Battery Powered DC Power Supply - CFD Setup Questions

[Anonymous]

I am trying to run a CFD simulation on a battery powered power supply. It is not possible to know how many Watts to assign to each component, but I do have a thermal camera and can measure component temperatures during battery discharge. Is it a valid technique to assign temperature as a boundary condition rather than Watts?

[Jon.Wilde]

Hi,

The important question for me is, what is the overall goal of the analysis?

It is OK to do this but really only really if we suppress the batteries from the mesh/analysis so that we do not have internal boundary conditions.

Thanks,

Jon

[Anonymous]

I need to create a reasonable thermal simulation of an existing product to use in the design of next generation products. Only a few components get hot during battery discharge; the rest are part of the charging circuit. I have calculated the wattage of one big resistor that is heating up, but I can't do those calculations for the transformer and transistors. With the thermal camera, I have taken the temperature of the hottest components. So I wonder if I can just assign those temperatures to the component faces. I can email the model to you if that will help you to help me set this up correctly.

[Jon.Wilde]

Hi Mark,

Assigning temperatures should be fine, so long as you are aware that this value will remain 100% fixed. Do suppress these components from the mesh too.

This is unlike applying a volumetric heat load, where the component temperature will be defined by the heat lost to the surrounding solids and air.

What do you need to see as a result from this simulation? What will define a pass/fail for the design?

If you run into other issues, sure share the CFZ, no problem.

Kind regards,

Jon

[Anonymous]

---

@wildej wrote:

Hi Mark,

Assigning temperatures should be fine, so long as you are aware that this value will remain 100% fixed. Do suppress these components from the mesh too.

This is unlike applying a volumetric heat load, where the component temperature will be defined by the heat lost to the surrounding solids and air.

What do you need to see as a result from this simulation? What will define a pass/fail for the design?

If you run into other issues, sure share the CFZ, no problem.

Kind regards,

Jon

---

Jon,

From your explanation, it sounds like I need to use heat generation vs. surface temperatures for the components. And that is the way I learned in the turorials. We will attempt to measure component Watts on the bench today.

As for pass/fail, we will be using the thermal data to aid us in future designs. This is a production unit we are trying to replicate in our CFD model.

[Anonymous]

I would like to run this simulation in a test chamber at ambient temperature and at elevated temperature. Please check my setup and advise if there is a better way.

[Jon.Wilde · Accepted solution]

Hi,

Check out the guide here for the best setup. This is for a ground based unit - which it looks like you are running here. If not, there is another approach, that you'll find in the guide.

I would think that fixing the floor to 21C is going to heavily impact your results as it appears to be on top of the condition.

Best regards,

Jon

[Anonymous]

Jon,

Thanks for the link. I'll try that setup on Monday.

[Anonymous]

I am getting this error in my CFD model now. Is it a mesh problem?

[Jon.Wilde]

Hi,

What changed between the previous analysis and this one?

What was in the output bar when this popped up?

One thought I did have - if this is to be table mountned - we might need to square off the edges of the case where it meets the table. Otherwise we are going to have some very skewed elements, or zero angles which are near impossible to mesh.

Kind regards,

Jon

[Anonymous]

Jon,

Here is the output on the failed model:


Mark McGee
Senior Mechanical Engineer
Acuity Brands
770 860-3128

[Jon.Wilde]

Hi Mark,

I am not sure that image worked OK.

Would you like to share the CFZ once more?

Thanks,

Jon

[Anonymous]

Here is the output:

[Jon.Wilde]

OK so this is past the meshing stage.

It says flow off as there is no fluid?

Is that right?

[Anonymous]

Here is the CFD file:

https://drive.google.com/file/d/0B8KVvS6tGzv_cEJiWURMRXVKOUE/view?usp=sharing

[Jon.Wilde · Accepted solution]

Thanks.

What I posted above is your solution. What I think should be your air domain is set to be graphite. Change it to air-variable and you'll be all set!

Thanks,

Jon

[Anonymous]

I'm back in business. Thanks Jon.

[Anonymous]

Please review my setup. I am trying to model a test chamber (oven) so I can test this product at 55° C. I modeled the chamber and let Sim CFD create the air mass. I set the boundary conditions on the top surface of the air mass as shown in the image. It seems to run alright, but I want to know if there is a better way to model the oven.