Steel Frame Out of Plumb

[pcronin]

We have a small two story structure that the steel frame is out of plumb and a total of 7 columns exceed the erection tolerances of l/500 in the AISC Code of Standard Practice. The misalignment was noticed by the gage metal contractor while installing the curtain wall. They kept installing the curtain wall and had the top and bottoms of the columns surveyed. Of course the floor slabs have been placed and cured and the contractor completed the curtain wall.

Now we have been called in to determine if any remedial work is required. Aligning the structure is out of the question because on the floor slab and the curtain wall. We are starting by reviewing the columns with the added eccentricity and p-delta, but where should we go from there. We do not know if the misalignment is from fabrication errors, the steel was never plumb, the steel was plumb but then shifted, etc. Should we be concerned about possible residual stresses if the structure was plumb after erection and then shifted?

The top of steel at the roof is 30 feet. The lateral system is masonry shear walls so the column load is gravity only. The worst misalignment is 1.5” in the weak axis of the column.

 

[jike]

Run thru your numbers and see what numbers you get.

In order to look at built in residual stresses you need to know the sequence of how and why things happened, otherwise you will be guessing.

 

[JAE]

Or, taking off from where jike left off...assume the worst residual stress condition and go from there.

Essentially, you will have additional lateral forces in the framing due to the P[Δ] forces developed from the leaning columns (both D and D+L combinations). This will add axial drag forces in floor and roof framing members, additional lateral force on your diaphragms and extra pull on the collector-shear wall connections.

 

[pcronin]

ok....I am almost done with my analysis. I am getting a few more items surveyed to help determine when the movement occurred including verifying the second floor displacements and seeing if any of the column have a bend to them.

If the displacement was during erection, everything is fine. Some stress increases due to p-delta, but nothing is overstressed.

If the movement occurred after the slab was cured and I model everything as nodal displacements, the columns still work fine, but I am having issues with beams that now see compression. Of course the beams are too slender for kl/r=200 for their full length. The top flange is laterally braced by the floor diaphragm, but the bottom flange is free to rotate.

In compression, I assume these beams need to be analyzed as unbraced for their full length and I may need to add diag bracing to limit there slenderness. Any thoughts on this approach?

 

[Ron]

Some of your answer will depend on whether you have localized "out of plumb" conditions or whether they are all out of plumb and "leaning" the same direction (imposed racking (wracking)). If you have the latter condition, directional bracing might be necessary.

 

[pcronin]

Ron - All different directions and magnitudes. I believe it to be fabrication errors (beams slightly too long/too short) since it is not uniform.

 

[msquared48]

If this is a steel moment frame, the lateral capacigty of the frame has been reduced, but, as suggested above, calcs need to be re-run to determie the stress related impact and new deflections/story drift.

Mike McCann
MMC Engineering

 

[JoshPlumSE]

Have you done a comprehensive "survey" of the plumbness of the buildings columns? The results of this may be interesting to some people I know.

I believe some of the AISC task committees were recently trying to get information erection tolerances issues. They were probably more concerned with taller structures.

However, it might be good for them to get their hands on a study of a smaller structure that may have not been built or constructed to the same level of care as your typical 10+ story building.