Not exactly Sir, I got some idea to decompose to total strain into mechanical and thermal provided the condition is isothermal. However, in therm-mechanical type loading where both temperature and strain changes with time, i am still exploring the way to decompose the strain. If you come across any such literature let me know. Thanks
I would suggest the standard TMF-testing apporach: Measure the thermal strain by cycling under Force-Zero-control for a few cycles when the system is stable. Then, substract the mean thermal strain cycle from your measured total strain in the experiment. This method should work, independent from the means of strain measurement.