|
[Sponsors] |
August 21, 2016, 12:01 |
Help with meshing
|
#1 |
New Member
jason
Join Date: Aug 2016
Posts: 11
Rep Power: 9 |
Hi CFD-online community,
I would like to seek some advice/help about meshing the geometry attached below. It is a cylindrical domain (single body) with ship propeller inside it that has been boolean subtracted away to model the flow domain of flow around a ship propeller in FLUENT. I am using the patch conforming tetra method as the body is not sweepable and patch independent give me a warning message: "The number of unconnected mesh volumes differs from the number of bodies in that part. ". Just by changing the adv size function to proximity and curvature and relevance to medium, it gives me a max skewness mesh of 0.999... However, i believe this is not good for FLUENT as it seems fluent prefer a quality of at least 0.98. I have tried different sizing controls, virtual topology but to no avail. I will upload a picture of the highly skewed elements later but they are mostly on the curved blades, hub and a few on the tip. Would appreciate any advice here or it is ok for me to go ahead with a max skewness mesh of 0.999. will this affect convergence of my results in FLUENT? Thank you. -From a CFD newbie |
|
August 21, 2016, 20:58 |
|
#2 |
Member
Omar jumaah
Join Date: Jan 2016
Posts: 36
Rep Power: 10 |
I suggest you to try slicing the cylinder into multi parts then select all resulted bodies to form new part next mesh.
|
|
August 22, 2016, 01:47 |
|
#3 |
New Member
jason
Join Date: Aug 2016
Posts: 11
Rep Power: 9 |
Hi juumah, thanks for the help. so i should slice the body into 3 sweepable bodies and put them in the same part right?
|
|
August 22, 2016, 02:12 |
|
#4 |
Senior Member
Join Date: Apr 2014
Location: Melbourne
Posts: 584
Rep Power: 14 |
yes create a part with those bodies in it.
|
|
August 22, 2016, 02:35 |
|
#5 |
New Member
jason
Join Date: Aug 2016
Posts: 11
Rep Power: 9 |
Hi, i have managed to slice the back portion of the body (see 3rd thumbnail) but I cant seem to slice the front portion (the portion of the cylinder ahead of the tip of propeller)
the error is "Operation result in non-manifold bodies". any idea why? EDIT: I realise i cannot slice the cylinder at the propeller tip and the end of the shaft as this will not make the other bodies sweepable. |
|
August 22, 2016, 10:42 |
|
#6 |
New Member
jason
Join Date: Aug 2016
Posts: 11
Rep Power: 9 |
Hi so after some advice from kapi and jumaah, i tried to split the part into multiple bodies and then form them into 1 part. (see pic 1)
Note: I am modelling the flow over the propeller in a single ref frame/domain not using multiple ref frame method. For the body containing the propeller (or should i say the body with a boolean subtraction of the propeller), i had to use patch independent method as patch conforming doesnt allow me to create a mesh there (for some reason, maybe due to bad geometry of the propeller). I used virtual topology to try and get around this. Managed to mesh successfully. Skewness is down to high 0.8 (which is great right?) but there was an error in the message box "The selective body meshing is not being recorded, so the meshing may not be persistent on an update. If you want to record the order of the body meshing, please use the Mesh Worksheet to track the meshing steps. Please see Selective Meshing documentation for more details." Not sure if this will affect anything in FLUENT. In addition, I also tried to split the part as in picture 2. As usual, i had to specify patch independent at the body containing the propeller. The rest are meshed automatically by Ansys Meshing as before. However, by splitting the part like this, I end up with a mesh around the propeller blade (see pic 3). Will this cause a major problem in FLUENT as well? Should I reduce the mesh size around those faces to avoid this problem? Thanks Jason |
|
August 22, 2016, 19:01 |
|
#7 |
Member
Omar jumaah
Join Date: Jan 2016
Posts: 36
Rep Power: 10 |
Hey, try to draw sketches on the top then use extrude : slice material, to avoid cutting the propeller. Select them in one form part after that you can mesh.
hope it will help |
|
August 22, 2016, 20:03 |
|
#8 |
New Member
jason
Join Date: Aug 2016
Posts: 11
Rep Power: 9 |
hi jumaah, thanks for your advice. would yoi mind drawing a side view to help me understand better? i think i did something similar as in picture 2 in the direct post above yours. thanks
|
|
August 22, 2016, 20:20 |
|
#9 |
Member
Omar jumaah
Join Date: Jan 2016
Posts: 36
Rep Power: 10 |
okay, just imagine slicing the cylinder from the top by extruding sketch through it, will get five parts
1. extrude sketch 1: slice material, select the cylinder 2. extrude sketch 2: slice material, select the two resulted parts finally will get five parts select them to form one part. |
|
August 23, 2016, 00:46 |
|
#10 |
New Member
jason
Join Date: Aug 2016
Posts: 11
Rep Power: 9 |
hi jumaah, could you explain why i must split it into 5 parts instead of just 2 (a smaller cylinder surrounding the propeller inside the bigger cylinder)
thanks |
|
February 5, 2018, 06:56 |
brick mesh non-sweepable parts ?
|
#11 |
New Member
Join Date: Jul 2017
Posts: 11
Rep Power: 8 |
Hi guys,
I am trying to brick mesh this geometry https://ibb.co/jj25gH. As you can see, I have split the body in several connected volumes which are all sweepable except for the 2 close to the lateral cylinders ( red zones in the pic) . Does anyone have any suggestion about how I can also hex mesh these remaining parts ? |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[ANSYS Meshing] only automatic meshing works, | kazra | ANSYS Meshing & Geometry | 2 | February 23, 2017 11:38 |
[Gmsh] 3D coil mesh: can't create the volume? | RomainBou | OpenFOAM Meshing & Mesh Conversion | 3 | July 18, 2016 05:09 |
[Gmsh] Vertex numbering is dense | KateEisenhower | OpenFOAM Meshing & Mesh Conversion | 7 | August 3, 2015 10:49 |
[ANSYS Meshing] Icem meshing queries | Mfaizan | ANSYS Meshing & Geometry | 21 | July 30, 2015 23:10 |