Hi everyone,

I am busy trying to evaluate the impact of using a 'structured' vs an 'unstructured' mesh, in highly-complex flow domains.

The assumptions used for the heat-tranfer coefficient at the walls of internal shapes seem to run into some accuracy problems for 'unstructured' meshes, when compared to 'structured' meshes.

I would really appreciate advice here.

My models typically contain around 494+ included internal shapes.

I was hoping to use the CFD package to provide me the surface heat-transfer coefficients, without having to calculate bulk heat-transfer coefficients.

Thanks for your advice.

Regards,

Des Aubery... (adTherm Technology - Virtual Thermal Engineering)

I think there was a pretty extensive debate of the structured - un-structured question recently here.

Try searching this forum, then the CFD-ONLINE archives. If your answer isn't there, you should at least have an idea of what to expect in the way of answers.

Hi Jim,

Thanks for your reply.

I had tried a search, but ended up with a lot of spurious data. Let me try again...

Best regards,

Des Aubery...