topic Meshing in Fusion Design, Validate & Document Forum

Meshing

Anonymous — Thu, 23 Mar 2017 16:09:15 GMT

Hi community,

I was hoping to find some help doing some static stress solving. I made a design off multiple triangular Truss-components, but the stress solving is taking a lot of time..

My rig:

i7 @2,4GHz

6GB ddr3

I was thinking the studies maybe take alot of time cause either:

-interferance of bodys

-many thin body rods with 60 degree angles, difiiculty of creating elements?

-6GB ram not enough for the number of elements?

So the question would be, can someone show some light on possibilities of fastening up stress simulation? maybe at the cost of the accuracy of elements. if someone has advice on mesh settings to use that would be nice aswell, as a single truss component did succesfully get meshed, but multiple seem to be a problem.

Thanks, Sjoerd

Re: Meshing

mcmillr — Fri, 24 Mar 2017 17:11:41 GMT

Hi Sjoerd.

If you click on the lightbulb on the mesh item in the browser you will see the mesh and the node and element count will be shown at the bottom of the legend. If you are getting large numbers here it will certainly take longer to solve. What numbers are you seeing?

Thin parts can certainly cause larger element counts - what is the length to thickness ratio of these rods? 20:1, 100:1?

Have you modified the default mesh settings? This could be causing an increase in mesh size.

It is also possible the long times are not entirely due to the mesh size, it could be some combination of loads or contact settings.

If you share your data or a smaller example of your geometry we might be able to provide some suggestions on speeding things up.

Re: Meshing

Anonymous — Wed, 29 Mar 2017 09:27:36 GMT

Hi Rob, thanks for replying,

I have added a picture of one of the components, Its the configuration I used to create components for an assembly, but the assembly (12 components) was taking to long to solve.

I did not find a legend of the mesh, but on its completion I captured the window that popped up.

The rods are

80x5 length +-5500 mm

60x2.5 +-900 mm (struts)

I modified the mesh settings as with default settings I received errors,

Model based size: 8%
Element order: parabolic
Max. turn angle: 50 deg
Max. adjacent size ratio: 2.4
Max. aspect ratio: 15
Min. element size (% of average): 10%

I just winged the mesh settings, as slight changes would create or resolve errors. If you could help me configure them that would be great, if possible with a result of reducing solving time.

Re: Meshing

mcmillr — Wed, 29 Mar 2017 14:38:57 GMT

Thanks Sjoerd for the information. Since you have long hollow tubes these will naturally generate a large solid mesh. Using solid elements are not ideal for this kind of geometry, and in the future Fusion will support different element types such as shells, and hopefully beams as well. You can try adjusting your aspect ratio to 20 which should reduce your mesh size further, but you will still probably have a few hundred thousand elements.

I will ask around and see if there are other suggestions on improving thesolve time with this type of model.

If you could export your model and share I would be able to see the actual timing and assess the impact of any changes to the setup.

This can be done under File|Export, Save to My computer, then you can attach the f3d.

Re: Meshing

Anonymous — Thu, 30 Mar 2017 19:39:11 GMT

Thanks overall!

But I don't know if the model is good solve study material as it has many interfering parts. When I have time I could improve the model, but the information about meshing will do.