[tacitosantos]

Hi,

I'm new in CFX. I'm starting the simulation with a disk brake and a heat transfer through the pad and there is airflow too.


I do not know how to make the heat transfer between the disk and the air.

I developed a new surface on the disk using the DM with imprint face.

Then I did an interface between the pad and the new surface, making the the divided disk with 2 interfaces, one with air and the other with the pad.

I also did, the boundry condition on the pad surface To provide heat flux Which will be transfered to the disk using the interface between the pad and the disk.

The problem is after simulation, the part of the disk Which makes connection with the pad does not make heat exchange with the airflow.

So how can I do this simulation?

And at the same time to Provide a heat source between the whole disk and the airflow?

Also providing the stationary heat source representing the pad?

Help me please!

Tacito

[ghorrocks]

First - do not post identical posts on multiple threads. I have deleted the duplicate posts.

You will need to connect the air to solid domains with interfaces to get heat to flow across the interface.

I think the ANSYS community webpage (on the ANSYS website) has an example of disc rotor modelling. If not then ANSYS support will have it. I recommend you use that as a starting point.

[urosgrivc]

You must create an interface that conects solid to fluid.
When interface is created you are than able to add a source on one side of the interface either (flux [w m^-2] or total power [W]) (in your case it doesent go like this as you daont hawe a real interface where the brake pad is)


[1] I sugest that you start without including the (brake pad)! as interfaces will be easier and you wont hawe errors, as you probably
dont yet understand the interfaces that much.
Start of with a simpler simulation that will work, you will than uderstand interfaces and will be able to add the brake pad, as it is not as straight forward as it seems.

[2] For your simulation:
For a brake disc that has a wery simple shape as this one (if you do not include holes for bolts (if you do this wont work), just cilindrical bodies), it is not that hard to include the brake pad as you are able to use
->(solid models)->(solidmotion) and two (fluid-solid) interfaces, 2 becouse the third one betwen pad and disc is not a true interface in your case as you didnot model the pad as a body.

[3] Your model can be even simpler if you dont include bolt holes and you will be able to include the brake pad.
it can just be a disc with imprinted face for brake pad.
do a bit of reserch on the ->(solid models)->(solidmotion) cfx
hint-> your mesh can be stationary the nodes do not need to rotate around an axis (but it must numericaly rotate so heat will go round and round) -> solidmotion.


I hawe probably complicated it a bit as there is multiple aproaches to solve this.
number 3 is probably the best option for you, but the model is just a bit diferent (simpler)
If you hawe any further questions just write them below but do some research asweal.

[tacitosantos]

Thanks ghorrocks and urosgrivc for your answers.

For this simulation, first I used the solidmotion with a imprinted face on the disk representing the pad and on this surface I inserted the heat flux.
But I did not trust in the result shown by the CFD-POST and because of this I used the domain motion for a new simulation.

Image 1 - Simulation using domain motion

Image 2 - Simulation using solidmotion

As you can see in the images there are differences for the temperature contours indicated on the disk.


This was the reason why I believed that domain motion was the right choise for the situation.


So my doubt is, should I use the solidmotion or domain motion?

Again, thanks for the help.

Tacito

[urosgrivc]

What do your heat transfer coeficient results look like?

[tacitosantos]

This is what do you want?

Tacito

[urosgrivc]

What I ment was, if you want to compare temperatures you need to compare all factors, HTC should be almost the same for both of your cases (as speed and shape are the same),
if it is not than you can not compare temperatures directly as you hawe diferent conditions.
And i wrote for htc becouse I wanted to see if you hawe included that the wall is rotating in the case of solid motion, but I cant see that from this result.
I can see contour lines but also everithing is blue, why.

Also your pad area,shape is diferent.
And as you have set HEAT FLUX [W m^-2]-> that means that the amount of work (A[J]=Power[W]*t[s]) that goes in to heating the disc
is padAREA dependant.
Larger area->more Power[W]=Area [m^2] *q [W m^-2]