Transient thermal analysis

Johan_S

I am trying to do a transient thermal analysis on a very simple model. It sort of simulates what will happen during the heating cycle of a thermal stress relieve.
A rectangular bar 100 x 100 x 500 mm.
A ramped temperature load on the one end.
Convective heat loss on the other end
No heat flux through the long sides.

The funnies I get are:
The temperature drops instead of increases and converges to 0 deg C in stead of 300 deg C.
The temperature oscillates near the heated end.

Any help will be appreciated.

DaveStupple

Johan,

Strange! It looks like it might be reading your temperature (specified in °C) in K. I changed the units here to K and added 273 to each temperature and the results are more sensible, I think.

disla

DaveStupple , you are Right.
Loads and Constrains table is reading temperature values in K when should be Cº.
It is clearly seen for time step t0 of solution

Victor

Wow, this is a fairly serious bug. Thanks a lot Johan_S, DaveStupple, and disla. It extends to some other features too. Here is the scope of what's affected:

Time dependent values defined by a table for:

• Temperature in thermal analysis with the internal solver
  
• Internal heat generation with the internal solver
  
• Pressure normal with the internal solver
  
• Pressure X,Y,Z with both the internal and CCX solvers
  

In those cases, it incorrectly uses the mksK unit (K, W/m^3, Pa) instead of the unit you specify in the table. So the workaround is to only use those units in tables.

It does not affect formulas, constant values, the time unit in tables, other loads or constraints, or temperature in mechanical analysis (for thermal stress).

Johan_S

Thanks for the prompt responses and the workaround.