Inventor Nastran Forum
Welcome to Autodesk’sInventor Nastran Forums. Share your knowledge, ask questions, and explore popular Inventor Nastran topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Linear Steady State Heat Transfer - CPU Cooling Fins

3 REPLIES 3
Reply
Message 1 of 4
Ruwan.B
100 Views, 3 Replies

Linear Steady State Heat Transfer - CPU Cooling Fins

Hi everyone.

I am trying to simulate a textbook problem that I performed the calculations for. See below:

RuwanB_0-1735387513478.png

RuwanB_1-1735387556572.png
For my simulation I am considering design 'A' only.

I have modelled the component as indicated above and performed a linear steady state heat transfer analysis with Aluminum 6061 as the selected material. I made sure to change the thermal conductivity to the specifications in the problem.

RuwanB_2-1735387911291.png


I applied a 75C temperature load to the bottom of the array as opposed to what the textbook problem indicated. This was to avoid a T2287 error.

Furthermore I changed the PCONV data in the NAS. file for both convection loads to form 2, which brought the solution closer to what I have calculated, but the results were still far from acceptable.

RuwanB_4-1735389032746.png


The issue I'm having is that the heat flux results are nowhere close to what hand calculations indicate. My calculations below:

RuwanB_3-1735388371339.png

The heat dissipation of the array would result in a heat flux value of:

Heat Flux = 113.857 W/0.022461 m^2 = 5 069 W/m^2 or 5.069 mW/mm^2

Taking the efficiency of the array as 100% for my simulation the heat dissipated by the array will then be 140.38 W, which results in a heat flux value of 6 250 W/m^2 or 6.250 mW/mm^2.

The resulting analysis shows 247 mW/mm^2, which is extremely high

RuwanB_5-1735389530519.png


Can anyone point me in the right direction here please.

Inventor Nastran 2025

 

 Thank you.

 

Labels (2)
3 REPLIES 3
Message 2 of 4
bdesiderato
in reply to: Ruwan.B

Hi,

 

Did you apply an initial condition temperature? Essentially your ambient temp. 

 

Additionally, I'm not sure what the problem is asking for. If you can explain what result you are looking for I may be able to help further. 

 

Typically you would set a heat generation load along with the ambient temp. condition and run the FEA, then the result would give you a maximum temperature. Why set the max temp on the bottom of the chip? If it gets too hot, you need to increase surface area via fins, or add a cooling fan to increase H value. 

 

Ensure you are using the correct material, namely using the correct thermal property values. 

 

 

Regards,

Ben 

 

bdesiderato_0-1735585759913.png

 

Message 3 of 4
Ruwan.B
in reply to: bdesiderato

Hi Ben

Thank you for your reply :slightly_smiling_face:

With a Linear Steady State Heat Transfer a load specifying "Initial Condition" is not allowed. You can only do this in a non-linear and transient environment. The problem is asking for the heat dissipation of the array. I did manual calculations indicating what values I am looking for. That is the screen shot of the calculation in the original thread. I'll paste it here again:

RuwanB_1-1735639917528.png

 

The heat dissipation of the array would result in a heat flux value of:

Heat Flux = 113.857 W/0.022461 m^2 = 5 069 W/m^2 or 5.069 mW/mm^2

Taking the efficiency of the array as 100% for my simulation the heat dissipated by the array will then be 140.38 W, which results in a heat flux value of 6 250 W/m^2 or 6.250 mW/mm^2.

And that is what I am looking for. A heat flux value in the range of 6.250 mW/mm^2 and not 246.829 mW/mm^2 as the analysis indicates.

Regarding applying the temperature at the bottom of the fin array... This has been done to avoid a T2287 error. It is not advisable to apply a temperature load and a convection load on the same face.

The correct thermal property values have been used, I have checked this along with the correct conversion of units.

Remember, the purpose of the analysis is not to improve on the design by increasing any cooling surface area or convection coefficient values. The purpose of this analysis is to validate manual calculations and see what is the percentage in error for this type of analysis.

Let me know if you have any other thoughts.

Inventor 2025

Thank you

Message 4 of 4
John_Holtz
in reply to: Ruwan.B

Hi @Ruwan.B 

 

The heat flux result of 247 mW/mm^2 is the heat flow in the solid. That has nothing to do with the 5.069 mW/mm^2 heat flux from the solid to the air that you calculated.

 

Can you provide the Inventor files? ( assembly files .iam and part files .ipt, whatever is used by your model) That will provide the answers more easily than trying to interpret the numbers in the Nastran file. For example,

  • Are you using temperature dependent convection coefficient? That would be the only time the PCONV needs to be changes for a different form.
  • Where does the area 0.022461 m^2  come from?
  • What is the total heat flow to the convection faces? That is what you want to compare to your hand calculation of 113.8 W. (Your heat flux of 113.8/0.022461 m^2  is an average heat flux from the solid to the air. You won't necessarily see that anywhere in the results. )

Out of curiosity, what is the hand calculated value of the temperature at the tip of the fin, and what is Nastran calculating? That will be a better indication if the results are reasonable or not.

 

John



John Holtz, P.E.

Global Product Support
Autodesk, Inc.


If not provided already, be sure to indicate the version of Inventor Nastran you are using!

"The knowledge you seek is at knowledge.autodesk.com" - Confucius :winking_face:

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report