Built up steel beam 6

[scottiesei]

Has anyone ever nested a channel against a web on a wide flange beam to strengthen it? I have a project with limited access in which an existing W (top flange loaded) needs to be strengthened, but not by much. The plan is to take a 12" channel and weld it to the center of the web as to provide access to weld above and below to the web. We can not get to the bottom of the beam else we'd just weld something to the bottom. I have not seen this done before, but the calcs with the new Sx and Ix are just what we need. The channel will be in the middle of the web, not smashed up against the bottom of the top flange so the load will be transferred by the welds. Comments?

 

[paddingtongreen]

I've done similar although I prefer alternating high and low slot welds in the channel web, I try to avoid overhand welds such as the bottom of the channel will need if done your way.

Can you unload the beam before you connect the channel? the channel will only share in supporting load applied after it is connected.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[asixth]

I have done similar with cold-formed purlins which were not lapped with installation. To strengthen the section we bolted a C-section to the web of a Z-section. The governing load case was uplift on the roof from wind so there was no need to prop the existing purlin to ensure full load sharing for gravity loads.

If you want the beams to act compositely you will need to check horizontal shear V*Q/(I*t).

 

[scottiesei]

This one is tough due to the limited access to only one side. I like the idea of using the slots and will suggest that to the contractor. We can install the channel, and then add the additional weight, but it might not be feasible to relieve the current load before installing the channel. The beam was design very efficient and is being used to its code calculated capacity. The channel is just to carry the additional weight we are adding. The channel will also extend just past the area which requires the increased section. Thanks so far guys, all helpful comments.

 

[Qshake]

Agree with paddingtongreen - welds may/will be hard to make pending the overall dimensions. There are some references out in the literature for what's required as for weld stick angle.

Angles are commonly used to strengthen beams.


Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[dhengr]

Scottiesei:
You say the existing WF beam is loaded and designed to its full capacity now, and you want to add more load, so you need to add the channel reinfm’t. What size WF vs. the 12" C, the access issue wrt welding, otherwise the welding position is about the same welding at the heel of the flanges or into web slots, which must be fabed. The slots should be nicely made, not just chopped out, and this will be the contractors expense, otherwise the slots do improve welding access. If you’re doing the design can’t you do more than “suggest” to the contractor, if you feel strongly?

I wouldn’t think the torsional problem would be significant. Isn’t the shear center of the channel on the other side of the WF web? And, you say the top flg. is well braced, 4' o/c. Nonetheless, the member is slightly unsymmetrical. Paddington wasn’t really talking about taking the load off of (removing them from) the existing beam, but rather jacking it up to de-stress it before you apply your channel. Thus the WF & C will truly act in unison wrt the total load. What you seem to be suggesting is a WF already at max. allow. stress; add a channel at zero stress; add the new load; now you have the WF over stressed and the channel, which might well be a lesser grade of steel, under stressed, not what you want. Jack the WF higher, to get compression in its bot. flg., and you will cause the C to work at a higher stress level when the new loads are added.

See design examples for cover plates, they do not terminate at the theoretical cut-off point. The channel must go past the point of need and be fairly well developed before it is needed. Check the new shear in the fully stressed WF, at its bearing points, with the new loads added, is this O.K.? What shear and first moment should you use in your shear flow calcs. to design your welds?

Good Luck

 

[scottiesei]

We were on the same page as far as load removal (or de-stressing) and to get the c working correctly we'll have to look more into it. My "suggesting" is that we are proposing this method, the whole design method is a suggestion right now. The existing shear and web crippling has been checked and is okay. I'll look a bit more into the development length for termination of the channel.
Thanks!

 

[dhengr]

Scottiesei:
Another, maybe more elegant, solution to your added load problem would be to use post-tensioning of the existing WF to pickup the new loads. Would that work, I don’t know, I can’t see the details from here: complete WF details, span length, bearing details, existing loading and new loading, etc. etc.

 

[JAE]

I know that the Iowa DOT has done quite a bit of research (and actual projects) using "under-the-WF" rod post-tensioning to increase old steel beam bridge's capacities along with FRP applications.

Here’s some links to WF strengthening using FRP strips:

http://library.modot.mo.gov/RDT/reports/Ri01016/RDT01017a.pdf

http://www.iowadot.gov/bridge/ibrc_projects/TRB_FRP_Boston.pdf

http://www.iowadot.gov/bridge/ibrc_projects/frp_tension_paper.pdf

Here’s a link to a bridge system using post-tensioning rods:

http://onlinepubs.trb.org/onlinepubs/trnews/rpo/rpo.trn127.pdf

 

[JAE]

Also keep in mind that with a channel on the side of the WF, you really don't need to worry so much about the connection between the two taking horizontal shear (i.e. VQ/I isn't really a concern) since the two shapes aren't really working compositely but simply in tandem with each other.

The connection should be adequate to transfer the channel's share of the load from the WF (which takes the load) to the channel in a vertical sense.

