I have an issue with the torque output from a 2D rotating surface for a while now and really need to have this figured out as quickly as possible. I wonder if anyone else has the same issue or could shed some light on this.
I take a single symmetric airfoil like a NACA0012 and put it in a rotating surface that is circular with the center sitting at the origin (see diagram). I do have a larger air domain around it. I set the material, RPM, mesh and put in a 3-degree-timestamp. The boundary condition is set to no wind speed since I am isolating the torque results. I run it for 4,800 iterations and the torque.csv is created at the end of the run.
Over the last full revolution with 30-degree interval, I try to match the hydraulic torque to the hydraulic forces (X and Y) in the csv file, but I find the calculated torque does not agree with the raw hydraulic torque output (see chart). I have accounted for the moment of the airfoil and added that to the calculated torque. The moment of the airfoil is obtained from the wall calculation at each 30-degree interval over a full revolution at the 25% point of the airfoil's chord.
I throw the wall calculation torque in the chart just for the sake of curiosity. I am well-aware that getting rotating torque results from the wall calculation is not recommended.
This torque issue is observed in both Simulation CFD 2012 and CFdesign 2010.
Hi,
what do you mean with no wind speed? A specified velocity condition of 0? Can you give us an complete overview of the boundary conditions or even attach a share file to your post?
Does the same difference of the "force-torque" and torque appear for different mesh sizes?
Marco, attached is the share file with the single airfoil. The scenario has a close-to-zero wind speed of 1e-06 mph in the positive X direction.
I was wondering if the hydraulic torque from the csv file is directly related to the hydraulic forces X and Y (or vice versa)? Or are these outputs solved implicitly from different sources?
I have not tried different mesh sizes yet, but I will definitely try that now. Thanks for the idea, Marco. I will post my findings once that is done.
Of course, they should be comparable. But to check this I'd eliminate some error sources. I've found that the rotation region is not perfectly rotating around its center, that's why there occur cyclical issues during the solution:
Also the mesh near the interface line can be finer since there is a (small) shift of the values around it:
A side note: I'd uncheck the output of all quantities that are not relevant for you. Especially the calculation of LMA needs additional ressources (and is anyway not necessary during the solution), so if you uncheck this you will see a tremendous speed-up of the simulation.
Yes, I also observed that strange gap between the rotating and static region. There was no intentional gap created and the rotating region center is placed at 0,0,0 in the CAD file. In the CAD file, the rotating region is just a surface that sit on top of the overall air domain surface.
Could there be something wrong that is throwing the rotaional center off?
I tried to make another CAD file where the rotating region was "cookie-cut" out from the rest of the region (separated between the regions), but again I saw the gap appearing when the simulation started to run.
Any suggestions to remove this unwanted gap and keep the rotating region at 0,0,0?
Hi Max,
Could you attach your native CAD files? I am not sure why we are not getting a perfect rotation based on what I am looking at.
Thanks,
Royce
What CAD package did you use to generate it?
-Royce
I use a CAD software named Shark FX. Do you prefer a parasolid file to work with?
Can you try to export parasolid or ACIS and then load them into Inventor Fusion which is part of the Simulation package and then start the simulation from Fusion?
I will try using Inventor Fusion to load the exported parasolid. Could this offset in rotation due to the inertia offset since there is only one airfoil in the rotation region?
This should not be the reason. The motion is independent of the geometry. One thing you can also try to get a more accurate solution is to move the rotating region further away from the airfoil.
Marco, I increased the rotating region radius by 20 percent, but I am still seeing the gap once the rotation starts.
Please see attached pictures. The first picture shows the enlarged rotating region and the second picture shows the closeup view of the gap at the top (12 o' clock) location of the rotating region.
Hi Max,
increasing the RR was just suggested to get a better mesh around the airfoil and keep the RR a bit away from areas of high gradients, not for the offset issue.
The gap seems to be around 0.1 ft only in y direction. Maybe you can try to move the RR in your CAD system by 0.04905 ft (got this value from the results) in negative y direction. This is for sure not elegant but may be a workaround if we stay in the tolerances for the RR calculation.
Did you have any luck trying to import either the parasolid, acis, or step format into Fusion?
I think Marco and I might agree that the issue with the 'gap' has to do with the accuracy of the geometry from your CAD package.
Might want to also try making the geometry in Fusion from scratch if importing translated geometry doesn't help.
-Royce
Importing the acis format to Fusion has not been successful. I probably need to figure out how to work Fusion first. I will also try to make the geometry in Fusion if the import method fails.
I will keep you guys posted.
Max
Do you have more questions on this vs what we discussed with your previous case?
Let me know
Apolo
Can't find what you're looking for? Ask the community or share your knowledge.