Total assessment of high strength, high clarity liquid silicone rubber.

Multiple companies, both global and regional, have been active in the production and sale of liquid silicone rubbers (LSRs) over the past two decades. Technology has advanced so rapidly in the field of LSRs that even within a single company's product line multiple changes have likely taken place since its inception. At this point in time, the larger players in the global LSR market - Dow Corning, Wacker, General Electric and Bayer - all have product lines of high clarity, high strength materials that look very similar in terms of physical properties. How, then, is a fabricator to decide which material best matches his/her needs? Complicating the situation is that very rarely is it possible to use products from different companies in the same application without making major changes to either the settings on the injection molding press, the mold itself, or both. This problem arises from the fact that, although physical properties are quite similar among the various product lines, rheological properties are grossly different. The ultimate physical properties in LSRs are achieved in a multitude of ways, resulting in products that handle significantly differently. For example, the Shore A hardness (durometer) of an LSR may be increased through the addition of filler, by increasing crosslink density, and/or by adding resins to the material. Each methodology has an associated implication with respect to the ultimate rheology of the system, and at least one major supplier of LSRs uses each of the aforementioned strategies to deliver the ultimate physical properties of their materials.

Experimental

Materials The LSRs used in this study were obtained through distributors of the materials. Samples of Silastic 9280 LSRs were taken from production lots that were intended for sale. All materials were used as is, with no additional processing other than the mixing of parts A and B in a 1:1 ratio (some rheological measurements required the inclusion of additional cure inhibitor). Due to the rapidly changing formulations of some producers of LSRs, materials available today may not be exactly represented by the studies described herein; the data contained in this study are based on materials obtained in early 1995.

Physical property measurements Physical properties of the specific lots of materials used for these studies were not measured; rather, "typical" values of each property reported herein were extracted from the product data sheets that accompanied the lots of materials. Data reported on Dow Corning Silastic were generated from press cured slabs (5 min., 150[degrees]C; 1.9 mm thick) using ASTM D 2240 to measure hardness and ASTM D 412 to measure tensile strength, elongation at break and 100% modulus. The tear strength of these materials was measured by ASTM D 624, and specific gravity was measured using ASTM D 792. From detailed SPC studies of the physical properties of LSRs from various manufacturers, it has been found that virtually any lot shipped as "approved" will possess properties similar to the reported "typical" values. Again, with the changes in formulations that take place over time, the "typical" values reported herein may not be reflective of the most current formulation at a given producer.

Cure profile measurements Cure rates were measured on an oscillating disk rheometer at 121[degrees]C (250[degrees]F]). "Liquid" dies were employed, and the sample size for measurement was five grams. Measurement was performed over six minutes using a 5.65 N-m (50 lb.-in.) torque scale. Samples were prepared by mixing equal amounts of each part A and part B, centrifugally mixing for 30 seconds, scraping the sample into a single mass, and then centrifugally mixing for another 30 seconds.

Sample preparation for rheological measurements In order to produce meaningful data, it was first demonstrated that the addition of extra cure inhibitor (necessary for elevated temperature measurements) did not appreciably alter the rheological profiles of any of the materials under evaluation. Thus, all samples produced for rheological evaluation were prepared by mixing equal amounts of part A and part B on a three roll mill (three passes), and then adding 3,000 ppm of 1-ethynyl-1-cyclohexanol (EtCH) to the LSR. The resulting mixture was milled for six additional passes to ensure good mixing. All samples were de-aired by centrifugation, and those not immediately used were frozen (-10[degrees]C] until measurements could be taken; studies showed that no differences were observed between "fresh" and "frozen-thawed" samples.

Creep measurements Creep was measured on a stress rheometer using a 25 mm parallel plate configuration. Measurements were performed at ambient temperature, and each sample was pre-sheared at 200 Pa for two minutes. Following the pre-shear, the material was allowed to relax for approximately two minutes. The plots contained herein were then generated by applying a stress of 100 Pa for two minutes, removing the stress, and monitoring the resultant strain throughout the entire process (total measurement time, four min. - see figure 1).

