I trust all is well.
I have a question related to MCalibration to calibrate TN-model for PEEK/CF composite as part of my PhD work.
I am trying to model PEEK/CF short and long behavior using TN-model to use later to predict Spring-in and warpage in my FE models in ABAQUS.
I have PEKK/CF samples consolidated at two different temperatures (350C and 400C) at different hold-times (15min, 60min and 120min). I have ran relaxation tests (relaxation time=10min) at temperatures of 30,50,70,90,110,130,150,170,190,210,230,250,270,290 and 310C using DMA (three-point bending fixture). So in total, I have 90 relaxation curves and I’ve constructed one Grand Master Curve for PEKK/CF that takes into account all the variations of consolidation temperature and time.
In addition to the 90 relaxation test results and the single Master Curve, I have stress-strain data (from three-point bending test) for PEKK/CF at 24,60,100,140,180,220,260 and 300C, all at the same strain rate.
So far, I’ve tried inputting all stress-strain data with the Grand Master Curve to calibrate TN-model but that didn’t result in a good fit. I’ve also tried using the stress-strain curve (at 100C) and the relaxation curve of a sample (consolidated at 350C for 15min ) at 110C as input and ran calibration for an hour but I got NMAD of 49.
I guess I am not sure what are the suitable curves out of the stress-strain data and relaxation data (Grand Master curve or the other 90 relaxation curves) that I should input in MCalbration to get a good fit with TN-model especially with this amount of experimental data.
Should I use them all and try another model? Or use a few selected curves from the experimental data and run calibration for TNM for longer hours?
Should I use The Grand Master Curve or the 90 single relaxation curves for the long-term behavior calibration?
Is it correct to even input the short-term data with the long-term data given that they are at different time scales and if not how do I combine them together to calibrate the model?
Your recommendation and guidance is much appreciated.