Increasing Beam Capacity 5

[DMWWEngr]

At work we are looking at changing the loading of a room. The dead load on the floor will be dramatically increased. I have checked and the primary steel beams supporting the load will fail in flexure. However, the beams are sufficent in block fracture, shear, etc. An initial quick check of the secondary steel beams WILL support the added load.

I'm looking for a good method to increase the flexural capacity of the primary steel beams.

What do you guys recommend??

There is a bit of room below the slab to increase the depth of the beams. I'm thinking about welding on a plate (perpindicular) to increase the depth (web). Then weld a plate at the bottom to create another flange. Is this a viable option?? Would the flange plate be necessary, use a tee instead?? What possible problems would this option bring on??

Using the above example, what if the additional plates had to be added in sections due to space limitations (say four 5' sections to make up the 20' span)?? What additional concerns would this bring up??

P.S. Removing the existing beams and installing new ones are not an option in this situation.

Any advice appreciated!! ---
Andrew

 

[JAE]

I would go with the WT shape to extend the depth. That is the key: extend the depth. You could also use a tube section as well.

Welds between should take the q = VQ/I horizontal shear. With 5' pieces, I would provide spliced connections capable of resisting the full tensile force developed in the added steel.

It sounds like you have static loading. If it is dynamic/repetitive you need to check fatigue on all these connections.

 

[valleyboy]

You have calculated that the beam will fail in flexure.

Has this calculation been based on the full moment capacity of the section, or a reduced moment capacity due to effective length?

It may be possible to reduce the effective length of the section, hence increasing the moment capacity of the section, by adding additional restraints to the primary beams.

Just a thought.

 

[DMWWEngr]

valleyboy,

Good Point!! That is something that I completely overlooked. I was being very conservative and assuming the whole span as being unrestrained.

Looking again, the beam (W16x31) may already be "restrained" by the 6" concrete slab that it is supporting. The problem is that I'm unsure if the metal decking is welded to the beam (and if so at what spacing). If the decking is attached, is it proper to assume the slab provides adequate lateral restaint??

If not, what is the most common way to provide lateral support?? I need lateral support at 4' spacings to get the allowable moment I need.

Sorry, I don't work with steel very often.

TIA!! ---
Andrew

 

[lsmfse]

I'm not familiar with LRFD, but it should allow greater capacity for the beam, however deflection may still be a problem. However "WT" welded to the bottom does work well.

 

[LPPE]

If the deck is welded to the beam, which it most likely is, then you should have sufficient lateral bracing. Also, try what lsmfse suggested with LRFD. Because you are only increasing dead load, you may get the result you are looking for.
If not, another gentleman in our office did the same thing JAE suggested and used a WT to extend the depth. I never heard from the contractor whether or not it was easy to do. I would suspect welding a WT to the bottom flange of an existing beam would be rather difficult. Make sure welder is AWS certified.

 

[JAE]

Good point pylko...Welding a WT is an overhead weld while adding a tube shape that is wider than the flange allows for downward welding on each bottom flange tip.

Agree that the deck should be adequate for lateral bracing (Lb = 0).

 

[airil]

I like to remind that lateral restraints only work if the design is controlled by the member stability. If the design still goes for the ultimate strength (i.e. plastic moment capacity) I suggest that any consideration should be placed on increassing the first moment of area of the section. It can be done by welding a very thick plate (or use high strength steel) to form the flanges.

Secondly, since you have overlooked the role of the metal decking as lateral restraint, you might forget to design the beam as a composite beam as well. Composite beam design method is definitely gonna increase the beams flexural capacities, both moment and stiffness.

 

[zliazmei]

JAE Pylko GOOD POINTS!
DMWWengr,
By me solutions for this problem are:
1. increasing the depth.

2. incresing the with of bottom flange.
Why not trying to ad a wide plate at the bottom flange.
this will increase the Moment resistance of the beam and also it's lateral stability.
For the good welder with corect equipment it will not be a problem, as the weld will be added downward.
Here where I am from it is not restricted by Codes.
It will help if you have minor problem with moment capacity.

In other case Jae and Pylko-THE best!

 

[Ron]

DMWWEngr...I would use the tube or plate approach..much easier to implement. Lateral support is likely there.

Try to use longest pieces possible...otherwise, since your addition will be in direct tension, the splice weld becomes a bit more critical.

Good comments..all. Are these forums great or what!!?

 

[evelrod]

DMWWEngr---I certainly won't disagree with the comments so far. All the solutions put forward here should work, but at what cost? We must keep an eye on the bottom line, I suppose. As to the difficulty in welding, the WT approach is BY FAR the easiest to impliment in the field. Overhead welding is one of the easiest welds to do correctly. A qualified welder will have NO problems. Continuous welds are not necessary to achieve you goals (alternate and stagger your welds to control warpage).
To weld on a sufficiently large plate to the bottom flange(or alternately, two smaller plates on the top of the bottom flange, both sides) is my next choice, but weight and handlling difficulties come into play. TS is not a good choice from the welding standpoint, although it would probably be the lightest option(If the TS section used exceeds the width of the flange, welding becomes much easier). What ever option you choose, make certain that you unload the beam to be welded on sufficiently (falsework or such added temp supports, watch the camber!!!) Never weld to a loaded beam like this!
For safety sake always use qualified union Ironworkers whenever possible. When in doubt, ere to the 'too strong' side.