 

[scottiesei]

All great info, everyone gets stars. The post tensioning is a really cool idea, but there most likely will not be a lot of access at one end. I would assume that some type of jack would be required to tension the cables. I still like the channel reinf to web idea the best. We will just need to work on a way to jack and shore the existing beam while installing the channel. Am I correct in stating that the only connection forces from the channel to the beam will be from the shear flow? The channel will not be supported at the ends, it will only be adding to the section where needed. So if the section is connected sufficiently, the act of deflection of the loaded W transfers the force to the channel. I really like the slot welds through the web as well. I am sure the welder will appreciate it plus deliver a better product. Thanks again everyone.

 

[larsacious]

Here are my thoughts on your issue.

1)It doesn't matter from an ultimate load (plastic design) if you add the channel before or after the existing WF section is loaded. From a plastic perspective, the ultimate capacity of your built-up section is based on the combined section. With this being said, I will still highly recommend you shore the existing beam prior to welding the new section since the heat of welding is going to temporarily weaken it. This is the same principle applied to composite construction.

2)You do not need to distribute your future load based on the relative rigidities of the WF and the channel. If you provide a fillet weld to connect the top and bottom flanges of the channel to the web of the WF then by strain compatibility they are forced to deflect equally and essentially become one section.

3)The connection of the channel to the WF should be such that you will be able to take full advantage of the section property of the channel. The ASD method of this was to solve the shear flow formula and provide more welds at the ends and gradually taper to minimal welds at the point of zero shear. The LRFD method is much simpler in that you only need to provide enough weld to transfer the plastic moment force on both sides of the point of maximum moment.

Finally, if you have access to one side of the flange I would recommend welding two rods to the beam. One rod should be located at the top flange and one should be at the bottom flange at the web to flange fillet locations. The eccentricity of the built up section should be minimal (torsion negligible), fabrication costs should be cheaper than drilling holes in the channel web, and your moment arm of the two channels is much greater than the flanges of the 12" channel.

Good Luck!

 

[JAE]

Well...not sure I agree with you 100%.

1) You are correct that for an ultimate load capacity check, the shoring doesn't matter in terms of strength. But the deflections would be different depending on whether you remove the load from the WF or not. This is true because if the WF goes plastic "earlier" then the whole section with deflect more.

2) "essentially become one section". Are you implying composite action? Think of this: what if the channel was made of flexible pasta...a wet noodle (i.e. take it to a ridiculous extreme)? The relative rigidities would then really matter because the load transfered through the weld from the WF to the channel would be zero. So the relative rigidities DO matter.

3) The ASD composite design method was in reality an ultimate design method using allowable strength safety factors. Both ASD and LRFD composite designs use the same concepts. There is really no composite action here because the two shapes are not stacked one upon the other (increasing "d") but rather are side-by-side in tandem.

 

[Ron]

A couple of comments to the excellent ones above...

If you have access problems, you'll likely have difficulty getting a channel in there as well. You might have to put it in as segments, so you lose some continuity and deflection will likely be worse, unless you can get a good weld in there.

Why not move the channel down to contact the bottom flange? That way you have easier welds at both locations.

As for the post tensioning idea, it does not have to be done with cables. It can be done with rods and turnbuckles...easy to work with.

With all, I would also suggest jacking to remove load prior to repairs.

 

[scottiesei]

We have not examined adding rods ilo the channel. Whether we use rods or the channel, it appears that we will have enough room to slide the new member. We can check the rod idea tomorrow to see if we can get enough section out of it as it would certainly be easier to get the rods up there versus the channel. If we use the slot welds and the channel, we can certainly drop the channel from the center, but of course we would be restricted by the fillets at the bottom flange.

 

[Ron]

scottiesei...grind the edge off the channel to take care of the fillet problem....not that difficult to do and you'll get a better fit, obviously.

The intent of the rod is not to increase the section, it is to "pull" the bottom flange together to reduce the tension in the bottom flange.

 

[paddingtongreen]

Post tensioning might not be feasible if the compression flange is controlling and jacking is out of the picture since the bottom flange is not accessible, but here is an idea that just might work if the beam is deep enough and depending on the forces involved. I assume for this that it is the bending moment and not shear that is the problem.

Weld gussets to the beam web at each end, as near the end as is reasonable. Weld some brackets, low on the web at one or more places near the center of the span. Bring in the channel with clip angles at the ends, and bolt to the aforementioned gussets, the channel must have holes in the bottom flange at the bracket locations. Run fully threaded high strength bolts or rod from the bottom flange of the channel into the brackets.

Pull the bolts up snug, and then tighten them a calculated number of turns of the nut.

The effect will be for the channel to put upward force(s) on the beam, relieving it of some of it's load and carrying those loads to the ends of the beam and dumping them back in.

In lieu of the calculated number of turns of the nut, you could use a strain gauge on the channel to see that it has the desired stress in it.





Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[paddingtongreen]

Oooooops, I hit the wrong button. I should have specified that there needs to be enough gap between the bottom flange of the channel and the bracket to deflect the channel a sufficient amount.

Michael.
Timing has a lot to do with the outcome of a rain dance.