Instability errors, incorrect bending moments accordingly and non convergence

[nali_ameen]

Hi, 

I would appreciate if someone can help me resolve the following error types:

1 type RZ

2 type RZ

2 type RY

I have used releases but customized it to restrain the RZ. However still I get RZ error. FYI, in my model I have used rigid links to transfer loads from main to secondary beams and used cladding to transfer UDL to the beams.

Also, I don't understand why I get instability error at node 95 in the mid span!!!

on other hand, I am getting very large unexpected BM at the support while no such load is applied to cause that:

Also, I get non convergence error when I use tension only braces. I tried to solve through setting the DL and LL load cases as auxiliary and "matrix update after each iteration" but still that didn't help.

FYI, I found some inaccuracy in my model (two members not exactly intersecting as I imported my model from CAD file) but used the tolerance option to overcome that.

Attached is my model for your reference.

Thank you in advance.

[Simau · Accepted solution]

Hi @nali_ameen 

Change:

- Supports pinned to pinned- RZ

- Release pinned-pinned as in the picture below

[nali_ameen]

Thank you so much @Simau . In reality I don't have moment connections for the beams! they are just restrained in transition as shown below: 

wouldn't restraining Rx result in inaccuracy for my model?

Thank you.

[Simau · Accepted solution]

@nali_ameen 

No i don't think so, its only for avoiding instability as tortional bending moment is very low

[Simau · Accepted solution]

@nali_ameen 

The torsional bending moment can easily be taken up by these 4 bolts.

[nali_ameen]

Thank you so much Simau. yes, exactly, torsional is restrained in my case but bending in minor and major is not restrained. I was confused between the local axis until I switched on the local axis display.

[nali_ameen]

Hi @Simau ,

I just noticed unexpected BM diagram at the edge beam!! I get very big positive bending at my column locations as shown below:

but if I remove the TRUSS property and simply make the braces pin-pin, then the result looks more realistic, as shown below:

have you got any clue for such behavior?

Thank you.

[Simau]

@nali_ameen 

You have created a new opening.
Add bars around this opening and the results will be as expected.

[nali_ameen]

Thanks @Simau, I added the new opening as I don't have UDL in this area. But I don't understand why BM in the edge beam is almost zero when I use pin-pin brace but BM increases a lot when I change the brace properties to TRUSS!!!

I added the beams as you advised (though in reality I don't have such beam) but BM become even bigger.

I did a simple 2D analysis for that edge beam, but the results are so different from Robot!! (as shown below). honestly, I don't expect to have big BM near the columns as the point loads are very close to the support!!

Attached is the model after adding the beams.

Thank you in advance for you support.

[nali_ameen]

The results make more sense when I change all brace properties to pin-pin (as shown below)

I don't know what's wrong with using TRUSS release option!!!

[Simau]

@nali_ameen 

I am a bit lost with so much modeles.

Here my 3D modele with results as expected

[nali_ameen]

Thank you so much @Simau, I had a look at your model and the results are so close with both options (pin-pin or Truss).

It look like something is wrong with my model and I don't know what is it!!!

would it be possible please if you have a look at my model (this attachment).

Thank you.

[Simau]

@nali_ameen 

Your model works fine if you add bar 69 103

Model attached

[nali_ameen]

Thank you so much @Simau , you are really amazing 🙂 may I ask why adding these two members change the result that much?! (honestly, I am interested in that as I don't have these beams in reality).

Also, in the absence of beams 69 & 103, the BM is very high in the edge beam with TRUSS modelling but somehow ok using pin-pin. I didn't understand the reason for this as well.

[Simau]

@nali_ameen 

Lets speak about the same model.

Opening 2 is different from opening 1, as there are beams over it.

Please correct your model and don't change it from message to another

[nali_ameen]

Thanks Simau and sorry if I have confused you with adding some members to the model. I have decided to do conservative 2D analysis instead as the results was very sensitive in this 3D model.

[Simau]

Then you didn't solve your issue.
Stay on 3D model and just correct it by creating a structure that allows an opening without member over it and see if its run correctly.

[nali_ameen]

@Simau I simplified my 3D model with following:

1. removing rigid links and using offsets instead.

2. removing the cladding and calculating the member loads manually.

honestly, I tried to save some time using the cladding option. but I think it is better for Robot to simplify the model.

I think the attached model make sense now (as I compare it with my 2D model). I would appreciate to have a quick look if results make sense for you as well.

Thank you again for your usual support.

[Simau · Accepted solution]

@nali_ameen 

The model seems correct, no inconsistencies in the results between truss and member pinned option.s

[nali_ameen]

Thank you so much Simau