|
[Sponsors] |
[ICEM] SAE Automotive reference model - 2D to 3D blocking concerns |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
September 12, 2015, 00:39 |
SAE Automotive reference model - 2D to 3D blocking concerns
|
#1 |
Senior Member
Ovi
Join Date: Oct 2012
Location: Sydney, Australia
Posts: 166
Rep Power: 13 |
Hello CFDOnline,
Hope you are all progressing well with your individual challenges in CFD. I have recently started to mesh the rear geometry of a simplified SAE car model with an underbody recess, which represents a simplified diffuser. I have tried a new blocking and meshing strategy for this geometry and initialised with 2D planar blocks. The difficulty lies in the following - a) Trying to extrude all the 2D blocks into 3D volume blocks for volume meshing. The extrusion using all three methods in ICEM 15.0 (interactive, fixed distance, along curve) fails sometimes and it's been very difficult to fill out the entire domain. b) Connectivity - When moving from a larger block for the entire domain and then splitting down to smaller regions, the connectivity and interface between blocks are automated. What happens when I have all of these adjacent blocks but they are essentially separate entities? Are they going to function well when creating volume meshes? I've attached the geometry and the blocking files and you're welcome to make some suggestions to improve the current approach. Ultimately, I would like to transfer to a full diffuser with a sharp junction and a kick angle. Thank you
__________________
-- Mechanical Engineering Sydney, Australia |
|
September 13, 2015, 01:22 |
|
#2 |
Senior Member
Ovi
Join Date: Oct 2012
Location: Sydney, Australia
Posts: 166
Rep Power: 13 |
Here are some more images of the new blocking approach.
I will eventually have to introduce the kick angle of the diffuser into the model however, the current version is facilitating this new blocking strategy with the 2D to 3D transformations. Please share your ideas here and your experiences of top-down or bottom-up approaches to blocking and meshing.
__________________
-- Mechanical Engineering Sydney, Australia |
|
September 20, 2015, 07:18 |
|
#3 |
Senior Member
Ovi
Join Date: Oct 2012
Location: Sydney, Australia
Posts: 166
Rep Power: 13 |
I have managed to split, move and reorient all the blocks so that they closely follow the geometry, however, there appears to be incorrect associations of points and edges. The blocking looks fine but the pre-mesh
is horrible. Can someone please point to any obvious errors? Is this the correct approach to generate volume mesh from the initial 2D blocking? Thanks in advance
__________________
-- Mechanical Engineering Sydney, Australia |
|
September 20, 2015, 08:09 |
|
#4 | |
Senior Member
Join Date: Feb 2011
Posts: 495
Rep Power: 18 |
Quote:
|
||
September 20, 2015, 08:23 |
|
#5 |
Senior Member
Ovi
Join Date: Oct 2012
Location: Sydney, Australia
Posts: 166
Rep Power: 13 |
I'm not really sure but that might actually be the case.
However, shouldn't this be fairly easy to correct if I just delete all associations and manually go through all the vertices and edges again? Additionally, I've got another question regarding 2D to 3D conversions. Image 1 - shows the model after 2D planar blocking and then 2D-3D conversion. Then I just extruded the faces around the cube to form the other blocks. Image 2 - revealed that although the blocks are there, it doesn't generate any meshes in that region! Really confused about this. Does this mean that we would need to fill an entire cross-section or profile with 2D planar blocking and then simply do 2D-3D conversion? Does that mean extruded faces aren't working? Thanks for your responses
__________________
-- Mechanical Engineering Sydney, Australia |
|
September 20, 2015, 08:44 |
|
#6 | |
Senior Member
Join Date: Feb 2011
Posts: 495
Rep Power: 18 |
Quote:
If you don't associate blocking to geometry (often edges to curves) the mesh won't be generated. |
||
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Wrong flow in ratating domain problem | Sanyo | CFX | 17 | August 15, 2015 07:20 |
[ICEM] SAE Reference Notchback model - diffuser mesh and index control | Crank-Shaft | ANSYS Meshing & Geometry | 1 | May 7, 2015 22:24 |
[ICEM] SAE Reference Notchback Model Blocking Strategy | Crank-Shaft | ANSYS Meshing & Geometry | 16 | January 27, 2015 23:10 |
Overflow Error in Multiphase Modelling with Two Continuous Fluids | ashtonJ | CFX | 6 | August 11, 2014 15:32 |
Error with Wmake | skabilan | OpenFOAM Installation | 3 | July 28, 2009 01:35 |