Discussion Groups

Simulation CFD

New Member
Posts: 1
Registered: ‎11-15-2013

Cannot get thermal distribution

100 Views, 1 Replies
11-15-2013 02:42 PM

I am kind of new to Autodesk tools. I want to simulate thermal distribution over a smartphone and we built very simple model for that.


I used Autodesk Inventor Pro to each component as a part (.ipt) and assemble them as a mold design file (.iam). Then I launched Simulation CFD and created a study pointing to the .iam file.


I am managed to assign materials for every components and specify the heat generation to IC chips.


Then I adopted automatic mesh size and clicked "solve". It solves within 1 minutes (it seems to be too fast... the simulation setup are shown in the right side of the figure) to give results. However, I am not able to see any thermal distribution on my device. Even the IC chips I assigned heat generation stay at 25C (the small blocks in figure are IC chips).


Since the smartphone is a closed case, I didn't specify any inlet and outlet air. Could this be the problem? I set the volume inside the device to be air.


Could anyone help? Thanks a lot.

Please use plain text.
Product Support
Posts: 934
Registered: ‎08-25-2011

Re: Cannot get thermal distribution

11-18-2013 01:16 AM in reply to: xqing



Do you have any reference to an ambient temperature anywhere? This is likely the issue. 

There are 2 approaches here.


1) Run with air around your device and let CFD calculate the air flow and heat transfer to it - you would need Boundary Conditions (BC's) on the top surface of the air, assuming yout phone is on a bench of somesort. I would use P=0 and T=ambient.

2) Use film coefficients - faster but less accurate - on the external surfaces of the material. Something like 5W/m2/K at ambient temp would be a good start


Typically the auto mesh will be too coarse - ideally you need to aim for 4-5 elements within air gaps and 2 elements through thin solids.

You could also consider running with air_solid (a material you would need to create yourself) as the airlow will be so minimal, you could assume conduction only and rule out any convection internally. This means we would not add a boundary layer to the fluid mesh, meaning a faster run time.


Hope that helps.


Kind regards,


Jon Wilde
Please use plain text.