Density reduction in auto PU RIM fascias.

In order for polyurethane materials to remain competitive in the automotive fascia industry and to alleviate some of the pressures put on by thermoplastic polyolefin (TPO) fascia suppliers, at least two issues must be addressed - recycling and weight reduction. The focus of this article will be weight reduction. The issue of recycling polyurethanes has been investigated by many and has, to some extent, been successfully commercialized (refs. 1 and 2).

Conventionally, weight reduction of RIM fascias was obtained through down gauging of part thickness. For example, many new fascias are designed at 2.7 to 3.0 mm wall stock thickness as compared to previous targets of 4.0 to 5.0 mm. Using today's technology, with the incorporation of hollow glass microspheres in RIM parts, the specific gravity of the substrate can be reduced to that equivalent of TPO materials.

A patented process has been developed (ref. 3) to reduce the density of RIM fascias by incorporating hollow glass microspheres which have a compressive strength greater than 4,000 psi and a maximum size of 120 microns. This new material retains the durability and styling capability advantages of polyurethane RIM materials. The low density RIM fascia satisfies OEM polyurethane material specifications and has a high glass painted outer skin comparable to conventional RIM products.

Although this technology is a proven process currently in production, we are still continuing to investigate the benefits of hollow glass microspheres in polyurethane RIM fascias. The objective of this article will focus on understanding the effects of microsphere loading and type on processing, physical properties and paintability. Also, the impact this technology has on economics will be discussed.

Reduced density polyurethane RIM fascias, using hollow glass microsphere technology, have been manufactured at Polycon Industries for well over a year. Therefore, a review of the advantages seen to date will be presented.

Experimental

Sample preparation

Samples were prepared on a 90 mt Cincinnati-Milacron RIM press. A current model year automobile rear fascia was used to evaluate the substrate physical properties, processing and the paint appearance. All test samples were manufactured at 10% above the minimum fill level and the resin component was nucleated to obtain a constant gas loading (43% by volume).

Hollow glass microspheres

Three samples of hollow glass microspheres were used in this study. The first sample, S-60, has a specific gravity of 0.60 and a compressive strength of 10,000 psi, the second sample, experimental K-55, has a 0.55 specific gravity and 8,000 psi compressive strength, while the third-sample, K-46, has a specific gravity of 0.46 with a compressive strength just over 4,000 psi. The main difference between the S-series and K-series glass is the type of feedstock used in the manufacturing process. Table 1 lists the particle size distribution for the different micrsospheres.

Table 1 - microsphere particle size distribution

                           Particle size distribution
Sphere                  10% vol.   50% vol.   90% vol.

S-60                    13[micro]  32[micro]  57[micro]
K-55                    15[micro]  30[micro]  53[micro]
K-46                    15[micro]  50[micro]  80[micro]

Testing methods

A list of the substrate physical properties tested and the corresponding ASTM procedures are presented in table 2. Ford Motor polyurethane material specification (ESB-M2D395A), listed in table 3, was used as a bench mark when analyzing the physical property results.

Table 2 - list of physical properties tested

Physical property          Test method

Density                      ASTM D792
Hardness                    ASTM D2240
Tensile strength             ASTM 412M
Elongation                  ASTM D638M
Tear strength                ASTM D624
Flexural modulus            ASTM D790M
Heat say                     ASTM D747
Izod impact strength         ASTM D256
Table 3 - material specification

Physical property                    Specification

Density (kf/[m.sup.3])                   ASTM D792
Hardness (durometer D)                  ASTM D2240
Tensile strength (MPa)                   ASTM 412M
Elongation (%)                          ASTM D638M
Tear strength (kN/m)                     ASTM D624
Flexural modulus (MPa)                  ASTM D790M
Heat say (mm)                            ASTM D747
Izod impact strength (kJ[m.sup.2])       ASTM D256

Raw materials

All formulations evaluated in this study included a 5,000 molecular weight tri-functional polyol, a crosslinker, zinc stearate based internal mold release, gel catalysts and a prepolymer methylene diphenylene diisocyanate (MDI). A reinforcing filler (wollastonite) was maintained at a constant loading of 6%, by weight, in the end product. The hollow glass microspheres were added at 2% and 4%, by weight, to the resin blend without any system modifications. The blends containing the microspheres were compared to the control sample which did not have any microspheres. Table 4 lists a sample formulation containing 2% hollow glass microspheres.

