Existing Open Web Steel Joists

[vmirat]

I found an interesting web site about the analysis of existing open web steel joists.

http://www.pdhonline.org/courses/s117/s117.htm

Mr. Stuart recommends against analyzing existing joists if you don't know the manufacturer because of the variability of manufacturer's products from the SJI catalog. What do you do if you have to add a load to an existing joist, such as a new HVAC unit, and all you have to go by is 16H5 joist designation (circa 1973 construction)?

 

[JAE]

There are a couple of avenues you can pursue. First, many joist manufacturer's leave small metal tags on the joists with project numbers and joist designations. Many times, the joist mfr. is still in business and you can contact them with the tag number and utilize their engineers to analyze the joist with the new loads (you can give them a load table).

Alternatively, you can use the SJI standard load tables to reconstruct a joist capacity diagram (for both shear and moment). Then compare that to your new shear/moment diagrams to verify whether you are OK or not. This usually suffices for small concentrated loads.

You can also add steel beams between joists to support new loads - connecting them to the supporting girders.

 

[Ron]

Early joist design was largely empirical. Equivalence assumptions have been made over the years with a modicum of analysis. One of the great difficulties is assessing in-place weld area and strength, thus analytical approaches are difficult.

JAE's approach is one that I have used over the years to evaluate new loads on old joists. This, obviously, gets tempered with greater safety factors as less information is known.

 

[PM]

JAE & RON beat me to the answer I wanted to offer. They're obviously not 'off' on July 4th.) I especially agree with JAE's comment ". . . many joist manufacturer's leave small metal tags on the joists with project numbers and joist designations." I'd add the following comments:
[ul][li]These 'shipping' tags are usually on both ends of only one joist in say 5 to 10 joist. The reason is because the tags identify a 'bundle' of one owsj designation during shipping. After the constructor installs the joists in their proper location the tags are often lost or removed.[/li] [li]The tags frequently do not explicitly identify the owsj manufacturer. The size and shape of the tags is a clue to the manufacturer and the font used is a give-away to those in the know, but most engineers would not know which manufacturers use which tags.[/li][li]Other clues to the identity of the manufacturer are: [ul][li]type of section used for top and bottom chords,[/li][li]type and size variation of web elements and[/li][li]type of weldment connecting web and chord elements[/li][/ul][li]If you can identify the manufacturer, you will need to tell him the tag information that will identify his contract file number. In my experience (ie. in excess of 300 buildings with owsj investigated), I've yet to find a manufacturer who can find their own file information for projects over a decade old. Your 29 year old building is a dubious proposition.[/li][li]Even if you find the manufacturer and he finds his documentation, it is unlikely he will have better load test capacity data.[/li][/ul]
JAE's other comment: "Alternatively, you can use the SJI standard load tables . . ." is not a bad idea to estimate the average capacity of a given size / span of owsj, but there are hundreds of fabricators, sometimes including Ma & Pa operations. Subtle differences in the connection details can result in substantial differences in joist capacity so I doubt the SJI tables alone can offer the requisite precision you need for your analysis.

RON's comment that: "Early joist design was largely empirical." is true, but for joists fabricated in 1973 the SJI standards or the CISC Standards in Canada should apply. The rest of RON's comments are right on the mark also.

I'd add a few more thoughts that may be relevant, namely:
[ol][li]If snow loads are significant in your area, then addition of major ventilation equipment on a roof may effect increased superimposed dead loads (ie. the HVAC Equip.) as well as higher live loads (ie. snow).[/li][li]The alteration of the roof may trigger a requirement to assess the existing roof according to current building code standards, rather than the original codes (circa 1973). This can result in a need to reinforce the entire roof. Don't laught, . . .this once happened to me![/li][li]A colleague in a neighboring jurisdiction solved the problem of unknown top chord connection capacity by developing an hydraulic machine that clamps to the web and chord elements, and then applies a calibrated force on each web element while measuring the deformation (strain). I'm not convinced of the efficacy of this 'home made' solution, but it's better than guessing at the weld condition.[/li][/ol]
Have you considered full scale in-situ load tests of a few joists? Obviously the size of the project dictates whether this is justified, but testing is a definitive solution to the assessment of load capacity.
Regards,