Visual Analysis software users out there that are good with plate modeling?

[Schambach]

I use the VA software for frame analysis and I get along pretty good. When it comes to modeling plates I struggle a bit. I'm curious if there are any users out there that would be willing and open to having dialogue and answering questions off-line and ideally be able to learn from each other. If not, I'm curious about the following:

I'm creating a bolted connection between two beams. For various reasons I need to place a 3" square piece of steel at the intersection of the flange and web (both sides of the web at each flange......so a 4-bolted connection) such that I can connect to another beam end-to-end. I will gusset the "free" side of that plate such that from a modeling perspective I have a 3-side fixed connection of a 3" x 3" plate. I am going to use a 1" diameter A325 bolt to resist around 30kips in tension.

When I model this plate and break it up into really small elements I'm showing stresses that are really high around the supported perimeter obviously. However, if I look at the stresses that are on the plate a mere 1/4" away from the very edge I find that I could make a 3/4" A36 plate work. If not, and I only use the stresses at the very edge I would need like a 1" thick plate!

My gut tells me that 3/4" plate should be fine, but I don't know what kind of latitude or engineering judgment that I can apply to ignore the stresses at the very edge and utilize stresses that are 1/4" from the edge.......or 1/2"......or whatever makes good engineering sense.

Can anybody provide some perspective on how to read the results of my output and apply it to my design in a reasonable way?

Thank you!

 

[struct_eeyore]

Sorry, a bit late to the reply. Last time I did plates in VA, I remember that it only checks stresses for out of plane loading. Hope this helps. So if you are trying to model a vertical tab connection, you'll need a different software.

 

[human909]

Also be aware of the abilities and limitiations of FEA software. Knowing when can ignore high localised stresses and when they need to be considered is very important. For simple steel plate structural connections normal hand calculation should suffice. Though simple scenarios can be good learning tools.

There are plenty of online articles and blogs on the matter of stress concentrations and singularities.

This one gives some reasonable advice regarding welds:

https://www.apolloedge.com/modeling-welds-for-finite-element-analysis-fea/

 

[Schambach]

Thanks for the advice.