Rod

 

[Briansch]

Just a side note on evelrod's comment to unload the beam prior to welding. I read a response to a question in Modern Steel Construction's "Steel Interchange", Feb 2002, (an AISC monthly) about this. The statement and question were; "It is a general rule that welding on an existing structural member is not permitted unless provisions are made to unload the member first...Is there a reference that discusses welding to existing structures?" The response from AISC's Charles Carter, PE, SE, was, "It is not a given that members must be unloaded prior to being welded or reinforced. Two good papers that address the issues are:

1. Ricker, David T. "Field Welding to Existing Structures." Engineering Journal, 1st Qtr 1988.

2. Tide, R.H.R. "Reinforcing Steel Members and the Effects of Welding." Engineering Journal, 4th Qtr 1990."

I purchased the Ricker paper from AISC. Excellent paper. The paper covers weld selection/design, reinforcing, heat input, weldability, and shoring and stress relieving among other things. Ricker mentions that adding a plate to the bottom flange of an existing overloaded beam and forcing camber in, may force the flange into compression. It doesn't sound like the beam that's being discussed is overloaded yet. Ricker goes on to say that welding a cover plate to the bottom flange of an unloaded beam tends to make the beam arch upward because of the heating and cooling from welding. This is one reason, he says, to keep welding as symmetrical as possible.

Briansch

 

[eureka]

Are mill certificates available with the actual yield of the steel? Usually steels, when tested, have a yield 10% to 15% higher than the yield specified.

Another option, if feasible, it to weld a plate at mid-span with with holes 6" or 8" down from the lower flange, and plates at the ends of the beams with holes close to the lower flange. Connect the plates with tie rods with turnbuckle thus creating a truss. When you tighten the turnbuckles you will put a compressive force in the lower flange and a tension force in the upper flange, thus gaining moment capacity for additional load.

 

[redhead]

Adding reinforcement to only the bottom flange crates an unsymmetrical section in which the stress in the compression flange can become critical because of the smaller Section Modulus. This condition will also change the maximum allowable unbraced length because of the change in the factor d/Af.

It is the practice in my office to attempt to balance the added reinforcement by also adding rods (not rebar) to the underside of the top flange. Thes are attached with intermittent, staggered flare-bevel welds. We have had very good success with this, and we have never bothered to relieve the loaded beam by jacking.

 

[JAE]

Welding onto a beam such as this does not create problems as the welds are usually:
1. Longitudinally placed on the flange edges (I agree that it is very unwise to weld ACROSS a tension flange as this would create an instantaneous plastic hinge).
2. The welds are usually stitch welds (such as 2" at 12" o.c. or similar)

Be sure to note the AISC LRFD (and ASD) Section B10 where it discusses requirements for connection plates. Usually the end sections of a cover plate are welded continuous for some distance with the staggered welds between.

 

[DMWWEngr]

Thanks to everyone for their great responses. After finding out that the metal deck was welded to the beam (@ 1' spacings). I went back and rechecked the beam with Lb=1'.

This time I was just barely off the moment capacity that I needed. There are no shear studs into the concrete so I cannot look at the beam as a composite section. I've went ahead with JAE's orginal suggestion and I'm going to "tube" the beam in with plates on each side. This will increase the capacity as well as provide some additonal limitation to deflection.

Since I've never worked with this type of beam, is there anything that would change dramatically with this beam that would need to be rechecked (i.e. shear flow, etc...) Any hints as to the best way to "orient" the plate sections for the easiest/safest welding??

Thanks again everyone. ---
Andrew

 

[Ingenuity]

DMWWengr,

i would re-read JAE's post/s - when he says "tube", i think he means a rectangular tube welded to bottom flange NOT boxing it ("tubing&quot with vertical side plates which is what i understand you are referring to in your last post.


HTH

 

[DMWWEngr]

Oh, is it acceptable to "tube" it in by adding two plates?? This would be easier to do since I wouldn't have to relocate the piping below the beams.

Capacity-wise it works fine I'm just concerned about overlooking/changing one of the buckling aspects??

TIA!! ---
Andrew

 

[valleyboy]

DMWWEngr

You state in your earlier post that the section nearly works, if considered as fully restrained. (The situation that you describe would almost certainly achieve full restraint.)

I assume that you have discounted other solutions, such as the addition of columns beneath the primary beams, to halve the span.

Another trick that I have used in the past, where the beam section is nearly adequate, is to add knee braces off the support columns, inclined at 45 degrees. This effectively reduces the span of the beam, and may get the result you're after. It does however encroach on your headroom local to the column positions. The fabrication and erection is however easier than adding a fabrication to the bottom flange of the beam. It also results in horizontal forces transferred into the columns, which you will need to check for.

Of course, it is possible that other criteria, such as deflection, governs your design.

 

[DMWWEngr]

Thanks for the suggestions on the bracing but I don't think the geometry that I have will allow that to work.

My deflection is just below an inch for this beam. Is that an accepatable amount for a beam supporting a concrete slab?? I have no feel for deflection....yet

What is the proper way to calculate Zx (plastic modulus)?? Is this shown in the LRFD manual somehwere and I'm not seeing it?? My EM books don't show it either. I know how to do it on a symmetric member but not on an unsymmetric member. I need to do this before I can calculate adding something to the bottom of the beam.

The plate idea will probably work fine for this situation but I'll need to know how to calculate (unsym) Zx in the future.

Also, still looking for any "ease-of-construction" suggestions on welding the plates??

Any comments appreciated. ---
Andrew