Silica reinforcement of oil field elastomers for improved decompression resistance.
The addition of fillers to an elastomer increases die viscosity of the elastomer, limits extreme chain mobility and thus improves mechanical properties such as modulus, tensile strength and tear strength. If the filler can be "wetted" by the elastomer, the filler particles may serve as physical crosslinks. The Guth-Smallwood equation correlates the increase in modulus to the volume fraction of the filler.
Gf/G = 1 + 2.5 C + 14.1 [C2.sup.2] (1)
Gf is the modulus of the filled elastomer,
G is the modulus of the unfilled material, and
C is the volume-fraction of filler.
The ratio of the modulus of the filled elastomer to the unfilled elastomer is called the "strain amplification factor," which is a measure of the reinforcing effect of filler. This equation is a derivation of a hydrodynamic relation developed by Einstein that describes the increase in viscosity of a liquid fluid with rigid spherical inclusions to the viscosity of the unmodified liquid.
These relations are strictly valid only for spherical particles that do not aggregate or interact with the fluid, but modifications of these equations by other workers such as Patel include a form factor for describing the effects of the structure and particle size of carbon black fillers. As carbon black loading is varied in an elastomer-black system, some of the rubber properties reach an "optimum" level at a certain carbon black concentration, as described by optimum loading equations developed by Patel (ref. 1). These rubber properties include hysteresis dynamic mechanical properties, heat buildup), abrasion and process ability for tire tread compounds (ref. 2). Additionally, substantial changes are observed in the dynamic mechanical properties and the tensile properties of the compound compared to the gum rubber (ref.3).
The optimum black loading is dependent on black parameters like particle size, structure and polymer-filler interaction. Peters, et al. (ref 4), have extended Patel's work on tire tread compounds to two different fluoro-elastomers, poly(vinylidene fluoride co-hexafluoropropylene), designated FKM, and poly(tetrafluoroethylene copropylene), designated TFE/P, plus hydrogenated nitrile rubber (ACN 37%), designated HNBR, the primary elastomers used in the oil patch. Furthermore, they also demonstrated the existence of optimally loaded black-FKM and HNBR compounds which exhibited both the greatest resistance to gas expansion combined with the highest strain energy or fracture toughness.
Ideally, elastic modulus and strain energy, which is the area under a load-displacement curve and is proportional to the work of deformation, are expected to peak at optimum loading. Therefore, an optimally loaded compound will have high inherent fracture resistance and the maximum resistance to gas expansion damage. Initial void formation in elastomers has been linked to a critical internal pressure, 51 6 E, where E is the elastic modulus of the rubber (refs. 5 and 6). The results reported here extend the work done on carbon black-filled FKM, TFE/P and HNBR to pyrogenically manufactured, fumed silica fillers pp-[Ssub.i, O.sub.2] - ER) whose particle shape, structure and interaction characteristics are expected to be quite different from that of carbon black.
Experimental Test slabs preparation
The test slabs were prepared by the respective manufacturers of the TFE/P, the FKM and the HNBR, and were used as received. Me specific materials used, designation and manufacturer are in table 1. The formulations in tables la, 1b and 1c consisted of several silica loadings below and one silica loading above the calculated optimum loading (OL) of each type of silica. The optimum loading was calculated as follows (ref. 7):
OL = 100.491-0.545(VV)-0.088([N.sub.2 SA]) (2)
Structure (VV) and nitrogen surface area (N2SA) values were taken from manufacturer's literature. Specifically, void volume values were based on measurements with triethanolaniine TEA) rather than di-butyl phthalate (DBP) titration liquid. Optimum loading levels for each of the silicas were adjusted to account for the specific densides of the TFE/P, the FKM and the HNBR. Note this relation does not include a specific polymer-filler interaction parameter.
Actual loadings were measured by thermogravimetric analysis (TGA) using the following test protocol:
50 degrees C (2 min) ~ 295 degrees C (10 min) ~ [N.sub.2] 510 degrees C (5 min) \_/ - [arrow down] <------- air 850 degrees C (10 min)
Tensile properties were determined per ASTM D412 on specimens die-cut from standard ASTM test slabs. Tension tests were performed on an Instron, Model 4202, at a cross-head speed of .008 m/sec. Tensile properties at 2050C were obtained using a forced air environmental chamber. Strain energies, as received and at 205'C, were calculated from the load-displacement curves using Simpson's Approximation. These data are tabulated in tables 2a-2c.
Exposure to an oilfield environment was carried out in a 3.8 liter stainless steel autoclave lined with Hastelloy and wrapped with a heater jackel The sour gas exposure conditions are shown in table 3a. The carbon dioxide gas impregnation and depressurization exposure conditions are in table 3b.
