I am modeling a bridge on ANSYS APDL using shell and beam elements. The beams and slab are connected using MPC184 (constrained element).
After I am done modeling and meshing I merge nodes with 0.001 tolerance yet I still get an error saying meshing will produce errors. even when i tried to merge all I still get a warning / error message
After Nummrg,all / Nummrg,NODE .... future commands which depend on the node to solid model connectivity (meshing, mesh, clearing, solid boundary condition transfer, etc.) may not operate properly.
My results are wrong. Is there a step that I messed?
Dear Zeinah,
I feel after doing NUMMRG,ALL only the nodes are getting merged not the associated lines/areas (there may be two coinciding lines/areas at the same location which are only sharing one common node.
To confirm this:
1. you may do FEM validity checks (https://femci.gsfc.nasa.gov/validitychecks/).
2. Select one node of the extreme end of the model and enter esln$nsle$eplot in the command line of ansys apdl. On repeating this command again and again (till the full model appear on screen) you may get the problem associated with model.
3. Even on seeing the displacement plot or stress plot you would be able identify the issue with the model.
Hope that is of some use.
Regards
NC
i had the same problem. Maybe its a bug or the nodes are exceeding the limit of ANSYS.You can use command allsel, below, area(any entity to which the nodes are attached) for selecting only nodes attached to the interested entity.
Naveen Bhatt thanks for your help,
this is my model without meshing, to model the deck slab area I joined many areas (some are small, some are large). then i merged nodes and got the error
do you think it is wrong if two lines are sharing one common node?
should i proceed by merging lines?
Sajid Ahammed thanks for replying, so you fixed the issue by selecting the specific nodes you want to merge?
because my reactions at the supports are wrong
Zeinah Mohamad Elnassar yes just try to select the nodes and merge it. You can watch some tutorials about this in you tube as a reference.
Zeinah Mohamad Elnassar Honestly it is difficult to understand the problem in your model from here. But between two adjacent areas all nodes which fall on common line should be shared (not one node of line), see attached figure. As Sajid Ahammed is mentioning you have to select everything (or the area of interest) before giving NUMMRG command (it is not a bug, ansys commands work only on active entities i.e. which are selected) and not on the inactive (unselected) entities.
Hope I understood your problem and gave some useful input.
Regards
NC
It is a “warning”, not an “error” message, so it only warns the user to pay attention because something is happening in the analysis. As far as understand, the message simply tells you that a node, which before merging belonged to only one CAD entity, after merging is shared by more than one geometrical entity – which is exactly the purpose of merge command. This was also explained very well by Naveen Bhatt.
I think that if areas are glued before meshing, nodes will be shared automatically at common boundaries. Anyway, besides this point, I have two comments:
- Why to connect shell/beams with MPC184 (constrained element)? Have you tried simply to connect them at a node? They should be compatible (i.e. same degrees of freedom);
- Why to use such small elements for the slabs? Normally, small elements should be only be used in regions where high stress variations are expected, otherwise they will increase the computational cost unnecessarily;
Denis Benasciutti thanks for your help.
1. I connected shell and beam with MPC184 to model the rigidity/composite action between the concrete slab and steel girders. the MPC184 rigid link is used to model the steel shear studs that are extended from steel girder to connect to the concrete slab.
2. the concrete railings (shell elements) in the concrete bridge slab (shell element) have small gaps (50mm) between them. I did not model the railings as one continuous shell element throughout the bridge length. they are five segments having 50mm gaps between them so I let ANSYS mesh it and this is the mesh that it gave the railings. Then I meshed the slab accordingly in a way that the slab mesh would match the railings mesh. ( I attached top view showing small gaps)
Thing is, although it says it is a warning it does affect the reaction forces at the nodes. The reactions I get seem wrong. This is why i think the issue could be merging the nodes.
Does gluing the areas before meshing work better than merging nodes?
Or do I do both? Glue areas pre-meshing then merge nodes post-meshing?
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity