Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

Is my method correct?

I want to do CFD run for a ceiling fan having 3 blades.

My system has components like wall, MRF and Interface. MRF is the cylindrical solid from which the Fan is subtracted. Walls form a room in which the air flow has to be monitored. For volume meshing I have used boundary layers at the interface of 2 layers and 5 layers on the MRF; whereas no BL is given to walls. Mesh is done using Tetras. My turbulence model is Spalart Allmaras.

Is my way correct? Please suggest better way.

Find more posts tagged with

AcuSolve

English

HyperMesh

v2021.0

Accepted answers

All comments

Junta_20361

It sounds right for your method.
but the walls should have the BL mesh.
You can show us some image for better visualization.

acupro

If the room walls are far away from the measurement region(s) of interest you may be fine without boundary layer meshing (thus, not capturing boundary layer flow properly). If you are interested in measuring near those room walls, you'll need the boundary layers in the mesh. Best practice is also to perform mesh sensitivity studies to see how different meshing techniques/mesh density affect the results of interest.

Upnishad Kasturi

Thank you for the reply.

Please suggest the turbulence model for the CFD simulation of ceiling fan. I have been using Sparlart Allmaras.

Upnishad Kasturi

Thank you.

Model is shared below.

Also want to share the mesh setting I have used. Boundary layer+tetra in present on the fan blades and the interface and no BL on the room walls. Mesh details are shared below

My mesh size is about 38 million.

My concern is regarding the residual ratio of pressure which is not converging. Graph is shared below.

BL+Tetra Mesh]

No BL Mesh

Model of fan with room and interface

Residual ratio result graph: Here velocity is converging, but pressure does not.

acupro

The Spalart-Allmaras turbulence model should be fine for this. The pressure behavior is likely due to boundary condition settings. Is this a completely closed domain - walls on all outer boundaries? Or - do you have an inlet and an outlet? Can you share the .inp and .Log files?