Elastomers are widely used as sealing materials due to their good elasticity and low cost. However, there are limitations to their use, such as they cannot be used at low temperatures or for very long periods of time, as the materials are subject to chemical aging. As they age, elastomers gradually lose their elasticity and their ability to recover elastically, which can lead to higher leakage rates than specified. In some applications, such as containers for radioactive waste, a service life of several decades is required, as it is not easy to replace seals. For this reason, an accelerated aging program of up to 5 years for O-Rings made of HNBR, EPDM and FKM was carried out.
A common problem in life prediction is the selection of the relevant expiration date criterion, which in ISO 11346 is a change of 5O% in the examined property, e.g. elongation in tension. For O-ring seals, compression set is commonly used as the life criterion, e.g. 57% (corresponding to 10% recoverable) or 85%. However, the O-Ring leakage rate is the only parameter that is directly related to the service life of the sealing system. Other material properties, even the compression set, can only be used as indicators and not as a basis for seal failure. Therefore, the entire component, i.e. the O-Ring, is aged, both uncompressed and compressed between flanges, to test the leakage rate of the aged seal. In addition, the determination of whether service life has been reached by leakage rate measurements is more easily related to testable properties such as compression set, hardness or viscoelastic loss factor as tested by Dynamic Mechanical Analysis (DMA). However, care should be taken to avoid the diffusion-limited oxidation (DLO) effect, which can lead to inhomogeneous aging and thus inaccurate aging data and life prediction.The DLO effect is related to oxygen partial pressure, specimen size, time, temperature and material (oxygen permeability). In addition to O-rings, 2mm thick film was aged for testing of material properties where O-ring construction is not suitable or where results would be inaccurate due to the DLO effect.
Results and discussion
1. Measurement with film material
1.1 Hardness
For HNBR and EPDM, the hardness is converted to Shore D hardness after aging for a period of time because of the large increase in hardness during aging. The correlation between the two hardnesses is determined by testing the Shore A hardness and the Shore D hardness of an unaged specimen. A Shore A hardness of 80 is equivalent to a Shore D hardness of 33.
The observed increase in hardness during aging for both HNBR and EPDM is due to the cross-linking reaction that occurs, the high polarity due to oxygen binding, and the loss of plasticizer for HNBR. For HNBR, cross-linking reactions occurring through alkyl, alkoxy or peroxy groups predominate. On the other hand, EPDM produces cross-linking (via monomer 3) and chain breaking (propylene fraction). Since there are more propylene links than ENB links, chain-breaking reactions dominate, although the propylene links are more resistant to oxidation than ENB links.The increase in hardness of EPDM may be due to the elevated polarity upon binding oxygen, such as the formation of ketone groups by β-breaking of the propylene links.
FKM shows essentially no change in hardness during aging.
2. Dynamic mechanical analysis
For EPDM and HNBR, the tanδ peak was shifted toward high temperature, indicating that the glass transition temperature (Tg) increased after aging. In addition, the tanδ peaks of HNBR and EPDM decreased after aging, indicating that the crosslink density increased and the molecular activity decreased. After aging at 150°C for 98 d, the butyl g of HNBR increased by 38 K, while the Tg of EPDM increased by 15 K. These conditions should be considered when applying near the low-temperature limiting temperature of the seals, as the Tg has a huge impact on the lowest working temperature of the material. On the other hand, FKM showed no significant change after aging at 150. C for 98d.
2. O-Ring Testing
2.1 Hardness distribution
Since the O-ring φ10mm) is much thicker than the film φ2mm), the aging is affected by the diffusion-limited oxidation (DLO) effect. This effect occurs if the rate of internal oxygen consumption is faster than the rate of diffusion inward from the surrounding air, resulting in less internal aging and uneven aging.
less, resulting in inhomogeneous aging, which distorts the aging data and overstates the expected life. Microhardness testing along the specimen section can reveal more information about the DLO effect. Since modulus and hardness have a non-linear relationship, the difference between the internal and external modulus of unevenly aged specimens is much greater HNBR aged at 125°C and 150°C shows less increase in hardness in the center of the specimen than near the surface of the specimen. This may be due to the DLO effect leading to inhomogeneous aging because less oxygen is available internally.The effect diminishes after 10 d, but is more pronounced at longer aging times. On the other hand, no hardness inhomogeneity was observed at the cross-section after 30 d of EPDM aging. However, non-uniform aging caused by the DLO effect was also observed after 101 d of aging at 150 °C.
