Adding Steel Beam Below Exist. Conc. Beam

[Eng_Struct]

Hi All,

I have an existing concrete floor beam that is designed for a certain floor live load. However, the client is installing a piece of equipment that will result in an area load larger than the design floor load. There is no information available on the reinforcing of the beam.

I am considering installing a steel beam directly below the concrete beam (supported by columns at each end with post-installed wall plates)to support the additional load from the equipment. The rationale is that the existing concrete beam will continue supporting the dead load and once it starts deflecting even a nominal amount due to the live load, it will engage the steel beam below that will provide the support for additional loading.

Is it correct to design the steel beam to take additional live based on the above rationale? Note that the conservative approach will be consider that the concrete beam fails and all the load (including both dead and new live) is supported by the steel beam. However, this approach causes issues with the floor slab as the steel beam will not be stiff enough and the two-way slab will start getting more deflection.

Regards

 

[kipfoot]

I give this a yesbut.

Yes. But... keep a very close limit on the deflection of the steel beam. My sense is that it needs to be very stiff relative to the concrete beam for this to work like you want it to. I think it's probably worthwhile to pick a point at the 1/4 span of the beam and chip away to find the reinforcing. (Then patch). Then your deflection and capacity calculations will have a better basis.

 

[CANPRO]

Once the beams start to deflect together they will share the load in proportion to their bending stiffness. You need to have a reasonable estimate on the bending stiffness of the existing beam to do this properly. Might be further complicated by the fact that the steel beam is simply supported and the concrete likely has some fixity at the ends. If you can't get a proper estimate on the stiffness of the concrete, I think the only safe thing to do is to size the new beam for the entire load. I don't think assuming the steel beam takes all the live load is appropriate and may be unconservative.

If the concrete beam has the shear capacity to support the new loads I wonder if you could at some external reinforcing to the beam. It could be a more efficient use of material and labor if you neglect the unknown reinforcing and add new plates to the underside of the existing beam to suit the new loads.

 

[JAE]

Agree with CANPRO,
I would do this:
1. Calculate the effective moment of inertia of the concrete beam, Ie (per ACI) using the FULL D+L service load as Ma in the derivation of Ie.
2. Divide the Ie you get by 2 (JAE safety Factor) to underestimate the concrete beam's stiffness.
3. Select a steel beam shape and analyze it with the concrete beam together - sharing load per CANPRO's suggestion - in proportion to their stiffnesses.
4. From this analysis get the Mu values (and shear of course) for each of the two beams.
5. Check the shear and moment capacities of each beam.
6. Check the deflection of the two beams together.
7. If either 5 or 6 are exceeding code limits then select a larger steel beam and repeat steps 5-6 until code is satisfied.
8. Now go back and re-analyze the two beams without that 2.0 safety factor on the concrete beam stiffness.
9. Check the concrete beam for this loading, which will be higher than the values found in step 5 for the concrete beam.

This allows you to use a portion of the concrete beam, but gives you some additional safety with a larger steel beam.

This is really a form of a parametric check as you are checking a "floor and ceiling" bracketed design:
1. Assume the concrete beam's stiffness is lower than it will be to get a larger, safer steel beam selected.
2. Assume the concrete beam's stiffness is higher than it will be to ensure that the concrete beam loading won't be exceeded if it doesn't crack and "loosen up" as much as you think it might.

 

[HTURKAK]

I got a similar case more than 25 yrs ago.
I checked the capacity of supporting columns and shear capacity of the beam and decided strengthening of the beam with adding steel plate to the bottom of RC beam. The slab propped then the cover at the bottom of the beam removed and the plate is welded to the links with point welds.
You may try the same set up.



He is like a man building a house, who dug deep and laid the foundation on the rock. And when the flood arose, the stream beat vehemently against that house, and could not shake it, for it was founded on the rock..

Luke 6:48

 

[WesternJeb]

I am going to take a different approach.. I have tried to do this before and the sizes I get are massive, too massive for the ceiling area constraints...

How about trying a composite beam? Send some post-installed anchors thru the flange of the beam and into the bottom of the concrete beam. You could potentially get away with using only a large steel channel. I would hire someone to scan the beam for reinforcement to make sure the anchors are missed during drilling. This is usually a low enough cost that it will still be less expensive than just using a huge beam.

If you are close on your code check, I would also consider wrapping the beam in carbon fiber. This can provide 20% additional capacity without being intrusive at all. Sika has some great products here in the states.

 

[EngDM]

How about trying a composite beam? Send some post-installed anchors thru the flange of the beam and into the bottom of the concrete beam.

Would you then not have to size those anchors for the shear flow across the interface of the concrete-steel? Might run into issues getting those anchors to work, and then you'd likely have to lap onto the beam longitudinal bars. Not sure how you'd flexural tension to the lap force, or if you'd even have to since it's an internal force.

I'm probably overthinking this, but I've typically just had the steel beam pick it up.

 

[WesternJeb]

EngDM, correct. Very similar to using headed studs on a steel beam into a concrete slab. I usually default to anchor spacings @ 12" o.c. and see what % composite that gets me. Keep in mind, I am only trying to gain the additional capacity needed, not get a fully composite section!

 

[Tomfh]

If adding a steel beam below, determining the load distribution between the concrete and steel beams is challenging due to the very staggered loading (the concrete is already carrying the dead load), and the uncertainty in load transfer to the steel beam during the wedging and grouting process. Given these uncertainties, relying on load-in-proportion-to-stiffness is impractical.

Instead, simply ensure that the combined capacity of both beams exceeds the total load, while also ensuring that the steel beam is fully restrained and capable of a ductile failure. This guarantees proportioning of the ultimate load to each beam's respective capacity.

 

[WesternJeb]

Tomfh, going to disagree all around with you here.

As far as your statement of it being impractical, that is why I choose to take the route of developing a composite section as well. It seems much more reliable and I don't have to count of the concrete beam potentially failing with flexural cracks opening up before the steel beam starts to take the load that I am hoping it does.

 

[Tomfh]

It's not difficult to ensure the steel beam achieves its capacity prior to the concrete rupturing. You just need to make it stiff enough, i.e. check its deflection, like you do with any steel beam. That, plus preloading the steel beam with wedges to ensure its mobilized, will ensure it's carrying its share of the load in an overload scenario.

Regarding anchoring a channel or flat bar to the underside of the beam to create a composite section, this doesn't work, as the plate is too slack, and there is too much slop in the bolt holes for meaningful composite action to develop prior to concrete failure. The plate only begins to meaningfully engage when the concrete has strained another few mm, by which time large cracks are opening.

 

[TLHS]

If you have dimensional constraints you could look at externally bonded FRP reinforcement