Analyzing an existing steel column encased with concrete 1

[abusementpark]

My firm is beginning design work on a major renovation of an existing 9-story steel structure (skeleton frame) that was built in the 1950s. The proposed modifications to the structure require us to analyze the existing steel columns for both gravity and lateral loading conditions. All of the columns were encased in concrete, which was reinforced with continuous wire mesh and discontinuous longitudinal bars. We initially thought to try and analyze these columns as composite columns per the latest AISC specifications. However, we do not meet the minimum criteria for transverse reinforcement and continuous longitudinal reinforcement. Also, there do not appear to be any connectors for load transfer between the concrete and steel. Thus, it seems to us that a composite column approach may not be possible per modern code provisions.

However, we believe the concrete should at least be somewhat effective in stabilizing the bare steel column for buckling. There used to be provisions in ACI (see attachment) that allowed you to increase the allowable compressive stress in the steel column based on the size of the concrete encasement. It appears that the existing columns were designed per this method. Also, we are unsure of the contributions that the concrete encasement may have on the bending capacity or lateral stiffness of the column. Basically, we are interested in any advantages that the column encasement may provide for axial and bending capacity of the columns.

Anybody thoughts or guidance would be greatly appreciated. At some point (I believe in 1971), the aforementioned ACI provisions for encased columns were removed. I'm sure if they were found to be invalid or just got omitted because it was becoming an archaic design approach.

 

[WARose]

If you have a wire mesh (as opposed to ties): I’m not sure you will have what it takes to meet any specification I am familiar with (in terms of area of steel or otherwise). The 1st edition of the LRFD manual had tables for composite column strength and a specification (which mandated a certain minimum amount of [total] longitudinal + transverse reinforcement). The good news is: they do not mandate the use of shear connectors.

Probably the best reference I can think of that goes in-depth on something like this is:

‘Composite Construction Design For Buildings’ by: Viest, et al. (1997; published by: McGraw-Hill/ASCE)

Chapter 4 deals exclusively with concrete encased/filled columns and has some worked out examples.

Another reference that may be of interest is:

'Composite Action Without Shear Connectors' By: Watson, et al. (AISC Engineering Journal, 2nd quarter 1974). This reference primarily focuses on beams (partially) encased in concrete developing composite action through natural adhesion & friction (and actually performing well). Not exactly your scenario, but an analogy could be made between your situation and this under the right circumstances (depending on what you are willing to risk).

In any case, hope that helps.

 

[Archie264]

The following is not analytical but rather antidotal. That is, I've spoken with the owner of a highly experienced demolition company and he relayed that he could "easily" demolish a steel building and he could "easily" demolish a reinforced concrete building, but buildings constructed of steel encased in concrete gave him fits. Apparently they are very tough, i.e., ductile. I later saw this same thing mentioned in one of the engineering trade magazines ("Structure" perhaps?) For what it's worth, in case it helps.

 

[ishvaaag]

First you analyze the whole outfit as fully compoosite, then you investigate if adhesion could reasonably be accepted extant at least at the service level. Then the same at the limit states, maybe it could go even as far.

Even a model of some substructure consistent with the resulting firces from analysis, ported to FEM, may help to ascertain better where one stands in terms of required capacity from bond ... chose some worse zone for that.

 

[grantstructure]

I've not tried this approach, but it seems to me you could treat the concrete as a continuous brace system for the steel column. You'd have to decide if it's cracked and what the Icr is going to be, and then prove that the concrete portion of the column alone can provide the strength to resist the force applied to it as a brace. It's certainly an interesting problem.

Failing that, you might be able to post-install shear anchors and provide the required vertical and shear steel in the concrete by enlarging them slightly, and then treat them as truly composite.

 

[abusementpark]

The 1st edition of the LRFD manual had tables for composite column strength and a specification (which mandated a certain minimum amount of [total] longitudinal + transverse reinforcement). The good news is: they do not mandate the use of shear connectors.

I'll have to take a look at that. Thanks.

'Composite Action Without Shear Connectors' By: Watson, et al. (AISC Engineering Journal, 2nd quarter 1974). This reference primarily focuses on beams (partially) encased in concrete developing composite action through natural adhesion & friction (and actually performing well). Not exactly your scenario, but an analogy could be made between your situation and this under the right circumstances (depending on what you are willing to risk).

Just read this reference. Good article. Gives me a little more confidence in the bond strength, but it is mostly anecdotal.

 

[abusementpark]

Archie - Interesting stuff.

grantstructure - I've thought about taking an approach where I figure out the required brace forces for the concrete to brace the steel at midspan and check the concrete section to see if it can span floor to floor. This would be a conservative approach because I would be assuming that the steel is resisting all the load and the concrete is only bracing the steel.

 

[jtx]

Taking a shot in the dark....

Do you need the contribution from the concrete to make the columns work? Being that the building was constructed in the 50's, is it possible that the steel columns are robust enough to take the gravity and lateral loads themselves? Without adequate concrete confinement, I'm not sure if the surrounding concrete would be sufficient to brace the column against buckling (just thinking from the perspective that concrete reinforcement can buckle due to loss of confinement).

Otherwise, if contribution from the concrete is needed could you establish composite action by the following?

-Drill through the concrete and post-install shear anchors, as Grant suggested, and grout the hole
-Add the required confining steel around the perimeter of the column to achieve composite action as required by AISC
-Cover the new reinforcement with new concrete

 

[abusementpark]

Do you need the contribution from the concrete to make the columns work? Being that the building was constructed in the 50's, is it possible that the steel columns are robust enough to take the gravity and lateral loads themselves?

With the proposed renovations and additions, the steel columns alone will not work. Part of the problem is the we are required to bring the thing up to code and the existing LFRS is lacking.

 

[Ron]

Agree with grantstructure's approach. Check around a bit...there's a chance the concrete was only for protection or aesthetics.

 

[johnbridge231]

Regardin the analysis of encased steel sections in concrete, you may find this example helpful.

Regards.

 

[SAIL3]

in my own engineering judgement the lack of shear connections and concrete confinement would be a major concern....could encase the concrete with FRP wrap or steel plate...would still leave the problem of shear connectors, though, so might only increase the overall axial load capacity ...

 

[ishvaaag]

I am not advocating using composite action for any case that one can, but this extract of one article may help to those that don't have connectors. See the shear connectors part.

Source is
NASCC Proceedings 1994
AISC
Composite Frame Construction
Lawrence P. Griffis, P.E.

Quoted Ref. 14 is
Bond Stress of Embedded Steel Shapes in Concrete, Composite and Mixed Construction.
Charles W. Roeder
ASCE 1985