the influence of part thickness and process variables on the replication of micro structured surfaces consisting of micro pillars. The height of these pillars increased more than 300% with increasing mould temperature and injection speed affected replication of features far from the injection gate implying variation and influence of the distance from the injection location. In general, higher surface roughness values imply a better replication of the tool surface.

The objective of the study was to develop a methodology to efficiently investigate how injection moulding process settings affect surface topography of manufactured components. In this study, the surface topography of textured injection-moulded samples that corresponds to automotive interiors was captured using an optical interferometer. Areal surface parameters, indexed in ISO 25178-2 [16], were used to characterise the study samples’ surface topography. These areal surface parameters were categorised based on their type, region, and distribution of the topographical feature they define. However, it is tedious, time consuming, and not imperative to use all the surface parameters to explain the variation in surface topography [17]. Hence, it is important to identify the significant surface parameters that discriminate sample surfaces and that explain the variation in surface topography. The novelty of this research lies in identifying the comparative scales of observation for injection-moulded textured surfaces and to shortlist a set of significant surface parameters that improves specification for comparison and effect analysis. Therefore, the surface topographical wavelengths of maximum difference were identified using scale-sensitive fractal analysis [18], and the statistically significant surface parameters were identified using multiple regression analysis with process variables as the independent variables and areal surface parameters as dependent variable. Using the significant surface parameters, the influence of process variables, tool temperature, injection speed, and melt temperature, on the surface topography of injection-moulded ABS and PP samples was investigated. The investigations on mould tool surface topography and surface gloss were not included in this study.

2 Experimental

2.1 Materials

The polypropylene (PP) used was a talc filled (15% by weight) high impact strength grade from Styron, denoted Inspire 1500 ESU. It is characterised by a density of 1.02 g/cm3, and a melt flow rate MFR (230 °C, 2.16 kg) of 9.5 g/10 min. The acrylonitrile–butadiene–styrene (ABS) used was a high impact strength grade from Elix, denoted Novodur P2M-V. It is characterised by a density of 1.04 g/cm3 and a melt volume-flow rate MVR (220 °C, 10 kg) of 18 cm3/10 min. The colour for both plastic materials is charcoal solid.

2.2 Mould tool

The mould tool was a rectangular cavity with a thickness of 2.7 mm, a width of 138 mm, and a length of 78 mm. The mould tool contained three different grain textures produced by EDM, and in the present study, the coarse texture away from the gate, shown in Fig. 1, is analysed. The mould had a film-edge gate 123 mm wide, 2 mm long, and 1 mm thick.

Fig. 1
figure 1

Illustration of surface measurement spots on the sample

Full size image

2.3 Processing

The manufacturing of test samples was carried out at IAC Group in Hjärnarp, Sweden. The ABS and PP materials were dried at 80 °C for 4 h prior to manufacturing. The injection moulding machine used is a DEMAG NC3 with a clamping force of 1000 kN and a general-purpose screw with a diameter of 30 mm. For the PP samples, the injection times were 4.5 s and 0.9 s, and for the ABS samples, the injection times were 5 s and 1 s.

2.4 Design of experiment

A full factorial design of experiments (DoE) was carried out for both materials (see Table 1). Each of the three processing parameters of interest, melt temperature, tool temperature, and injection speed, was varied at two different levels.

Table 1 Overview of the factorial DoE, with 23 experiments
Full size table

2.5 Characterisation

The surface topography was measured using Zygo’s Nexview 9000 coherence scanning interferometer. The vertical resolution or the repeatability of the RMS for the measuring instrument is 0.008 nm, and the surface topography repeatability is 0.08 nm. The spatial sampling is 0.95 µm with 916 points in x- and y-axes. The measurements consist of coarse grain surface topography relocated on the injection-moulded ABS and PP samples. Five surface measurements on each sample were relocated, on top of the coarse grain as shown in Fig. 1, in proximity to avoid the variation caused by the distance from the gate, as observed by Masato et al. [14, 15].

