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

[GAMBIT] Mesh refinement for melting during welding

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

Reply
 
LinkBack Thread Tools Display Modes
Old   February 17, 2015, 16:30
Default Mesh refinement for melting during welding
  #1
Member
 
Rafal
Join Date: Aug 2013
Location: CK
Posts: 36
Rep Power: 5
vekh is on a distinguished road
Hello, I have problem with mesh creation for laser welding simulation. I'm working with 2D rectangular samples and I want to create model with VOF for conduction mode welding (so I need to exploit volumetric (w/m3) heat source). I wrote suitable UDF which taking into account both laser power and laser penetration depth (with respect to extinction coefficient and refractive index).
The obstacle is a shallow depth of laser penetration. It is the order of ~900 nm - that is approximate depth at which laser power extinguish. So I need to create mesh at certain area with cell size 100 nm size. Considering that whole sample have 4x4 mm output mesh will be enormous big. I tried to use grading, manualy dividing etc. but there is always high aspect ratio during mesh chceck in Fluent or mesh is "too big" to grab the heat source...
Do you know how to create 2D mesh with certain area mesh refinement which will be the "base" for mesh generation for outer regions?
vekh is offline   Reply With Quote

Old   February 18, 2015, 02:19
Default
  #2
Super Moderator
 
-mAx-'s Avatar
 
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,160
Rep Power: 32
-mAx- will become famous soon enough
you may use size function.
Can you post a picture of your geometry?
__________________
In memory of my friend Hervé: CFD engineer & freerider
-mAx- is offline   Reply With Quote

Old   February 18, 2015, 10:46
Default
  #3
Member
 
Rafal
Join Date: Aug 2013
Location: CK
Posts: 36
Rep Power: 5
vekh is on a distinguished road

This is my best - at the moment - mesh with my geometry. It's rectangle with symmetry at X-axis. Red dotted line is border between air and steel. 1 is air 2 is steel (steel is patched after initialization). Blue arrow is laser spot center and gravitational acceleration works in X+ direction.
At the moment 2D model with 2x3 mm dimmension is meshed with ~125k elements. I don't have huge experience with such kind of simulations but It's seems to that there is too many element like for that element size... Beside that I have weird pressure jumps from ~3e-5 to 6e8 and problem with courant number - I think that it's mesh fault.
-mAx- I do not want to being pushy but can You give me some advice where to find similar example or something of which I could learn size functions? Thanks in advance.
vekh is offline   Reply With Quote

Old   February 18, 2015, 11:19
Default
  #4
Super Moderator
 
-mAx-'s Avatar
 
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,160
Rep Power: 32
-mAx- will become famous soon enough
If your problem is mesh ratio, then I would refine in y-direction, untill aspect ratio is ok
125k elements isn't a big mesh.
You may have problem, with mesh over several millions of elements.
Let me know
__________________
In memory of my friend Hervé: CFD engineer & freerider
-mAx- is offline   Reply With Quote

Old   February 18, 2015, 20:05
Default
  #5
Member
 
Rafal
Join Date: Aug 2013
Location: CK
Posts: 36
Rep Power: 5
vekh is on a distinguished road
OK. I'm training with size functions... It's weird a bit but maybe will help ;-)
Quote:
125k elements isn't a big mesh.
You may have problem, with mesh over several millions of elements.
Agree but it's 2x3 mm testing case for VoF preparation. Target is to create 3D model with moving heat source. Approx. model size will be ~50x50x3 mm like on the picture (sorry I'm bad at paint )
The red area is the "refined" mesh for heat source path... That could be kinda huge mesh. I wonder if there will be a need to use moving/dynamic mesh to limit cells amount or other trick...
vekh is offline   Reply With Quote

Old   February 19, 2015, 02:00
Default
  #6
Super Moderator
 
-mAx-'s Avatar
 
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,160
Rep Power: 32
-mAx- will become famous soon enough
in worst case, you will work with tri elements and size function.
If your final work will handle moving mesh, you should think about a strategy to treat moving domain (remeshing, smoothing or layering), because you won't prepare a geometry with moving mesh (remeshing) like a geometry with moving mesh (layering)
Can you provide your dbs file?
__________________
In memory of my friend Hervé: CFD engineer & freerider
-mAx- 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
SnappyHexMesh for internal Flow vishwa OpenFOAM Native Meshers: snappyHexMesh and Others 24 June 27, 2016 08:54
problems using snappyHexMesh 2.1.0 on a supercomputer Sunxing OpenFOAM Native Meshers: snappyHexMesh and Others 9 September 20, 2014 09:30
Unwanted mesh created by snappyHexMesh--baffle bieshuxuhe OpenFOAM Native Meshers: snappyHexMesh and Others 2 April 23, 2014 01:59
snappyHexMesh sticking point natty_king OpenFOAM Native Meshers: snappyHexMesh and Others 2 April 17, 2014 01:24
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 11:40.