Cancel
Showing results for
Did you mean:

# Get deformations in coordinate system deflected by deformation results

Phenom

Hello Experts,

is there a way to get deformation results transformed into a coordinate system base which is also deflected by deformation results?

Aim: I have a deformed structure which is loaded with different temperature values yielding in temperature deformations. In my structure there is a (former) cylindrical hole and I want to get information about ovalization and out-of-plane-deformation of the hole. I have defined a local cylindrical coordinate system to describe my center of deformation in undeformed state. But my structure suffers from global deformations yielding in movement and rotation of my hole. That means the supposed center of my hole is shifted out of center of coordinate system and the "bases axes" of hole tilt and I determine a global movement in the radial deformation which I'm not interested in.

Can I define an "entity" or a coordinate system to get rid of that superposition of global movement and rotation? - I don't want to do it in Excel because I want to save pictures of the relative deformation state to visualize the effects I 'm interested in.

Any suggestion will help. With best regards, Michael

4 REPLIES

# Re: Get deformations in coordinate system deflected by deformation results

Siemens Phenom

Michael,

Here's something to try. Create a field of your displacements about the cylindrical face. First generate a contour plot of displacements. Then use Identify Results to create a Cartesian-Cartesian field. You can limit the data queried in Identify Results to just the nodes on the cylindrical face to minimize the size of the field. If you have the FEM as the work part when you created the contour plot, then the field generated in Identify Results will be written to the FEM. That's important for the next step. If you don't, you will have to export the field and import it into the FEM.

In the FEM, surface coat the cylindrical face elements, then copy and translate the surface coat elements. Translate them by (0,0,0) so that you have coincident nodes. You can delete the surface coat mesh at this point. Next, translate the nodes in the copied mesh using the displacement field. This will give you the deformed position of the nodes on the cylinder face. Now you can construct a coordinate system using the deformed nodes as a guide. Once you have the coordinate system, you can delete the copied mesh. Go into post processing again, and now use the coordinate system you created from the deformed mesh.

Regards,

Mark

Mark Lamping

Simulation Product Management

Simulation and Test Solutions

Siemens Industry Sector

Siemens Product Lifecycle Management Software Inc.

mark.lamping@siemens.com

www.siemens.com/plm

# Re: Get deformations in coordinate system deflected by deformation results

Phenom

Mark,

thank you for response.

I wrote a long description of my attempts but my browser window crashed. I'm going crazy with that stuff at Friday afternoon!!!!!

So here a short summary:

1. I could not find a way to create cart-cart-field from "identify results". Is it a new function in NX10 or 11?

I use NX8.5.2.

2. I made a model with RBE2-triad to visualize the deformation of my circle in a simplified shell model.

created 2 "moved-by-deformation"-triads according to 2 Load cases in my model and defined deformed coordinate systems with the moved (deformed) triads.

By scaling 1:1 I found in Post processing my original RBE2-triad exactly at the position of my moved (deformed) triad, okay so far.

After transforming results into the appropriate deformed coordinate system I could determine the modification of nodal coordinates but the displacements got values I did not expect. I expected constant values (in best case of zero).

Could you have a look at my model and excel sheets? I zipped it (KS_deformed.zip; 2MB).

Thank you and best wishes, Michael.

# Re: Get deformations in coordinate system deflected by deformation results

Siemens Phenom

Michael,

I can’t remember when the Cartesian-Cartesian field was added to the Identify Results command, but I’m sure it isn’t in NX 8.5. I think it was added in NX 9 or 10. But you got past that limitation.

Testing more with your model, and reading your replies again, it seems like you want a contour plot of the deformed coordinate values (i.e. original node location + displacements) in terms of the deformed location coordinate system. For example, if you had a cylindrical coordinate system at the center of your deformed hole, you would want to know the deformed radial dimension variation of the nodes on the hole. Interrogating just displacements won’t give you what you want. So my original suggestion was incomplete. Having the deformed coordinate system is important, but post processing just displacements isn’t enough.

Now, I’m testing with NX 10. I know some behavior with respect to what I’m about to suggest has changed in recent releases. What I have in 10 may not be in 8.5. My suggestion is to use the Identify Results output in Excel to do your evaluations. Create a displacement contour plot using the deformed location coordinate system. Then use Identify Results for select nodes. Send that to Excel and you will get results like that shown in the attached spreadsheet. Notice it states that the results are in a local coordinate system.

1. Adding the X+DX will indicate how far off the deformed coordinate system’s YZ plane a given node is. If it is on the plane, then this sum will be zero.
2. Sqrt((Y+DY)^2 + (Z+DZ)^2)) is the radial dimension in the deformed coordinate system’s YZ plane. This will indicate how far out of round the hole nodes are.

Maybe this is what you are doing already in Excel, but I’m not positive. I also don’t know if the node location is written in the global coordinate system in NX 8.5 or in the local coordinate system as it is in NX 10. I do know that we had an issue with that in the past. The displacements would be written in the local system, while the node location was written in the global (absolute) coordinate system.

Regards,

Mark

# Re: Get deformations in coordinate system deflected by deformation results

Phenom

Mark,

thank you for response. As you assumes correctly, I'm interested in "...  a contour plot of the deformed coordinate values (i.e. original node location + displacements) in terms of the deformed location coordinate system...".

In Abaqus there you can specify deformed nodal (grid point) positions to transform the total displacement situation onto that deformed displacement triple (triad) as base for deformations with removed deformation values of that base. That means the nodes building the base have zero displacements if their position remains in a state of orthogonal, rectangle base. That’s why I wrote that I expected in best case zero displacements at my RBE2-nodes building "my" deformed orthogonal vector base.

But I didn't keep in mind that I have to consider and include the coordinate values of the nodes in Excel additionally. That's an important point.

As you asked, I think in NX 8.5.2 the coordinate values are also transformed with respect to local coordinate system as in my case the values change too (in excel) after changing the result coordinate system.
By the way, one good indicator here is the NX failure, that all annotations have wrong positions if the result coordinate  system is modified to different origin with differing directions of base vectors (in NX 8.5.2).

Last point: Our entire ideas and considerations yield in my supposition, that a contour plot visualization of the deformed displacements in "deformed coordinate system" is not possible, right? - Manipulations in excel is the preferred treatment.

Best regards, Michael