Dear Mr. Bergström,
does anyone know or have the same problem like me and know how to solve it? Im testing my seals directly via an uniaxial tension test, where i put the rings on cylindrical mounts directly. Reason is that I don't want to manufacture specific test specimen but just use the product itsself. I have already managed to get reproducible results in strain/stress even for different O-Ring sizes of the same material.
The only problem i haven't solved so far is the uncertainity of the initial moduli during cyclic loading (unloading going to stress=0 so far). It's not possible to put the ring on the mounts without beeing prestressed and at the same time also beeing pulled straight to directly measure material behaviour. This means in stress curve we can see a exponential increasing when the ring is beeing pulled straight but then it switches to material behavior.
Simply defining an offset stress of e.g. 1N somehow isn't sufficient because forces differ depending on specimen. Some rings are already straight at this point and some aren't. The other thing is starting at a higher stress skips my initial modulus right? I was also thinkting about unloading to a point nothing near to the specimen getting loose on the mounts.
I can somehow prepare the data and cut it at the inflection points where pulling the seal from loose to straight ends and material behavior begins but thats not the best way i think. A sketch is attached to better depict the situation.
I would really appreciate if anyone can help me with that and/or has experienced the same issues and solved it (in a better way).
Thanks a lot!