Filtern
Dokumenttyp
- Zeitschriftenartikel (2)
- Vortrag (2)
Sprache
- Englisch (4) (entfernen)
Schlagworte
- HNBR (4) (entfernen)
Organisationseinheit der BAM
- 3.4 Sicherheit von Lagerbehältern (4) (entfernen)
Eingeladener Vortrag
- nein (2)
Elastomeric O-ring seals are used in a wide range of applications due to their excellent elasticity. However, like all polymers, elastomers degrade under the influence of e.g. time, temperature, oxygen, radiation and mechanical stress. Especially in applications where a replacement of seals after a certain time is hard or almost impossible, it is important to know the seal lifetime. Therefore, an extensive ageing programme with elastomers made of EPDM, HNBR and other materials was started with the aim of determining suitable methods for accelerated ageing and lifetime prediction.
In order to determine the lifetime of polymeric parts, the time-temperature superposition principle is commonly used to shift property changes obtained by accelerated ageing at higher temperatures to lower temperatures. If the shift factors yield a straight line in an Arrhenius diagram, a corresponding activation energy can be determined. However, we have found that the shift factors and thus the predicted lifetime depends on the test method that yielded the shifted data. For example, the shift factor between 125 °C and 150 °C ageing temperature was roughly the same (5/5/4.5 respectively) for hardness, density and maximum of loss factor tan measured on HNBR, but different for elongation at break (8) and compression set (2.2, excluding DLO-affected data). A possible explanation might be that while the different oxidation reactions proceed with a fixed activation energy, they have differing impact on the measured properties. For example, hardness is lowered by chain scission reactions, and increases by crosslinking reactions during ageing. As usually both chain scission and crosslinking reactions occur during ageing, the measured hardness increase reflects only the net effect of both reaction types. On the other hand, compression set is influenced additively by both reaction types: chain scission leads to an increase of the remaining deformation, as broken chains lose their recovery potential, and crosslinking reactions during ageing fix the compressed geometry, which also leads to less recovery. Thus, compression set reflects the total number of changes in the network and shows degradation effects much faster, which results in lower shift factors and lower activation energy. This phenomenon was observed for EPDM as well.
In order to verify our hypotheses and to gain further insights, temperature-dependent oxygen consumption measurements are currently being performed and results will be presented at the conference. The shift factors and activation energy determined by oxygen consumption measurements are expected to be close to the values for compression set.
Determining a suitable and reliable end-of-lifetime criterion for O-ring seals is an important issue for long-term seal applications. Therefore, seal failure of ethylene propylene diene rubber (EPDM) and hydrogenated nitrile butadiene rubber (HNBR) O-rings aged in the compressed state at 125 °C and at 150 °C for up to 1.5 years was analyzed and investigated under static conditions, using both non-lubricated and lubricated seals. Changes of the material properties were analyzed with dynamic-mechanical analysis and permeability experiments. Indenter modulus measurements were used to investigate DLO effects. It became clear that O-rings can remain leak-tight under static conditions even when material properties have already degraded considerably, especially when adhesion effects are encountered. As a feasible and reliable end-of-lifetime criterion for O-ring seals under static conditions should include a safety margin for slight dimensional changes, a modified leakage test involving a small and rapid partial decompression of the seal was introduced that enabled determining a more realistic but still conservative end-of-lifetime criterion for an EPDM seal.
At BAM, which is a federal institute for materials research and testing in Germany, it is one of our tasks to evaluate the safety of casks designed for transport and/or storage of radioactive material. This includes assessment of elastomeric seals applied in the casks. One of the aims is to identify a suitable method for estimating the service lifetime of the seals with regard to the requirements for long-term safety (40 years and more) of the containers. Therefore, we started an accelerated ageing programme with selected rubbers often used for seals (HNBR, EPDM and FKM) which are aged at four different temperatures (75 °C, 100 °C, 125 °C and 150 °C) for up to five years. In order to assess sealability, O-rings were aged in compression by 25 % (corresponding to the compression during service) between plates as well as in flanges that allow leakage rate measurements. For comparison, uncompressed O-rings were aged as well. In order to understand the underlying ageing mechanisms in each material, material properties such as hardness, density and tensile properties were examined. Additionally, compression set (CS), which represents the recovery behaviour of a seal after release from compression, is measured. For obtaining results closely related to practical conditions, O-rings with a full-scale cord diameter of 10 mm were aged. However, this set-up can lead to heterogeneous ageing caused by diffusion-limited oxidation (DLO) effects especially for HNBR, resulting in distorted bulk properties such as compression set. However, if DLO-affected data is excluded, extrapolations of CS data are possible using time-temperature shifts.
For selecting an appropriate end-of-lifetime criterion, leakage rate measurements were performed, since leakage rate is the only characteristic directly correlated to the performance of the sealing system. A significant increase in leakage rate was considered as the end of the lifetime. However, the O-rings remained leak tight under static conditions even though material properties had already degraded strongly. For this reason, a modified, more demanding leakage test involving a fast small decompression of the seal was developed that allowed determining a more conservative end-of-lifetime criterion with a safety margin for EPDM seals. FKM, which is a very heat and oxidation resistant material, exhibited only little degradation, even though it had the smallest activation energy.
O-rings made of HNBR, EPDM and FKM were aged in the compressed and uncompressed state at 150 °C, 125 °C, 100 °C, 75 °C, 60 °C and 23 °C for aging times of up to five years. Hardness was measured and increased with aging time and temperature for HNBR and EPDM, but it remained practically constant for FKM. Indenter modulus measurements were performed on the lateral O-ring surface (that was free of DLO effects) to assess an influence of the compression during aging, but none was detected. The equilibrium compression set (CS) exhibited faster and stronger degradation than hardness and was used for lifetime predictions using the time-temperature superposition (TTS) principle. With an end-of-lifetime criterion of 70 % CS, lifetimes of 4.5 years, 50 years and 526 years at 75 °C were estimated for HNBR, EPDM and FKM, respectively. The activation energies derived from an Arrhenius plot of the shift factors from the TTS were 85 kJ/mol, 99 kJ/mol and 78 kJ/mol for HNBR, EPDM and FKM, respectively, revealing that a higher activation energy does not necessarily mean that the material has a higher lifetime at lower temperatures. Furthermore, the measured lifetime of EPDM O-rings at 100 °C (5 years) was compared to that predicted on the basis of the lifetime at 150 °C as well as 125 °C using the corresponding shift factors. The error of the prediction was only ± 4 %. However, this precise prediction could only be achieved using the five-year long-term aging data. When using only data from aging times up to 0.5 years and 2 years, the lifetime of EPDM O-rings at 100 °C was underestimated by 31 % and 22 %, respectively.