CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > ANSYS Meshing & Geometry

[ICEM] turbine mesh refinement

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   June 10, 2011, 14:20
Default turbine mesh refinement
  #1
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
Hi,

I have been trying to mesh a fairly highly twisted turbine blade using a standard o-grid configuration and have managed to get a fairly good coarse mesh. However, i need to simulate using a low y+ mesh and so have been trying to refine the pre-mesh, especially in the o-grid (the hub and casing should be relatively straightforward ...).

however, in setting the edge pre-mesh values to sizes which i know are close to what i need for y+ = 1, i start to get inverted cells especially near the trailing edge. I wondered if my blocking and associations are not conducive to this (see o-grid) at TE, where the o-grid starts quite far to the rear of the physcial TE.

was wondering if you guys have any advice, or comment on the mesh / blocking in general.

PS have previously been using gambit and so am still coming up to speed with icem, so any advice would be appreciated.

cheers
jonathan
Attached Images
File Type: jpg screen.jpg (95.1 KB, 89 views)
File Type: jpg screen1.jpg (96.3 KB, 68 views)
File Type: jpg screen2.jpg (97.8 KB, 71 views)
File Type: jpg blovking.jpg (83.0 KB, 56 views)
Jonathan is offline   Reply With Quote

Old   June 12, 2011, 18:34
Default "corners"
  #2
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 35
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
The Ogrid works best when at the "corners" of the rounded portions. For the leading edge, you have this correct. But for the trailing edge, your Ogrid connects along the airfoil long before the corners. This leads to distorsion at the actual corners...

My suggestion would be to slide those verts back...
__________________
-----------------------------------------
Please help guide development at ANSYS by filling in these surveys

Public ANSYS ICEM CFD Users Survey

This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)...

CFD Online Users Survey
PSYMN is offline   Reply With Quote

Old   June 13, 2011, 03:08
Default
  #3
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
Hi Psymn

thanks - originally i had an issue with the blocking / premesh not recognising the trailing edge if i unassociated the block vertices from the points they are associated with on the blade geometry. I could not figure out why the mesh 'collapsed' there when i did, and did not follow the curvature of the trailing edge surface. So, to work around this - i have splined out the blocking and it seems to stick to the blade TE surface ok now.

in refining the premesh now to get BL elements (1st cell thickness 0.01mm), i am getting inverted cells in the tip gap region. see attached - dont really understand why, if both the upper and lower surfaces of the tip gap block are associated to the top of the blade, and the projection of the blade onto the casing?

again, near the inlet where i have specified a first element thickness of around the same (0.02mm) the premesh appears to collapse the first cell in a isolated region for some reason. Again, the upper and lower edges of the block are associated with the curves wich are defining the tip gap region and i have specified the edge mesh refinement to copy to all parallel edges, so it seems a little odd ...

otherwise, this is a great tool - esp the hexa tool for complex geometries!

cheers
jonathan
Attached Images
File Type: jpg inverted1.jpg (93.8 KB, 50 views)
File Type: jpg inverted2.jpg (95.1 KB, 50 views)
Jonathan is offline   Reply With Quote

Old   June 13, 2011, 05:49
Default
  #4
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
hi Psymn,

had a bit of free time so just thought I would have a go at moving the o-grid vertices back towards the TE as you suggested - but still get a degenerate TE mesh for some reason - have attached a few images - so not sure why this happens ...

i am fairly certain this must be an association issue, but i dont understand why this should happen? The associations are that the pressure / suction / TE curves are associated as a compound entity to the three corresponding hexa edges. this is different to the LE where the geometry curve corresponds better to the size of the hexa block, and i can associate the LE curve directly with the front edges of the hexa block. having moved the verts near the TE backward, i cannot do this at the TE - perhaps this is the issue??

any advice is much appreciated! thanks v much,
cheers
jon
Attached Images
File Type: jpg new.jpg (94.8 KB, 39 views)
File Type: jpg new1.jpg (95.2 KB, 31 views)
File Type: jpg new2.jpg (95.8 KB, 27 views)
File Type: jpg new3.jpg (73.3 KB, 20 views)
Jonathan is offline   Reply With Quote

Old   June 17, 2011, 11:17
Default Associations...
  #5
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 35
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
Yup, it is all about the associations...

They are surface projected edges, which usually means they project to the nearest surface. But in your case, you had probably had issues with mesh projection jumping across your blade (from the convex to concave side), so you went in and applied a "Face to Surface" association. The problem is that surface didn't go all the way back to the rear of the airfoil.

To check these things, right click on edges to display "Projected Mesh Shape". You will probably see that those corner edges suddenly jump back to where you ended up placing the vertex... sliding the verts worked as a psudo fix for you because you happened to move those edges to where they were projecting... You can confirm this by turning on "Faces" and right clicking to display "Face Projection". At 13.0, it just shows a big colored face, but in 14.0, it will actually be color coded with more meaning.

Anyway, lets fix it. There are many ways... You could split that face along the edge of the surface so things line up... You could remove the association and fix the origial prjection problem with edge splits... You could break up the two sides of the airfoil by part and use associate to PART (each part can have multiple surfaces)...

