Maintaining Column Stability

[cetiger]

I am working on demo plans for an existing structure. There is an interior mezzanine that the building owner would like to remove. The demo is simple for the most part, but I do have a few areas of concern. There are several locations where a continuous steel beam spans across columns running from the foundation level up to the mezz level. Directly above some of the foundation level columns are columns as well that bear on the continuous steel beam and go up to support the main roof structure above. So that the contract does not have to shore up and remove the columns above the continuous beam, I'm thinking about having the contractor cut the continuous steel beam (supports mezz only) on each side of the column and leave the beam section between the existing two columns providing some type of column splice detail that can be fabricated in the field. The columns are gravity only columns and are not part of the lateral system. Has anyone done anything like this in the past, if so, would you be willing to share a detail or describe the column splice detail?
The columns and beams are both W sections.

 

[Triangled]

have not done this, but some, perhaps obvious, thoughts.....
assuming of course that you're confident your new KL/r will work,
there are two items you might want to get an evaluation on prior to fully designing your retrofitted column.. 1) does the contractor prefer this method, and 2) is the owner happy with the appearance. You could achieve this by sketching a rough and overstated idea of the connection for their consideration.

 

[cetiger]

I've thought about the KL/R factor...I'm not sure that would be an issue, as I'm not changing the L value of either column section, but rather trying to create a stiffened point that would act as a braced location where the continuous beam exists now. In my mind, essentially the columns will continue to perform under the same load and section characteristics as they have seen over the last 30 years, as there will be no new loads introduced. The loads have always traveled from the upper column section, thru the w section of the beam to the lower column section below, but the beam did provide a brace point where the columns bolted to the upper and lower flanges of the continuous beam. What I think I need to do is stiffen the w section of the beam so that the beam section will not have the tendency to buckle or rotate once each end is cut free.

Currently there isn't a demo contractor on board, as the CDs will be sent out to bid once completed. If I were the contractor, I think I would prefer keeping the existing columns in place vs the option of shoring up 30+ feet of roof above, removing two columns and trying to retrofit a new steel section in place.

The final appearance is not an issue since the project architect plans to wrap all the interior columns within the space once the mezz has been demoed out.

 

[Triangled]

I may have gotten tangled up in your original post, but in my mind's eye I'm envisioning a retrofitted column spanning from grade to roof, in which case your L has dramatically increased.

 

[BAretired]

A sketch might help clarify the situation. You stated "I'm thinking about having the contractor cut the continuous steel beam (supports mezz only) on each side of the column". If you do that, you remove lateral support provided by the beam. Why wouldn't KL/r of the column be a consideration?

BA

 

[Triangled]

".... but rather trying to create a stiffened point that would act as a braced location where the continuous beam exists now. "

This is what is confusing me. I have no idea how to create an unbraced point that acts like a braced point. And if it's unbraced I think your L is from grade to roof. Your profile indicates registration in multiple states, so quite likely I may be misunderstanding what you're saying. Perhaps a sketch? Others here are brighter than me.

I am a contractor in addition to a registered engineer, and honestly, I think I'd rather pull the two columns and install a new one. But that's just me and, again, I think I may have a confused idea of what you're trying to do.

 

[cetiger]

I have attached a sketch of the existing conditions. Understood, I realize I'm going to loose the bracing point in one direction (no bracing exist in the other direction out of the page). The connections are bolted (4 bolts each flange of web per column). In it's existing state per the sketch, assuming one of the columns were to try and buckle due to being overloaded, would we be in agreement that the only element to resist the buckling at the column/beam intersection would be the shear capacity of the bolts at the beam's flange? If that is the case, then could we not design some type of collar or built-up section that encompasses both the top and bottom columns locally at the intersection of the continuous beam that could provide the same amount of stability to resist the buckling as the capacity of the bolts?

Triangled, I agree in that I would prefer to install a new column vs retrofitting the existing, but I'm not sure that's an option with the existing conditions. The new column would need to be 30+ feet. The only way in the building is a double door, so no room for large pieces of equipment for lifting into place. Also the owner would prefer not to have to shore up the roof framing above for the sake of three columns. The remainder of the mezz is free standing, so it can be removed with minimal impact to the rest of the structure. Just trying to think outside the box on this one vs a more traditional approach.

 

[paddingtongreen]

Those cap and base plates and no stiffeners in the beam look like pinned connections, are they able or can they be beefed up, are the bolts sufficient to transfer the stability moment?

Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin

 

[BAretired]

I am surprised the original designer didn't put stiffeners in the beam aligning with the column flanges. That could be done now. Install welded plate stiffeners of the same thickness as the column flange. Then weld the upper and lower column to the beam so that the column is effectively continuous from top to bottom.

You do not show the cutting of the beam and you don't show any new bracing at mezzanine level. If the beam is cut each side of the column, the effective length for both major and minor axis of the column will be the height from foundation to roof. Make sure the column is adequate to carry roof load over a two story height. If not, you may have to plate the column each side to improve its radius of gyration.

BA

 

[JAE]

cetiger - I would agree with the previous posts - I think you Lx and Ly are both increased here. The out of plane situation may be that the mezzanine, through lateral bracing of your cut beams, is bracing your column too via the weak axis bending of your beams.

In any case, what I would start looking at is splicing on a pair of large channels on either side (near and far sides in your sketch) that would lap the existing columns above and below some distance - say 3 feet each way....maybe much longer if needed for hour new KL/r values. The channels could be welded onto the existing column flange tips to create a sort of box shape with extended channel flanges pointing outward. These could possibly be installed prior to any demolition other than some holes in the mezzanine floor.

I would prior to this also check on the ACTUAL alignment of the upper and lower columns - they very well could be mis-aligned a bit in which case you would have some possibly serious out-of-plane second order effects developing.

 

[a2mfk]

We don't know all of the other details, but maybe cutting the top of the column off right below where the presumed roof beams frame in is a better and easier option? Then you are out of the buckling danger zone and can go with basically a pinned connection from that stub column to a new continuous column that you have designed unbraced for the entire height. This may be about the same cost if you have other steel being fabricated on this job.

Otherwise I agree with the last couple of posts about basically creating a moment connection to take care of the stability issue.

 

[Triangled]

" In it's existing state per the sketch, assuming one of the columns were to try and buckle due to being overloaded, would we be in agreement that the only element to resist the buckling at the column/beam intersection would be the shear capacity of the bolts at the beam's flange? "

I can't think of any mode in the existing structure where column buckling would occur at that base plate location. Therefore I don't think you can make any deduction from the shear strength of those base plate bolts that leads to your solution. The bolts restrain the base plate because they're connected to a rigid floor which is presumably rigidly connected to the planet, although I think you mentioned that the continuous beam is not a part of the lateral system and that the beam is part of a free standing mezzanine. There must be some lateral stability to that mezzanine.......

The bolts at the beam line which are a pin connection for the bottom of the second floor column and the top of the first floor column, become a hinge when the remainder of the second floor is removed. The task then becomes to unhinge the hinge.

I could imagine the rx and ry of the first floor column, having been designed to support floor plus roof loads, could be sufficient to support a roof load only with the L increased. If the rx and ry of both story columns are happily both the same and also sufficient to support the roof load, then I would see the solution as being along the lines of creating continuity across the floor beam section as BA is saying.

If rx and ry are different between floors or insufficient at the revised two story L, then you've got a column with varying KL/r along the length, and things are getting tricky. You'll be beefing up the section as described above... "some distance - say 3 feet each way....maybe much longer if needed for hour new KL/r values. " That "some distance" item..... I'm not sure how to calc that. I'd love to hear a remark on that item.