cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Integrated mode FEA problem solving with WF3...

PAULKORENKIEWIC
1-Newbie

Integrated mode FEA problem solving with WF3...

Guru's,

I am wondering if there is a way to determine the geometry that is
causing an model not to be able to AutoGEM a mesh. I have a fairly
simply pressure load on a somewhat complex, only partially done part and
my AutoGEM seems to be having difficulty. When I tried to just run an
analysis, it failed with a fatal error regarding not being able to
create all the mesh elements. So I tried to AutoGEM the mesh before
running in hopes of it SHOWING me the problem area which I then could
most likely simply fix (as I mentioned the model is only in a
preliminary state...). Unfortunately, it's been chewing on this for
about 30 minutes now, seemingly stuck on "99.6% complete"... I have
never had any luck doing anything in independent mode, not even problem
solving geometry like Integrated mode always seems to recommend when
something goes wrong. So, I am wondering if there is a way to visually
find out just where the mesh is having it's difficulty in Integrated
mode?

BTW: is there any truth to the statement I read, or heard a while back
from somewhere, that AutoGEM a mesh beforehand, will speed up your
analysis run? Seems to me, in the end, it would be the same total time
but I've never done any comparisons...

thanks in advance...

Paul Korenkiewicz
FEV , Inc.
4554 Glenmeade
Auburn Hills, MI, 48326-1766


This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.
3 REPLIES 3

Paul,

The only way I know of in WF3 to search for meshing problems is to use the Independent mode of Mechanica. It is not as easy, but you can perform the meshing there and it will highlight the surfaces that are bad. I don't normally run the solution in the Independent mode, but I do use it often to debug the mesh or geometry. Also, check on the accuracy of the part. A large accuracy setting can pose problems for the mesher.

Wildfire 4 is suppose to fix this by giving mesh and geometry diagnostics inside ProE while using the integrated version of mechanica.

Regarding meshing beforehand, you are right that the total time is the same. Unless I am wanting to examine the mesh in a particular region, I typically let the solver create the mesh during the solution.

Hope this helps,

Steve



Paul,

If you mesh in integrated mode, you only need to do a surface mesh. You
will see nodes created densely in the problem area.


e
SPINE
Dana Coombs
Analytical Engineer
1301 Goshen Parkway
West Chester, PA 19380

Hi Paul,

In my experience, anytime the meshing process hangs at a value (like your 99.6%) your mesh will not complete correctly. The % complete should always be moving upward towards 100% at a reasonable rate. If it just keeps repeating the same % over and over, it is not likely going to complete. In some cases, I've seen it actually start moving backwards (99.6.....99.2.....88). If this happens, you're guaranteed that it won't complete the mesh. Even if your mesh completes after the 99.6% hang, your analysis will not run properly. Usually, this mesh will result in a high singularity zone and analysis results will be inaccurate in and around that zone. Also your convergence criteria might not be met.

I seem to remember a stop sign in the bottom right hand corner of Pro/E when creating the mesh. Once you see it "hang" like this, you can try hitting the stop button (granted you may have to hit it a bunch of times for the system to respond). When you hit the stop button, a menu should appear that tells you your mesh is incomplete and it asks if you still want to keep it. I've selected yes to keeping it and a partial mesh will display on the screen. In my models, I've seen chunks (technical term here) of the mesh missing like empty voids. The areas of the model that are not meshed are very likely the areas that are causing you the most trouble and these are the areas that you need to clean up.

Technically, saving the mesh before running the analysis will save time if your analysis settings are set to use the mesh that is already created. In most cases, you'd only really be saving a few minutes of time. I worked on a large assembly once that took ~40 minutes to mesh and this saved quite a bit of time to reuse the mesh but in three years of using Mechanica at that company, it was the only project where I thought it saved me time. In actual analysis time, I don't think it changed how long it took the software to solve.

As I remember it, WF4 was supposed to have some major improvements to the meshing process (much more robust). It also has some other nice enhancements that might benefit you outside of this particular issue.


Mike -


Subject: [proecae] - Integrated mode FEA problem solving with WF3...Date: Thu, 13 Mar 2008 07:13:08 -0400From: -To: -

Guru's,

I am wondering if there is a way to determine the geometry that is causing an model not to be able to AutoGEM a mesh. I have a fairly simply pressure load on a somewhat complex, only partially done part and my AutoGEM seems to be having difficulty. When I tried to just run an analysis, it failed with a fatal error regarding not being able to create all the mesh elements. So I tried to AutoGEM the mesh before running in hopes of it SHOWING me the problem area which I then could most likely simply fix (as I mentioned the model is only in a preliminary state...). Unfortunately, it's been chewing on this for about 30 minutes now, seemingly stuck on "99.6% complete"... I have never had any luck doing anything in independent mode, not even problem solving geometry like Integrated mode always seems to recommend when something goes wrong. So, I am wondering if there is a way to visually find out just where the mesh is having it's difficulty in Integrated mode?

BTW: is there any truth to the statement I read, or heard a while back from somewhere, that AutoGEM a mesh beforehand, will speed up your analysis run? Seems to me, in the end, it would be the same total time but I've never done any comparisons...

thanks in advance...

Paul Korenkiewicz FEV , Inc.
4554 Glenmeade Auburn Hills, MI, 48326-1766
-----End Original Message-----
Announcements