topic Re: Question about boundary conditions in CFD Simulation in CFD Forum

Question about boundary conditions in CFD Simulation

Anonymous — Sat, 01 Jul 2017 11:28:41 GMT

Hello, Im working on a simulation model, and Im not sure how to use the boundary conditions correctly to get the scenario I want.
Its a formula student car that have an aero box, everything is created in CREO and saved to an .step file.

When I've run the simulation, the car somehow gets lift from the undertray so my thoughts is that I dont have an correct "ground". The undertray should build up an pressure difference that would generate down force instead of lift force.

The conditions that I've used is
Front side; Velocity 15 m/s
Ground side; Velocity component 15 m/s
Back side; Pressure 0 Pa.

Can I fix this with boundary conditions? I've heard that Slip/Symmetry could fix this if i put it on the "ground side", but I haven't found anything that confirms it.

Any information/tips is great!

Re: Question about boundary conditions in CFD Simulation

Jon.Wilde — Mon, 03 Jul 2017 07:59:18 GMT

Hey. welcome 🙂

Here are my thoughts!

You can import Creo geometry directly if you wish to, might be faster 😉
Be sure to run a 1/2 symmetry model, you can mesh it better and hopefully in a shorter time period
I think I would use slip on the underside, although to be fair, if you know you have 15 m/s there it does sound very sensible and I suspect it is not your issues here
Is your mesh good enough quality? Have you run a sensitivity study to understand this?
You might need a FAR better mesh, check this out.

Thanks,

Jon

Re: Question about boundary conditions in CFD Simulation

Amal.C — Mon, 03 Jul 2017 08:04:58 GMT

Hi @Anonymous,

I would like to add to Jon's points this simple tutorial of a sport car simulated in Autodesk CFD.

There, you can see how we defined half of the model and the right boundary conditions + mesh.

I hope this helps.

Thanks,

Amal

Re: Question about boundary conditions in CFD Simulation

Anonymous — Mon, 03 Jul 2017 14:51:52 GMT

Thank you! 🙂

Yes Im using a 1/2 symmetry model.
Ive only used auto-size and adaptations, thought that would make the mesh good enough but ill try the sensitivity study!

And by the way, for some reason the adaptions sometime "bugs(?)" when it comes to the 2nd mesh, on its last iteration. All the results gets converted to "nan".

Thank you for the reply!