Dynamic mechanical properties of the various materials were measured using a Rheometrics mechanical spectrometer, Model RMS-800. Measurements were made in forced dynamic shear using torsion rectangular geometry. Temperature sweeps from 520C to 2050C at a constant strain of 0.2% and rate of 0.2 Hz were performed on 45 mm x 12 mm x 2 mm rectangular samples. The complex moduli in MPa, G*, were then recorded versus temperatures.
Results and discussion
Experimental strain energy data at 205'C obtained with FKM, TFE/P and HNBR vulcanizates filled with various loading of A200, R972 and A130 are plotted in figures 1, 2 and 3, respectively. There is no significant strain energy enhancement with the use of silica fillers in these formulations, which could signify limited polymer-filler interaction.
Dynamic torsional rheometry was used to derive a viscoelastic property called the reinforcement factor - the ratio of dynamic moduli, G*, for filled and unfilled polymers. In a previous report (ref.9) on SBR-1500 using capillary rheometry, it was found that reinforcement factor increases with increased carbon black loading. Above a critical loading level there is a rapid rise in reinforcement factor. A. Pouchelon, et al. (ref. 10), showed the dynamic storage modulus (G') in uncrosslinked silicone rubber compounds as a function of volume fraction of a fumed silica, A150, filler. Above a critical filler loading of 20phr, the slope of the curve increased rapidly. Whereas neither the FKM nor the HNBR compounds gave stress-strain property responses typical for a reinforcing filler (figures 4 and 6), the TFE/ P compounds as plotted in figure 5 showed a relatively low slope section where additional filler produces an improvement of viscous reinforcement plus a steeper portion of the curve beyond a threshold concentration, called optimum loading" where modulus increases dramatically but fracture properties are substantially reduced.
The plots of strain energy after H2S exposure versus weight percent of silica for FKM, TFE/P and HNBR show the toughening characteristics imparted to these polymers in terms of their ability to stand up to H2S or sour gas. When contrasted to FKM-carbon black compounds (ref. 11) which degraded and became very brittle after H2S exposure, the FKM-silica compounds retained roughly 70% of their strain energy after sour gas exposure. The TFE/Psilica compounds retained 52% of their strain energy after sour gas exposure. The HNBR-silica compounds retained 65% of their strain energy after sour gas exposure. Surprisingly the hydrophilic silicas, 200 and 130, were more reinforcing than the hydrophobic variant, R972, for all three polymers m sour gas service. Generally, hydrophobic silicas are expected to have greater polymer-filler interaction and superior mechanical properties.
The plots of strain energy after C02 exposure versus weight percent of silica for FKM, TFE/P and HNBR show the toughening characteristics imparted to these polymers in terms of their ability to stand up to CO2 service. The FKM-silica compounds retained 90% of their strain energy after CO2 exposure. The TFE/P-silica compounds retained 96% of their strain energy after C02 exposure. The HNBR-silica compounds retained 70% of their strain energy, after C02 exposure. Similarly to sour gas data, the hydrophilic silicas were more reinforcing than the hydrophobic variant for all three polymers in C02 service. It is worth noting that the filler manufacturer (ref 12) recommends the use of pretreated hydrophobic silicas like R972 since a formation of bubbles can develop during the vulcanization or during the post vulcanization treatment when hydrophilic silicas are used. Ferro (ref. 13), et al., also reported on the higher stability of white fillers coated with hydrophobic chemicals when compared to the hydrophilic variety.
In figure 7, the 205 degrees C normalized strain energy curve and the 205 degrees C normalized elastic modulus curve of the TFE/P/A200 compound were overlaid. The point where the elastic modulus or the reinforcement factor at highest strain energy intersect is considered the optimum loading of silica. With TFE/P-silica compounds, the highest available elastic modulus coincides with the highest strain energy. Therefore, the silica filler in TFE/P is reinforcing, which is defined as a filler which raises the modulus while still maintaining the rubber-like qualities of the base material and which at the same time increases the strength of low-strength rubbers (ref. 14). The FKM/silica and HNBR/silica characteristics are less apparent. Gent's theory (ref. 15) concerning the dependence of bubble formation on the elastic properties of the polymer-compound and critical flaw size seem to apply to silica filled TFE/P compounds. Subsequent crack propagation after bubble formation will depend on the strain energy (refs. 16 and 17) or fracture toughness of the compound. Relying on elastic considerations alone will not give the optimum decompression resistant compound, but the added consideration of strain energy or fracture toughness must also be taken into account. Strain energy falls off dramatically beyond the optimum loading, so a compound with the highest elastic modulus will not always perform the best. In 1982, Hull (ref. 18) showed that pretreated hydrophobic silica imparts very good mechanical characteristics to TFE/P polymer vulcanizates in terms of aging behavior.
