I have a problem which is annoying me every now and then. Mostly when I have no time to deal with.
I create different groups to collect nodes or elements or both for prost processing, i.g., I want to create a list of results after analyzing for using in EXCEL.
Occasionally I can't use that groups because aftersaving and reloading the complete structure of model the groups are empty or did lost most of the objects.
Can anyone tell me, what's the reason of that and what shall I do to dodge it, because I have good reasons to create the groups and want to keep them. The nodes or elements do not belong to the same meshes or so and sometimes they are of different type, too.
My model is mostly a structure with PRTS, iPRTs, FEMs, AFMs and SIM.
Sometimes I slightly guess that groups in AFMs or SIM get lost faster or more often.
Am I the only one with that problem? - Thank you for your support. Every hint is welcome.
Best wishes, Michael
First question is how you group your nodes / elements: by assigning meshes elements or by choosing the bodies or with automatic grouping?
- group your elements / nodes in FEM files by automatig grouping, only this option will ensure you that your groups will be associative any time
- remember that when you not use automating grouping, you always will loose associativity of the elemetns during remeshing or 'resolving nodes/elements' in AFM
- if you want to create your own unique group which can not be covered by automatic grouping, do it in assembly fem when your elements and nodes are 'resolved' and just before running the Solver.
here some hints:
I hope it will help you
Assuming that your meshes aren't changing, your group contents shouldn't change either. However, if meshes update, you will lose any reference to those nodes/elements. That includes references to them with respect to groups and loads/boundary conditions. Here's an example:
1. A mesh exists on a face that contains node 10
2. A group exists that contains node 10
3. The geometry changed, or the mesh was edited such that an update of the mesh was required. This means that the mesh is first deleted, then recreated. Node 10 no longer exists. Therefore it can't be in the group.
Meshes on the other hand are retained after updates. That is why the previous suggestion of using automatic groups to maintain update behavior works. Automatic groups add meshes to groups, not nodes/elements. A user has the ability to add meshes to any group, not just automatically created groups.
I can't think of any other reason why nodes and/or elements would be removed from groups without the user actually removing them.
Simulation Product Management
Simulation and Test Solutions
Siemens Industry Sector
Siemens Product Lifecycle Management Software Inc.
thank you for your response. I will answer both answer threads of you in the same entry.
First of all, of course I already know how to build groups of elements and mostly I use
- "manually created MESH groups", because they do not change after renumbering the elements or nodes of a mesh or
- "Automatically created groups of material or property", as you suggest, dear TomekZeWschodu. These groups mostly help to work in post processor in a fine way.
But sometimes I need special groups with elements which do not belong to the same property or material or mesh. Then the methods described above do not fit. Here I select the nodes or elements into a group manually by picking it or giving their labels.
I know that remeshing or mesh updating will destroy groups based on labels of elements or nodes. But I did not manipulate my model. If the model would be modified in a "lower" part like FEM or PRT I should loose RBE-elements and nodes defined in AFM by node-to-node-definition using "native FEM nodes", too. And the "native" elements and nodes from FEMs should have new numbers. But the numbers of nodes and elements did not change and my RBEs are still there.
Furthermore I have constraints defined on NODES directly in SIM and they are still existing, too. They should vanish by a modify process. These elements and nodes are the same which vanish in my groups after "Save, Close and Reopen".
Therefore I do not believe that my model was updated ore modified. But nevertheless the groups are empty.
As a new hint:
I have found some elements created in AFM directly which were still in their group. (In my case I did create 23 groups with thousands of elements).
Maybe FEM elements are affected more often?! - I do not know.
Is there another mechanism which destroys the membership of an object in a group?
EDIT: 2016-08-25, 10:57 a.m.
I did load my model in NX10 and there all groups a filled with elements. It must be an error in NX8.5.2, which I use normally.
You're explanation makes sense to me and I agree that if nodes/elements in a group are being lost, but not in loads/boundary conditions that use those same nodes/elements, then there is a problem with groups. Hopefully this problem is addressed in NX 10 (per your testing compared to 8.5.2). Should you observe this problem again in NX 10, then please open a call with GTAC.
Mark, all right, I will do so. Michael