[rouco75]

Hello everyone,

Today I am here to ask you how can I compute the center of pressure in a CFD simulation, in my case in Star CCM+, but I think this is a general thing and therefore it can help people in other softwares.

The problem I am facing consists of finding the CoP through the use of the momentum given a reference point, so where this momentum equals 0, this should be the CoP.

After researching for a way to find it, I came to the conclusion that using this method along the height of the ground doesn't allow the user to know the actual aero-balance of the car, as if done like mentioned above, the drag is not computing to cancel the moment, and therefore, the position of the CoP along the longitudinal axis is not correct.

Moreover, if the reference point moves upwards, the position of the CoP changes, giving as a result, a line in the direction of the forces where the equation in the top of the image is satisfied.

As I didn't know how to find the real CoP, my next move was to try to compute it as the NASA states, https://www.grc.nasa.gov/WWW/K-12/airplane/cp.html , but I can't see how can I select all the mesh elements on the surface of the car in Star-CCM to compute it through this method.

Here you can see the basic sketch with the way I am seeing it.



I think that if anyone can share a good explanation of how to find the CoP it would be helping a lot of people that may be facing the same problem as me.

Thanks in advance to this amazing community.

[LuckyTran]

Originally Posted by rouco75

The problem I am facing consists of finding the CoP through the use of the momentum given a reference point, so where this momentum equals 0, this should be the CoP.

After researching for a way to find it, I came to the conclusion that using this method along the height of the ground doesn't allow the user to know the actual aero-balance of the car, as if done like mentioned above, the drag is not computing to cancel the moment, and therefore, the position of the CoP along the longitudinal axis is not correct.

If what you say is true, then you have somehow discovered a counter-example for the parallel axis theorem. Or clearly, you are doing something wrong.

The most convenient way is to calculate the net force and moment about the origin and use the parallel axis theorem to shift to find the CoP.

If you want to do it the NASA-way, then you need to define three custom field functions as Pressure*Position[0] (and also Position[1] and Position[2] for y and z). Then do 4 reports for the 4 needed surface integrals. Select all the relevant surfaces when you define the report.

After performing all of this, you will hopefully realize that this is how Star computes the Forces and Moments in the first place and you need to somehow un-conclude your own conclusion that this method doesn't work.

[rouco75]

Originally Posted by LuckyTran

If what you say is true, then you have somehow discovered a counter-example for the parallel axis theorem. Or clearly, you are doing something wrong.

The most convenient way is to calculate the net force and moment about the origin and use the parallel axis theorem to shift to find the CoP.

If you want to do it the NASA-way, then you need to define three custom field functions as Pressure*Position[0] (and also Position[1] and Position[2] for y and z). Then do 4 reports for the 4 needed surface integrals. Select all the relevant surfaces when you define the report.

After performing all of this, you will hopefully realize that this is how Star computes the Forces and Moments in the first place and you need to somehow un-conclude your own conclusion that this method doesn't work.

I may have not expressed myself properly in english, but I am here to try to solve my problem. Of course I have probably made lots of things wrong.

This said, I don't see any of your points:

-Isn't the parallel axis theorem applied only for inertia moments? In this case I know the force and the moment in a defined point, but as the moment equals F*d, depending on where I select the moment it is going to change the distance d.

-Regarding the NASA way, I don't know what 4 surface integrals are you refering to. As I see it, I would like to compute the pressure and X-position of each tiny cell over the surface of the car, to integrate it and then divide it by the integral of all the pressure over the car, giving as a result the X-position of the CoP.

Best regards and sorry if I sounded rude, it wasn't my intention

[LuckyTran]

The four integrals are: the integral of pressure in the denominator; three integrals in the numerator, one each for pressure * x, pressure*y, pressure*z


The parallel axis theorem is not limited to only moments of inertia. Any weight function can be applied. If you know the force and moment about any point, you can find the equivalent moment about any other point. It establishes the equivalency and gives the methods. The pressure*x integral is a moment.


When you compute these integrals, you're just calculating the moments about the origin which is right back to where you started having trouble.

[rouco75]

Originally Posted by LuckyTran

The four integrals are: the integral of pressure in the denominator; three integrals in the numerator, one each for pressure * x, pressure*y, pressure*z


The parallel axis theorem is not limited to only moments of inertia. Any weight function can be applied. If you know the force and moment about any point, you can find the equivalent moment about any other point. It establishes the equivalency and gives the methods. The pressure*x integral is a moment.


When you compute these integrals, you're just calculating the moments about the origin which is right back to where you started having trouble.

Ok, I think that I understood it now.

So could you share any method to find the CoP or give me some indications to research for it?

Because I think that the parallel axis theorem is what I was doing in the picture, where the moment is cancelled with the total force and the lever arm perpendicular to this force. (Represented with the force in grey, any of the moments in yellow, and the perpendicular lever arm in blue) [note that the numbers don't make sense as I just wanted to explain myself as clear as possible]

[rouco75]

I have been thinking about it and I had the idea of making an histogram where the x axis represents the longitudinal position along the car and the y axis represents the static pressure.

If once this is done I compute the centroid of the area under the histogram, will this be the longitudinal position of the Center of Pressure?

Thank you