[Abdullah khan]

hi guys !! I hope you are doing good. I am working on a two domain problem of conjugate heat transfer between fluid with certain velocity and solid domain as pipe thickness. I have attained mesh quality above .5 but the problem lies in fluid domain. THE solid domain mesh is ok but when i check the fluid domain mesh it shows some surfaces not proper meshed. when I open it in cfx I can’t define proper boundary condition because of that empty windows or slots in fluid domain. I am using ICEM blocking method to create unstructured mesh. Please refer to pictures for better idea. solid domain is made hidden only fluid domain is shown in pictures. any suggestion or methodology of problem solving ??? thanks in advance Attachment 520502.jpg

[SPH_CFD]

Originally Posted by Abdullah khan

hi guys !! I hope you are doing good. I am working on a two domain problem of conjugate heat transfer between fluid with certain velocity and solid domain as pipe thickness. I have attained mesh quality above .5 but the problem lies in fluid domain. THE solid domain mesh is ok but when i check the fluid domain mesh it shows some surfaces not proper meshed. when I open it in cfx I can’t define proper boundary condition because of that empty windows or slots in fluid domain. I am using ICEM blocking method to create unstructured mesh. Please refer to pictures for better idea. solid domain is made hidden only fluid domain is shown in pictures. any suggestion or methodology of problem solving ??? thanks in advance Attachment 52050Attachment 52051

Did you use Ogrid type for this problem? Could you please show me cross-section of pipe with say clearly fluid and solid domain, I can give you my advice

[Abdullah khan]

yes i have used o grid and also modified o grid in the fluid domain with different edge distances

[SPH_CFD]

Originally Posted by Abdullah khan

yes i have used o grid and also modified o grid in the fluid domain with different edge distances

Maybe my experieence for blocking is not so high. However, I try it and feel it seems more diffifcult to control with complicated geometry. I recommend to generate 2 different meshes and using non-conformal or you should try other generation sofwares.

[Abdullah khan]

Originally Posted by SPH_CFD

Maybe my experieence for blocking is not so high. However, I try it and feel it seems more diffifcult to control with complicated geometry. I recommend to generate 2 different meshes and using non-conformal or you should try other generation sofwares.

kindly if you can explain your answer in a bit detail i am student working on my research i am not professional can you guide me how to generate 2 different meshes and using non- conformal method and also which software for meshing you can suggest

[Antanas]

Originally Posted by Abdullah khan

hi guys !! I hope you are doing good. I am working on a two domain problem of conjugate heat transfer between fluid with certain velocity and solid domain as pipe thickness. I have attained mesh quality above .5 but the problem lies in fluid domain. THE solid domain mesh is ok but when i check the fluid domain mesh it shows some surfaces not proper meshed. when I open it in cfx I can’t define proper boundary condition because of that empty windows or slots in fluid domain. I am using ICEM blocking method to create unstructured mesh. Please refer to pictures for better idea. solid domain is made hidden only fluid domain is shown in pictures. any suggestion or methodology of problem solving ??? thanks in advance Attachment 52050Attachment 52051

Check associations. Also execute Check/Fix blocks and fix inverted blocks.

[SPH_CFD]

Originally Posted by Antanas

Check associations. Also execute Check/Fix blocks and fix inverted blocks.

The problem is that associtation of ICEM is working not very well. We can see this geometry with many curves, it means you must cut many blocks and association to catch curves at body. I also have a similar problem but ICEM seems not to understand my association and that is a point.

@Abdullakhan: If you still want to try blocking like that, I suggest you to be sure about association edges along tube (not only cross-section)

[Abdullah khan]

Originally Posted by SPH_CFD

The problem is that associtation of ICEM is working not very well. We can see this geometry with many curves, it means you must cut many blocks and association to catch curves at body. I also have a similar problem but ICEM seems not to understand my association and that is a point.