3 Methodology

3.1 Surface relocation

Same measurement spots are relocated for all samples, and this surface topography relocation was achieved in two steps: coarse relocation using the coordinate relocation on the instrument’s measurement table and fine relocation using co-localization analysis in MountainsLab with surface topography of ABS1 sample as reference [19]. The string of operators applied to these surface measurements using Digital Surf’s MountainsLab software included removing the measurement noise and filling non-measured points. Using the complexity plots from scale-sensitive fractal analysis, the wavelengths of maximum difference between the samples were identified [18, 20]. A robust Gaussian filter of order 0 was applied to extract these wavelengths of interest by applying a bandpass filter with nesting indexes of 5 and 50 µm. The relocated and filtered surfaces were characterised by areal surface parameters indexed in ISO 25178–2.

3.2 Surface parameters

Areal surface parameters were categorised into height, functional, spatial, hybrid, volume, and feature parameters.

Height parameters characterise the surface features normal to the surface and provide information on the surface amplitudes. Height parameters include root mean square height (Sq), arithmetic mean height (Sa), skewness (Ssk), Kurtosis (Sku), maximum peak height (Sp), maximum valley height (Sv), and maximum height of surface (Sz).

Spatial parameters include autocorrelation length (Sal), texture aspect ratio (Str), and texture direction (Std). Spatial parameters provide information on the spatial distribution of topographical features.

Hybrid parameters include root mean square gradient (Sdq) and developed interfacial area ratio (Sdr). These provide information on complexity of the surface topographical features.

Feature parameters are used to characterise specific features on the scale-limited surfaces using pattern recognition toolbox, and three main types of features include areal (hills and dales), line (coarse and ridge), and point (peak, pit, and saddle). Feature parameters include density of peaks (Spd), arithmetic mean peak curvature (Spc), mean dale area (Sha), mean hill area (Sha), mean dale volume (Sdv), and mean hill volume (Shv).

Material ratio-related parameters include material ratio parameters, volume parameters, and areal parameters for stratified surfaces.

3.3 Analysis

To identify the robust set of parameters that discriminates the measured surface topography, multiple regression analysis was applied with material and process variables as independent variable and areal surface parameters as dependent variables. The methodology adapted in this paper is an extension with shortlisting by correlation coefficient to the methods used in [21] to identify the significant surface parameters. The regression statistics and their implication on the relationship between the process variable and surface measurements are explained below [21, 22]:

  1. 1.

    Coefficient of determination, R2: R2 provides the variability of the surface parameter with respect to the process variables. The higher the R2, the higher will be the variability in the measurements explained. In this study, the surface parameters with R2 higher than 85% are considered for discriminating the samples and explaining the variation in the surface topography.

  2. 2.

    Significance F: probability of the measurements with respect to the process variables is not random, if the p value (significance F) associated with the F-test is less than α (0.05).

  3. 3.

    p value: independent variables (regressors), regression coefficients are reliable, and comparisons between the samples are valid, if the p value is lower than 0.05 for a confidence level of 95%.

  4. 4.

    Shortlisting by correlation coefficient, R: in the last step, the surface parameters within the same category that have higher correlation coefficient, R, with other parameters are shortlisted. Parameters with higher correlation coefficients enable to filter the interchangeable parameters for analysis and interpretations [23]. Surface parameters that have higher R2 but lower correlation coefficient, R, with other parameters in the same category are also included. These parameters emphasize the independence between parameters explaining a unique property of the surface and is important to include these lower correlated parameters for analysis (Fig. 2).

Fig. 2
figure 2

Surface analysis workflow

Full size image

4 Results and discussion

Optical microscopy photos showing the coarse grain surface texture of the test samples are presented in Fig. 3. The differences in surface topographies are negligible in macro scale, though the scattering of light is evidently different, indicating the differences in micro-scale features.

