Open moosebuild opened 2 years ago
Imported from idaholab/thm#164 by @andrsd at Nov 30, 2020 10:36AM MST
Note that this is a property of the heat conduction equation on a fixed-volume domain; thus, it doesn't have any solution in the case of THM (fixed-volume). The recommended response is to disallow temperature-dependent density in heat structures.
By Hansel, Joshua E on 2020-11-30T17:36:01 (imported from GitLab project)
Imported from idaholab/thm#164 by @andrsd at Nov 30, 2020 10:33AM MST
Bug Description
If the user specifies a temperature-dependent density for a heat structure, energy conservation is violated. This can be shown theoretically as well as numerically. The reason stems from the fixed control volume that does not expand with density - you don't even conserve mass when changing density.
Bug Impact
Energy balances are violated in this case, which can severely affect the accuracy of the solution.
Steps to Reproduce Bug
Run this input file:
Note
E_change_rel_err
is O(1) instead of O(-8).