Characterization of hard-to-process oil field application materials.

Polymer compound/composite materials for oil field applications are usually known for high hardness and extreme mechanical strength. These materials are very tough to process in the manufacturing plant. Mixing of these materials requires high shear rates for optimal dispersion. These compounds are produced from an internal mixing process that requires long mix cycles. Also, these materials are compounded with the usage of more scorchy ingredients for optimal properties. Consequently, these materials demand above average work history and are typically hard to process.

Currently in the rubber industry, processability is primarily measured using the Mooney viscometer. Unfortunately, this seventy-year-old method is not sensitive enough to measure subtle processing differences among factory batches of the same compound. Sometimes it cannot even distinguish com pounding variations among several compounds because they yield the same Mooney value.

Other instruments such as ODR (oscillating disk rheometer) and MDR (moving die rheometer) can be used to measure scorch times tsl and ts2, which are the times until one torque unit or two torque units rise above the minimum torque, respectively, is achieved. Although these data are good at the curing temperature, they do not address the state of viscoelastic properties at the processing temperatures when the temperature is continually rising.

With traditional rubber test methods, available technical information on uncured polymer compound/composite materials for processing purposes was rather limited until acquisition of the RPA 2000 rubber process analyzer. This instrument provided unique capabilities through a sealed, pressurized cavity that enabled us to more effectively test the processability properties of uncured compounds.

Cured rubber compounds used in down-hole applications are engineered materials especially designed for this application. This vulcanized rubber must meet many specified requirements, including dynamic performance requirements. These dynamic performance requirements can be determined by measuring their elastic and viscous qualities at different frequencies, strains and temperatures with the RPA 2000. If a given rubber compound does not possess specific dynamic properties, then the rubber product could fail prematurely while in service. Rubber, by its very nature, is a viscoelastic material. That means it possesses both viscous and elastic properties. The balance of these viscoelastic properties determines how well a blow-out preventer, for example, will perform in service. The RPA can measure these cured viscoelastic properties of the rubber compound by in situ curing of the "green" compound, followed by measuring the stress responses from a sinusoidally applied strain to the test sample at a lower temperature after the cure.

Figure 1 demonstrates this measured torque response from the sinusoidally applied strain. As can be seen, the torque stress response from this oscillating strain is not exactly in phase. In fact, the angle that this response is "out-of-phase" is known as the phase angle (8). The reason there is a phase angle at all is because cured rubber is still a viscoelastic material. While the crosslinking from vulcanization has greatly enhanced the elastic quality (G') of the cured rubber, it still retains a viscous quality (G") as well. The reported tangent 8 is simply equal to G"/G'. So a cured rubber that retains a higher tan [delta] is more energy-absorbing, usually dissipating that energy as heat.

[FIGURE 1 OMITTED]

The factory problem

A polymer composite material, which was being used in production, had a higher rate of scrap when compared to other production compounds because of splits at the splice during the compression molding operation. Compression molding inherently requires a higher viscosity compound for processing a small cross-section o-ring part.

This is at odds with high shear rate mixing and mixing times that are required to get the desired dispersion. In the process of achieving good dispersion of a fully reinforcing carbon black, a high level of work history must be applied to this compound. However, with an increase in dispersion, the compound's viscosity is inadvertently reduced. The danger is that the compound's viscosity may be lowered too much and result in compression molding problems, such as non-fills, because of insufficient back pressure. Therefore, a much more sensitive measure of compound viscosity and theology was required because of the uniquely tight viscosity window that this compound had to meet. If the compound possessed too low a viscosity, there would be molding problems such as non-fills or pock marks. On the other hand, if the compound possessed too high a viscosity, this could indicate dispersion problems or poor polymer breakdown. If careful viscosity control is exercised utilizing the RPA, then better knitting can be achieved at or before curing temperatures, yielding parts without splice problems. Figure 2 shows the part without quality problems. Figure 3 shows a common example of the quality defect caused by poor control of the compound viscosity.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Experimental

A series of Design of Experiments (DOE) runs using the RPA were conducted to develop the most sensitive protocol to as sure that the subject compound would consistently perform within the viscosity/rheological window (discussed earlier) to assure good quality control of this compound.

From the DOE conducted, it was determined that the variable temperature analysis (VTA) feature of the RPA worked best in controlling the relevant quality aspects of this compound. This VTA simulates the actual plant processing and curing temperatures. A close examination of all RPA outputs was performed. It was found that the minimum S' (elastic torque) at various processing temperatures correlated the best with the processability issues (splits, etc.) in the manufacturing of the parts. The best RPA test conditions are shown in table 1.

The RPA outputs that were examined were S' (elastic torque) and tan [delta] at [M.sub.L] (min. torque), at [M.sub.H] (max. torque), and at 175[degrees]F, 200[degrees]F and 235[degrees]E The specific values of the S' at these three temperatures were found to give a direct correlation with the processability and acceptability of the manufactured parts. This RPA test protocol has now been adopted as the quality control standard for testing this compound. Statistical process control (SPC) principles were applied, and control charts are used routinely to control this process using the RPA.

