on only one half of the surface, had the lowest MSE and absolute area, which means that both methodologies agree that this specimen is the best in regards of homogeneity. N26624-1 had the most and strongest tiger stripes among the plaques, and thus both outcomes agree that this is the worst surface. The rest of the ranking correlates similarly, with the only exception being between N26298 and N26624-3. This could have been caused by the difference in the quality of the pictures made for the absolute area method. Nevertheless, Borealis verified the validity of the absolute area results, and thus it can be concluded that the absolute area method is sufficient to characterize the surfaces of the parts originally produced for this master thesis.
4.2 Injection volume flow rate
It can be seen, that for the polished surface ("S1") the grayscale values are lower in general. This can be attributed to the roughness of this surface: it can be assumed, that the smooth surface does not cause the light to reflect differently. On the other three, the small bumps and valleys, caused by the bubbles or the texture of the mold, cause the exterior to have a slight silverish color. This is generally true for every part produced.
Looking closely, the only significant difference is on S4, when the melt temperature is 220 °C, the SCF content is 0 %, and conventional heating is employed, increasing the injection volume flow rate makes the surface quality worse, as it can be seen at figure 4.5. It could be assumed, that this is caused by the high shear forces between the relatively low tempered melt and mold. Increasing the volume flow rate enhances this, and that is why grayscale is higher at 800 cm3/s for these settings.
The other surfaces also show a similar trend, but at a smaller scale. It should be kept in mind, that the difference shown here relative to the whole grayscale scale is only around 2 %, thus, an error in measurements could have made this change.
50 70
40 50 60
Mold Temperature [°C]
Grayscalelevel[-]
400/220/0 800/220/0
Figure 4.5: Difference in grayscale levels when changing the injection volume flow rate from 400 cm3/s to 800 cm3/s at 220 °C and 0 % SCF content at 50 °C and 70 °C mold temperatures on S4
As mentioned before, three parts were produced aside from the original DoE to test how the surface reacts to extreme injection volume flow rates. Here, the other three settings were 260 °C for the melt, 130 °C for the mold temperature, and a 0.4 % SCF content. At first look, it can be seen that using extreme flow rates make the grayscale levels much worse, but also indicates that using very low flow rates are better for the surface, as shown in figure 4.6. However, when looking closely at the pictures, at 200 cm3/s, a new type of defect can be seen in a circular shape, as it can be seen in figure 4.7. Guo et al. assumes that at low volume flow rates, the bubbles will nucleate at the flow front, while at high volume flow rates, the bubbles will stretch on the surface too much. [Guo07]
Thus, it can be assumed that these circular shapes are caused by bubbles that came in contact with the mold surface, and ruptured. However, as only one part was produced for 200 cm3/s and 1900 cm3/s, further research would be needed at this conditions.
S1 S2 S3 S4 30
40 50 60 70 80
Grayscalelevel[-]
Run 50[200/260/130/0.4] Run 49[800/260/130/0.4]
Run 51[1900/260/130/0.4]
Figure 4.6: Difference in grayscale levels when employing 200 cm3/s, 800 cm3/s, and 1900 cm3/s injection volume flow rates
Figure 4.7: The difference between using 200 cm3/s, 800 cm3/s and 1900 cm3/s injection volume flow rates
4.2.2 Homogeneity
In general, all of the homogeneity results have large deviances. This could indicate four things.
First, the method of evaluation might carry a large amount of error with it. As the average of five pixels were calculated on the middle line of the plaque, and then these values were averaged for the five parts of one run, the result will intrinsically have a large deviance. Secondly, the so-called
"alignment issue" might had its workings here for homogeneity. As the different parts had to be put in and out of the photobox during the digitization session manually, small alignment issues could have happened, which could have led to the fact that the positions of the pixels were not perfectly
aligned when they were compared to each other. Third, manual cropping had to be done for some pictures, and this could have also led to the same alignment issues. Finally, looking at the parts, it could be generally concluded, that the parts themselves were also not the same, which indicates that the reproductability of the specimen could also be questioned. Nevertheless, the results for homogeneity could still be assessed, as these deviances were not too extreme.
Changing the injection volume flow rate from 400 cm3/s to 800 cm3/s does not show a large difference in the overall scope, but in general, it could be concluded that increasing the volume flow rate helps homogeneity. There are a few exceptions where this was not observed, but these results were random, as visible in figure 4.8.
400 800
0.4 0.6 0.8 1 1.2 1.4 ·104
Injection volume flow rate [cm3/s]
Homogeneity[-]
Figure 4.8: Main influence graph of injection volume flow rate for homogeneity
Looking at the graph where the homogeneity is divided by the surfaces, on S2, S3 and S4, the higher value is favored. This also aligns with the articles reviewed, where an argument was made that high enough volume flow rates will make less defects on the surface, due to less contact time of the melt with the mold, and that it prevents bubble formulation in the flow front. In an example, it can be seen that for Run 19 [400/220/70/0.7], silver streaks dominated the surface, but for Run 20 [800/220/70/0.7], these defects were smoothed out, as it can be seen in figure 4.9 and 4.10.
S1 S2 S3 S4
0.6 0.7 0.8 0.9 1 1.1 ·104
Homogeneity[-]
400 cm3/s 800 cm3/s
Figure 4.9: Main influence graph of injection volume flow rate for homogeneity, divided by surfaces
Figure 4.10: The variance in grayscale levels between Run 19 [400/220/70/0.7] and Run 20 [800/220/70/0.7]
When using extreme volume flow rates, a huge difference can be seen in figure 4.11. This is a prime example that showing the homogeneity is necessary, because for the overall grayscale levels, what could have been concluded is that low volume flow rates are better, but here, it is visible that homogeneity is worse for S2 and S4 compared to 800 cm3/s. The silverish color is still dominant at 200 cm3/s which could mean that bubble formation started at the flow front, and many of them indeed ruptured at the mold wall, but the color is more smoothed out, because the low shear stresses did not smear these bubbles. On S1 at 1900 cm3/s, tiger stripes-like defects can be seen, with the exception that the lines are horizontal, and not curly shaped. As mentioned in the literature review, one cause of this type of defect could be too high shear stresses between the melt and the mold.
[Heu97] Interestingly, S4 showed a different behavior compared to the other surfaces. At the top area, no surface defects were observed. This might have been due to the structure of that mold texture, or that the mold heating conditions were better around this area.
S1 S2 S3 S4
0.5 1 1.5 2
·104
Homogeneity[-]
Run 50[200/260/130/0.4] Run 49[800/260/130/0.4]
Run 51[1900/260/130/0.4]
Figure 4.11: Difference in homogeneity when employing 200 cm3/s, 800 cm3/s, and 1900 cm3/s injection volume flow rates