Table 4 - sample formulation

Component                               Concentration (pphp)

Polyol (5,000 molecular wt.)                   100.00
Crosslinker                                     21.23
Internal mold release                           9.15
Gel catalyst                                    0.13
Cure catalyst                                   0.16
Wollastonite (15.1 aspect ratio)               12.94
2% hollow glass microspheres                    4.31
MDI prepolymer @ 1.08 I                        68.00

Introduction of microspheres increases the cost per unit mass of the blend; however, the weight savings in the final part provides an economic benefit to justify their use.

Paint process

All samples tested were primed with a conductive primer and top coated with a 1K base coat and clear coat. The end product appearance was evaluated by measuring the surface gloss and DOI. A TI Systems Appearance Data Collector model #1854 SQC was used to measure the results.

The painted samples were investigated for surface defects that may be induced by the glass microspheres, particularly around regions of high shear flow, i.e., around or over contours in the mold or areas of thin cross-section.

Results and discussion

Processing

One of the most significant parameters in the processing of polyurethane fascias is the resin specific gravity. A reduced specific gravity facilitates the mold filling process and also reduces the part density. The resin specific gravity is typically reduced by the introduction of an inert gas through a porous media. The process of reducing the resin specific gravity is referred to as nucleation. There are many factors, including the type of fillers used, which impact the ability to nucleate. Fillers which have a positive influence on the nucleation process (usually porous and irregular in shape) are commonly called nucleating agents.

In order to evaluate the effect of the microspheres on nucleation, each formulation was subjected to nitrogen being injected through a sparger tube until a maximum gas loading was achieved. Figure 1 is a graphical representation of the results. It is evident that hollow glass microspheres do not substantially increase nucleation. This is most likely due to the extremely smooth and spherical nature of these particles. It is interesting to note that the gas loadings exhibited by the blends containing the K-46 microspheres were marginally higher than the other blends. A possible explanation can relate this phenomena to the greater likelihood of microsphere rupture due to the lower compressive strength of the K46 microsphere. Therefore, the broken glass, now irregular in shape, behaves as a nucleating agent. For example, by introducing sites for bubble formation.

[Figure 1 ILLUSTRATION OMITTED]

In order to determine the influence microspheres have on mold filling or flow, minimum fill studies were conducted for each system at a constant gas loading. The data reveal (figure 2) that the resin blends with K-46 microspheres were able to fill the mold cavity with less material than the other blends. The blends which incorporated the S-60 microspheres were marginally better than those using the K-55 microspheres, which were subsequently better than using no microspheres at all. It should be noted that 4% blends provided greater weight savings than the 2% blends for both K-46 and S-60 microspheres; however, this was not observed with the K-55 blend. It is unknown why the K-55 blend deviated from the expected results. In all cases, except for the 4% K-55 blend, the material savings provided an economic benefit to justify the use of the microspheres.

[Figure 2 ILLUSTRATION OMITTED]

It appears as though the microspheres with the broadest distribution provided the most benefit for flow. Small microspheres are beneficial as they inhibit the increase in viscosity (ref. 4), while the larger microspheres are known to enhance the flow. Studies by Strand (ref. 4) made reference to an analogy where larger microspheres have a greased ball-bearing effect on the material as it flows through the,mold cavity. Enhanced flow of the components has an additional benefit on processing by eliminating the need to over pack parts.

One interesting observation made during the sample preparation was the samples which contained the glass microspheres (independent of type and loading) did not have any nucleation blisters, while the control samples exhibited sporadic blister formation. Nucleation blisters are small inert gas bubbles which coalesce into larger bubbles. These larger bubbles subsequently ruin the fascia's appearance and are typically located away from the gate and near mold vents, i.e., low pressure regimes. Consequently, the minimum fill for the control sample would have increased if an equivalent quality part was produced, i.e., at times, processors will over pack the parts to compress the larger gas bubbles in order to produce a blister free part.