@Abdullakhan: If you still want to try blocking like that, I suggest you to be sure about association edges along tube (not only cross-section)

the method i use is i create mid line and extrude face along the curve and then associate every edge to the curve in both fluid and solid domain so that no block is remain without association the second o grid in the fluid domain is created after association .the question is that if i attain mesh quality up to standard the why is those empty slots in my mesh in fluid domain more ever along tube you mean that longitudinal edges to surface or curves association ?

[Antanas]

Originally Posted by Abdullah khan

the method i use is i create mid line and extrude face along the curve and then associate every edge to the curve in both fluid and solid domain so that no block is remain without association the second o grid in the fluid domain is created after association .the question is that if i attain mesh quality up to standard the why is those empty slots in my mesh in fluid domain more ever along tube you mean that longitudinal edges to surface or curves association ?

Enable 'Show associations' for edges and make sure that everything is associated properly. Also 'Run Check/Fix blocks' and 'Fix Inverted Blocks'. Did you do it?

[Abdullah khan]

Originally Posted by Antanas

Enable 'Show associations' for edges and make sure that everything is associated properly. Also 'Run Check/Fix blocks' and 'Fix Inverted Blocks'. Did you do it?

yes i have tried it still not working

[Antanas]

Originally Posted by Abdullah khan

yes i have tried it still not working

Share your files

[Abdullah khan]

Originally Posted by Antanas

Share your files

https://www.dropbox.com/s/j3axqq311z...2016.wbpj?dl=0
here is the drop box link i shared work bench file including geometry and icem meshing

[Abdullah khan]

Originally Posted by Antanas

Share your files

https://www.dropbox.com/s/8tjqgi1pfa...ect_cache?dl=0 here the remaining files

[Abdullah khan]

Originally Posted by Antanas

Share your files

https://www.dropbox.com/s/40i7lnt3w6...Geom.agdb?dl=0
https://www.dropbox.com/s/t2uow8oalwwcs8w/ICM.blk?dl=0
https://www.dropbox.com/s/l1bivwg8di6qb8c/ICM.prj?dl=0
https://www.dropbox.com/s/g1ushg8t8m6o0p4/ICM.tin?dl=0
https://www.dropbox.com/s/7aps9bvea02g14g/ICM.uns?dl=0
https://www.dropbox.com/s/8tjqgi1pfa...ect_cache?dl=0
https://www.dropbox.com/s/j3axqq311z...2016.wbpj?dl=0
here is the files you can check the file you want those are drop box links let me know if you have any problem in downloading these files

[Abdullah khan]

Originally Posted by Antanas

Share your files

i hope to hear back from you soon

[Antanas]

Originally Posted by Abdullah khan

i hope to hear back from you soon

You have bad topology. Unnecessary and inconsistent surface splits, different curve directions (see first image). It's hard for mesher to decide how they all should be connected. I worked a little on your geometry and got good mesh. (see second image).

You can download my workbench 17.2 project here.

[Abdullah khan]

i cant open the work bench file you sent i am using ansys workbench version 15 if you can email me the mesh file 08pwmec2782@gmail.com more ever what you suggest how can i fix the problem how can i improve typology,curve directions and make it consistent surface split i design the geometry in PRO -E i am unable to design it in ansys design modeler because of the sweep didn't work in the second plane perpendicular to the first plane it couldn't sweep all at once upload the mesh file file only so that i can check and see if i can proceed future i am attaching my geometry dimension if you can model it for me ???

[Abdullah khan]

Originally Posted by Antanas

You have bad topology. Unnecessary and inconsistent surface splits, different curve directions (see first image). It's hard for mesher to decide how they all should be connected. I worked a little on your geometry and got good mesh. (see second image).

You can download my workbench 17.2 project here.

i am waiting for your kind reply

[Antanas]

Originally Posted by Abdullah khan

i am waiting for your kind reply

Fluent Mesh File

[Abdullah khan]

Originally Posted by Antanas

Fluent Mesh File

i am using CFX for my analysis i cant define fluid and solid domain separately in CFX .the mesh file you sent is for fluent .