I want to insert a gusset into a frame and have it taken into account when running a static test.
Everything I have read and watched about bonded connections speaks of using this tool or that tool to do this or that. None of them tell you where to find these tools and how to operate them.
Would somebody please break the deadlock and write me a blow-by-blow of how to perform this task?
It will be much appreciated.
Parts come in Globally Bonded when you start an analysis. It's automatic.
Just make sure that the surfaces are in contact (flat).
The problem is they don't!
When I work through the test I am instructed to ensure that beams are bonded to solids. No matter what I try, I cannot get past that phase.
I have inserted a gusset into my model
When I Calculate Joints I get
I do a Curvature based mesh at a medium Density setting. While the Model view shows the gusset properly seated, the mesh test does not
Not surprisingly, the test has problems
I click 'No' and get
I click 'Yes' and get
I click 'OK' and get
and, again unsurprisingly, no results.
Either I am doing something wrong or something is rotten in the State of Dassault.
Would someone please tell me how to get this right.
Thanks
Ah, I see, when we here use gussets, etc. on weldments (such as this) we will model the tubes as solids (eliminating the radii, squaring things off, etc. instead of modeling weld mat'l). It seems to give better results and most times the elements are too short to really be considered 'beams' by the software in any case.
There obviously is something problematic with using solid gussets and beam elements. Would it work to model the gusset using shell elements and see if it would then join with the beam elements?
How would a Static Test (or any other) cope with having a Shell gusset? I mean, a shell of a gusset would have little strength. Or do I have the wrong concept of Shell?
Shells have thickness defined therefore do have strength related to such.
However, since your model doesn't appear to be all that big (from the pictures) why don't you model the entire thing with solid mesh and see how that runs.
It would be interesting to see if SWX Student gives a different result, but it would not provide test results appropriate for a hollow-section structure. Would it?
I use the commercial version every day to do just that. Seems to work OK for us. Of course I model it cleanly so I don't have to model the welds (square the corners on steel tube).
You might have hit the nail on the head. Modify the RHS to have square corners. However;
1. SWX does nice work mating compound corners.
so there shouldn't be a problem.
2. Modifying the cross-section of a beam must affect the way it reacts to a stress test. This would not be much of a problem if you are designing a trailer to tow behind your car, but in aerospace design it could be an issue.
We trade off between 100% accurate geometry, weld metal, etc., and good results (the difference between mere design and engineering). What you show in your pic would have some nasty elements which can really skew the accuracy of the results and/or severely increase the solve time. We could have a whole philosophical discussion on aspects of each but that wouldn't be too productive and is best done over a few beers anyway.
BTW, we do do aerospace work. Again, it's an engineering thing. We can't afford to have things break because it could cost lives and potentially endanger our company and all the people dependent on us for their livelyhood (which I also see as an 'engineer' thing - http://www.order-of-the-engineer.org/).