Originally, it was believed that the microspheres had some soil of stabilizing effect on the gas bubbles; therefore, a froth stability test was conducted (note: froth stability tests are commonly used to evaluate the stabilization effect of surfactant materials). The different resin blends were nucleated in the process day tank and then were subjected to ambient conditions. The resin blend specific gravity decay was then measured for a period of five hours. From figure 3 it is clear that there is relatively little difference in the specific gravity decay among the different blends, thus indicating that the microspheres do not stabilize the gas bubbles.

[Figure 3 ILLUSTRATION OMITTED]

By considering classical bubble dynamic theory, one may pose the following hypothetical explanation of the observed stabilizing phenomena. At equilibrium the gas pressure in a spherical bubble is larger than the pressure in the surrounding liquid by a difference of [Delta]P, given by the equation:

(1) [Delta]P = (2[Gamma])/R

where R is the bubble radius and y is the surface tension. Therefore, the gas pressure in a small bubble is greater than the gas pressure in a large bubble. Consequently, there is a tendency for gas to diffuse from small bubbles into large bubbles. The diffusion of gas from one bubble to another induces a stress on the bubbles. For instance, low pressure differences between different size bubbles provides better bubble stability (ref. 5). Since the microspheres displace the liquid component, the pressure exerted by the liquid on the gas bubble would be reduced. This reduction in the pressure differential, between the liquid and the gas, may decrease the pressure between different size bubbles which is favorable for bubble stability, as previously mentioned. Therefore, by starting out with a gas-liquid system with an improved stability, the end product should see the same improvement in bubble stability.

Physical properties

A summary of the substrate physical properties is shown in table 5.

Table 5 - summary of physical properties

Sample    Specific gravity      Tear (kN/m)    Tensile (MPa)
Control        0.98               100.5            22.3
2% S-60        0.97                98.8            22.4
4% S-60        0.96                85.8            18.0
2% K-55        0.97                85.7            18.5
4% K-55        0.98                84.7            17.6
2% K-46        0.96                88.1            18.1
4% K-46        0.94                81.7            16.9

Sample       Elongation         Flex. mod.       Hardness
                (%)               (Mpa)            (D)
Control        150.3              400.1             53
2% S-60        138.7              355.7             53
4% S-60         96.7              339.1             53
2% K-55         88.4              367.2             52
4% K-55        101.5              367.9             53
2% K-46         97.1              362.3             54
4% K-46         96.4              373.4             53

Sample               Heat sag (mm)          Izod (kJ/[m.sup.2]
Control                 8.1                         28.5
2% S-60                 6.5                         27.8
4% S-60                 5.9                         26.0
2% K-55                 8.3                         25.7
4% K-55                 6.1                         28.2
2% K-46                 4.7                         25.6
4% K-46                 7.2                         25.5

Due to the spherical shape of the glass microspheres, one can expect little enhancement in the reinforcing properties of the composite material. As a result, tear strength, tensile strength, flexural modulus and elongation will decrease with density. Thus, this trend becomes more apparent at higher microsphere loadings. However, it is important to note that the microspheres do benefit other properties such as part shrinkage and heat, sag, which is consistent with the findings of Suddaby et al. (ref. 6). Generally, part shrinkage decreases with an increase in microsphere loading (figure 4).

Although microspheres reduce part shrinkage, it does not alter the fascia's dimensions appreciably. Therefore, this technology is well suited for direct implementation into existing programs, and does not require any additional tooling costs.

[Figure 4 ILLUSTRATION OMITTED]

Since glass microspheres have no bearing on hardness, one may deduce that a non-interrupted skin is formed, which of course is a very desirable feature for a painted surface.

Impact strength, measured by the notched Izod impact test (at room temperature), generally decreases with microsphere loading. The blends containing the S-60 microspheres appear to be marginally higher then the blends containing K-55 and K-46 microspheres at equivalent loadings. This trend is consistent with the compressive strength of the microspheres.

In summary, the 4% K-46 material provides the greatest weight savings, based on density reduction. However, the author believes that the 2% S-60 blend is the preferred composition based on weight savings, substrate physical properties and processing enhancements.

Painting summary

Figure 5 summarizes the painted substrate appearance. Gloss values for these compositions with primer, black base coat,and clear coat were all above 90, which is typical for this paint and substrate combination. The interesting point to note is that the DOI values showed an improvement over the control sample with 2% and 4% microsphere loading, while relatively little change in the gloss values was observed.

