Problems finding realistic torque to rotate a body

Using SimScale Fluid Flow / CFD
1

Thanks in advance for reading all of this and any help is much appreciated!

I’m attempting to use SimScale to find the torque required to spin a lawn mower blade to compare it to experimental data I’ve gathered, mostly for the purpose of making sure I’m using the program correctly and accurately, however the process has given me a number of problems along the way.

The relevant simulations should be public and available at SimScale

I attempted to model my simulation off of SimScale with a compound modeled in solidworks based off of the lawn mower blade, but any attempts at transient simulation ran into CFL issues. I tried again with a less refined mesh, but still had to decrease the Delta T too far to be feasible in order for the simulation to actually run. Attempting to reduce the refinement level of my mesh further led to the blade geometry being lost in the rotating region before I could get it to run correctly. Simulating a blade with a simpler geometry seemed to work fine (Incomp basic transient – run 1), but using the actual blade geometry continued to produce CFL errors or require far too much computational time, even switching to a MRF zone instead of AMI.

To shoot for an easier target, I’ve switched to a (EDIT 20 Feb, 0924 CST: steady-state, not transient) simulation, which I think I’ve finally got dialed in to a point where it runs consistently (Incomp SS Real Level 2 Mesh – Runs 2,3). However, the torque values (pressure moment y) that I’m getting from the results control are still over 30% higher than what I’m getting experimentally and what I’ve found to be consistent with academic data

I’m still relatively new to these types of simulations and have likely missed some simple things along the way. That said, if someone wouldn’t mind taking a look at my simulation labeled ‘Incomp SS Real Level 2 Mesh’ (run 2 or 3) and wouldn’t mind giving me a few pointers or things to look into that might be affecting my simulation’s accuracy, it would be much appreciated!

Thank you in advance for any and all assistance, and please let me know if I need to provide more information on some part, I’ve attempted to provide what is needed while maintaining brevity, but I can’t be sure.

2

Hi @abl!

So few things that caught my attention in your simulation. Imho, the fineness of the blade and rotating zone is not fine enough as the features are not well resolved and very “edgy”. Also for the inlet I would place the zone a bit away from the inlet and increase the box in general to make sure boundaries do not influence the flow in a negative sense. Additionally do not use no-slip for the outer walls to mimic the atmosphere but the slip condition, that should increase the accuracy.

Also are you sure that you want the propeller to “suck in” the air from the atmosphere? Just curious about the setup. Naturally when refining the Mesh the CFL will go up so to keep the upper bound and stability you have to decrease the time step (look at the definition of the CFL number).

For any additional input I am tagging our PowerUser Barry (@Get_Barried).

Cheers,

Jousef