Why are Forces Results so different between Custom Inlet-Outlet ZeroGradient and FixedPa=0 conditions? :)

[DaleKramer]

I noticed that when using Custom Inlet-Outlet with ZeroGradient pressure condition, that the air does not generally enter and exit the mesh at 0Pa. This is bothersome since I would really like to investigate the internal pressure inside a paraglider wing with respect a farfield air gage pressure of 0 Pa.

Until I investigated how to ensure that the air entering and exiting the mesh is at 0 Pa, I simply have to determine the relative difference of the inside airfoil Pa and a farfield pressure (which is generally the pressure at the lower corner of the entrance face). Here is an example (Note: the farfield pressure and the internal pressure are only coincidentally the same here because of the low AoA of the run):

ZeroGradient -30Pa farfield 3aoa 10mps OPENorg2 -100to60Pa scale.png

Next I have investigated using the FixedPa=0 condition and yes, the farfield pressure does move to zero and everything LOOKS fine as you can see in this image:

FixedPa=0 0Pa farfield 3aoa 10mps OPENorg2 -100to60Pa scale.png

BUT, looks are deceiving , the Forces Results of these two runs are about 15% different!

Any ideas out there as to why, in the ZeroGradient run, the Pressure Moment y is 161 Nm and the Pressure Force z is 80.7 N, while for the FixedPa=0 run, the Pressure Moment y is 140 Nm and the Pressure Force z is 70.7 N

Here is the project SimScale

Thanks,
Dale

[jousefm]

Hi Dale,

These links might be helpful to you:

• Inlet-Outlet Boundary Condition

• Outlet Boundary Condition

---

Best,

Jousef

[DaleKramer]

Perhaps I should have mentioned that I am again investigating (I have had very little luck achieving satisfactory results in the past with this method) using Custom Inlet-Outlet condition that is supposed to allow me to change the AoA of oncoming air to a geometry.

I have seen ZeroGradient mostly used on setups like this but again, I want a proper 0 Pa reference on the BMB faces I mentioned.

My Top, Bottom, Front and Rear BMB faces are set to this condition and my initial conditions match the conditions for U, k and w. Initial conditions for Gage Pressure was 0:

For +ve AoA’s I will try Bottom and Front BMB faces set to InletOutlet-FixedPa=0 and the Top and Rear set to OutletInlet-ZeroGradient and report back…

[DaleKramer]

I can’t explain why having one Custom Inlet-Outlet set to FixedPa=0 is so different to the more commonly use ZeroGradient condition but I have found a way to achieve the entrance pressure to 0Pa with almost exactly the same Force Results.

Custom Inlet-Outlet atZeroGradient gives: Pressure Moment y is 160.9 Nm and the Pressure Force z is 80.7 N
My new method gives ------------------------ : Pressure Moment y is 160.4 Nm and the Pressure Force z is 80.6 N

Here is the new methods pressure plot which is now what I wanted:

Farfield=0Pa 3degIN 10mps 4768311v OPENorg2at0aoa Front_Bottom=InletOutlet-FixedPa=0 Top_Rear=OutletInlet-ZeroGradient.png

Here is the trick which uses both a Custom Inlet-Outlet and a Custom Outlet-Inlet (only tested for +ve inlet air angles):
.
EDIT Nov 6, 2023: Looking back on this over 3 years later, I see that the Ux and Uz in the 1st image below were incorrect, they should match the Ux and Uz in the 2nd image below OR more simply the Ux and Uz in both images Ux and Uz should have MATCHED (fixed now).
.

I hope others can benefit from this.

I had the idea to piece together this method after I reviewed the OpenFoam documentation on InletOutlet and OutletInlet boundary conditions…

Dale

EDIT: I also tried to confirm that this method of changing the angle of inlet air, provides similar Force results as the traditional approach which has slip walls above and below the geometry. The traditional always uses 0 degree inlet air angle on a number of meshes, each mesh with the geometry rotated to a different AoA in CAD program.

To get results within 5% of each other, I had to make the BMB four times as high (The actual Force values for 2, 4 and 8 times higher BMB are in the project…).

The 5% difference is significant, but when one weighs how much easier it is to run different AoA simulations, I am OK with it. Who knows maybe it is the traditional approach that gives the wrong Force Results.

Someday I may try to see which approach is closer to real world Force results.

[DaleKramer]

@dylan , if you are still around, what do you think of my solution here ?

Hope you are well…
Dale