[Figure 5 ILLUSTRATION OMITTED]

As microsphere loading increased, a marginal depletion in DOI was observed. Perhaps at higher microsphere loadings the microspheres accumulate near the part surface, thus affecting the part appearance. As a result, painted samples were investigated for surface defects. There were no visible defects in any of the samples evaluated. It is quite possible that the microspheres are present close enough to the surface such that they affect the appearance but have no bearing on the surface quality.

Field evaluation

The incorporation of hollow glass microspheres in RIM fascias at 2% loading, by weight, can reduce the end product density by as much as 10%. The reduction in density is primarily based on the displacement of polyurethane with a less dense material. Furthermore, with the incorporation of microspheres, RIM parts can be processed with a lower resin specific gravity, i.e., can increase the gas loading without negatively affecting the part quality. Without the use of micro-spheres, acceptable parts were processed with a resin density of 700 kg/[m.sup.3.] At a 2% microsphere (S-60) loading, by weight, the same quality parts were processed with a resin density of 550 kg/[m.sup.3]. The ability to process parts at a higher gas loading (lower polyol density) is a result of the better quality of nucleation. This in turn has reduced the number of molding defects seen. For example, over the past year, Polycon Industries has seen a 2.2% increase in molding yields due to the microsphere technology. Blister defects have decreased by 25%, while flash defects have dropped 15%. The flash defect decrease can be attributed to the enhanced flow which eliminates the need to over pack parts. Over packing creates unnecessary flash, which increases the possibility of flash defects. The combined benefit of weight reduction and increased yields corresponds to a 9.3% decrease in raw material usage, which is a significant savings.

The one negative aspect of this technology is the increase in porosity defects. Porosity defects have increased by approximately 3.2%; however, this is less than 0.5% of all molding related defects. There are at least two reasons why microspheres are believed to cause porosity problems. First, the microspheres may orient themselves close to the surface of the finished product which make them susceptible to being pulled out of the part during the finishing step, thus leaving a hemispherical pit on the surface of the part. Secondly, ruptured microspheres near the surface may provide sites for solvents from primers and paints to diffuse into which are subsequently flashed during the baking stages. The sequence of events which leads to the solvents evaporating while the primer or paint is curing may lead to the formation of porous defects.

Conclusions

* 2% S-60 blend provides a good balance of weight reduction, physical properties, and processing enhancements.

* Microsphere technology can be retrofitted into existing programs without system modifications and any additional tooling costs.

* DOI improves over the control sample for the different microspheres and loadings investigated.

* Processing enhancements are independent of micro sphere type and loading, for the microspheres and loadings investigated.

References

(1.) Morgan, R.E., et al., "Recycling RIM thermoset polymers into automotive fascia," SAE 1994 Proceedings, Detroit.

(2.) Sinopoli, D. and P. Farkas, "Ecostream process in automotive polyurethane foams," SPI-Polyurethanes 1996 Proceedings, Las Vegas, pp. 68-76.

(3.) Allen, M., et al., Conix Canada, 1996, "Process for the production of lightweight polyurethane moldings," U.S. Patent No. 5,532,280. U.S.A.

(4.) Strand, N. S. "For RIM - a filler that acts like a reinforcement," Plastics Engineering, pp. 35-37, 1980.

(5.) Frisch, K.C. and Saunders, J.H., Plastic Foams Part I, New York, Chapter 11, 1972.

(6.) Suddaby, B.R., et al., "Reduced density RIM with improved impact properties," SAE 1994 Proceedings, Detroit.

Acknowledgements

"PU rubbers, vulcanizable via dimerized TDI, to produce wear- and tear-resistant roll coverings" is based on a paper given at the May, 1997 Rubber Division meeting.

"Denisty reduction in auto PU RIM fascias" is based on a paper given at the September, 1997 Polyurethanes World Congress.

"Tire tread compounds with silica/CB blends" is based on a paper given at the May, 1997 Rubber Division meeting. "Polyurethanes in vascular grafts" is based on a paper given at the October, 1997 Rubber Division meeting. In the November, 1997 issue (vol. 217 no. 2) we omitted that the authors of "Role of Mooney-Rivlin constants in seal formulation under non-equilibrium conditions," - V. Subramanian, S. Ganapathy and K.R. Seshadi are all affiliated with Anna University in India.