Since you have already split the edges to fit the airfoil, changing the surface associations back to "Closest" seemed to be best solution.

1) Reduce your Index control (make it easier to select faces). Right click on Blocking (or use hotkey i) to turn on the index control. Increase O3 to 1-1 (assuming you created the OGrid in that direction).

2) To reduce clutter, turn off everything else in the tree. You don't need to see surface or curves or mesh or even faces to select a face, they just get in the way. All you should be looking at is edges.

3) Blocking => Associate => Face to Surface, select Closest and pic the two faces that were associated to surface before.

4) Turn on the premesh to make sure it is working properly now...

Post a pic if it worked ;^)
__________________
-----------------------------------------
Please help guide development at ANSYS by filling in these surveys

Public ANSYS ICEM CFD Users Survey

This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)...

CFD Online Users Survey
PSYMN is offline   Reply With Quote

Old   June 17, 2011, 14:57
Default
  #6
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
Hi Psymn,

great - thanks - reassociating the faces to the surfaces using the 'closest' condition worked well.

attached are a couple pictures of the remeshed TE ...

cheers
jonathan
Attached Images
File Type: jpg fixed1.jpg (100.7 KB, 58 views)
File Type: jpg fixed2.jpg (93.6 KB, 41 views)
Jonathan is offline   Reply With Quote

Old   November 8, 2011, 12:09
Default
  #7
Far
Super Moderator
 
Far's Avatar
 
Sijal Ahmed Memon (turboenginner@gmail.com)
Join Date: Mar 2009
Location: Islamabad Pakistan
Posts: 3,914
Blog Entries: 6
Rep Power: 38
Far will become famous soon enoughFar will become famous soon enough
Send a message via Skype™ to Far
is this a J-type grid?
Far is offline   Reply With Quote

Old   November 8, 2011, 12:51
Default
  #8
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
no, o-grid
Jonathan is offline   Reply With Quote

Old   November 8, 2011, 13:06
Default
  #9
Far
Super Moderator
 
Far's Avatar
 
Sijal Ahmed Memon (turboenginner@gmail.com)
Join Date: Mar 2009
Location: Islamabad Pakistan
Posts: 3,914
Blog Entries: 6
Rep Power: 38
Far will become famous soon enoughFar will become famous soon enough
Send a message via Skype™ to Far
O-grid in the vicinity of blade and what about the rest of the domain? Is it a H-type, J-type, L-Type or mix?
Far is offline   Reply With Quote

Old   November 8, 2011, 15:06
Default
  #10
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
yes, as you say, o-grid around the blades and h-grid in the rest of the domain. there are a couple pictures of the coarse mesh at the start of the thread as far as i recall.

regards
jonathan
Jonathan is offline   Reply With Quote

Old   November 8, 2011, 15:09
Default
  #11
Far
Super Moderator
 
Far's Avatar
 
Sijal Ahmed Memon (turboenginner@gmail.com)
Join Date: Mar 2009
Location: Islamabad Pakistan
Posts: 3,914
Blog Entries: 6
Rep Power: 38
Far will become famous soon enoughFar will become famous soon enough
Send a message via Skype™ to Far
did you try to reach the mesh quality produced by TurboGrid? at the moment I am trying to do this, since there are many geometries where automatic templates may not work.
Far is offline   Reply With Quote

Old   November 8, 2011, 15:13
Default
  #12
Senior Member
 
Join Date: Mar 2010
Location: Cape Town, SA
Posts: 153
Rep Power: 8
Jonathan is on a distinguished road
hi - not sure exactly what you mean - but if i understand correctly you are asking whether i was able to reproduce equivalent mesh quality with ICEM as one would have by using TurboGrid ... ?

No, i meshed everything in ICEM for this project, and in actual fact, for the bulk of the work, i used Gambit ...

rgds
jon
Jonathan is offline   Reply With Quote

Old   November 9, 2011, 10:08
Default J Type grid
  #13
Far
Super Moderator
 
Far's Avatar
 
Sijal Ahmed Memon (turboenginner@gmail.com)
Join Date: Mar 2009
Location: Islamabad Pakistan
Posts: 3,914
Blog Entries: 6
Rep Power: 38
Far will become famous soon enoughFar will become famous soon enough
Send a message via Skype™ to Far
@PSYMN

Can we create the J-type for this geometry, and what should be the blocking strategy
Attached Images
File Type: jpg J-blocking.JPG (21.9 KB, 34 views)
Far is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Moving mesh Niklas Wikstrom (Wikstrom) OpenFOAM Running, Solving & CFD 122 June 15, 2014 06:20
snappyHexMesh not refining surfaces Hydro1004 OpenFOAM 3 August 29, 2012 11:56
Convergence moving mesh lr103476 OpenFOAM Running, Solving & CFD 30 November 19, 2007 15:09
Turbogrid Shroud-Hub Mesh Measure Limits. Turbine Santiago Orrego. CFX 0 July 27, 2007 13:30
Icemcfd 11: Loss of mesh from surface mesh option? Joe CFX 2 March 26, 2007 18:10


All times are GMT -4. The time now is 02:28.