Fig. 3
figure 3

Optical microscopy photos showing the surface texture for PP1, PP8, ABS1, and ABS8

Full size image

4.1 Scale-sensitive fractal analysis

The scale-sensitive fractal analysis is used to identify the scale limited datasets at which the differences between the samples’ surface topography are maximum [24]. Complexity plots, or rate of change of the surface relative area, is plotted as a function of scale as shown in Fig. 4. From the plots, it is observed that the complexity of surface topography produced at higher tool temperature, injection speed, and melt temperature, PP8 and ABS8, is high compared to PP1 and ABS1, respectively. From the complexity plots, the scales of maximum difference are identified between 10 and 1000 µm2, and the nesting indices corresponding to the identified scale range are calculated as the length of the side of an equilateral triangle which computes to approximately 4.47 µm and 44.72 µm. Therefore, the corresponding surface topography is extracted applying a robust Gaussian filter with nesting indices of 5 µm and 50 µm. The purpose is to enhance scale ranges with texture features of interest and attenuate other scale ranges, similarly as [24, 25].

Fig. 4
figure 4

Complexity plotted as a function of scale to identify the most important scales for differentiating between the surfaces

Full size image

4.2 Surface topography measurements

From the surface profiles in Fig. 5, the difference in surface topography is visible between the samples of polypropylene (PP) and acrylonitrile–butadiene–styrene (ABS) produced at different process variables. Relocated profiles are presented to make the comparison easier. The amplitudes are slightly higher for PP compared to ABS, which is expected since the PP material had lower melt viscosity than the ABS material and better melt flow at the selected processing conditions to replicate the high-frequency features from the mould cavity. Also observed from the surface profiles in Fig. 6 is that the variation in surface topography of ABS varies differently for the different process variables, tool temperature, and injection speed, compared to PP. The quantitative analysis using areal surface parameters was performed to investigate the influence of process variables in more detail.

Fig. 5
figure 5

Relocated surface profile measurements after bandpass filtering with nesting indices 5 and 50 µm of ABS1, ABS8, PP1, and PP8 samples

Full size image
Fig. 6
figure 6

Relocated surface profile measurements after bandpass filtering with nesting indexes 5 and 50 µm of ABS and PP samples

Full size image

4.3 Significance of surface topography parameters

The areal surface parameters represent the surface topographical features and their statistical distributions. In general, the parameters of greatest relevance often depend on the surface topographical information needed to study the process or function. However, in this study, all 3D-parameters were evaluated to determine which set of parameters best describes the variation in surface topography between the injection-moulded samples. In this study, the areal surface parameters with coefficient of determination, R2, higher than 85% and for α = 0.05, significance value F less than 0.05 are selected as significant parameters, shown in Table 2. Interaction effects of process variables are cautiously added in the model considering the adjusted R2.

Table 2 Coefficient of determination, R2, adjusted R2, and significance F for areal surface parameters (green values indicate surface parameters with R2 > 85%) and significance F < 0.05)
Full size table

From the methodology, significant surface parameters, highlighted in Table 2, include root mean square (Sq), arithmetic mean height (Sa), root mean square gradient (Sdq), developed interfacial area ratio (Sdr), material volume (Vm), peak material volume (Vmp), core material volume (Vmc), core roughness depth (Sk), and reduced summit height (Spk). Surface parameters, Sq and Sa, belong to the same category of amplitude parameters and have high correlation; hence, the surface parameter, Sa, with higher R2 is considered to evaluate the influence of process variables. Similarly, the volume parameter, Vmc, with higher R2 is considered for the analysis and interpretation among the volume parameters.

4.4 Influence of process variables on significant surface parameters

By studying p values, the significance of effect on surface topography by melt temperature, tool temperature, and injection time is determined. If the p value is lower than 0.05 (which relates to a confidence level higher than 95%), the effect is considered to be significant. The p values for the different processing parameters used in the present investigations are presented in Table 3.

