Mesh Quality Reduction after Inflation

  • 566 Views
  • Last Post 31 August 2018
  • Topic Is Solved
aCVP posted this 15 August 2018

Hi all - I was able to successfully mesh a heart from a CT scan using the "shrink wrap > convert to solid" method in SpaceClaim. However, whenever I add inflation, the mesh quality vastly reduces (on occasion, I will get a "stair-stepping" error). I've included pictures for comparison.

I've played around with the inflation settings (including changing the gap factor) per the user manual + tutorials, but I can't seem to improve the mesh quality.

What can I do to improve the mesh quality after adding inflation?

Thanks!

 

Order By: Standard | Newest | Votes
Kremella posted this 16 August 2018

Hello,

Could you please elaborate the nature of your analysis? Are you solving laminar or turbulent flow in this geometry? 

What are your settings for the inflation layers and what is the maximum skewness you are getting?

I am attaching a playlist of videos here which take a deep dive into some inflation techniques on Ansys WorkBench meshing.

I hope this helps.

Thank you.

Best Regards,

Karthik 

aCVP posted this 17 August 2018

Hi - thank you for your response!

My major problem is how do you know when you have enough inflation layers? I am modeling blood flow within a heart. 

A Fluent tutorial for Cornell said that your layer thickness should be 1/10th of the diameter of your inflow - however, they did not give any explanation as to why. Is this standard?

Thanks.

kkanade posted this 31 August 2018

Hello, 

For improving mesh after prism generation, you can use Fluent Meshing. 

In WB Meshing please export .msh file from File --> Export

Start Fluent in mehsing mode. Read .msh file. Go to Mesh --> Auto Node Move. 

Select all boundary zones and click apply. 

On prism layer thickness: it depends on what y+ you are targeting, what velocities you are using etc. 

In general 5 layers of prism layers should be enough for regular flows. Please try to reach aspect ratio of 5-10 for first prism layer. 

Regards,

Keyur

rwoolhou posted this 31 August 2018

The other thing to check is where the poor cells are. With scan derived geometry you can get spikes and crevices which persist through the skin process. It may be you need to refine the surface mesh in this area. 

 

 

Close