Structural Mechanics Simulation Failure (Imported Displacements)

  • 99 Views
  • Last Post 19 March 2019
  • Topic Is Solved
aCVP posted this 14 March 2019

Hi everyone! I am currently trying to run a 3D transient structural simulation using ANSYS Mechanical, but I am running into a few problems. 

Here is my scenario (see picture below for more detail): 

I have a section of myocardium (body1, flexible) with an implanted cannula (body2, rigid). I also manually defined the contact between the two bodies as a joint; please note that the contact is simply an edge (circle). I imported a displacement load from a .txt. with xyz displacements that map correctly to the named selection (i.e. nodal selection of the myocardium wall). This displacement load is set for 0.03s into the simulation.

The goal: I want to see how the cannula moves (in terms of translation/angle/rotation) as the myocardium deforms.

However, when I run the simulation, it almost immediately crashes. I am not sure what I am doing wrong, but any advice would help!

Thanks.

Order By: Standard | Newest | Votes
peteroznewman posted this 14 March 2019

If you can Attach your Workbench Archive .wbpz file to your post, if the file is < 120 MB, that would make it easier for some of us to look at your model. If the file is too large, you can Clear Generated Data on the mesh and redo the Archive.

The ANSYS members are not permitted to open attached files, so for them, you should insert screen snap images of your model.  Show the Analysis Settings and the Load.  What is the load between 0 and 0.03s?  After the solver fails, cilck on the Solution Information folder and look at the Solver Output. Search (Ctrl-F) for the word Error and copy paste some of what you see there.

  • Liked by
  • Jackely
  • aCVP
aCVP posted this 17 March 2019

Hello and thank you for your reply!

My .wbpz file is attached (version 19.0) - please let me know if you have any trouble accessing it. It will give a warning that external data cannot be found, but it is actually embedded in the "Setup" module, so you can disregard the error.

My main problem is with contact connections. I need to see how the implanted cannula moves as the tissue deforms. Please note that the implant is defined as a rigid body, while the tissue is defined as flexible. Unfortunately, my knowledge is pretty limited on contacts/connections as I am new to Mechanical, so I am sure I am overlooking something...

Please see the screenshots below for more information. Thank you again for all of your help!

Attached Files

peteroznewman posted this 17 March 2019

The above archive is from AIM 19.0 not ANSYS Workbench. I only have AIM 19.2 installed, so I can only reply with words and pictures. I will be looking at this today. It looks like you work in the Mechanical interface, even though you started AIM.

I tried to Restore Archive using AIM 19.2 and got this error.

Try making a zip archive of the entire folder.  Is that < 120 MB?  If so, try clearing the mesh, and doing File, Save As to a new folder to leave behind any extra files.

  • Liked by
  • aCVP
aCVP posted this 17 March 2019

Hello and sorry for the inconvenience. I have made a .zip archive (attached) of the entire folder that is <120MB. Please let me know if you are able to open it. Again, thank you so much for your help! 

Attached Files

peteroznewman posted this 17 March 2019

You made the zip file while the files were open in AIM.  I know this because after I extracted the files from the zip file and used AIM 19.2 to open the file, I got this warning message:

You should make zip files after you close all files that are open in any application. That is the benefit of the Project Archive, .wbpz file, since the application is making the archive, it can close all the files before it makes the archive.

I clicked Unlock and hoped for the best, but was disappointed that I could not open the project.  I got the error that no bodies transferred.

I suggest you get a project that has no errors, and make a Workbench Project Archive .wbpz file and try again.

  • Liked by
  • aCVP
aCVP posted this 18 March 2019

I went ahead and remade the project using workbench and made an archived project (attached). I was able to open it successfully on another computer using version 19.0. Please let me know if you have any trouble with importing the project - I apologize for all the trouble (hopefully it works now)!

Thank you for your continued support - it is much appreciated!

Attached Files

peteroznewman posted this 18 March 2019

I opened the .wbpz file without error using ANSYS 19.0

There is a problem because the displacements are showing values of 1e+308 mm

It seems you have specified the displacements of all the nodes. What is there left for ANSYS to solve for? What is not known?

  • Liked by
  • aCVP
aCVP posted this 18 March 2019

That is very odd - the displacements should NOT be those numbers (in the text file, they are very small, e.g. 0.5mm). The unknown is how the device would move in relation to the surrounding moving tissue (i.e., since the tissue/device are bonded, as the tissue moves, the device should move).

EDIT: I just downloaded the project and this was the displacement load that I received after right-clicking and loading the displacement. I am not sure why the displacement isn't transferring over computers correctly...

aCVP posted this 18 March 2019

I have attached a .zip file with the project in case it is the archive that is messing up. The .txt files used for displacement are also located within the "user files" folder.

Again, I apologize for all of the trouble.

 

Attached Files

peteroznewman posted this 18 March 2019

I finally got the same display you have for imported displacements. It looks like there is a defect in the Archive program with this model.

Below is the small zone around the rigid canula. The problem is the displacements are not compatible with a rigid body bonded to that same edge.  Imagine what shape that hole will be after those displacements have occurred. It won't be a circle, yet those edges are bonded to a rigid circular object. There is no solution.

You have to change the Tabular Loading from Program Controlled to Ramped, otherwise, it is a Step input, which you don't want.

Have you tried running the simulation without the rigid body?  That is what I did in the attached zip file. I have an avi movie in there too.

Attached Files

  • Liked by
  • aCVP
aCVP posted this 18 March 2019

Thank you for your update! I think I will have to model both parts as flexible bodies. To give you more background, I've attached a picture of the device:

The attachment point is technically flexible. However, the very top of the device is made out of a very stiff material. It is very important that the tip of the device does not deform (which is why I originally defined the whole body as "rigid"). 

Is there a way to ensure that the tip of the device only translates/rotates?

Thank you again for your continued help!

peteroznewman posted this 19 March 2019

I'm guessing you have data on the displacement of a myocardium without a rigid canula attached, and that you cut a circular hole in the geometry at one point in time, and are trying to "plug in" a canula to that data. Am I guessing correctly?

The displacement of the myocardium is a result of the pressure acting on its face that is supported by the tension/stiffness in the muscle and mass of the tissue. A different approach might be to search for the tension/stiffness of the tissue so that a known pressure results in the displacement that was measured. Do you have the pressure vs time signal?

Complete FEA heart models have be described in the literature.

  • Liked by
  • aCVP
aCVP posted this 19 March 2019

You are correct! Thank you for the link - I will take a look at it.

Thank you again for all of the help!

Close