Viscosity as a function of shear rate measurements All viscosity vs. shear rate measurements were performed on a capillary rheometer. Samples were loaded into the barrels of the rheometer at 0.14 Mpa (20 psi) nitrogen, and those measured at elevated temperatures were allowed to equilibrate for one minute before measurements were begun. Long (32 mm x 2 mm diameter) and short (0 mm x 2 mm diameter) dies were run in parallel using pressure transducers rated at 10.3 MPa (1,500 psi) and 3.5 MPa (500 psi), respectively. The die entry angle chosen for each die was 180[degrees]. Viscosities were measured for shear rates ranging from 25 sec-sup1 to 2,000 sec-sup1 for each material studied, with measurements at shear rates less than 100 sec-sup1 highly variable because of the transducers chosen for the study.

Results and discussion

Physical properties Physical properties of high strength, high clarity LSRs now in production at the major manufacturers are very similar. Tables 1 and 2 illustrate this point quite clearly, showing data for 40 and 60 durometer (Shore A) materials available from Dow Corning, Wacker, General Electric and Bayer. A few trends and minor advantages are observed for specific product lines, but there really are no data available that suggest a clear advantage for a given producer's material that spans an entire product line.

Cure rates Figure 2 shows representative cure profiles (ODR) for Dow Corning, Wacker, General Electric and Bayer 50 durometer (Shore A) high strength, high clarity products. The slopes of the curves are relatively consistent, while the times at which cure initiates and the shapes of the cure curves vary somewhat. Initiation of cure is quite easy to control, most often by adjusting the catalyst/inhibitor ratio in the LSR. Cure shape can be altered through varying concentrations of both catalyst and inhibitor; the suppliers in this study have obviously used somewhat different technologies and/or catalyst/inhibitor ratios. An interesting phenomenon is observed when measuring cure rates of the same materials at higher temperature; cure curves become nearly indistinguishable when temperatures of 200[degrees]C or higher are employed. Basically, any material in this study can be made to look like any other from a cure perspective simply by varying the inhibitor and/or the catalyst levels. Considering the thickness, profile and surface area of an injection molded LSR part, a fabricator may favor one material over another based on cure characteristics. However, as in the case of the physical properties of these materials, there is no single product line with an obvious advantage over any of the others based on cure profile alone.

Creep profiles While some producers of LSRs report viscosity data for their products (either using a Brookfield viscometer or a cone and plate rheometer) (ref. 1), it is not uncommon for a user to have little or no information about low shear viscosities of the different available materials. It is commonplace for users to simply manipulate materials by hand and/or monitor the amount of pressure necessary to pump materials out of drums/pails to form opinions of relative viscosities. Some producers and users have traditionally used the property extrusion rate (amount of material that is extruded from a 3.2 mm nozzle under 0.62 MPa pressure), but this value is relatively meaningless unless LSRs with very similar viscosities are compared. Since the amount of shear that a material will experience in an extrusion rate measurement will vary greatly as a function of the material's viscosity, dissimilar materials might shear thin to a significantly different extent. Thus, a thicker material might look equivalent to a thinner one if it shear thins more.

In looking at creep, a measurement of the strain at constant stress, one can gain a feel for relative viscosities of materials at low stresses (the higher resultant strain measured as a result of stress, the lower the viscosity of the material). Figures 3 and 4 show creep profiles for 40 and 60 durometer (Shore A) products, respectively, that have experienced a stress of 100 Pa and then were allowed to relax.

Material viscosity at low shear is certainly important to a user of LSRs. High viscosity materials will certainly take more pressure to move from drums/pails than lower viscosity materials. Materials that are thicker will also be more likely to cause pumps to cavitate, especially if the materials do not shear thin significantly once they are subjected to shear stress.

Evaluation of the data generated in the creep experiments yields two important points. First, there is a relatively large difference in viscosity among the various available product lines. Second, the shape of the creep profiles are very similar, regardless of which producer's materials are evaluated; this is not surprising when considering that all of these materials are based on relatively similar technologies.

