|
[Sponsors] |
March 22, 2011, 10:01 |
ICEMCFD Meshing
|
#1 |
New Member
Sumeet Kumar
Join Date: Mar 2009
Posts: 21
Rep Power: 17 |
Hi ALL,
I am trying to mesh a geometry with hex dominant cells, Please go through the snaps I have posted here. I have a cicular pipe on one end and a rectangular cross section box on the other end. The transition is from a circular cross-section to a rectangle. Hence there is a planar diffuser in between. I am trying to mesh this geometry using the blocking method in ICEMCFD. Geometry http://i1228.photobucket.com/albums/...CEM/Geo1-1.png http://i1228.photobucket.com/albums/.../ICEM/Geo2.png Premesh http://i1228.photobucket.com/albums/...EM/Premesh.png Blocks with O-grid http://i1228.photobucket.com/albums/...idBlocks-1.png Hex Cells - Mesh Quality - Aspect Ratio http://i1228.photobucket.com/albums/...spectratio.png Hex Cells - Mesh Quality - Quality http://i1228.photobucket.com/albums/...EM/Quality.png I am getting very high aspect ratio hex cells in the transition part, Can anyone assist me in having a better quality mesh. I could not figure out a way to mesh it properly in the diffuser part. Fluent solver reports failed mesh checks: Checking wall distance. WARNING: The mesh contains high aspect ratio hexahedral or polyhedral cells. The default algorithm used to compute the wall distance required by the turbulence models might produce wrong results in these cells. Please inspect the wall distance by displaying the contours of the 'Cell Wall Distance' at the boundaries. If you observe any irregularities we recommend the use of an alternative algorithm to correct the wall distance. Please select /solve/initialize/repair-wall-distance using the text user interface to switch to the alternative algorithm. P.S. : I am quite new to ICEMCFD Meshing. |
|
March 22, 2011, 15:11 |
|
#2 |
New Member
Claudio
Join Date: Feb 2011
Posts: 20
Rep Power: 15 |
I am quite new to Icem but I'll try to help.
It looks to me that ICEM is not correctly placing the nodes on the O-Grid internal edges. It can be more clear if you would: right click on Pre-mesh>Scan planes then you should swap the planes and find out what is going on inside it. I can guess some solutions... but keep in mind that I am not an expert: 1)Extend the split on the first O-grid (the one in the round pipe) and connect it with the square section trough a more clean block structure. Check for example the one which ICEM does automatically when you create an O-Grid by selecting blocks. 2) Simply move the internal vertices around and see what happens 3) Use a different law for the nodes on the internal edges. I hope that this was helpful. |
|
March 24, 2011, 14:40 |
|
#3 |
New Member
Adrian Dunne
Join Date: Jan 2011
Location: Ireland
Posts: 26
Rep Power: 15 |
you might want to try put in tetrahedral elements in that transition region.
|
|
March 24, 2011, 16:46 |
Fix topology and snap to fit...
|
#4 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Sorry, your topology is wrong and then you didn't fit your bad topology before creating your Ogrids...
Lets look at topology first... You said it yourself, it is just a pipe with a transition... Therefore you just need a pipe type topology... Create a single block for the length of the pipe... Associate the ends of the block to the ends of the pipe... (one end is associated to the circle, the other to the square at the opposite side...) Then split the block twice across its length... Associate these new splits, one with the circle on the one side of the transition and the other with the rectangle on the other side of the transition... Actually snap fit these verts into place before creating the Ogrid... The quality will already be decent except for the "corners" of the round pipe. Then apply the ogrid... Select all three blocks for the Ogrid, also select the inlet and outlet faces... Apply. Now your blocking is done. Adjust your edge parameters to have the right mesh distributions, etc... |
|
April 21, 2011, 01:22 |
|
#5 |
New Member
Join Date: Jan 2011
Posts: 24
Rep Power: 15 |
hi,
i too have same type of queries.. i have an inlet port in a tank through which i give my feed.. and also an outlet port. once i do hexa meshing, first i finish it for the tank region. then i went for the two ports. i have meshed it this way.. like created inidividuals blocks for the two ports and meshed it. do i have to include a block for the same in the tank cos i have the outlet attached to tank and then extruded out. so is it right that i do meshing for the ports separately and then the curved part on the tank surface separately.. i am new to this meshing thing.. please figure it out. and also i have a bend in my inlet port which is then extruded straight for few mm. now how do i mesh this?? |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Question about meshing 2D airfoil in icemcfd | cxcxcx0505 | ANSYS | 2 | May 26, 2010 19:09 |
ANSYS ICEMCFD Streaming Meshing Demos | ANSYS | Main CFD Forum | 2 | October 22, 2004 07:55 |
APL/A* environment in ICEMCFD HEXA/TETRA meshing | naga | Main CFD Forum | 0 | July 23, 2004 07:06 |
Singularity of grid?Volume meshing vs face meshing | Ken | Main CFD Forum | 0 | September 4, 2003 11:09 |
Volume Meshing & Face Meshing? singularity of grid | ken | FLUENT | 0 | September 4, 2003 11:08 |