PLUG ASSIST MATERIALS FOR IMPROVED FORMING OF TRANSPARENT POLYPROPYLENE By Kathleen Boivin and Noel Tessier CMT s Inc., Attleboro, MA Introduction A new class of syntactic foam with a copolymer base, available in grades HYTAC -FLX and HYTAC -FLXT, has been designed to improve clarity when forming transparent materials. Copolymer syntactic addresses concerns about scratching associated with syntactic foam and is a potential replacement for solid engineered polymers. The performance of copolymer syntactic was evaluated and compared to the performance of an engineered solid polymer and a traditional grade of thermoplastic syntactic foam. excellent release of tacky materials and minimal mark-off. PEEK is a solid engineered thermoplastic polymer while B1X is a customary grade of thermoplastic syntactic foam. The following plug designs were run: original, modified and 1.5, as illustrated in Figure 1. The modified plugs were Marbach designs and had a greater degree of taper along the length of the plug in an effort to minimize degree of scratching. Typically, syntactic foam is the plug material of choice when thermoforming thin gauge parts due to low heat transfer and excellent surface finish. Syntactic foam contains a high fraction of air since pre-formed hollow spheres are a main component. The air content gives syntactic foam its key property of low heat transfer which allows it to pre-stretch the sheet being formed with minimal chilling. The spheres are commonly glass, but can be ceramic or polymeric, and are bound together by epoxy, urethane or thermoplastic. Since clarity is an essential attribute when forming transparent parts, there is concern that the glass content in syntactic foam will cause scratching. PP is extremely susceptible to scratching during plug-assisted thermoforming, especially when forming deep draw parts such as drink cups. Therefore, solid engineered polymers such as polyetheretherketone (PEEK) and polyetherimide (PEI) are commonly used as plug assist materials for forming transparent PP. Copolymer syntactics have been designed to minimize scratching and improve clarity when forming transparent materials. In contrast to solid engineered polymers, copolymer syntactic can be easily machined and polished, and offers the benefit of lower cost. Experimental To validate the performance of copolymer syntactic for forming transparent materials, plugs made of HYTAC -FLX (FLX), HYTAC -FLXT (FLXT), polyetheretherketone (PEEK) and HYTAC -B1X (B1X) were evaluated for forming clear PP drink cups. FLX and FLXT are both epoxy copolymer syntactic foams. FLXT is impregnated with polytetrafluoroethylene (PTFE) for Figure 1. Plug design. PEEK was only run in the original design and was considered the control since it had a history of running well for this particular cup design. B1X, FLX and FLXT were run in the original and modified designs. To assess the effect of plug surface finish, a set of polished plugs and a set of as machined plugs were run in the modified design. In addition, B1X was run in the 1.5 design. A summary of the eleven plugs that were run, along with surface roughness data, is shown in Table 1. Table 1. Plug Summary. Run# Plug Plug Design Surface Ra(μm) 1 PEEK Original Polished 0.51 2 B1X 1.5 Polished 2.54 3 B1X Modified Polished 2.44 4 FLX Modified Polished 1.42 5 FLXT Modified Polished 1.47 6 B1X Modified As Machined 2.57 7 FLX Modified As Machined 1.85 8 FLXT Modified As Machined 1.93 9 B1X Original Polished 1.70 10 FLX Original Polished 1.40 1
11 FLXT Original Polished 1.04 Forming trials were conducted at Milliken and Company on an Illig RDM 54K thermoformer with in-line extrusion. The sheet formed was a 1.9 mm (0.075 in) thick clarified homopolymer PP. The part formed was a 140 mm tall drink cup with a draw ratio of 3.7. Each plug was run for 45 min and cups were collected at 5 min intervals. For Runs # 4 and #10, the plugs were only run for 15 min due to issues with machine jams. Sheet surface temperature was measured and adjusted slightly from 155 C to 157 C to achieve acceptable material distribution. The cups collected at 15 min and 45 min were measured for the optical properties of gloss, haze and clarity. Gloss measurements were made on the inside of each cup, above the bottom ridge on the sidewall. Haze and clarity measurements were taken at the top sidewall of each cup, outside the scratch area, and at the bottom sidewall of each cup, in the scratch area. With all of the data included, the correlation was weak with an R 2 value of 0.58 (an R 2 value of 1 indicates a perfect correlation). However, two