Low temperature sealing capabilities of fluoroelastomers.
Fundamentally, elastomer seals that are under compressive deformation work by applying a pressure against their respective housing (ref. 1). This sealing force (pressure) is time and temperature dependent and will gradually decrease due to physical and/or low chemical effects. The retention of this sealing force is dependent upon the stress relaxation of the rubber compound. The result of stress relaxation is that compressive stress eventually drops below system pressure resulting in seal leakage (ref. 2). The sealing force at lower temperatures depends on the extent of recovery of the elastomer at that temperature. As temperature decreases, the recovery from an applied deformation becomes slower. The rubber becomes progressively stiffer and leathery, loses its elastic recovery capability and finally becomes hard and brittle like glass (ref. 3).
The rubber chemist uses several tests to define low temperature characteristics of elastomers. Traditional tests such as stress relaxation, brittleness, temperature retraction, compression set, stiffness and glass transition temperature are meaningful to a knowledgeable rubber chemist. However, these tests and their associated results are less than clear to a design engineer unfamiliar with elastomers. Furthermore, the engineer concerned with seal design usually cannot make full use of these low temperature data because there is no direct correlation between laboratory tests and the actual service performance of the seal due to the varying nature of end use conditions.
While the fluids and heat resistance of fluoroelastomers has been well documented (refs. 4, 5 and 6), the ability of these elastomers to seal at low temperature has not been discussed as extensively. The purpose of this article is to examine some traditional rubber methods of measuring low temperature properties and compare those results to a test designed to evaluate the low temperature static sealing of o-rings. Testing is performed on a variety of fluoroelastomers to document differences and establish trends. By evaluating the performance of various fluoroelastomers in a practical low temperature o-ring test, the design engineer may gain a better understanding of the low temperature sealing capabilities of these polymers.
Although the test results disclosed in this article are of short duration, it is the intention of the authors to evaluate the effects of longer term compressive deformation, stress relaxation and hot/cold cycling with the static o-ring test device in future work.
Low temperature tests
Laboratory low temperature tests evaluated were brittle point, TR-10 and Tg by DSC. The brittle point, which is a mechanical deflection of the rubber in a cold bath, was run per ASTM D2137 (ref. 7). The temperature retraction test, TR-10, which is a mechanical stretching and recovery of the rubber in a cold bath, was run per ASTM 1329 (ref. 8). Glass transition temperature (Tg) by differential scanning calorimetry (DSC) is not an ASTM-defined test, but the test method used is standard for rubber labs and is explained in detail by Laird and Liolios (ref. 9).
Low temperature o-ring leakage test
Test specimen: O-rings conforming in dimensions to SAE Aerospace Standard 568-214 were used (ref. 10). The approximate size of the o-rings are 25mm I.D. by 3.54mm cross section (0.984" I.D. by 0.139" C/S). These will be referenced as -214 o-rings through the balance of this article.
Test apparatus: The o-ring test apparatus consists of a stainless steel block and plugs as shown in figure 1, along with associated fasteners, valves, piping, fittings, quick-connects and means of applying nitrogen at 1,380 KPa (200 psi) pressure. Leakage detection was provided by a linear mass flowmeter.
This apparatus was adopted, with modifications, from SAE Aerospace Material Specification 7273B (ref. 11). Actual design and prints were provided by AC Rochester Division, General Motors Corp. (ref. 12). The gland design conforms to SAE Aerospace Recommended Practices 1231 and 1232A (refs. 13 and 14). A radial squeeze of 10, 20 and 30% is provided within three separate cells of the test block. The surface finishes are industry standard (0.07-1.52 um). Cooling of the apparatus is provided by a low temperature refrigerated freezer capable of reaching -90 [ degrees ] C.