Table 3 p values for statistically significant surface parameters with respect to injection moulding process variables (green p values indicate a significant effect of the process variable on the surface parameter)
Full size table

Based on the results in Table 3, there is statistically significant difference in surface topography between the PP samples and the ABS samples. There is also a statistically significant influence of tool temperature and injection speed on all the significant surface parameters. The effect of melt temperature is statistically less significant on surface parameters Sa, Sdq, Vmc, and insignificant on the surface parameter, Sk.

From Fig. 7, it is observed that surface parameters, Sa, Sdq, Vmc, and Sk, are higher for PP samples compared to ABS except sample ABS3. ABS3, manufactured at higher tool temperature, lower injection speed, and lower melt temperature, has higher values of surface parameters compared to PP3. Tool temperature and injection speed have significant effect on the surface topography represented by surface parameters, Sa, Sdq, Vmc, and Sk. Comparing samples PP2 and PP3, the differences in the surface parameters are negligible indicating that tool temperature and injection speed have similar effect on the surface parameters. Similarly, this is also observed in samples PP6 and PP8. Figure 8 shows the Abbott curves calculated on the surface with two bearing ratio thresholds defined using vertical lines at 10% and 80%. It is observed that the differences in material volume is higher in peak (Vmp) and core regions (Vmc). The surfaces of ABS1 and ABS5 have lower material volume in peak (Vmp) and core regions (Vmc) compared to ABS4 and ABS8 suggesting a positive influence of tool temperature and injection speed. A similar trend is observed for PP surfaces.

Fig. 7
figure 7

Mean values and standard deviations of surface parameter readings for the ABS and PP samples manufactured at different injection speed, tool temperature, and melting temperature

Full size image
Fig. 8
figure 8

Abbott curves for ABS and PP surfaces

Full size image

4.4.1 Functional influence of the significant roughness parameters

Surface parameters represent the surface topographical features and its distributions. These features affect the surface functional behaviour, and the functional importance of different surface parameters is summarised in [26,27,28,29]. Here, the selected significant surface parameters and their functional interpretations are briefly discussed.

  • Arithmetic mean height, Sa, represents the average of amplitudes and is often associated with surface roughness and asperities.

  • Root mean square gradient, Sdq, represents the slope measurement of features which comprise the surface and is associated with applications related to surface wetting, surface aesthetic and sealing properties [26, 27].

The higher the values of Sa and Sdq, the higher the amplitudes and complexity in the micro-structures responsible for diffuse scattering which might result in lower surface gloss.

  • Core material volume, Vmc, represents the volume of material between the heights p (10%) to q (80%) in the material ratio curve.

This parameter represents the material available for load support after the top layer (10%) is worn out.

For material with similar mechanical properties, the higher the Vmc, the higher is the load withstanding capacity in the material ratio p (10%) to q (80%) [27].

  • Core roughness depth, Sk, represents the height of the core surface with dominant peaks and valleys removed.

Similar to Vmc, this parameter is associated with surface wear resistance.

4.5 Main effects and interaction effects

The main effects and interaction effect plots of injection speed and melt temperature are shown in Figs. 9 and 10.

Fig. 9
figure 9

Main effects of material, injection speed, tool temperature, and melt temperature of surface parameters. a Arithmetic mean height, Sa. b Root mean square gradient, Sdq. c Core material volume, Vmc. d Core roughness depth, Sk

Full size image
Fig. 10
figure 10

Interaction effects of melt temperature (MT) with a material (M) and b injection speed (IS) of surface parameters Sa, Sdq, Vmc, and Sk

Full size image

4.5.1 Effects on roughness by material choice

From Fig. 9a, the differences in surface topography between ABS and PP are observed to be around 18% in Sa, 36% in Sdq, 10% in Vmc, and 35% in Sk. This is primarily due to the higher melt sensitivity and shear thinning of PP compared to ABS at higher injection speed and melt temperature [7, 8, 30]. Shear thinning is characterised by the decrease in viscosity with increase in shear rates. PP is a semi-crystalline material which exhibit higher reduction in viscosity for a range of reducing shear rates compared to ABS, which is amorphous material.

