Convergence of Load combination

[civeng15]

Hi everyone,

Trying to analyse an industriel steel buiding made of 8 stories and 32m hight (using VDesign).
Once I run the analysis some of the combinations are not converged, what does this mean and how to correct it.
Also how to check the P-delta effect if converge or not.

Thanks

 

[Lion06]

It means that it's unstable. If the program is running an iterative analysis it applies the loads to the undeformed shaped, this in turn causes displacements with cause secondary deformations and forces (this is the second order analysis). It will continue to iterate until the analyzed shape is the same as the previous iteration. That is the convergence. If it just continues to move (i.e. it doesn't converge), then the structure is unstable.

I don't think this means that it's going to fall down tomorrow (especially if it's been standing for a while), but it does mean that, on paper anyway, it's unstable. I would double check boundary conditions and end restraints. Try to use an analysis method that might have been employed at the time of original design as a sanity check (e.g. use k factors without reduced stiffness and notional loads instead of the DAM for steel).

What kind of lateral system does the building have?

 

[civeng15]

Thanks StructuralEIT for this clarification,

it is an open structure to support equipments for refinery plant where a combination of X and chevron bracing are used.

 

[frv]

Str.EIT is right.

I would add that excessive internal (p-small delta) second order effects may also lead to divergence.

 

[steellion]

See the "Computer Modeling" thread from yesterday for more tips. The O.P. was having similar issues.

 

[delagina]

is there anything in the code that requires you to use pdelta for steel?

weird, i work in Refinery/Petrochemical/Oil and Gas industry and we dont usually use pdelta analysis but linear perform analysis on steel and these are major EPCs on multi-billion projects. we dont use LRFD either but ASD.

 

[steellion]

delagina, 2nd order analysis is mandatory under AISC 13th Ed. provisions.

 

[frv]

delagnia..

AISC requires that you account for second order effects (see AISC Section C1.1). This may be accomplished in several ways, not necessarily carrying out an actual rigorous 2nd order analysis (e.g. "amplified 1st order elastic analysis), but there are limitations on when you can use methods that don't explicitly account for the effects.

 

[Lion06]

There are obvioulsy some members that won't have any second order effects, but where they do occur you have to account for them! Check out the 13th Ed. AISC Manual (AISC 360-05) as others have noted.

 

[civeng15]

I appreciate your help guys