I'm working on an Instron 5969 tester, and I'm trying to measure the residual strain in materials within their elastic zone. I'm using the test profile, which allows me to stretch the material a specific amount, hold it for a moment, release the force, and record the change in length. The problem I'm having, is when the load is removed (slowly and continuously), the machine seems to panic. It starts shaking, and the recorded values jump all over the place. I've tried multiple ramp rates (the rate at which the load is removed) to no avail. Any ideas on what else I could try?
It sounds like you have a PID control issue which will give you an unstable machine. I wouldn't recomend playing with the settings if you don't know what you are doing, but tweaking them may fix the problem. The below may fix your problem but it would probably be good for the PID values looked at.
If you are completely removing the load then you need to make sure your specimen is supported properly so the pull bars and the specimen can be treated as a solid bar, capable of through zero fatigue testing. As a test for this if you ramp from a positive load to a negative load, you should have a nice smooth line in both load and position/strain. If you have any slack it the load train you not have a smooth line in one or more of the transducer channels.
An alternative would be not to go all the way to zero. If you keep a small near zero load (x) on the specimen when it should be at zero load (depending on what you can control to) and ramp to and from x. Then as long as x is small enough and the specimen is elastic in that range the relative strains + the strain at x may equal the absolute strain. I have used this method with metals many times, but have not had much experience in testing plastics so cannot say or sure.
Hope that helps