Etabs wall-column modelling

[surya teja]

Hi everyone,
I am trying to model a building where I have a wall that runs in the first two storeys and then only columns continue from the second storey to roof. I have modelled my wall as a shell element in first two storeys and column as line element from second storey to roof. Also I made sure that at the point of connection of the column to the wall, the wall has been meshed manually. Now the issue I am facing is that when I see the axial load diagrams for vertical load, I see that just from the wall-column junction the axial load value actually starts reducing. I tried to check in the analysis base reactions of etabs which gives me almost expected value by a similar model where there is no wall and only columns have been modelled. However, when I try to add manually from the each of joint reactions of vertical reactions at the base, the total value seems a lot less.
Your inputs are really appreciated.

 

[Trenno]

Is the case that some of the vertical load is bleeding into the return walls also, which is what I'd expect to happen with a two-storey box.

 

[surya teja]

Yes treno, I understand that the vertical load is getting dispersed in the wall. I am trying to see where is this going into further. If I am trying to look at the joint reactions at the base and then add them up which comes around 15000kN, it is not matching the intended value. However, in the analysis tab under base reactions, I see the total vertical reaction which is about 23000kN, matches with the intended value. I also wanted to understand what is the approach followed by others for the design of the wall in this case? Do you ignore the presence of the wall and design as if column is continuing to the base and provide a higher reinforcement? or you design as per the pier results provided by etabs?

 

[EZBuilding]

I'd recommend creating a "dummy" load case and applying a point load of 1 kip to the roof of that column. This should help you see how the load path is travelling across your building without looking at results that are jumbled across different load cases and load applications.