Discussion of results

Figures 2 and 3 show pictures of a good part and a defective part, respectively. Figure 4 compares the RPA variable temperature analysis S' (elastic torque) results at the stated temperatures for various in-house production compounds. From experience in production, the compounds that are not excessively high for S' values were found to process better. However, if these S' values are too low, then non-fills can result.

[FIGURE 4 OMITTED]

Figure 5 compares the RPA variable temperature analysis tan [delta] at the stated temperatures for these same compounds. Again, the compounds with higher tan 5 at their respective temperatures were also found to process better (fewer splits, fewer knit lines).

[FIGURE 5 OMITTED]

Figure 6 shows S' values at the respective temperatures for the developmental replacement compounds for the high viscosity material that was producing high scrap rates. Experimental data from study A show that the compounds with higher S' torque values produced more defective parts in the form of split, knit lines, etc. On the other hand, the compounds with lower S' torque values also generated parts that were defective in the form of underfills, because the materials flow a little too much. Compound N-90 was found to process better in manufacturing, and S' torque values measured from the RPA were adopted for quality control in production mixing.

[FIGURE 6 OMITTED]

Figure 7 gives tan [delta] values at the respective temperatures of the developmental replacement compounds for the high viscosity material that was producing high scrap rates. However, the data were not as discerning as the S' value that was shown in figure 6.

[FIGURE 7 OMITTED]

Figure 8 shows the SPC three sigma charts for S' of the RPA viscoelastic properties for the developmental replacement compound with optimal elastic torque.

[FIGURE 8 OMITTED]

Figure 9 shows the SPC three sigma charts for tan delta of the RPA viscoelastic properties for the developmental replacement compound.

[FIGURE 9 OMITTED]

Conclusion

The methodology developed using the RPA 2000 enabled us to predict the processability of the critical compound to reduce scrap related to splits and non-fills.

Neither the oscillating disk rheometer (ODR) nor the moving die rheometer (MDR) were useful in detecting and characterizing this problem, as these are not capable of running variable temperatures in a short programmable cycle.

The usage of the variable temperature analysis cure procedure and special attention to the S" values at 175[degrees], 200[degrees] and 235[degrees]F indicate that this is an excellent way to characterize materials for processing.

Higher hardness materials, 85 durometer A and above, can definitely be characterized using this VTA protocol for processing purposes.

Sources

1. ASTM D1646--Viscosity, stress relaxation and pre-vulcanization characteristics (Mooney viscometer).

2. ASTM D2084--Standard test method rubber property vulcanization using an oscillating disk cure meter.

3. ISO 3417--Standard test method for elastomeric materials to determine values of viscosity, cure characteristics and temperature at various points in the cure cycle using an oscillating disk cure rheometer.

4. ASTM D6204--Standard test method rubber property--measurement of unvulcanized rheological properties using rotorless shear rheometers.

5. J.C. Vicic, "Testing of polymers for oil and gas applications," Energy Rubber Group, September 1984.

6. J.S. Dick and H.A. Pawlowski, "Applications of the rubber process analyzer in predicting processability and cured dynamic properties of rubber compounds," Educ. Sympos., Rubber Division, ACS, May 18-21, 1993.

7. J.S. Dick, C.A. Sumpter and B. Ward, "New effective methods for measuring processing and dynamic property performance of silicone compounds," Educ. Sympos., Rubber Division, ACS, Sep. 29--Oct. 2, 1998.

8. J.S. Dick and H.A. Pawlowski, "Rubber characterization by applied strain variations using the rubber process analyzer," Rubber World, Vol. 211, No. 4, January 1995.

9. J.S. Dick, M. Ferraco, K. Immel, T. Mlinar, M. Senskey and J. Sezna, "Utilization of the rubber process analyzer in Six Sigma programs to improve quality and reduce production costs," Educ. Sympos., Rubber Division, ACS, Oct. 16-19, 2000.

10. J.S. Dick and H.A. Pawlowski, "Applications for stress relaxation from the RPA in characterization and quality control," Rubber World, Vol. 215, No. 4, January 1997.

11. J.S. Dick and H.A. Pawlowski, "Applications for the rubber process analyzer," Educ. Sympos., Rubber Division, ACS, Nov. 3-6, 1992.

12. H.A. Pawlowski and J.S. Dick, "Measurement of the viscoelastic properties of elastomers with a new dynamic mechanical rheological tester," Educ. Sympos., Rubber Division, ACS, May 2-5, 1995.

13. H.A. Pawlowski and J.S. Dick, "A new mechanical tester designed for testing rubber," Educ. Sympos., Rubber Division, ACS, May 19-22, 1992.

14. J.S. Dick and H.A. Pawlowski, "Alternate instrumental methods of measuring scorch and cure characteristics", Educ. Sympos., Rubber Division, ACS, Oct. 26-29, 1993.

Jogesh Arora and Hamid Salem, Cameron Elastomer Technology, and John Dick, Alpha Technologies