We use cookies to operate this website, improve its usability and track visits. Please note that by using this site you are consenting to the use of cookies.
Learn More |
Stress Simulation Mesh Failure
SOLIDWORKS Forums
3 months ago

Hi everyone,

 

For my second year engineering project, I've had to design a bike frame (long-john type cargo bike), and am now trying to get some stress calculations done on it to see whether this design would be safe.

 

I looked up some tutorials on how to run a simulation, followed the steps, only to get stuck at the point of meshing the model.
Every. Time.
I lost count of the simulations I've attempted to run.

  • I tried using the standard mesh. I tried setting it to the finest setting. I tried manually setting the element size to an order of magnitude less than any geometry in the model, tried using mesh control on the failing edges and faces, all to no avail. The furthest I've gone with standard mesh was letting Solidworks try to mesh the thing for 12 minutes, at which point it crashed with no results.
  • I also went and edited the model, originally built from surfaces then joined and then thickened, to remove some unnecessarily complex geometry. No improvement.
  • I then moved on to curvature driven mesh, and using that I can get the mesh progress indicator to start, linger for a while on 48%, move up to 96%, and then instantly fail.
  • On using "Mesh failure analysis", all I get in these last few attempts under "failing bodies" is a dash with nothing written beside it.
    I've again tried setting the mesh to its finest setting, then lowered it manually again, leading only to increased memory consumption and still no mesh.

 

I then resorted to online tutorials and posts, without finding much in the way of useful insights.

 

The university's resident CAD technician has also been useless, for the second time now.

I am now resorting to the Solidworks forums in hopes to find an expert and figure this out.
I think I may have to edit the model further but I need to know what exactly is going wrong.

 

The frame model is attached.

 

Thanks in advance.

 

GBergamelli.


Categories: Simulation

Comments
Last comment By: Chris Clouser   Tue, 21 Mar 2017 16:32:22 GMT
Re: Stress Simulation Mesh Failure

Hi Gabriele,

I opened up your file but it did not show any studies setup in it.  Did you save your work in progress?  Right now, just looking at the model as it is, you will have a great deal of trouble meshing this with solid mesh elements as the tubing is 0.60 mm thick and the overall diameter of the tubing is ~32 mm.  For just an okay mesh you are looking at >1 million elements and for a good mesh far beyond that.  The solve times will be very high and it will be days between testing iterations of your design.

 

You don't mention whether you tried using shell elements but if you have not tried using them you would want to for this model.  I rolled your feature tree back up to just above "Thicken1" where it is still a surface and the program will will recognize it for use with shell elements automatically.  It seemed to mesh very readily in that state.  You may want to stay in surface design for this where possible rather than changing over to solid design and solid mesh elements.

 

By: Ryan Dark  Mon, 20 Mar 2017 23:25:25 GMT
Re: Stress Simulation Mesh Failure

To be honest, I would remodel the frame using "weldments" and the proper structural members.  This might require a slight learning curve to figure out weldments, and some of your features are not typical structural members, but they can still be included in a multibody weldment part. 

 

Once done in weldments, as I've been preaching in other places on the forum and catching a lot of crap for, there are more things that people really don't think about on the surface.  For instance, this will become a multi-body part in weldments which GREATLY helps to control mesh when there is a failure.  Right now you have to apply one element type to the entire body, where it is thick and where it is thin.  NO BUENO!

 

Approach this model as if you are going to manufacture it!  Tubes and fabricated shapes fit to each other but not monolithic.  Ensure there are no interfering bodies or gaps, and then when you go to mesh, only the part or parts that is giving you trouble will fail and you can focus in on that.  Most of this should mesh no problem.

 

ANOTHER WIN FOR WELDMENTS!!!

By: Chris Clouser  Mon, 20 Mar 2017 23:29:19 GMT
Re: Stress Simulation Mesh Failure

Ryan, there are structural members missing in your picture (the lower chain-stays in the rear and the cross members of the basket), and shell elements will have to assume all materials are the same thickness, which I don't think is the intention.  If you'll notice the model, there are a variety of material thicknesses.

By: Chris Clouser  Mon, 20 Mar 2017 23:31:52 GMT
Re: Stress Simulation Mesh Failure

Hi Chris,

Yeah, I was just looking through the model for where the design went off the rails.  I'm not going to fix the model but that spot seemed to be where the model turned into an un-mesh-able monster.  If it were taken from that point forward it could probably be made to mesh and be analyzed.  I'm sorry that was not clear in my initial response.

By: Ryan Dark  Mon, 20 Mar 2017 23:39:38 GMT
Re: Stress Simulation Mesh Failure

I will try analyzing the surface frame, after redoing the dropouts and the cargo rims with surface modeling.

 

The frame isn't meant to have varying thickness anyway, that was only done as part of my frantic attempts to get the simulation to work by removing edges. I just modeled the dropouts and cargo rims as solids for that seemed the fastest option, but now I see how it can be incompatible with simulation.

 

I had assumed Solidworks would have trouble dealing with zero thickness surfaces, but I have no other options by now.

 

Still a preferable solution than redoing theentire thing with weldments.

By: Gabriele Bergamelli  Tue, 21 Mar 2017 05:06:50 GMT
Re: Stress Simulation Mesh Failure

I'm glad that helped.  I was looking at how long you took to make the model (looks like you started in February on it) and it seemed like you put a lot of time into it.  I'm glad you don't have to start over.

By: Ryan Dark  Tue, 21 Mar 2017 15:27:01 GMT
Re: Stress Simulation Mesh Failure

Gabriele Bergamelli wrote:

 

I will try analyzing the surface frame, after redoing the dropouts and the cargo rims with surface modeling.

 

The frame isn't meant to have varying thickness anyway, that was only done as part of my frantic attempts to get the simulation to work by removing edges. I just modeled the dropouts and cargo rims as solids for that seemed the fastest option, but now I see how it can be incompatible with simulation.

 

I had assumed Solidworks would have trouble dealing with zero thickness surfaces, but I have no other options by now.

 

Still a preferable solution than redoing theentire thing with weldments.

If you want to run a cursory check, you can assume that the frame is uniform thickness, but in reality there will be various thicknesses due to the different structural members being used.  For instance, I'm assuming you don't want your bottom bracket to be paper thin as you are showing.  That also goes for the steer tube, and perhaps the seat stay.

 

Obviously on this you just want to get it done and turned in, but I would recommend in the future, if the project is a welded structure, you might use the weldments functionality of SWX and you will have a much wider variety of options when turning to simulation.

By: Chris Clouser  Tue, 21 Mar 2017 16:32:22 GMT
You are not authorized to view this page No results found! Suggestions: Check spelling, try a different search, or browse topics below.