About heat dissipation

Hi SimScale super user!
I’m Simulation two LCoS models
One is the original named RDP370FR, the other is RDP370F_J
There is only one difference between the two models.
That is RDP370F_J has a hole. And It’s size is same as active area of RDP370FR die active area.
Both models are in the same condition.
Generally, shouldn’t the RDP370F_J’s die temperature be lower than RDP370FR?
However, Result of the simulation, RDP370F_J’s is about 2~3 degrees higher.
Am I set the properties wrong? or not, I really don’t know what I missed.
Please review it and reply if you have any questions.

Best regards,
Jake

Hi jake_lee

when i checked your simulation I noticed that the effect of the hole seems to be not really significant.

Temprature and velocity vector for LCoS with hole

Temprature and velocity vector for LCoS without hole

My idea is that the hole is not big enough so that the air can’t really exchange with the air at the surface of the LCoS. Maybe you can increase the hole in the direction of the airflow.

Another option to see if that concept works you can increase the air stream, with that the influence of the hole should get more significant.

2 Likes

Thank you so much!
I’ll review what you said. Than, I have some questions.

  1. There is no problem about my setting to compare them. Is that right?
  2. I understand the problem may be that the temperature of the panel does not decrease because the hole is small. But no matter how small the hole is, shouldn’t the temperature be lowered even a small size? Do you know why the RDP370F_J temperature increase about 2~3 degrees?

Thank you,

Hi jake_lee,

  1. from what I’ve seen you have only done geometry changes therefore you can compare the results.

  2. 2-3 degrees seem very small therefore it still might be a numerical noise. From what I’ve seen of the simulation which has been performed after that test. Please have a look that the simulation is fully converged . Especially the temperatures are still rising and have not converged to a steady result yet.

Best regards
Sebastian