2.2 Compressive stress relaxation (CSR)
The maximum effect can be observed from the results of the compressive stress relaxation test performed at 150°C, as shown in Figure 8. The relaxation is expressed as the ratio of the measured
force F to the initial force F. The ratio of the force F to the initial force F. F0 is defined as the force after 30 min of relaxation at the test temperature. Three specimens of each material were tested, showing good reproducibility. the test was terminated after 55 d, as EPDM reached the 1O% residual force criterion. The observed relaxations were due to physical effects (e.g., entanglement sliding, relaxation of suspended chain ends) and chemical reactions (e.g., oxidative chain breakage).FKM showed the advantages of its high temperature resistant material, with only a small relaxation of 75% after 55d. In contrast, EPDM had only 10% residual force in the same time. hNBR started to drop faster than EPDM, but the drop no longer varied after about 20d. This is due to the DLO effect resulting in less internal aging of the specimen, which retains more force.
2.3 Compression set (CS)
CS generally increases with aging time because cross-linking forms new chemical bonds that are in equilibrium with the compression state. And chain breakage to bond breakage, loss of recovery ability. Compared with the hardness and Tg. In the aging temperature is lower than 150 ℃, EPDM CS value also increases significantly. This is because the hardness and Tg by the aging process of chain breaking and cross-linking reactions in the opposite direction, only resulting in a slight change in the measured value, but the superposition of the two types of reactions lead to CS are increased. If the specimen shrinks during the aging process. The CS value can increase to more than 100%, e.g. due to cross-linking.
2.4 Measurement of leakage rate
The leakage rate is measured with an O-ring aged in the flange. A significant increase in the leakage rate is considered to be the end of seal life. Figure 11 shows
shows that the leakage rate of an aged O-Ring is slightly improved (reduced) compared to an unaged O-Ring. One reason for this is that the rubber adapts better to the roughness of the sealing surface as affected by time and temperature. In addition, the significant reduction in the leakage rate of the aged HNBR may be due to the increase in crosslink density during aging of the material, resulting in a decrease in the number of gas molecules penetrating through the material After 98 d of aging at 150°C, the O-rings remained leak-tight, even though the hardness and the DMA have shown a significant reduction in the values of the mechanical properties and the compression permanent deformation (cS) to 8O (HNBR) and 94 (EPDM) (it should be noted that the the hardness and DMA values shown were tested from film).After aging HNBR at 150°C, the O-rings showed a smaller aging effect compared to film, due to the apparent DLO effect in the O-rings. However, after aging for 184 d, the CS values for both HNBR and EPDM exceeded 100, indicating that the recovery height of the O-rings was less than the flange spacing of 7.5 mm. The EPDM O-rings were completely leaking after aging at 150°C for 184 d, indicating that vacuuming was not possible before measuring the leakage rate because there was air freely flowing between the O-rings and the flanges. For HNBR, the O-rings were non-leaking at 2O°C and 60°C. This may be due to the fact that the O-rings were still sticky. This may be due to the fact that the O-Ring is still adhering to the flange. This adhesion effect was evident when the compression-aged half O-rings were removed from between the two flat plates. one of the three HNBRO rings leaked when cooled to 30°C after aging at 150°C for 184d, but did not leak when retested at 2O°C. The O-ring was not leaking at 2O°C. It is possible that thermal contraction opens a pathway between the adherent surfaces. The measured HNBR leakage rate has a large dispersion, which explains why only one O-ring leaks, and it can be assumed that the O-ring is no longer suitable for further use after aging at 150°C for 184 d. The results show that the O-rings are no longer suitable for further use even if other properties have been compromised.
The results show that the O-rings are able to remain leak-free even though other properties have been greatly affected by aging. Similarly, even with a residual sealing force of only 1 N/cm, the O-Ring remains leak-tight. This shows that the leakage rate under static conditions is not as sensitive as the change in material properties, and that even a significant decrease in material properties does not have a significant detrimental effect. When correlating seal failure due to increased leakage with material properties, a more precise determination of the time to failure is required.
3. Conclusions
Aging tests of HNBR, EPDM and FKM O rings and films showed significant changes in properties during the aging process. tan(3), the viscoelastic loss factor of HNBR, decreased, and the hardness and increased significantly, probably due to the crosslinking reaction dominating the aging process, resulting in an increase in crosslink density. In addition, the diffusion effect of HNBRO shaped rings limited oxidative effects during aging at 125°C and 150. C, resulting in distorted compressive stress relaxation results. The hardness and DMA test results of EPDM when aged at 150°C were similar to those of HNBR, but with less variation. While the hardness of EPDM aged at 75. C, 100°C and 125°C changed slightly, the compression permanent deformation (Cs) increased significantly at all aging temperatures, probably due to the chain breaking and cross-linking reactions that increased the CS, and the hardness was affected by these aging reactions in the opposite direction. fkm had better aging resistance than the other materials, with no change in hardness, and higher residual force in the relaxation test at 150°C,. The compression permanent deformation does not change much after aging. The leakage rate test shows that even if other properties are greatly reduced, the O-ring remains leak-free. This suggests that the choice of end-of-life criteria has a significant impact on the predicted life, and that standard indicators involving material properties may not necessarily be relevant to component function such as static leakage rate.
