Help Identify Unusual Steel Roof Joist 1

[mfstructural]

I have a project where the client is planning to add a 1,700 pound RTU on an existing roof. The proposed RTU is 115"x64" with the long direction perpendicular to the roof joists preferred. I've attached a sketch of the layout. There is a 8" CMU wall that can be taken advantage of if necessary. joists spaced at 60". They are open to orienting the RTU in the other direction if necessary. The RTU is to rest on two rails spanning perpendicular to the joists.

I am having trouble determining the capacity of the joists. I submitted the information and measurements to SJI and am awaiting a response. In the past I have not had good luck with joists with these section shapes as far as getting definitive data.

Has anyone seen joists like this or have any documentation on these types of joists that might help?

I'm also thinking about orienting the long dimension of the RTU the same direction as the CMU wall so a good portion of the load is transferred into the CMU wall.

I've included some photographs.

 

[mfstructural]

I also came across this Butler joist truss purlin catalog from a friend. Some of the bottom chord thicknesses and web members thicknesses are what I measured but the depths are not. I wouldn't think that the joists would be overstressed sharing 1700# across three joists but they are. I'm basing analysis on 20psf DL, 30psf LL. based on the attached pdf. see blue box for 40'-50psf. But the joists I have do not have end reinforcing as specified in the table and my joists are 25" tall. This is not making any sense.

 

[SWComposites]

Your 25" measurement looks like the truss depth, not the web depth. See section D-D.

 

[lexpatrie]

Check these for signs of corrosion or water being held in the top of the hat channel.... and probably make the owner aware of said item:

https://www.p12.nysed.gov/facplan/BCSitems/Structural_advisory_letter.html

 

[mfstructural]

Hey everyone, I received a response from SJI.
Their response:

"The joists you have presented look most similar to Butler (non-SJI) products. Attached is everything we have for this company. The information provided and the comments made concern a company that was not a member of the Steel Joist Institute. As such, we are simply providing information that was or is available to the public. SJI has not reviewed this information for engineering accuracy and take no responsibility for what is contained therein."

I received some documents from them but one is for a 30' joist and one for a 40'. I also received the same table I posted above. None of this makes sense. The last column of the table indicates that reinforcement of the web members and joints is required for 40' spans, except for 20psf LL. The web members size for 40'-20psf is 1.05" consistent with what I measured. The bottom chord thickness for 40'-20psf is indicated as .078" and I measured .108". The .108 is the bottom chord thickness for the 50'-40psf joist. Is it possible they mixed and matched?

The analysis shows that the 40'-20psf joists need to be fully reinforced for shear and moment for placing two 300 pound point loads on the joist (near the middle of the span), but no reinforcement is required for the same loading for a 40'-30psf. That doesn't make sense that there would such a large difference between the two as far as reinforcing is concerned.

I wanted to see if anyone had experience adding load on these type of joists.

Thanks,

 

[lexpatrie]

I've dealt with proprietary joists once (Palram), and in that case we discussed it with the owner and went with new joists fabricated with a splice in the middle.

Bear in mind partly that was due to an increase in snow loads since the original building, the age of the building, a few new mechanical units, what looked like a fire-egress walkway across the roof for apartments, some new restaurant equipment being hung from the joists as well as an open question about a snow drift due to a parapet extension (or not) on the front of the lower building, and a question as to whether the original joists were designed for snow drift from the two story building the roof in question landed into.

When it comes to reinforcing a joist like this (and all joists, strictly), you are at least partly vouching for all the welds and all the elements and materials as good enough (if you leave them as-is). While the reinforcing can be reasonably designed (and I'd done actual reinforcing with well-identified Vulcraft or other SJI joists), the overall analysis of the in-situ joist, particularly when it is proprietary, feels like there's a lot more involved, or a lot of potentially iffy assumptions involved (weld size, weld length, weld electrode) as well as how these were modeled (true pins is unlikely, when it comes to the web elements). Steel Joists are not strictly designed per AISC, they have their own standards and their own ways of assessing weld strength, and strength of the elements.