4.5.2 Effects on roughness by injection speed and tool temperatures

From Fig. 9b, c, the main effects of increasing tool temperature and injection speed suggest similar incremental effect on the significant surface parameters by approximately 40% in Sa and Sdq, 35% in Vmc, and 50% in Sk. As mentioned in [15, 30], tool temperature is critical for replicating the micro surface topography which helps to maintain the thermal gradient between mould cavity surface and the polymer melt and eliminate the formation of skin layer. Injection speed is directly proportional to shear rate, and the viscosity of the polymer melt decreases with increase in injection speed [8, 30]. With higher injection speed, the duration of polymer melt and mould cavity reduces, this prevents the skin layer formation. Hence, an increase in tool temperature and injection speed resulted in higher values of Sa, Sdq, Vmc, and Sk.

4.5.3 Effects on roughness by melt temperature

As observed in Fig. 9d, effect of melt temperature on Sa, Sdq, Vmc, and Sk is observed to have less significant main effect. In general, melt temperature has a positive effect on the replication of the surface texture. By increasing the melt temperature, the viscosity of the material decreases, thereby favouring in filling micro surface topography. But, as explained in [30], increase in tool temperatures at different melt temperatures has produced varied replication results, with positive effect at low melt temperatures.

4.5.4 Interaction effects on roughness

As observed from Fig. 10a, b, there are significant interaction effects with material PP and injection speed. The increase in melt temperature by 30% for PP increases the surface parameters Sa, Sdq, and Vmc by 20%, and Sk by 16%. A distinct dependence of the melt temperature for semi-crystalline material, PP, compared to amorphous material, ABS, is observed due to the differences in melt sensitivity and shear thinning behaviour discussed in [6]. PP exhibits good flow behaviour with larger shear thinning at higher melt temperatures, and hence, the influence of melt temperature on PP is significant compared to ABS.

Contrarily, for ABS, an increase in melt temperature is observed to have no significant effect on the surface parameters. At high injection speed, an increase in melt temperature increases the surface parameters Sa, Sdq, and Vmc by 17% and Sk by 10%. The interaction effects of melt temperature and tool temperature, injection speed, and tool temperature are found to be insignificant and not added to the model.

5 Conclusions

The results and analysis are specific to the injection-moulded ABS and PP samples manufactured at the process settings stated in Table 1 but the implemented methodology can be used in general to discriminate and analyse the scale-limited surface topography.

The methodology developed helps to:

  • Identify relevant scales for discriminating between surface topographies and definition of appropriate filters.

  • Identify the significant surface parameters that discriminate the samples surfaces.

  • Understand the effect of process variables on surface topography.

In general, process conditions have strong influence on the surface topography replication in injection moulding. In the present study, it was concluded that:

  • The tool temperature and injection speed have a large effect on the surface topography of injection-moulded samples.

  • The identified significant surface parameters, Sa, Sdq, Vmc, and Sk, filtered for the relevant scales, discriminate between the sample surfaces and can be used for process optimization.

  • ABS samples’ surface topography is less affected by the process variables compared to PP samples due to the differences in melt sensitivity.

  • A high tool temperature and high injection speed resulted in increased average amplitudes (Sq), complexity in features (Sdq), core material volume (Vmc), and core roughness depth (Sk).

  • Generally, increasing melt temperature limits the increase of the ‘frozen layer’ and replicates the structures of the tool effectively; however, a rather small effect on the surface parameters is observed.

Future work includes investigations on the resulting gloss of textured injection-moulded samples and degree of replication by evaluating the surface topography of the mould tool.