ETABS Modelling a long column

[Redtelis]

Hi guys,

Within my ETABS model, I got the ground floor & 1st floor and a mezzanine floor between them. As you can imagine the columns are modelled at each level but the thing is that I want to model the columns which are not within the mezzanine, therefore these columns will be long column with double height as they are formed between the ground & first floor as they are not passing from the mezzanine level.

The thing is that ETABS likes to break the columns at each level and I cannot find a way to model these columns as one single long column. Moreover, this will be wrong during the design check as the height of these columns is double than the others.

Is there any way to stop ETABS break these columns between the levels?

Thanks

 

[Trenno]

From an analytical perspective, there is no difference whether the columns are broken or not. Only difference may be the mass distribution of the column, which affects modal performance.

You could try setting the frame automesh options, so that it doesn't break at storeys, intersections etc.

ETABS should recognise the length of the column as double height when it runs the design module, as there are no mid-height restraints connected to the node.

 

[Redtelis]

Hi Trenno,

Thanks for your help.

What do I have to edit within the frame automesh options?

 

[Trenno]

Try setting them to "no auto meshing."

Failing that, delete the top column, then Align Joint to the Z coordinate of the 2nd storey. should be one continuous column then. ETABS won't join frame members (columns) when it passes through a storey.

 

[jayrod12]

You also must ensure that the node at the middle of the column (floor height) isn't attached to your diaphragm or else your results will be skewed.

 

[PKonnerup]

I would typically overwrite column length in design parameters, for these selected columns, and let meshes be at levels.

 

[JoshPlumSE]

I agree with PKonnerup.... Easiest thing to do is adjust the unbraced length of the column in the design parameters.

 

[estructe]

I think that the unbraced length of the column shall be adjusted, as you said, but priot this action you will have to decide if the column is sway or non-sway. Non-sway columns will have an ubraced length smaller than the actual one. On the other hand sway columns have greater unbraced length than their actual length.

Structural Analysis of 2D Structures software
Static, Dynamic,Linear and Non-Linear Analysis
of frames, trusses, beams etc.

 

[dadomago75]

Hi Redtelis,
easy to resolve in the proper way without affecting the quality of the model.
[ol 1]
[li]open a new window displaying the 3D model.[/li]
[/ol]
[ol 2]
[li]select the upper column (the half column connected to the upper slab/beam)[/li]
[/ol]
[ol 3]
[li]select the top end node and drag it to the node of the upper slab/beam[/ol][/li]


This will allow to get single value of the internal actions for the column, and most important, the node at mid height in your case will lead to instability during the modal analysis.
This is because the mass associated to that node is basically zero as well as the stiffness. This may lead to zero value into the mass and stiffness matrix creating convergence issues, especially if you use Ritz vector (should be the default algorithm to use).

Hope this help.
Cheers
Davide