Thermal test

[MNS747]

Hi
appologies if i am repeating.
I am trying to carryout a warm and cool test.
After specifying boundary conditions as 100 degrees and predefined fields as 21 degrees in step 1 and run for 10 minutes I can see the temperautre distributions across my model. Now I want to perform a cooling test for same 10 minutes in step 2. The only way I am doing is by changing the boundary conditions to 21 degrees at the same place where I applied 100 degrees previously in step 1. Now the output I am getting is a sudden drop in temperautre at the place from 100 to 21 (that ofcourse is not right). there should be a gradual decrease in the temperautre from 100 to 21. Can any one help me how this can be done. I have tried giving amplitudes but i am getting errors. Do I have to give any formula in analytical fields for thermal distribution in boudary condition dialog box? Sorry If my question is confusing.

 

[corus]

Perhaps the boundary conition is being applied instantly rather than being ramped over the time period? Seethe *step option to change amplitude to ramp to apply the load/displacement linearly over the time step

corus

 

[MrMyers]

What are you trying to model? I find thermal BC's as too 'rigid' and prefer to use SFILM (aninteraction) which, in your case, would give you the gradual increase/degrease.

The hard part is figuring out what SFILM coeficcient to use. It can e calculated a number of ways, or simply find ball park figures from a good paper or other reference.

 

[MNS747]

thanks corus and MrMyers for your replies.
Corus ...I did switched on the Ramp instead of instateneous in step2 and I got the time dependant decrease from 10 to 20 minutes. where as I expect it to come down exponentially from 100 to 21 from 10th minute to 20th minute. ( i have also used the decay function in amplitudes. no success)

MrMyers...I am trying to work out the spindle of CNC machine and applying temperature on the bearings. from start to 10th minute it goes fine as an exponential ramp. But for cooling down it doesnt despite I have stopped the first boundary condition of 100 degrees by modifying it in second step. Thanks for the Surface Film interaction idea and I will try it.

I have tried a lot ways to do it ..used amplitudes as tabular and time decay, modifying BCs, sometimes played with analytical fields for distribution. Any other idea you come up by reading will be appreciated.