Inaccurate propeller data?

Hello again, while I have figured out the post proccess results and mlhow they look the force plots arnt accurate, I know this because I have tested this in real life. The propeller should spin at 5000 rpm and should be enough to propel this aircraft. Do you know why the force plots are so wrong along the y axis?

Project link- https://www.simscale.com/workbench/?pid=4645129274202983280&mi=spec%3Aebe1285f-142d-41ea-a8d0-6ca89fc646c3%2Cservice%3ASIMULATION%2Cstrategy%3A23

Thank you for your time and help,
Frank lucci

Hey Frank!

Could you please elaborate on why are the plots wrong? Do you have measurement data to compare?

No, but I know that at this rpm the plane excess thrust at 6000 rpm because this propeller produces 300 grams of thrust, added with the .6 newtons of drag should be at least less . 6 newtons of drag but is actualy double the drag which doesn’t make any sense because I ran multiple simulation of the drag of just the that part and it all came out around. 5 to .8 newtons? Could it be the mesh of the propeller?


Screenshot_20210930-152914_Chrome|690x439

You can see how the thrust should be enough for the aircraft even considering the little drag from the propeller

I still don’t know what is wrong?

Hi Frank,

I see many possible sources of error here, please review all of them, perhaps you can improve your results:

  1. Your computation of thrust and relation to drag. You are considering an operational point right? That means the device is flying at constant speed so thrust = drag. Review this thoroughly: should it be the case? Are there other factors affecting this computation? Check what others have done in the past.
  2. The inlet speed. In conjunction with (1), the inlet speed should be exactly the flight speed.
  3. The rotation speed of the propeller: 6049 rpm is not equal to 600 rad/s.
  4. The mesh: as you mention, it could induce error if not fine enough. Perform a mesh independence analysis.
  5. Measure the thrust in the propeller alone, the body alone and compare. This could reveal some hidden issues with the model.