Procedure: O-rings of a good visual surface quality are lightly lubricated prior to installation in the test plug gland. The plug is fastened into the test block with socket head screws until finger tight. A secondary backup -214 o-ring conforming to AMS-7273B is positioned directly above the leak port. The apparatus is placed into the low temperature cabinet and the temperature is lowered at a rate of 0.25 [ degrees ] C/minute until nitrogen leakage is detected (0-10 standard cubic centimeter per minute; SCCM). The temperature is monitored continuously by RTD platinum thermocouple. The test is terminated when 10 SCCM leakage is obtained. The apparatus is allowed to warm back to room temperature and the procedure is repeated with another test specimen. The time to run a test varies between 3-6 hours depending on how low a temperature is needed to detect a leak.
All test samples were prepared using standard rubber processing techniques. Compounds were mixed either in a laboratory internal mixer or on a two-roll mill. The FKM vulcanizates tested were prepared using commercial fluoroelastomers from Du Pont. Table 1 provides a general description of the different types of FKM polymers tested.
Standard test results
Slabs and -214 o-rings were prepared from the fluoroelastomers noted in table 1 using standard carbon black type formulations. Shown in table 2 are the compound formulations used and stress/strain and compression set data along with results from traditional low temperature test methods: TR-10, glass transition by DSC and brittle point. All fluoroelastomer compounds in table 2 are a nominal 75 [plus or minus] durometer stocks.
The compound based on FKM-E60C is designed to meet the major o-ring specifications, MIL-R-83248 and AMS-7280. Many fluoroelastomer o-rings in service today are based on this type of "standard" fluoroelastomer, so it is useful to compare other FKM types to E60C. The low temperature data in table 2 shows the historical trends discussed in earlier literature (ref. 15). That trend notes there are certain property "tradeoffs" made, when the fluorine content of a fluoroelastomer is increased. Although the fluids resistance of a fluoroelastomer is improved by increasing fluorine content, dynamic low temperature flexibility as measured by TR-10 and Tg (DSC) is compromised. This trend is observed with FKM-B600 and FKM-6191 in table 2 when compared to FKM-E60C. In the case of FKM-B70, a modest improvement in both dynamic low temperature flexibility and brittle point is observed. This is accomplished by maintaining the fluorine content of FKM-B70 at 66% through the use of a terpolymer rather than a dipolymer backbone.
Specialty grades of fluoroelastomer, like FKM-GLT and FKM-GFLT, contain PMVE monomer that improves dynamic low temperature performance as measured by TR-10. This unique PMVE technology, seen in FKM-GFLT, enhances both fluids resistance through higher fluorine content, and improves dynamic low temperature flexibility as compared to "standard" FKM-E60C.
Having documented the performance of the compounds in table 2 to traditional rubber lab low temperature tests such as TR-10, DSC and brittle point, the same compounds were used to make o-rings which were tested in the static sealing device. Of particular interest was to ascertain if the sealing test results would parallel brittle point or glass transition data.
Low temperature o-ring test results - phase 1
Shown in figure 2 are the results of the same six fluoroelastomer compounds tested sealing 200 psi nitrogen at 10% squeeze. The o-rings tested were "dry"; they were not treated or immersed in any fluid after postcure. The results indicate a significant difference in their ability to seal at low temperature. All the FKM o-rings sealed well at room temperature and continued to hold pressure to below -20 [ degrees ] C. However, between -20 [ degrees ] C and -30 [ degrees ] C, high fluorine types 6191 and B600 developed leakage. When leaks occurred, they generally increased rapidly from 0.1 through the limit of the flow meter (10 SCCM). The graphs show this rapid rise from 0 to 10 SCCM leakage in a very short time and temperature span. Testing showed E60C leaked at -31 [ degrees ] C, B70 leaked at -33 [ degrees ] C and GLT leaked at -44 [ degrees ] C.
Traditional vs. non-traditional low temperature tests
When comparing results from the various traditional lab tests (TR-10, DSC, brittle point) to the results from the o-ring tester, the following trends were observed: (All results are best to worst):
* Glass transition tests: GLT [ arrow ] GFLT [ arrow ] B70 [ arrow ] E60C [ arrow ] B600 [ arrow ] 6191
* Brittle point tests: GLT [ arrow ] GFLT [ arrow ] B600 = B70 [ arrow ] 6191 [ arrow ] E60C
* Low temperature o-ring tests: GLT [ arrow ] GFLT [ arrow ] B70 [ arrow ] E60C [ arrow ] B600 [ arrow ] 6191
The trends observed in the low temperature o-ring test are the same as those noted in glass transition tests such as TR-10 and DSC as seen in table 3. However, in every case, the o-ring seals at a lower temperature than TR-10 and DSC data indicate. Examining the difference between a 5
SCCM leak and the TR-10 results in table 3, the o-ring seals at an average 13 [ degrees ] C lower temperature than the TR-10 result. The trends noted in the brittle point test do not always match the trends seen in the low temperature o-ring test.
