Vehicle Aerodynamics
 

Hello all,
I have recently started working with CFD softwares and have a question. I'll be simulating a car and analysing the drag co-efficient and the downforce.

I believe there are two approaches for this simulation

1) Keep the car and the road at rest and give inlet velocity equal to the velocity of the car

or

2) Directly give the car it's velocity without creating a velocity inlet.

I wanted to ask you if the values of drag and downforce would be different for these two approaches ? Or would they be same?
Also, if I want to give the velocity to the car itself, how can I give it, in Fluent ?

Thanks !

i think results will be same.:cool:

You do not want to keep the road at rest unless you are comparing results to a non-rolling road windtunnel.

the road isn't stationary relative to the car
 

the road actually moves at the same speed as the air (if the car is assumed stationary).
this would cause a difference in results between a moving car, stationary road+air simulation and a stationary car+road, moving air simulation.
I would suggest simulating the road as a moving wall.

I am not sure, but I think results will be only slightly different for moving and stationary road, since, rougly speaking, the size of road bl is much less than the distance between road and a car.

So are you implying that the distance between road and a car is negative? :-)

In fluent, you can consider the stationary air at time zero and you should consider moving wall boundary condition for external wall of your car. You should mesh the outside part of your car that you guess there is boundary layer, the part of air around the car that is influenced by car moving.
Be successful.

Hello smhosseini,

thanks for your reply. Unfortunately, I'm a bit confused. You mentioned that I should consider air as stationary at time zero.
Can you please explain this point ?
Do you mean that I should do a transient simulation ?
I thought that for this case, a time steady simulation would work.

Thanks, again,

Harshal

I'm sorry because I didn't explain completely.
You can solve this problem steady or unsteady. I mean you don't need any inlet or outlet flow. you can start iteration with stationary air for steady state solution. If you need the variation of boundary layer with time, you can solve unsteady problem.

Hello smhosseini,
thank you for your reply. I plan to do a time steady state simulation. In fact, I have several doubts regarding the setting in Fluent and have posted the same in the Fluent forum. Would you be so kind to view my post and to answer it ? I have posted the link below. It would be really helpful, if you could answer or suggest some solutions for my doubts.

http://www.cfd-online.com/Forums/flu...downforce.html

Thank You,
Regards,

Harshal

Simultaneous simulations for different speeds
 

Hello all,
I need your help. I have to find out the Cd and Cl values for a car using k-epsilon and k-omega models for multiple speeds (about 15 different speeds). Can some one please tell me how to do that ?

Thanks a lot,

Harshal

Your coefficients shouldn't change with speed.

Hello Plucas,
thanks for your reply. Yes, you are right. Ideally, the Cd and Cl values should not change w.r.t speed. However, since the accuracy of results in CFD depend on numerous factors, it is quite possible that one does not get constant values for different inputs (here speed). So, I want to find out if the values change w.r.t speed and if they do, by how much.

Thanks again,

Harshal

Cl and Cd depend on reynolds number, i.e. on speed

As somebody already said, you need your road as a sliding wall if you want your car to be simulated in operating road conditions


How would you do that? Sliding mesh?

Maybe the results would not change (I'm not completely sure though) but the cost of your simulation will change a lot.

Hello truffaldino,
thanks for your reply. I'll look further into this and see if and how the Cd and Cl will change with increasing speeds.

Hello,
could you please tell me how the Cd and Cl values will change for increasing speeds (30 m/s, 40 m/s and 50 m/s).

Will both or any one of Cd and Cl increase / decrease with increasing speeds ?
Which parmeter of the two will change more than the other ?

Anything like that would be very helpful.

Thanks,

Harshal

Yes it does at 100% theory level. In application, a vehicle at those speeds he requested, the coefficients should be the same.

For which speed does a car manufacturer specify car's Cd value ?
 

Hello,
I have a question regarding Cd value.
As I learned recently, the Cd value depends on the speed.
However, when car manufacturers declare the Cd value of a vehicle, I noticed that they simply mention the Cd value, without specifying the corresponding speed.

I have carried out several simulations for various speeds ranging from 30 m/s to 50 m/s and need to determine the accuracy of my results. For this, I am comparing my simulation values to the Cd value of 0.35 declared by the manufacturer.

However, I do not know for which speed the Cd value should be 0.35. Hence,I am unable to draw any conclusion from my results.

So, does anyone know how the Cd value of a vehicle is specified ? Is there any standard test cycle procedure like in case of fuel consumption, to determine the Cd ?

I would appreciate any insight or information regarding this.

Thanks,

Harshal

Hi


Simply: No, the drag coefficient should be completely independent from the speed. It is a dimensionless coefficient and is made for comparing the aerodynamics of different shapes, even if speed, Reynold number and even size is different, think of airplane models in experiments (btw. these three always correlate anyway). This doesn't hold for all speed regimes but for a quite large regime the drag coefficient is indeed more or less reynold independent, which is surely what plucas meant, too.
The coefficient is defined as the proportion of the force on the body and the "possible force" from the flow around it, measured by the dynamic pressure.

cd = F / (rho/2 u^2 A)

The area A is some area which is characteristic for your body. Aerospace engineers use the wing area, car engineers use the projected front area of the car, usually without mirrors as far as I know.

Long story short. Let fluent calculate the force on your body aka on all body surfaces, divide the force by half of your density, the square of your inlet velocity and the projected surface of your car. Plot the cd for different velocities and you'll see they will be roughly the same. -> done :)

Have fun