Cutting off bar joists; replace bottom with I-beam 2

[Thoughtclaw]

I'm the staff engineer for a smallish contractor. A customer has come to us needing more head room over a machine they want to bring in. There are 30" bar joists in the area, and they were hoping to replace those with 6" beams if possible. My analysis shows they need at least a W8x15 (it's a 20' span), and they said they can work with that. My initial approach was to put a new frame up with columns just inside each wall. Our chief estimator came up with a much less costly option, and it seems to me it could work, but as I'm still fairly green, I thought I'd run it past you folks. His idea is to weld a plate to the top chord of the bar joist, then seat the beam on an angle that's bolted to that plate. I've never analyzed such a thing and am not really sure how to go about doing so, but I'm confident I can figure it out. But before I dug into it, I thought I'd seek some outside opinions from those more experienced than I. I've attached his very rough sketch of the idea. Any thoughts?

 

[phamENG]

Attachment didn't work. Sounds fishy - like he's trying to support the bar joists on a beam that is, in turn, supported by the same bar joists. Sketch should clear that up, though.

Though I must say your initial design doesn't feel right, either. Replacing 30" bar joists with a W8x15? 20' span? Have you considered deflection? Is it a roof? If so, did you look at ponding? Did you go from a 6' joist spacing to a 2' beam spacing?

Do you have a senior engineer there or are you flying solo?

 

[Thoughtclaw]

Flying solo, though I do have a senior engineer I can consult if I need to. Sorry about the attachment; I believe I've done it correctly this time.

My plan was to support the top chords of the joists directly on the beams.

 

[lexeng18]

I don't see an attachment so its hard to visualize from just your descriptions alone. I will say I recently learned the hard way how difficult it can be to successfully weld to bar joists. Unless they are heavily loaded or long spans, good chance the chord members are ~1/8" thick which is very easy to burn through and susceptible to heat input damage. Requires a skilled welder and technically speaking one that is certified to AWS D1.3 since it is less than 3/16" thick.

It can be done, but I personally will be exploring every alternative but welding to joists into the future. Not worth the headache.

 

[structSU10]

Your sketch seems to show this joist spanning from girder to girder - if that is the case how is dropping in a new beam on each side of the existing joist and connecting to the girder not more economical than that modification to the joist?

Also, as state before, you should be trying to match the stiffness of the existing joist with your new framing as well as meeting strength requirements.

 

[JLNJ]

30” joists imply a pretty large load for such a small span..

Something is missing from this story.

 

[Thoughtclaw]

Sorry for being unclear. The joists are spanning 40', but there's only 20' of it in which they need the extra head space. Thus, the problem with dropping in a new beam on each side and connecting to the existing girder is that it would have to span the 40' and would therefore have to be a deeper beam.

 

[r13]

I suggest to go more traditional way - shore the area, then remove and install the new beam. The shoring company should be able to design a temporary support system with opening that allows personal passage.

 

[jayrod12]

Yeah, you won't be able to modify the joist the way you're thinking. And in fact, that is so not correct without significant reconfiguration that I question the ability of your office to come up with a feasible solution alone.

You should consult a local structural engineer to help, however for starters, steel joists are generally pretty uneconomical to modify/reinforce due to the complexity of the configuration, the generally thin steel used, and that they are usually optimized right to the limits for the original design loading. If I were the engineer here, I'd be putting in new 40ft long beams, even if they had a moment splice somewhere to allow for fit up. And I would be putting as many of them as possible into the space to keep the depth down. But 40ft is a long way, L/D = 60, that's extremely high. You need more depth.

 

[phamENG]

I don't like this at all. You're cutting out the middle of a joist right where it needs to be its deepest to resist the maximum moment and prevent excessive deflections.

What is the joist spacing? What is the level above - floor or roof? What's the construction? Wide rib roof deck? Form deck with concrete? Composite deck with concrete? What's the live/snow load on the roof?

The safer and more economical option will probably be to remove the joists entirely and redesign this section. Whether that's through new framing between existing girders or addition of new columns to support new framing will depend on the particulars you haven't shared with us.

Unless you are intimately familiar with the SJI specifications under which these joists were design, the current specifications, the reasons they've changed in the time between them, and experience designing and detailing joists, you shouldn't attempt this in a real project. Fun theoretical design to challenge your abilities, sure - but not something to stake the lives of the workers in this facility on.

This is a pretty advanced piece of retrofit/structural modification. I'd get that senior engineer involved sooner rather than later if I were you.

 

[Thoughtclaw]

Thank you all for your ideas. I will consult the senior engineer immediately.

 

[Thoughtclaw]

Oh … and it's a standing seam roof above. Snow load is 30 psf. Spacing is 5'.

 

[structSU10]

Sound like framing a pop up roof over the bay they need this extra headroom would be a way to go. Its certainly pricy, but I doubt you can get something to retrofit the existing roof plan without dropping in a lot of columns and footings.

 

[phamENG]

Standing seam? With no structural deck? I doubt that....

 

[r13]

On top of your senior engineer, you should bring in the contractor experienced in demolition and erection as early as possible, as he might know how to do the job safely, in an efficient and cost saving manner.

 

[jayrod12]

I would bet it is structural deck that he means when he says standing seam. It's probably a flat roof system.

I think structSU10 has the right idea. Put new beams on top of the girders, all work done from the roof. New roofing, and then remove the existing joists entirely. Give your client more headroom then they are asking for. Think of the attached.

 

[Ron]

phamENG.....it is likely an older building with no insulation at the roof level. A standing seam roof can be attached directly to joists just as it is attached to purlins on a PEMB.

Thoughclaw...open web joists have often been "designed" by empirical testing, not direct analysis. For this reason you would "plowing new ground" in respect to your standard of care....your liability goes up in the event of a failure.

 

[Thoughtclaw]

Thanks so much for all the ideas, guys. Unfortunately, going up any higher is not an option. We could tear off the roof and go from there if necessary, and that's an option we're looking at.

Anyway, I have brought my senior engineer into the loop. I will let you folks know what we wind up doing. Again, I greatly appreciate all the input from all of you.

EDIT: It is in fact a PEMB (Butler), with insulation under the roof panels.

 

[phamENG]

Interesting. I've seen PEMBs with joists, but they usually employ something a bit better than their standing seam roof. Just another thing to watch out for, I guess. Thanks.

 

[msquared48]

Interesting that Butler used open web joists. Usually use “Z” purlins. Must be an older building, or the 40 foot span made the open web joists cheaper. Butler usually uses 25 to 30 feet for their mainframe spacing.

Mike McCann, PE, SE (WA, HI)