Effect of deflection ("squeeze") on o-ring sealing
Another factor which can affect the low temperature sealing of fluoroelastomers is the percent deflection or "squeeze" that the o-ring is subjected to in its application. In figure 3 results are shown for the six fluoroelastomer compounds when "dry" o-rings were tested with 10%, 20% and 30% squeeze. The results indicate there is a minor positive benefit with more o-ring squeeze when attempting to maximize low temperature sealing. For example, at a 10% squeeze, E60C showed a 5 SCCM leak at -31 [ degrees ] C vs. -35 [ degrees ] C at 30% squeeze.
However, in many applications, o-rings normally operate with only 10-20% squeeze due to other design considerations.
Effect of fluid immersion on o-ring sealing
Still another factor which can affect low temperature sealing is the plasticizing effect the fluid being sealed can have on the fluoroelastomer. An example of this is the effect gasoline immersion can have on fluoroelastomer o-rings, such as those presently used in automotive fuel injectors and fuel line quick-connects.
To simulate this, the fluoroelastomer compounds from table 2 were immersed one week at room temperature in 91 octane unleaded gasoline and one week in Reference Fuel C. The results of these tests are shown in figure 4. The results show the swelling effect of the gasoline and Fuel C have on the fluoroelastomer compounds. Without exception an improvement in sealing ability at lower temperatures was noted.
Shown in table 4 are the volume swell and o-ring leakage results of the fluoroelastomers tested in 91 octane gasoline and in Ref. Fuel C. There appears to be a trend evident whereby the FKM compound that swells the most also exhibits the most improvement in low temperature sealing. High fluorine FKM-6191 exhibited the lowest volume swells and the least low temperature improvement while lower fluorine containing polymers showed higher volume swell and greater low temperature improvement.
Low temperature/low hardness o-ring results - phase 2
Another factor that can be varied in fluoroelastomer o-rings is the hardness of the compound. Most fluoroelastomer o-rings come in a hardness range of 70-80A durometer as shown in table 2. It has been suggested that lower hardness compounds may be "softer" at low temperatures and therefore seal better. To explore how lower hardness fluoroelastomer o-rings seal at low temperatures, three of the polymer types used in the 75 durometer study were reformulated to be 60 durometer. The formulations and properties of these compounds are shown in table 5. The compounds are based on FKM-E60C, FKM-GLT and FKM-GFLT. It should be noted that in order to obtain low hardness, the amount of reinforcing filler was reduced in these formulations. As a result, these compounds were not only softer, but also showed lower modulus and tensile strength than the 75 durometer compounds.
Shown in figure 5 are the sealing results with the lower hardness compounds. In this specific example, a modest improvement of 2-4 [ degrees ] C in low temperature sealing was observed with the lower hardness compound as compared to the standard 75 durometer FKM.
Fuel/methanol resistance and low temperature sealing
One of the most challenging sealing applications under development in the automotive industry are o-rings for vehicles which can use any possible blend of unleaded gasoline with methanol. This type of fuel is commonly referred to as "flex fuel" and places new demands on many types of polymers, including fluoroelastomers. In addition to flex fuel resistance, it is desirable for seals such as fuel injector and quick connect o-rings to have at least -40 [ degrees ] C static sealing capability.
By combining data from figure 4 of this article and volume swell data previously published (refs. 15 and 16), it is possible to generate a graph which reviews both flex fuels volume swells and low temperature sealing of the fluoroelastomer compounds shown on table 2. Such a graph is shown in figure 6.
