SolidWorks FEA; Simulating weld beads

[StuckOnSolidworks]

Attached in an image of a model I'm trying to accurately do a static simulation on (in Solidworks).

I have two slats in the horizontal plane and fourteen slats in the vertical plane - there is a cut out in each of the slats to allow them to overlap as shown. However, there is always a 0.5mm clearance gap between each of the slats. Consequently the only contact there is the model is between the weld beads – the weld beads joining the slats together, without the weld beads there is no contact. This is all as intended. I have applied the desired external loads and fixtures to my model, and it is now just a case of deciding on the correct method of 'contacts' to use - which is where I'm having problems. If I chose 'bonded' it will treat the contact areas as if I welded the slats together with infinitely-strong super glue. This show’s in my results as the weld show no sign of stress. My other option is to choose 'no penetration', then the simulation won't run - this is to be expected as there isn't any friction and my whole assembly would just fall to pieces. I don't want the contact to be infinitely strong, nor do I want it to be frictionless - I want it to accurately resemble how welds would behave under force. To do this the only idea way I can think of is to apply a coefficient of friction to 'no penetration' contact's (this as I can't see to get the 'Edge weld' connection to do what I want either; it too seems to treat my welds as infinitely strong-bonds). Has anyone else ever simulated the stress on a weld joint? If so am I going about it the right way? Thanks in advance, Tom.

 

[EdDanzer]

Create a solid model part with small gaps between the parts where the welds do not penetrate. Do not use an assembly.

Ed Danzer

http://www.danzcoinc.com

http://www.dehyds.com

 

[kellnerp]

There is a SW Simulation forum for this type of post also.

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

http://www.linkedin.com/in/engineeringtransport

"Node news is good news."

 

[sema79]

Sounds like you may not be familiar with contact sets. You can leave the global model as "bonded" and still set individual contact sets to "no pentration" for two touching faces. If you set up "no penetration" contact sets for the faces that are only touching then the other faces will still be bonded under the global contact "bonded". You can setup multiple contact sets for the touching faces. Remember to check the deformed shape to see if the model is deforming the way you expect it would. I apologize if this is not helpful or if I missed the issue entirely.

 

[rstupplebeen]

It all depends on the end goal of the analysis. I would use sema79's suggestion then extract the loads on the joints and use these values in hand calculations or detailed joint models. I hope this helps.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/