CFD - PERRINN F1 Car (w/ spec mods)

[jjn9128]

Asking for advice. My simulations have non rotating wheels because of restriction with the current setup I have. As mentioned I am getting about -2 CzS which seems quite short the -3.45 CzS that Perrin and other used to get. Do you think is sensible to think that the deficit comes from the non rotating wheels? I know the rotating wheels should generate more downforce however -1.5 CzS seems a lot to gain.

What is your experience with non-rotating wheels? I did compare contours and they look very similar to the available literature.

What numbers do you get with your setups?

Another quick question: do you account for momentum change between inlet and exit flow planes (radiators intake and outlet, engine intake and offtakes) in the CxS calculation? It seems in my setup they account for an additional 0.55 of CxS on top of the pure aero CxS of 1.15 roughly..

What ride height are you running at? Front and rear. When you say no rotating wheels, I understand no MRF or sliding mesh condition for the wheel rims but you've at least given the tyre surface a rotating wall boundary condition?

[ggaggio19]

I am running roughly 20-50mm at front and rear respectively. I think the setup is pretty much the same as Perrinn one.
No unfortunately my current setup does not allow even to impose a rotating wall boundary condition. The velocity is zero on the tire surface :S

I am in the process of switching to OpenFoam but not quite there yet...

Another possibility is that I am calculating the forces in the wrong way. At the moment I am assuming atmospheric pressure on the contact patches to counterbalance the void there. Is this the best way?

[jjn9128]

I am running roughly 20-50mm at front and rear respectively. I think the setup is pretty much the same as Perrinn one.
No unfortunately my current setup does not allow even to impose a rotating wall boundary condition. The velocity is zero on the tire surface :S

I am in the process of switching to OpenFoam but not quite there yet...

Another possibility is that I am calculating the forces in the wrong way. At the moment I am assuming atmospheric pressure on the contact patches to counterbalance the void there. Is this the best way?

No. There is no pressure on the contact patches because there's no air there. Rotating boundary condition barely costs anything to implement or are you saying it isn't possible in the CFD you're using??!?!

[ggaggio19]

It is not possible in the CFD I am using at the moment for this. It is bad I know, but I am not sure how bad it is for the qualitative flow around the car. Not sure if it is costing me the -1.5CzS I am short of or there is something else wrong in the setup.

In literature it seems that the difference between isolated rotating and not rotating wheels is roughly 20% for the Fz, which is quite less to what I have.

As for the contact patches, how do you compute the CzS if I can ask? If I do not assume anything on the patches I get enormous vertical forces because of the unbalanced surface... Am I doing something wrong in my thought process?

[jjn9128]

CzS is calculated using the surface pressure integral?! Or am I confusing the question?

[ggaggio19]

CzS is calculated using the surface pressure integral?! Or am I confusing the question?

Possibly. Let me explain what I have tried to do.

1. I calculated integral(p_static * Cell_Area * nz *dA) where nz is the zeta vector of the normal versor to the cell. What I get is a CzS of about ~-10, that is because I have "unbalanced" surface where the tires contact patch is.

2. I calculated integral( [p_static-p_amb] * Cell_Area * nz *dA) and in this case I get CzS ~-2.2 which seems more reasonable. In this way I am counterbalancing the effect of not having any surface on the contact patches and "sort of accounting for the ground reaction".

Though -2.2 is way short of the -3.4 that I should get with my geometrical setup and my question basically is: "am I getting the wrong CzS because I calculate it in the wrong way or because of the not rotating wheels or because my CFD is wrong somewhere else?"

Not sure you can help, but i hope you can give me some hints

[jjn9128]

Not sure you can help, but i hope you can give me some hints

Just double checking, you don't have a rotating wall condition on the wheels, do you have a moving wall or slip plane boundary condition on the ground plane?

[ggaggio19]

Not sure you can help, but i hope you can give me some hints

Just double checking, you don't have a rotating wall condition on the wheels, do you have a moving wall or slip plane boundary condition on the ground plane?

Wheels are stationary and velocity is set to 0 on them (like the anywhere else on the car).
Ground is set with same velocity as the air (Uinf = 50m/s)

[Dynamicflow]

Which flow solver are you using if you do not mind me asking?

[ggaggio19]

Which flow solver are you using if you do not mind me asking?

I am using an implicit compressible density based flow solver.

[jjn9128]

Which flow solver are you using if you do not mind me asking?

I am using an implicit compressible density based flow solver.

Fluent/ Star/ Cfx/ OpenFoam?

[ggaggio19]

Which flow solver are you using if you do not mind me asking?

I am using an implicit compressible density based flow solver.

Fluent/ Star/ Cfx/ OpenFoam?

A "custom version of OF" that does not support a rotating velocity boundary condition. It does not allow to set a rotating velocity of an object around an arbitrary axes. I know you will not be able to help on debugging the solver side of things...

[Dynamicflow]

I am using an implicit compressible density based flow solver.

Fluent/ Star/ Cfx/ OpenFoam?

A "custom version of OF" that does not support a rotating velocity boundary condition. It does not allow to set a rotating velocity of an object around an arbitrary axes. I know you will not be able to help on debugging the solver side of things...

Was it written specifically for launch and re-entry applications?

[ggaggio19]

No it was not written for that. It is a code we developed for other specific stuff that does not support rotating wheels.

I guess my question more than debugging my solver was if what we are observing in terms of CzS loss is sensible when wheels do not rotate (V=0)

And in general if the way we calculate forces is the right way...

[Dynamicflow]

How big is your CFD model?

With wheels rotating, your likely to lose more down-force since the wake would be bigger.