of the outliers were data for cups formed with the PEEK plug. With the PEEK plug, the heaviest scratch area occurred above the location where gloss was measured resulting in artificially high gloss values. With the PEEK data removed, the R 2 value increased to 0.71. Since plug design impacts where the scratches occur, the correlation was further evaluated using only the data for the modified plug design. The R 2 value for the correlation was 0.83, as shown in Figure 2. Since there was a significant correlation between scratch rating and gloss, gloss can be used to quantitatively measure degree of scratching. The cups were also evaluated visually for scratches, haze and plug mark. Visual attributes were rated on a scale of 1 to 5 with 1 being the worst rating and 5 being the best rating. The cups were measured for thickness along the sidewall and bottom to assess material distribution. Scratches and Gloss Results and Discussion The gloss and scratch rating data is summarized in Table 2. One of the goals of the study was to determine if gloss measured on the inside of the cup in the scratch area would correlate to the visual scratch rating. Since gloss is a measure of how light reflects off a surface, it should relate to degree of scratching. If a correlation exists, then gloss could be used as a quantitative measure of severity of scratching. Table 2.Gloss and Scratch Rating Values. Run# Plug Gloss (GU) Scratch Rating 15 min 45 min 15 min 45 min 1 PEEK 88.5 82.5 3.5 3.0 2 B1X 42.8 41.1 3.0 3.0 3 B1X 52.7 51.3 3.5 3.0 4 FLX 77.0 ----- 4.0 ----- 5 FLXT 91.9 79.5 4.5 4.5 6 B1X 44.2 41.4 3.0 3.0 7 FLX 61.4 65.5 3.5 3.5 8 FLXT 69.4 71.5 3.5 3.5 9 B1X 47.9 37.2 2.0 2.0 10 FLX 91.9 ----- 4.0 ----- 11 FLXT 87.0 79.4 4.0 3.5 Figure 2. Gloss data for modified plug design. To evaluate the effect of plug surface finish on scratches and gloss, the data for the polished and as machined plugs in the modified plug design was used. Charts showing the effect of plug finish on scratch rating and gloss are shown in Figures 3 and 4, respectively. Figure 3. Scratch rating for modified plugs. 2
With the syntactic plugs, the effect of plug material on scratching was related to surface roughness (Ra). With the modified plug design, the average Ra for the FLX and FLXT polished plugs was 1.45 µm while that for the B1X plug was 2.44 µm. The considerably smoother surface of the FLX and FLXT plugs reduced the degree of scratching. The B1X data was used to determine the effect of plug design on scratches. The average scratch ratings for the modified, original and 1.5 designs were 3.13, 2.00 and 3.00, respectively, and statistical analysis indicated a significant effect. Since the modified design resulted in the least degree of scratching, the increased taper seemed to minimize scratches. Figure 4. Gloss for modified plugs. Polishing the plugs increased both the cup scratch ratings and gloss readings by about 20%. Statistical analysis showed that the effect was significant. Therefore, polishing the plugs dramatically reduced the amount of scratching in the cups. The effect of reduced scratching with a polished surface was related to surface roughness (Ra). As shown in Table 1, polishing the plugs significantly reduced the surface roughness. Polishing reduced the Ra of the FLX and FLXT modified plugs by 23% and 26%, respectively. With B1X, polishing lowered the Ra by 5%. To evaluate the effect of plug material on degree of scratching, the data for the polished plugs was used. Table 3 gives the average scratch rating and gloss reading for cups formed with the polished plugs. Table 3. Gloss and Scratch Ratings for Polished Plugs. Average Average Scratch Rating Gloss (GU) B1X 2.63 47.3 FLX 4.00 84.5 FLXT 4.13 84.5 PEEK 3.25 85.5 ANOVA P-value 0.023 0.000 Significant Yes Yes The ranking of plug materials for scratch rating from best to worst was FLXT > FLX > PEEK > B1X. The gloss for B1X was only 47.3 GU while that of the other plug materials ranged from 84.5 GU to 85.5 GU. Based on both scratch rating and gloss, B1X caused the greatest degree of scratching while FLXT scratched the least. Although the scratch rating for PEEK was only 3.25, the gloss was 85.5 GU. Again with the PEEK plug, the heaviest scratch area was above where the gloss measurements were taken resulting in artificially high gloss values. Haze To determine the effect of plug surface finish on haze, the data for the modified plug design was used. In the scratch area, polishing the plugs reduced the average haze value from 7.3% to 5.8% and the difference was statistically significant. Above