The FKM compound formulation in table la shows that all the compounds were formulated with the coupling agent, gamma-chloropropyl triethoxy silane (1:1 N330 black), except for compound V17 which is identical to compound V16 without the coupling agent. Based on the 40% decrease in tan delta and 12% decrease in elongation going from no coupling agent in compound V17 to compound V16, with the coupling agent, it would appear that no fracture toughness advantage was realized due to the potential increase in chemical crosslinks between the FKM polymer and the silica filler imparted by the coupling agent. Perhaps this surface treatment did not improve polymer filler coupling which could explain the behavior in this case. On the other hand, this could be due to slight differences in structure or surface area. Additional work will be needed to define the reasons for this result.
* FKM/silica, TFEP/silica and HNBR/silica compounds can be optimized for superior fracture toughness in [Hsub.2 S] and CO2 service. The fumed, hydrophilic silicas, A200 and A130, were more reinforcing than the hydrophobic variant, R972, for these three polymers in both H2S and CO2 simulated service.
* TFE/P type fluoroelastomer compounds show significant reinforcement effects among the different grades of fumed silica. For A200, optimum reinforcement levels calculated from equation 2 show that it is achieved at a critical volume fraction, Dc, of 0.11 which corresponds to a critical concentration of 20 phr. This [phi.c] is very close to the Dc of 0.09 determined for A150-filled silicone gums ref. 19). Contrast this to a [phi-c] of 0.22 for carbon black filled FKM-type fluoroelastomers (ref. 20).
Based on this work and prior investigations by the authors (ref. 21), carbon black fillers appear superior to silica fillers for reinforcement of FKM and HNBR elastomers. In contrast, silica fillers show better reinforcement of TFE/P than do carbon black fillers.
Although these investigations did not specifically measure polymer-filler interaction, the slopes of the reinforcement factor plots may provide some indication of this interaction. Further analysis is required to quantify this.
Quantification of this polymer-filler interaction term win improve the accuracy and predictability of equation 2.
1. A.K. Sircar, "Optimum loading of carbon black in rubber by Monsanto oscillating disc rheometer," Rubber World, November 1987.
2. A.C. Patel and J.T. Byers, "The influence of tread grade carbon blacks at optimum loadings on rubber compound properties, " Rubber India, April 1982.
3. J.M. Funt, presented at a meeting of the Rubber Division, ACS, Montreal, Canada, May 26-29,1987. 4. LA. Peters, J.C. Vicic and S. Haeberle, "Optimum loading of carbon black for explosive decompression resistance of elastomer compounds for oilfield service, " presented at the International Rubber Conference 88, Sydney, Australia, October 10-14, 1988.
5. R.L. Denecour and A.N. Gent, J. of Polym. Sci., Part A2,1853 (1968).
6. M.J. Doyle, How elastomers fail in fatigue," Machine Design, March 1988.
7. Reference 2, loc. cit.
8. Reference 4, loc. cit.
9. J.T. Byers and A.C. Patel, "Selecting optimum carbon black loading for properties affecting rolling resistance and performance, " Carbon Blackboard.
10. A. Pouchelon, P. Vondracek, Sendempirical relationships between properties and loading in filled elastomers, Rubber Chem. and Tech., 62, 788 (1989).
11. Reference 4, loc. cit.
12. H. Ferch, A. Reisert, Aerosil R972 as filler influoroelastomers," Degussa Technical Bulletin No. 73.
13. R. Ferro, G. Giunchi, S. Aloisio and C. Lagna, Effect of environmental conditions fillers on cured fluoroelastomers," Rubber and Plastic News, 40 February 19,1990).
14. F. Bueche in Physical properties of polymers," Inter-science Publishers, New York, 1962, p. 47 and p. 232.
15. A.N. Gent and DA. Tompkins, J. of Appl. Phys. 40, 2520 (1969).
16. A.N. Gent in Fracture, volume 7," H. Liebowitz, Ed., Academic Press, New York 1972, ch. 6, p. 315.
17. F.R. Eirich and T.L. Smith, ibid, ch. 7, p. 351.
18. D.E. Hull, Elastomerics, 114, No. 7,27,(1982).
19. Reference 10, loc. cit. 20. Reference 4, loc. cit. 21. Reference 4, loc. cit.
|Printer friendly Cite/link Email Feedback|
|Date:||Dec 1, 1990|
|Previous Article:||Carousel system aids ingredient batching.|