The upper half of the chart in figure 6 depicts volume swells of the six fluoroelastomers in five test fuels: Fuel C, 85/15 Fuel C/methanol, 70/30 Fuel C/methanol, 50/50 Fuel C/methanol and 15/85 Fuel C/methanol. The bottom half of the chart shows the temperature where each of the FKM compounds experienced a 5 SCCM leak in the o-ring test. Note the o-rings were conditioned one week in Ref. Fuel C at room temperature which is also representative of the volume swell test conditions. In examining the data in this chart, one can evaluate the tradeoffs in obtaining a fluoroelastomer that is resistant to any Fuel C/methanol blend and effective at low temperatures. The data indicate that FKM-E60C, B70 and GLT have high volume swells when high methanol content is present, thus they are not selected for this type of service. The higher fluorine types, FKM-6191, B600 and GFLT, show a much improved volume swell profile in all the test fuels, but their low temperature sealing ability varies. Of the fluoroelastomers evaluated, the best combination of low volume swell in the Fuel C/methanol blends and good low temperature sealing ability at -40 [ degrees ] C is exhibited by GFLT.
Summary and conclusions
* The ability for o-rings to seal at low temperature follow those trends observed in glass transition (DSC) and temperature retraction (TR-10) testing. They do not follow trends seen in brittleness tests. * The low temperature sealing performance of the fluoroelastomer o-rings tested were as follows when tested without fluid or heat age exposure (best-worst): GLT [ arrow ] GFLT [ arrow ] B70 [ arrow ] E60C [ arrow ] B600 [ arrow ] 6191. * Following immersion in Reference Fuel C or gasoline, all fluoroelastomer types exhibited improved sealing at low temperature and follow the above noted trend. * The degree of o-ring compression (squeeze) appears to exhibit a minor positive effect on low temperature sealing above 10% deflection. * O-rings compounded for low hardness show a moderate positive improvement in low temperature sealing when compared to a standard 75 durometer fluroelastomer. * The ability for fluoroelastomer o-rings to seal at low temperature, as evaluated by the test method and apparatus discussed in this article, is predominately dependent upon the type of base polymer.
Table 1- description of the various fluoroelastomers evaluated
Polymer Percent Type of Cure designation fluorine monomers system FKM-E60C 66 VF/HFP Bisphenol FKM-B70 66 VF/HFP/TFE Bisphenol FKM-B600 68 VF/HFP/TFE Bisphenol FKM-6191 70 VF/HFP/TFE Bisphenol FKM-GLT 65 VF/TFE/PMVE/CS Peroxide FKM-GFLT 67 VF/TFE/PMVE/CS Peroxide
VF - vinylidene flouride; HFP - hexafluoropropylene; TFE - tetrafluoroethylene; PMVE - perfluoromethylvinylether; CS - proprietary cure site monomer
Table 4 - effect of fuels on volume swell and low temperature leakage
FKM E60C B600 6191 GLT GFLT % Fluorine 66 68 70 65 67
Volume swells - 168 hrs. @ 23 [degrees] C
91 Octane unleaded, % 1.5 1.6 1.4 1.3 2.7 2.1 Fuel C, % 3.7 3.6 3.1 2.0 6.6 4.3
Low temperature o-ring sealing test - 5 SCCM leak temperature
"Dry" o-rings -31 -33 -26 -25 -44 -36 168 hrs./unleaded gas -34 -37 -33 -27 -47 -41 168 hrsa./Ref. Fuel C -39 -40 -37 -29 -54 -46
Table 5 - a comparison of low hardness FKM compounds - formulas and data
E60C-60 GLT-60 GFLT-60 FKM-E60C* 100 - - FKM-GLT - 100 - FKM-GFLT - - 100 Magnesium oxide - high activity 3 - - Calcium hydroxide 6 3 3 MT black (N990) 10 10 10 Process aid no. 3 1 1 1 Triallysocyanurate (TAIC) - 3 3 Organic peroxide** - 3 3
