AnsweredAssumed Answered

problem with simplified electronics thermal model

Question asked by Richard Miner on Apr 27, 2016
Latest reply on Jul 18, 2020 by Bill Dempsey

Like • Show 0 Likes0
Comment • 5

Hi All,

New user here both with Solidworks Flow Simulation (2016) and this forum. I've modeled a simple resistor in flow simulation as follows:

1. materials for three components - conductor (Nichrome); conductor package material (Alumina); heatsink (aluminum)

2. Heat Source applied to the conductor (1 Watt)

3. Contact resistance between components - no resistance

3. Goal - Global Goal Max Temperature

4. Mesh - Global Mesh (moved level of initial mesh to 4); Local Mesh (Regional cuboid, refining cells at 3 for all three choices, all 3 advanced refinements are set at 2)

My issue deals with 'verification'. I want to see if Flow Simulation is at least 'ball park'. I've run the model several times using different mesh refinements, different materials, with and without contact resistances, etc. In all cases my max temperature is 73C +/- 3C on the heat sink. In the lab I applied exactly 1 watt to the resistor and waited for steady state. Then I used a very good (and well calibrated) thermal camera to measure the max temp of the heatsink...which was 53C. As you can see my results aren't even close.

Not sure where to go from here. Any suggestions would be welcome.

Thanks,

rich

Attachments

resistor_04-27-16.zip395.5 KB

No one else has this question

Outcomes

Amit Katz Apr 27, 2016 1:48 PM

If you want to repeat your results from the lab, you need to set up your simulation likewise. I'm assuming the heatsink in your lab isn't floating in air, correct? Make sure you replicate your experimental conditions, including the temperature of the ambient air. Also when simulating natural convection you must make sure that your enable gravity in your calculations, it's in the "general settings" menu.

Actions

Richard Miner @ Amit Katz on Apr 27, 2016 3:15 PM

Hi Amit,

Thanks for the response. yes, the resistor is suspended in air and the ambient temp is 25C, also I did invoke gravity in the -y direction with the wizard upon set-up. I would be happy to 'pack and go' and send you the model for a quick review.

Thanks!

rich

Actions

Richard Miner @ Richard Miner on Apr 27, 2016 3:47 PM

Hi Amit,

I edited my first post to include a zipped pack-and-go file.

thanks,

rich

Actions

Amit Katz Apr 28, 2016 8:28 AM

I see, well that is odd behavior. How does the temperature of the hot cylinder compare with your lab data? Are the temperatures at the tips away from the heat sink comparable? Are they close to the 87C you see in the simulation? This is a fairly simple model so I'm not sure where the error could be coming from, other than maybe in the lab you weren't running at 1 W.

Actions

Bill Dempsey Jul 18, 2020 9:23 PM

Richard

I know you posted this years ago but I just saw this... I just ran your model in SWX2018 and it comes in pretty close to your measured. I am posting really for anyone else wishing to examine the same kind of problem.

Potential errors on measurement are:

1) Emissivity of IR camera not correct and thus misreads. You have to correlate with a temp sensor which you might know to do but others may not

2) Airflow is generally never "zero" in a lab unless you are in a closed chamber.

3) Heat source via leg is not correct. Add volumetric source to main body. You can see in the attached pictures that the legs are dissipating heat

4) Make sure radiation is turned on if "zero" airflow as at your temps it will dominate over local convection cooling.

Regards

Bill

Actions

5 Replies

Amit Katz Apr 27, 2016 1:48 PM
Mark Correct
Correct Answer
If you want to repeat your results from the lab, you need to set up your simulation likewise. I'm assuming the heatsink in your lab isn't floating in air, correct? Make sure you replicate your experimental conditions, including the temperature of the ambient air. Also when simulating natural convection you must make sure that your enable gravity in your calculations, it's in the "general settings" menu.
Like • Show 0 Likes0
Actions
- Richard Miner @ Amit Katz on Apr 27, 2016 3:15 PM
  Mark Correct
  Correct Answer
  Hi Amit,
  Thanks for the response. yes, the resistor is suspended in air and the ambient temp is 25C, also I did invoke gravity in the -y direction with the wizard upon set-up. I would be happy to 'pack and go' and send you the model for a quick review.
  
  Thanks!
  
  rich
  Like • Show 0 Likes0
  Actions
  - Richard Miner @ Richard Miner on Apr 27, 2016 3:47 PM
    Mark Correct
    Correct Answer
    Hi Amit,
    I edited my first post to include a zipped pack-and-go file.
    
    thanks,
    
    rich
    Like • Show 0 Likes0
    Actions
Amit Katz Apr 28, 2016 8:28 AM
Mark Correct
Correct Answer
I see, well that is odd behavior. How does the temperature of the hot cylinder compare with your lab data? Are the temperatures at the tips away from the heat sink comparable? Are they close to the 87C you see in the simulation? This is a fairly simple model so I'm not sure where the error could be coming from, other than maybe in the lab you weren't running at 1 W.
Like • Show 0 Likes0
Actions
Bill Dempsey Jul 18, 2020 9:23 PM
Mark Correct
Correct Answer
Richard
I know you posted this years ago but I just saw this... I just ran your model in SWX2018 and it comes in pretty close to your measured. I am posting really for anyone else wishing to examine the same kind of problem.

Potential errors on measurement are:
1) Emissivity of IR camera not correct and thus misreads. You have to correlate with a temp sensor which you might know to do but others may not
2) Airflow is generally never "zero" in a lab unless you are in a closed chamber.
3) Heat source via leg is not correct. Add volumetric source to main body. You can see in the attached pictures that the legs are dissipating heat
4) Make sure radiation is turned on if "zero" airflow as at your temps it will dominate over local convection cooling.

Regards
Bill
Like • Show 0 Likes0
Actions

problem with simplified electronics thermal model

Attachments

Outcomes

Related Content

Recommended Content