Frame analysis-rigid link

[yta10]

Hello

I have two channels side by side. The surfaces where the channels are met are welded together.

What would be the correct way to have rigid link? I have two combinations as below. 

1)

2)

[Cris-Ideas]

Your model is not really in scale of the beam model.

So generic FEA would be better suited for this task.

If yo want to go with beam model either add links like you did, or define 3 beams. two on both sides on the connection and one with the twinned cross section for connection lenght.

[yta10]

Hello

Thank you for your reply.

I eventually have to see beam stress in the I-beam. I think FEA would take too much time to work on it. 

So I have rigid link between I-beam and channels. In reality, they are clamped. 

if I want to use rigid link between the two channels, what configurations of rigid link would be proper between 1) and 2)? 

[Cris-Ideas]

I do not understand what is 1) and 2).

You need to force links between channels.

But your model is so far away from assumptions of the beam theory that you can get similar quality result with 1 min hand calc.

I do not know what is the function of your structure so obviously I am not able to say much about whether this analysis as adequate at all.

Generic FEA would be actually much quicker for this model, and results would be closer to reality.

[admaiora]

I agree  with @Cris-Ideas 

You would like to see local stress/deformation in your case