(*) Contains incorporated bisphenol cure system (**) 45% active 2,5-dimethyl1-2,5 bis (t-butyl-peroxy) hexane
Vulcanizate conditions: Press cure: 10 min. @ 177 [degrees] C Post cure: 24 hurs. @ 232 [degrees] C Physical properties at R.T. - original
100% Modulus, MPa 2.8 2.8 3.2 Tensile strength, MPa 10.5 10.4 6.8 Elongation @ break, % 233 205 147 Hardness, durometer A, pts. 61 57 60
Compression set, method B, o-rings
70 hrs. @ 150 [degrees] C, % 10 21 24 70 hrs. @ 200 [degrees] C, % 19 32 34
Low temperature properties
DSG, Tg, initial, [degrees] C -13.5 -32.1 -27.0 , Inflection , [degrees] C -15.3 -31.0 -24.3 , Final, [degrees] C -11.1 -28.2 -20.0 Temp. retraction, -17.2 -30.6 -25.3
TR-10, [degrees] C
Brittle-point, C -23 -49 -41 [Tabular Date 2 & 3 Omitted]
PHOTO : Figure 1 - low temperature o-ring tester
PHOTO : Figure 2 - FKM o-rings (10% o-ring squeeze)
PHOTO : Figure 3 - low temperature sealing of FKM o-rings (leakage at 10%, 20%, 30% squeeze)
PHOTO : Figure 4 - low temperature sealing of FKM o-rings (after fuel immersion - 168 hrs. at 23 [degrees] C - 10% o-ring squeeze)
PHOTO : Figure 5 - low temperature sealing of low hardness FKM o-rings (standard 75 vs. 60 durometer compounds - 10% o-ring squeeze)
PHOTO : Figure 6 - flex fuel resistance vs. low temperature sealing of FKM o-rings (aged 168 hrs./23 [degrees] C - 10% o-ring squeeze)
K. Nagdi, "Correlation between laboratory tests and service performance of elastomeric seals at low temperature," Kautschuk+Gummi-Kunststoffe 41, Jahrgang, Nr. July, 1988. D.L. Hertz Jr., "O-rings for low pressure service," Machine Design, April 12, 1979. SAE AIR 1387A, "Designing with elastomers for use at low temperatures," Society of Automotive Engineers, 1985. J.D. MacLachlan, "Fluorocarbon elastomer: A technical review," Polym.-Plast. Technol.Eng., 11(1), 41-53, 1978. J.M. Wilson, W.S. MacLaughlin Jr. and R.G. Peck, Jr., "New fluoroelastomer types for automotive sealing applications," SAE Paper #860496, February 24-28, 1986. R.E. Knox, "Viton: A high performance fluorocarbon elastomer for use in hostile environments," SAE Paper #770867, September 26-30, 1977. ASTM D 2137, "Standard test methods for rubber property - brittleness point of flexible polymers and coated fabrics," 1989. ASTM D 1329 "Standard test method for evaluating rubber property-retraction at lower temperatures (TR test)," 1988. J.L. Laird and G. Liolios, "Thermal analysis techniques for the rubber industry," presented at the Rubber Division meeting, American Chemical Society, Detroit, October 17-20, 1989. SAE AS 568A, "Aerospace size standard for O-rings," 1984. SAE AMS 7273B, "Rings, sealing, fluorosilicone rubber," 1984. Private communication - AC Rochester Division, General Motors Corp., 1986. SAE ARP 1231, "Gland design, elastomeric o-ring seals, general consideration," 1973. SAE ARP 1232A, "Gland design, elastomeric o-rings seals, static radial," 1977. R.D. Stevens, T.L. Pugh, D.I. Tabb, "New peroxide curable fluoroelastomer developments," presented at the Rubber Division Meeting, American Chemical Society, Los Angeles, April 23-25, 1985. R.D. Stevens and E.W Thomas, "Fluoroelastomer developments for automotive fuel systems," SAE Paper #880022, February 29-March 4, 1988.
|Printer friendly Cite/link Email Feedback|
|Author:||Revolta, William N.K.|
|Date:||Oct 1, 1991|
|Previous Article:||Markets, news.|
|Next Article:||Long term sealing ability of butyl o-rings.|