Learning Forum

[skdubey]

Hi friends,

I am currently working with a complicated fluid domain that has been decomposed into several control volumes, which can now be fully swept.

CASE A: When I run the model using the "merge" topology, I encounter difficulties in meshing the body to achieve a complete conformal body. This is crucial because the mass flux needs to behave continuously at the various interfaces of the control volumes. I face challenges in meshing this, as each part of the control volume needs to be meshed differently, and Ansys indicates that it cannot track the record. I have an important question: "HOW CAN WE TEST THE MESH INDEPENDENCY IN SUCH CASES?"

CASE B: If I run the model using the "none" sharing topology, it simplifies my work and allows me to mesh the entire body properly compared to CASE A. The only issue I face is that when I obtain the results and take cross-sections, the mass flux does not match. I believe this is due to the non-conformal meshing. One way I found Ansys can handle this is by using the interface option and creating coupled walls. However, as you can see below, there are numerous contact options available, which makes the entire process laborious. Is there any easier way to achieve this? At least in this case, I can reduce the mesh size and compare mesh independence.

Thank you for your help.

 

Thanks

 

[Rob]

If you don't use the Merge Topology option chances are you won't (and can't) get a conformal mesh. However, if the domain is meshed well and you don't have massive jumps in cell size of the interface zones the fluxes should balance fine with a nonconformal mesh. 

[skdubey]

 

@Rob thanks but I know and this is bit annoying becasue If I choose the merge option, ansys mesh is not able to track my record and if there is any change in the meshing setting which generally take atleast 7 to 8 hr to build whole get spoil. I am not sure if there is any better way to do grid independence for such case..... there is no big jump bit it is noticeable. 

 

How does the coupled wall setting (fluent) in the case of none sharing topology will help :https://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node186.htm

Any suggestion? 

 

 

[Rob]

Coupled walls are for when you have cells on both sides of a wall, ie it's part of a conformal mesh. Rather than use ancient documentation I strongly recommend using the Help system within the software. 

Without seeing any of the mesh & geometry I can't comment. Assuming you do have many cell zones and also added several local sizing commands and/or size functions Ansys Meshing has to evaluate all surfaces/settings to all other surface/settings prior to creating the mesh: that tends to slow things down. 

Re mesh independence, have you read up on mesh adaption (refinement) in Fluent? For mesh creation, will Fluent Meshing speed things up?