the scratch area, plug surface did not impact the haze level. Since haze is a measure of how transmitted light scatters, haze values in the scratch area were related to the degree of scratching. As the amount of scratching increases, the transmitted light scatters more resulting in increased haze. To determine the impact of plug material on haze, the data for all of the plugs was used. As shown in Table 4, the differences in haze both in and above the scratch area were significant. The ranking for haze in the scratch area from least to worst was FLXT<FLX<PEEK<B1X. FLXT resulted in the least amount of haze in the scratch area. Again, the haze values were related to the degree of scratching. Since FLXT scratched the least, cups formed with it had the lowest haze in the scratch area. Table 4. Haze Values vs. Plug. Average Haze Value (%) In Scratches Above Scratches B1X 8.80 5.38 FLX 5.38 5.54 FLXT 4.78 5.32 PEEK 7.73 3.70 ANOVA P-value 0.001 0.093 Significant Yes Yes At first glance, the effect of material on haze seems to contradict the gloss results. However, the gloss measurements were made at 1.5 cm above the bottom ridge while the haze readings were taken 3.0 cm above the ridge. With the PEEK plug, scratches were not prevalent at the lower location but were heavy at the upper position. The difference in measurement location explains why the cups formed with PEEK had high gloss values but low scratch ratings and high haze in the scratch area. 3
The differences in haze values above the scratch area were also significant. The haze values for B1X, FLX and FLXT were very similar with an average of 5.41% while the value for PEEK was lower at 3.70%. However, there was evidence that the lower haze with PEEK was due to sheet surface temperature not plug material. Statistical analysis was used to determine the impact of sheet surface temperature on haze. There was a significant effect on haze above the scratch area but no effect on haze in the scratches. As shown in Table 5, the average haze value above the scratch area increased linearly as sheet surface temperature increased. Therefore, haze in the scratches was dependent on plug material while haze above the scratches was affected by sheet temperature. Table 5. Haze Values vs. Sheet Temperature. Sheet Surface Average Haze Value (%) Temperature ( C) In Scratches Above Scratches 155 8.34 3.47 156 6.93 4.82 157 6.62 5.55 ANOVA P-value 0.782 0.033 Significant No Yes Clarity To assess the effect of plug surface finish on clarity, the data for the polished and as machined plugs in the modified design was used. Polishing the plugs significantly increased the clarity in the scratch area from 93.5% to 96.2%. Above the scratches, plug surface finish did not affect clarity. The increase in clarity in the scratch area with a polished plug surface was related to the degree of scratching since clarity is dependent on how much light is scattered. The data for the polished plugs was used to evaluate the effect of plug material on clarity. As seen in Table 6, the ranking for clarity in the scratches from highest to lowest was FLXT=FLX>PEEK>B1X. FLXT and FLX resulted in cups with significantly greater clarity in the scratch area compared to PEEK and B1X. Above the scratch area, clarity was not affected by plug material. Table 6. Clarity vs. Plug. Average Clarity Value (%) In Scratches Above Scratches B1X 89.2 97.8 FLX 97.6 97.8 FLXT 97.5 98.0 PEEK 92.3 98.4 ANOVA P-value 0.078 0.408 Significant Yes No Based on the data for the B1X plugs, plug design had a significant effect on clarity in the scratch area but did not impact clarity outside the scratch area. As shown in Table 7, the clarity for the modified design was 91.6% while the average value for the other designs was only 84.6%. Again, this difference was dependent on the degree of scratching. The modified design resulted in the highest clarity in the scratch area and also exhibited the least amount of scratching based on the visual test. Table 7. Clarity vs. Plug Design for B1X. Average Clarity Value (%) In Scratches Above Scratches Modified 91.6 97.2 Original 84.9 98.1 1.5 84.3 98.3 ANOVA P-value 0.032 0.123 Significant Yes No Plug Mark Plug mark was evaluated visually with a higher rating for minimal mark off. Based on the data for the modified plug design, the plug mark rating increased from 3.25 to 3.70 with a polished surface. Polishing the plugs resulted in cups with considerably reduced plug mark. Although plug material would be expected to have an effect on plug mark, analysis of this set of data did not show a significant effect. The