Dear colleagues
the problem consists of a block in 2-D which measures 0.1x0.02m thick, with an oxide layer patch 2.5 µm thick . It is modeled in SI units. A temperature shock is applied on the oxide layer patch. When the whole system is cooling down it behaves quite normal until the temperature differences between the elements have reached only a few degrees Celsius. Then stresses start to rise extremely to a level of several GPa without ending tendency. The elements seem to blow up. It is not a boundary conditions issue because the system can expand – contract almost freely. Also the material data has been checked thoroughly.
Is there any idea why this happens? The system has also been modeled with axisym elements with the radius of 0.05m. The result is almost the same. Any ideas appreciated… Thanks a lot!
Hans, as you are pointing out, it is a fast quenching problem, did you try checking the max. temperature increment option in coupled temperature displacement step. You might want to control the delta_T_max value such that your solution is stable and does not take large temperature increments by making the strain increments to diverge. also check and play with the max. strain increment error tolerance value. hope it helps! best,
Try to find out which strains are creating these stresses (look at different components and different stresses like LE, THE, PE etc.). If the temp. difference is small, these should not be thermal strains.
If you upload your model and/or some results plots showing the distortion, it might be easier to help.
Thanks for the suggestions. It seems that the maximum temperature increment lead to the divergence.
- Badges
- Science topic