You are correct; when T=0, CAESAR II will not run that thermal load case. The temperature field can hold two values: 1) temperature and 2)the actual thermal strain. You use #2 for materials with no thermal strain data stored in the program. CAESAR II decides what value you are entering by the magnitude of the number. If the number is less than (the default value of)0.005, then it is strain, otherwise it is temperature. You can change this default value in the configuration file. So when T=0, that's strain and a strain of 0 creates no load, therefore no load case.
If I want to run a job at zero temperature, I would probably set T=0.1.
Your point about expansion range - if I have one temperature above ambient and a second temperature below ambient; yes, I guess the total stress range would be the sum of these two excursions from ambient. But I would probably still run the difference between these two operating sets. It is more direct and avoids any question regarding summation technique.
_________________________
Dave Diehl