While the SJI ones have load tables going back quite a way, and the 25% (or 50% on older) shear minimums, there's not a lot of information on the proprietary ones, like for Palram I got some details on the sections but nothing on the welds, and no information on the grade of steel involved.

 

[mfstructural]

I wanted to provide an update on what I've done so far:

I used TEDDS to calculate the Ix of my joist using the section builder:

Then I did the same thing for 18H8 out of the 1967 Ceco catalog, for which I have the section properties of the chords as a sanity check:

The Ceco catalog has the same Ix as calculated in TEDDS:

The SJI manual has the following equation for calculating Ix.

I then did a hand calc to calculate Ix for the 18H8 using the SJI equation. Calculated as 174 in4, pretty close:

Our W is 125plf (5psf DL and 20 LL, joist spacing of 5'), while calculated allowable PLF based on SJI equation is 162 plf. I then put these in my spreadsheet:

There is a masonry wall right below the center of the RTU that I am planning to use to carry some of the load, so the 418 point loads I'm applying are conservative.

What does everyone think about this methodology?

I think the only thing I'm not sure about is the Fy of the chords. I have a computer output I received from SJI for Butler joists. The properties under the "diagonals" list indicate Fy=55 ksi. I'm not sure about the bottom chord Fy, I gather it may be 33ksi from this section of the computer calculation:

 

[bones206]

I think the 33 ksi is the allowable (0.6 x 55 ksi).

Would you be willing to upload your butler material? I found myself in the same situation with a new RTU on Landmark purlins. Fun times!

 

[lexpatrie]

I guess I'll go all spectrum on this, that's a Tomanovich spreadsheet that's "yours" in that you are the EOR and you're using that tool. Any bugs or errors belong to you when you use it, which you most likely know. I'm not aware of any retraction due to an error but I couldn't say which version is the current version either.

As far as allowable stress, sure, fine, I would lean on the SJI for the grades of steel they used around the time of original construction, and I think there were some hat channel SJI producers way back when, (no more, as of say 2012), but what always makes me hesitate is the iffy nature of the welding and or the difficulty of measuring the welds and not knowing the weld strengths/electrodes.

Most engineers probably don't go "that deep" in the analysis and look at the elements themselves, but that's half the picture (and the whole "lets neglect the gusset it's too complicated to analyze" thing on the I35W bridge collapse always wanders into my head in these situations. You could look at what would have been needed at the time, perhaps, and base it off the minimum weld throat and whatever electrode was popular, but SJI uses different rules for weld strength than AWS because they are (I guess, better controlled, shown by testing to perform fine)?

As a side note, I've been meaning to post this -

This is a direct link into the internet archive (which is working again) for slideruleera's web site, so quite a few of these catalogs were contributed to the site and you can download them here. There are other threads, of course.

I'm not a huge fan of proprietary joists, and I had a receptive audience on the one project they showed up on, so we bypassed them. I do a lot more 1990s Vulcraft style stuff and that I get more into the guts on because there's better documentation. I also work in an area where this era of construction had the snow load on the roof reduced since, so there's some margin for additional load, especially if the design drawings say it's designed for 40 psf snow load, which is often.

Engineering Tips Web Page

I'll also mention that the webs don't necessarily have the same yield stress of the chords, you can notice that in some of the SJI specifications back when.

 

[EngDM]

I know I'm a bit late, but the picture of the joist on page 1 of the supplemet document looks an awful lot like your condition. I've uploaded all I have for Anthes joists.

 

[TRAK.Structural]

Why not just reinforce the existing joist for flexure and/or shear depending on the analysis? Joist reinforcing isn't terribly difficult or expensive in my experience and just a little bit of reinforcing buys you a lot of extra strength.

 

[bones206]

@EngDM I feel pretty confident these are Butler’s Landmark joists. The fastener holes in the top chord, the continuous crimped webs, and the shape of the joist seat are all characteristic of Landmark joists.

@TRAK.Structural Easier said than done, I think. These joists are super thin, funky cold-formed shapes that don’t lend themselves well to traditional reinforcement. The webs are literally just spot-welded to the chords. I don’t think they are good candidates for reinforcement IMHO.