INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Jobs

Is ANSYS as cumbersome to use as it seems?

Is ANSYS as cumbersome to use as it seems?

(OP)
I previously worked for a government contractor that had its own proprietary FEA code, and we generated the mesh for these models in Cubit (on which Trelis is based) and visualized the simulation results in VisIt. I remember complaining about Cubit and other pieces of software not working the way I wanted it to back then, but now that I'm working with ANSYS, those complaints seem trivial. I've only been working with ANSYS for a few weeks now, so if I'm mistaken about any of this, I would love to understand what the correct way to do things is.

In Cubit, I could select any piece of geometry and instantly see its mesh characteristics (i.e. the number of elements being assigned to an edge, the mesh size of a surface/volume, etc.). Furthermore, the solid body could be moved with relative ease and the mesh would move with it. Additionally, elements and nodes could be individually selected to be added to sets. Finally, the algorithms for generating a mesh seem significantly slower and produce poorer quality results than the corresponding algorithms in Cubit.

In our simulation software, I did not realize how nice we had it to not have to worry about contact since the simulation software automatically detected when two bodies came in contact with each other and handled the resulting response quite well. Furthermore, when we did need to define areas from two bodies that, for example, needed tied/bonded together, we had the ability to select the nodes that needed to be connected instead of being required to select surfaces from the underlying geometry. The combination of these two things apparently not being possible in ANSYS seems to result in either far more complicated meshes (due to the surface needing to be cut up to limit the size of the surfaces in contact) or areas being selected that are far larger than necessary being included in contact connections. For example, if I have bolts connecting two plates together, and want to prevent the bolts from just passing straight through the plate, unless I'm mistaken, I need to create a frictional connection between the bolt head and the surface of the plate. Since this results in the large surface of the plate being included in this connection unnecessarily, I could split the surface so there is a smaller surface surrounding the bolt. However, this is far more work than is really necessary, and extremely tedious in cases where there are multiple bolts.

To say the least, ANSYS may be great at performing simulation, but from what I've seen so far, its preprocessing tools are at best rudimentary and at worst extremely crude. Admittedly, I'm ranting here somewhat, but if someone could point out something that I've missed that make these things easier, I would greatly appreciate any help that could be provided.

RE: Is ANSYS as cumbersome to use as it seems?

Are you using Mechanical APDL or Workbench?

Rick Fischer
Principal Engineer
Argonne National Laboratory

RE: Is ANSYS as cumbersome to use as it seems?

(OP)
Workbench (ANSYS version 18.1)

RE: Is ANSYS as cumbersome to use as it seems?

If you are not already, I suggest using SpaceClaim for preprocessing geometry. For example simplifying geometry and splitting large surfaces (like your example of bolt nut/plate contact)
is in my opinion quite easily achieved.

For example you can project all the bolt head surfaces (parallel to plate) to plate. Then select all projected surfaces and increase the diameter and you are ready.
Does not matter if there is five or 100 bolts, time taken is the same.


Imprint command in SC is also very practical. It goes through the model and if a smaller and larger surface are touching, it will split the larger surface
to match the size of the smaller one.

RE: Is ANSYS as cumbersome to use as it seems?

(OP)
I do use SpaceClaim. Splitting the surface may be easy when there is only a handful of bolt holes to handle, but when there are more than that, it becomes extremely tedious. All I want is for ANSYS to understand that solid elements can't pass through each other and to detect the contact automatically. That doesn't seem to complicated, and ANSYS is supposed to be one of the most advanced FEA packages out there. It shouldn't have a problem handling something as basic as preventing two objects from existing at the same place at the same time. Since it apparently can't handle something that simple, I am required to handle it myself by manually telling it which surfaces to handle contact with.

RE: Is ANSYS as cumbersome to use as it seems?

I think some of the explicit codes can automatically detect contact, but I was not aware of any implicit codes that do that. I think a problem with that approach is that allowing contact on all surfaces of all bodies could produce a huge computational burden. Workbench does by default place contact on all surfaces that touch or are within a user specified distance (pinball diameter). I typically turn this off because I prefer to create contact manually, and because the default is to generate contact every time you update geometry, which creates redundant and unnecessary geometry and is something of a pain.

Rick Fischer
Principal Engineer
Argonne National Laboratory

RE: Is ANSYS as cumbersome to use as it seems?

(OP)
The limitations with the contact algorithm are annoying, but somewhat understandable (though I do think ANSYS' implementation could be greatly improved if it allowed the specification of contact sets based on nodes and elements in addition to surfaces). The really frustrating thing is how poor of an experience meshing is. The UI is awful, and the algorithm produces poor quality results. It's not uncommon for me to end up with body's that are supposed to be meshed with an all quad sweep have wedge elements that have a Jacobean of negative one. Similarly, it's also fairly common for a mesh to end up with elements that have an aspect ratio of 100 or higher. So far I've been able to tweak the mesh to prevent this, but the only time this happened to me when I was using Cubit was when I missed a small edge in the geometry, and that forced the algorithm to generate an element that thin. ANSYS' meshing algorithm is so terrible that stuff like that happens routinely for no apparent reason.

RE: Is ANSYS as cumbersome to use as it seems?

Yeah, I dont like the meshing in WB either.

You can apply contact to nodes an elements in MAPDL (classic Ansys). Check out the GCGEN command. This could be implemented in WB with a command snippet.

Rick Fischer
Principal Engineer
Argonne National Laboratory

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources


Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close