thank you so much for your ideas!
1. The absolut value function was only used because otherwise NX would not have accepted the formula. Perhaps because of the exponent, NX was afraid that the argument of the power function might be negative ;-). In practice, however, as you have already mentioned, this is irrelevant since the wall temperatures must always be higher than 184.07 ° C.
2. I tried to switch to micro-Watt, mm and Celsius. Funnily, the numerical values of the heat transfer coefficient and the heat flux have not changed with respect to W, m, Celsius ;-). Unfortunately that did not help.
I don't know exactly which file contains the solution deck. I found a file called INPF. There is a table called Convection to Enviroment(1) obviously with values for the the wall temperature and the corresponding heat transfer coefficient. These values are correct.
I used a formula field, please refer to the picture below:
Thank you so much again, best regards
ok, then I'm out now.
Last hint: I've found in your input desk, Line 1035:
$ ----- $ Units $ ----- $ PARAM UNITS 5 1.000000E+03 1.000000E+03 1.000000E+00 -2.731500E+02 $
Maybe there is an unit conversion. Check for desciption of NX Thermal/Flow input desk to understand the values.
"INPF" looks like an input desk.
And you were right, of course W/m² is the same as micro-W/mm². It was an idea of an INTERNAL ERROR of SimCenter in transferring values into input desk of NX Thermal/Flow. But you did find your table as expected. And I made a mistake as I thought you use a real formula, sorry.
Now I'm done.
Best wishes, Michael
I promised to keep you informed.
The solution is that Thermal / Flow expects the mean temperature of the wall and the fluid. If one uses a table with the mean temperatures instead of the wall temperatures, Thermal / Flow calculates correctly using both Convection to Enviroment and Thermal Coupling.
NX 12 gives you the option to choose between mean temperature and wall temperature when using Thermal Coupling. That does not work. In both cases the mean temperature must be used.
Many thanks to the colleagues of HTS / RUAG Coswig, who have solved the riddle.