Open Web Steel Joist - Top Chord Analysis

[Serhiy2]

Good day. I'm analyzing the existing Open Web Steel Joist for new load. The existing joists are made by Omega Joists, Canada and are dated 1999. I'm trying to determine top chord capacity which is in the shape of hat channel and is 2.36mm thick. Using the approach given in steel and cold-formed design codes, flexural-torsional buckling is the governing mode of failure in compression. I'm assuming that top chord is laterally and torsionally restrained at panel points. Given that my webs are flattened at ends, the assumption about torsion-restrained top chord at panel points feels to be a bit stretched. Wondering, how do you guys normally deal with something like this?

Also, given that corrugated roof deck is welded to top flanges of the top chord hat channel, would it be reasonable to assume that roof deck provides continuous torsional restraint to my top chord member?

Thanks

 

[KootK]

I deal with these thing in Alberta a lot as well and they can be a challenge to assess (particularly the seats).

I'm assuming that top chord is laterally and torsionally restrained at panel points.

Yes to torsional restraint; no to lateral restraint. Your lateral restraint will normally be assumed to occur at the deck fastenings and / or bridging.

Also, given that corrugated roof deck is welded to top flanges of the top chord hat channel, would it be reasonable to assume that roof deck provides continuous torsional restraint to my top chord member?

I would normally say no unless I could somehow be guaranteed fasteners on both top chord flanges somehow so as to have them form a torque resisting couple.

 

[Serhiy2]

Thanks for response KootK. Since you've analyzed these before, how did you estimate Cw (warping constant) for their hat channel chords? I tried using the formula given in AISI cold-formed manual for hat channels but that formula is for hat channels with 90 degree bents and with these joists it's not the case so I have to create somewhat similar hat section to analyze it. My joists have type A chords if it helps.

 

[KootK]

Since you've analyzed these before, how did you estimate Cw (warping constant) for their hat channel chords? I tried using the formula given in AISI cold-formed manual for hat channels but that formula is for hat channels with 90 degree bents and with these joists it's not the case so I have to create somewhat similar hat section to analyze it.

I'm afraid that my recollection of what I've done in the past is poor. About eight years ago, I discovered that we have former joist supplier engineers floating around locally that will do these exercises as delegated engineering. And I've been loving that model ever since.

My vague recollection of past investigations:

1) It seems reasonable to me to me to estimate Cw based on a faux section with the vertical-ish webs made truly vertical and some other minor adjustments to ensure a conservative estimate.

2) I'd done some tinkering with the direct design analysis method of cold formed design using the free software CUFSM.

 

[jayrod12]

I've also modelled these in RisaSection to get a bit more accurate of properties. Our office has most of the old books so we're often able to pull the dimensions and yield stresses straight from the old info.

 

[Serhiy2]

I used Risa Section but it's not able to calculate Cw and shear center for this type of section and those are the two values which I'm missing in order to get Cr. Thanks for your input guys, we've decided to reinforce the chords.

Also, when analyzing crimped tube webs, I'm analyzing them as regular tubes. The fact that tubes are crimped at their ends shouldn't matter, correct? I'm using k=1 when analyzing them.

 

[KootK]

Also, when analyzing crimped tube webs, I'm analyzing them as regular tubes. The fact that tubes are crimped at their ends shouldn't matter, correct? I'm using k=1 when analyzing them.

Agreed. You're modifying the stiffness of the cross section right where it doesn't much matter at all.

 

[JoshPlumSE]

Serhiy2 -

Flexural Torsional Buckling (in my experience) is not something that is considered in joist catalog capacity ratings. I could be wrong, but when I worked for RISA text support I remember a number of times where RISA showed a failing top chord under existing loads due to FTB and the joist catalogs said it was okay.

Granted, if this is a cold formed truss that would be a different story than I remember dealing with. Also, I think RISA's FTB calculations were not as "open" as they are now. By that I mean that you could not enter in a user specific Lt value.

Regarding Cw, I don't think there are many "general" section property programs that will calculate this. But, most cold formed steel programs will do it. Even if the section you enter is not EXACTLY the same as what you're looking at, the Cw you get out of it should be pretty good.

 

[Serhiy2]

Gents, thanks a lot for your help.

JoshPlumSE, current risa version allows to input L torque and gives a pretty good breakdown on how the capacity of the cold-formed member is calculated. Too bad it only works for AISI standard but it doesn't do that when CSA S136 is used.

 

[Ron]

Most open web joists of that period were empirically designed so doing an analytical replication is quite difficult.

 

[milkshakelake]

I agree with Ron. It's hard to design these analytically with precision and accuracy. From what I know, a lot of these were load tested rather than designed. I generally submit a joist investigation sheet to the Steel Joist Institute and they send me a load table.

https://steeljoist.org/find-existing-joists/

A basic exercise I tried a few years back was to calculate it and see how close I got to the load table. My calculated load was much, much higher than the published load. Maybe I suck at it, but it shows at least one example where the calculation didn't work, or maybe I missed something crucial. Either way, a load table will be more reliable and protect you more from liability.