Viscosity, measurements as a function of shear rate Probably the most important information that an LSR fabricator needs to make a decision as to which producer's material will best suit its needs is how a given material will handle under the shear imposed by its particular injection molding machine, meter/mix component and the mold itself. Obviously, the pressures used in the operation, the sizes of the pipes, gates, vents, etc., will all impact the overall shear experienced by the material. No two set-ups are likely to impart exactly the same shear history on materials. From discussions with industry experts (ref. 2), it has been gleaned that most operations in an injection molding system are likely to impart shear rates of between 100 and 1,000 sec-sup1 on a liquid silicone rubber. It is certainly possible that shear conditions outside of this range can be experienced depending on configuration; however, using this order of magnitude as an approximation on which to base a study is certainly used to help us gain an understanding of how materials will behave in this likely range of shear rates.

Figures 5 and 6 show viscosity as a function of shear rate for 40 and 60 durometer (Shore A) LSRs. Obviously, all products from all producers are shear thinning to a significant extent. Also as obvious is that there are sinificant viscosity differences between products, even under high shear rate. In evaluating the data from these experiments, it is important to concentrate on the information at shear rates greater than 50 sec-sup1; because of the transducers chosen for use on the rheometer in these experiments (to optimize the accuracy of the measurements in the shear rate range of interest), viscosity data for lower shear rates are not reproducible.

Since injection molding operations do not take place entirely at ambient temperature, viscosity vs. shear rate data were also generated at elevated temperature to see if any of the observations made at low temperature are not valid when the materials are heated. Figures 7 and 8 show the same experiments detailed in figures 5 and 6, but with measurements taken at 100[degrees]C rather than at ambient temperature. What is observed from comparison of the data at different temperatures is that trends remain the same, both in terms of relative viscosities of materials and in degree of shear thinning. The only real difference between the data generated at high temperature vs. that collected at ambient temperature is that the high temperature viscosities are significantly lower, as expected; LSRs are well known to exhibit lower viscosity at higher temperatures, much in the same manner as a "non-filled" polymer melt (ref. 3).

With the viscosity data as a function of shear rate in hand, it is possible to generate some interesting plots to evaluate product lines from the various producers. Figure 9 shows the relationship between viscosity of materials at a shear rate of 100 sec-sup1 and the hardness (Shore A durometer) of the materials. This plot simply shows how a given producer's product line would be expected to behave at a shear rate of 100 sec-sup1; it is a useful way to evaluate whether adjustments in injection molding hardware will be necessary when changing from one product to another within a product series. The difference between product lines shown on this plot is substantial; while the S-9280 LSR series remains relatively constant in viscosity regardless of durometer, most of the other product lines vary substantially as a function of hardness. Figure 10, which represents the same type of data at a shear rate of 1,000 sec-sup1, shows a similar viscosity discrepancy between products in several of the product lines, although the differences between products made by different manufacturers are smaller. Other tests show that increasing the temperature of the materials to 100[degrees]C yields the same type of information that is drawn from the ambient temperature measurements.

Conclusions Differentiating between high strength, high clarity LSRs produced by different manufacturers using physical properties and cure rates is not easy. Multiple technologies can yield similar results even though the materials may handle grossly differently in an injection molding application. Rheological measurements are an excellent way to explain differences in material performance in an application; low stress experiments yield information on relative viscosities of materials as well as relative elasticities of products. Even more important, capillary rheometer measurements of viscosities as a function of shear rate can yield information that accounts for different behavior by dissimilar materials in a molding operation. Gross differences in viscosity between two materials at shear rates encountered in molding operations can easily explain why recalibrating of an injection molding press and/or fine tuning a mold might be necessary before changing materials. The fact that no two LSR manufacturers use exactly the same technology to develop their LSR products accounts for the observation that it is often difficult, if not impossible, to change suppliers without major "retuning" of an injection molding process.

References [1.] See Dow Corning STI or Wacker LSR Product Data Sheets for details. [2.] Private communications with mold makers, Dow Corning rheologists and injection molding press manufacturers. [3.] C.W. Macosko,"Rheology principles, measurements and applications," VCH Publishers, Inc., New York, 1994, ch. 11, P. 510.