Prediction of uniaxial tension and relaxation
I am currently working on a material model (Yeoh-3xBergstromBoyce, Abaqus PRF) to predict tensile tests and relaxations test.
It has a really good match for the tension test at different strain rates 3 1/min, 1 1/min and 0,1 1/m. But the relaxation behavior is absolutely wrong so I also included relaxation data to fit. Now my model has a quite good match to relaxation data but the prediction of the tensile test for small strain rates like 0,1 1/min has worsened in the rise at tension. The modeling for larger strain rates 1-3 1/min has become even better.
So I tried a lot of calibrations and I have always the same problem. Either the relaxation or the tension curve for small strain rates fits.
My suggestion is that the model can be fittet in different ways for reproducing tensile test at different strain rates. The parameters of every network could be changed for that. But I think for the relaxation the parameters of the networks can't be easily changed. Because Abaqus PRF=generelized Maxwellmodell=Pronyterm and in Prony every network is describing its own time decade.
Do I have wrong ideas or is it not possible to model both at the same time?