data for the polished plugs in the modified and original designs was analyzed to determine the effect of plug design on plug mark. The average plug mark rating for the modified design was 3.70 while that for the original design was 4.00. When comparing just the modified and original designs, the original design tended to reduce the appearance of the plug mark in the cups. Distribution The material distribution curves for B1X, FLX and FLXT are shown in Figures 5 through 7. As shown in the plots, polishing the plugs had little impact on material distribution. As expected, plug design did have a significant effect on thickness profiles. Compared to the original design, the modified design resulted in thicker rims and thinner bottoms. 4
design, the B1X and FLX plugs were cut to the same exact geometry since the materials were expected to have similar friction. However, the FLXT plug was made slightly larger to compensate for the lower friction of the material since it contains PTFE. As friction at the plug/sheet interface increases, more material tends to be pulled into the cup. The thickness profile achieved with FLXT was similar to that of the B1X. Therefore, the larger geometry of the FLXT plug compensated well for the lower friction. Figure 5. distribution with B1X. Figure 8. distribution comparison. Figure 6. distribution with FLX. Compared to the other plugs, the FLX plug resulted in significantly thicker rims, thinner lower sidewall and thinner bottoms. Runs #4 and #10, both with FLX plugs, had to be cut short since the thin lower sidewalls caused jams on the machine. The results indicated that the friction at the plug/sheet interface with FLX is lower than that of B1X. To compensate for the lower friction, plugs in FLX should be made slightly larger than those in B1X. Compared with the syntactic materials, PEEK resulted in thinner rims and thicker lower sidewall. The results indicate that plug geometry of the syntactic foams needs to be further modified to increase the thickness of the lower sidewall. Conclusions There was a strong correlation between visual scratch rating and gloss measured on the inside of the cup. Therefore, gloss can be used as an objective means of quantifying differences in degree of scratching. Figure 7. distribution with FLXT. As depicted in Figure 8, plug material had a major effect on material distribution. In the modified plug Polishing plugs greatly improved overall quality of the cups through reduced scratching, increased gloss and reduced plug mark off. Polished plugs also resulted in cups with less haze and greater clarity in the scratch area indicating that these properties are related to the degree of scratching. 5
Plug material had a significant effect on overall cup quality with FLXT performing the best and followed closely by FLX. FLXT resulted in cups with minimal scratching and high gloss. In addition, FLXT minimized haze and maximized clarity in the scratch area. PEEK led to moderate scratching while B1X had the greatest degree of scratching. Compared to B1X, the copolymer syntactics can be polished to a much smoother finish which results in fewer scratches. The cups formed with PEEK had the lowest haze values above the scratch area but the effect was due to sheet surface temperature. The average haze value increased linearly as sheet surface temperature increased. Plug design influenced scratch rating, clarity and mark off. With the modified design, the greater degree of taper along the plug length resulted in cups with the minimal scratching and greatest clarity in the scratch area. However, the modified design resulted in slightly greater mark off compared to the original design. Both plug design and plug material had an effect on material distribution. Compared to the original design, the modified design brought less material to the bottom of the cups. The results with the modified plug design showed that the larger geometry of the FLXT plug compensated well for the lower friction. Compared to B1X, FLX brought less material to the bottom of the cups indicating plugs in FLX should be made slightly larger than those in B1X to compensate for lower friction. Based on the study, syntactic copolymers are an alternative to solid engineered polymers when forming transparent PP. For optimum results, plugs should be polished and the material should be run at minimum sheet temperature. Acknowledgements The authors would like to thank Milliken and Company for running the thermoforming trials and measuring optical properties. Thanks to Marbach for help with plug design and manufacture. 6