Thermal DOF coupling (CP's on TEMP) are introduced in a restart of a thermal transient analysis having nonlinear (temperature dependent) properties. It doesn't look like ANSYS is honoring the couples. What's wrong?
CP`s will work in a transient thermal restart if the model is linear (material properties not a function of temperature). They will NOT work if the model is nonlinear (properties ARE temperature dependent). This is illustrated by attached input file test02.inp.
Workaround: USE CE`s instead of CP`s. This works when the thermal transient model is nonlinear (material properties are temperature dependent). Proof is provided with attached input file test03.inp.
Read the input files into an interactive ANSYS session. Note that each has a `toggle` parameter (nonlinear = 1 => run nonlinear properties).
Additional background (Knowledge Base 18088):
This is a non-standard use of the program. CP's will not work in a nonlinear analysis when introduced in a load step in the middle of a run (hence why the CP family of commands are noted in the doc as "PREP7" commands and not "SOLUTION" commands. [We do not, however, block this use because CP/CE's are one area that we give users enough rope to hang themselves in case they want to attempt to do something tricky (like this!).]
However, CE's will work (we account for the temperature (or displacement) discontinuity when they are introduced in the middle of a run).