CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   car simulation - boundary layer street (

sanchezz January 12, 2010 00:40

car simulation - boundary layer street

I'm simulating a car and I have a question about the boundary layer:

Do I also apply a boundary layer on the ground surface?? In reality, there is of course one but how important is this?
If I do need it, can i just use the same settings as for the boundary layer in the car (since I can't even calculate the reynolds number of the street - obviously in reality, the air on the street is not moving at all).

Would be great if someone had some advice.


rikio January 12, 2010 01:19

In my opinion, it is not necessary to build BL of ground if the ground effect is not important. BL on car body should be taken into account because usually we need to obtain the drag force, lift force, or even the thermal characteristics, which were greatly affected by the boundary layer.
Just for your reference, because I do not have study on this field. :)

sanchezz January 12, 2010 01:45

I need drag and lift of the car body, that's right. But some devices are exploiting ground effect, so I guess this is an important aspect.

rikio January 12, 2010 09:22

Are the devices parts of the car? Anyway, I think you do not need to model ground boundary layer too. Because coarse mesh could solve the flow field near the wall, though it is not so exactly as fine mesh. We do not care the velocity distribution in this very thin layer, right? Maybe you just want to know the effects ground BL bring to car body.

sanchezz January 12, 2010 09:33

"Devices" will probably be diffusers. But with a ground clearance of only one inch, I thought that the boundary layer could have some effect. I don't care about perfect modelling the velocity distribution in the BL of the ground, just if it has any impact on the diffuser performance.

Attesz January 12, 2010 14:40

I'm not experienced it, but maybe a thin layer with one element will be not a big problem. Because with coarse mesh, you can get more thicker bl. But I don't now, how much importance has it...Anyway, if you don't use bl.mesh, setting those walls "Free slip" walls is a better choice, i think.


ghorrocks January 12, 2010 18:23

The best thing is for you to do a mesh sensitivity analysis and determine for yourself what mesh is required in the area. Otherwise you are just guessing.

sanchezz January 13, 2010 08:12

Thanks ghorrocks, you're right. That would at least be the most scientifical approach, even though you can save lots of work by learning from other people's experiences - it's not like nobody has ever modeled a race car before...
Unfortunately those specifics aren't mentioned in many research papers, that's why I'm asking here.


ghorrocks January 13, 2010 19:05

Not only will you find out once and for all whether you need BL meshes under the car, but in my experience going through this sort of exercise always shows up interesting flow features or physics which cannot be seen in the larger model, or you get a feel for the accuracy of the simulations. It has benefits beyond the obvious as you get a deeper understanding of what is going on.

All times are GMT -4. The time now is 10:21.