Minimum Steel Joist Load

[jochav52802]

I've got a project where we're looking to add some new rooftop equipment and need to evaluate whether or not the existing steel joists can handle the load.

The load diagram on the original drawings specify loads that are lower than what is specified in the joist load table for the period in which the joists were designed.

Is it possible that the joists were designed for the lower loads specified on the drawings or does SJI require that the joists be designed for the load table values if they control over the specified loads?

Any help is appreciated, thank you!

 

[r13]

You should establish the strength of the onsite joist by measuring member size, length, then check against the joist table, which you should have one already.

 

[dik]

As r13 notes do a quick design check... if short span then the webs may be a single bar bent in a

http://WWWWW

shape and welded to the top and bottom chords... steel strength may vary between the chord sections and the web members. Try to determine which is which. and then check chord capacities using something similar to wl^2/8 and using 0.9d or 0.95d as the effective chord depth... use ql/2 for the reactions and increase this by 1.41 (or whatever) to get the tension in the web members. If there is a little latitude, you may be able do a refined analysis. You may have to get steel samples for lab testing. The above may show you quickly if you are in trouble before you start... use current and 'real' dead loads and live loads.

With adding mechanical units, the greatest problem I've encountered is usually caused by the snow accumulation... in these areas, it is often greater than the new RTU. You might check to see if there is added load due to the obstruction first... I usually do this just to see what I'm walking into. If there is a problem check to see if an RTU of a smaller physical size can be used.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik

 

[JoshPlumSE]

jochav -

This is actually a good question (at least to me). My quick thoughts:
1) If there were no load diagram given in the drawings then I would be pretty certain that the joist listed on the drawing were based on the standard SJI tables.

2) The joists listed on the drawings are probably still based on the standard SJI tables. However, there are times when you want to show a load diagram to joist manufacturer so that they know where to add reinforcement for a point load or such.

3) There are also times when you loads may not be uniform. Where the max moment doesn't appear at mid span and the shear force at the end of the joists may be unequal. In these cases, it may also make sense to show a load diagram to the joist manufacturer. In today's day and age there are some relatively standard design procedures to deal with this. But, some may want to give load diagrams anyway.... just to put the onus / responsibility on the joist manufacturer. I would probably use one of these standard procedures to check the joist listed in the drawing against the load listed in the drawing. Make sense?

The reference I've seen cited the most for this procedure is:
Designing with Vulcraft by Fisher, West and Van de Pas


4) Ultimately, it may be best to check the joist like others have mentioned. Maybe even contact the previous engineering firm or the joist manufacturer regarding what you're seeing on the drawings vs what you see in the field.

 

[jochav52802]

Thank you r13,

Unfortunately, our client won't allow us to access the joists over a certain area since it's in a secure lab space, so measuring the member sizes isn't possible. I was hoping that SJI would require standard joists such as 24K7's to be designed as minimum for the SJI load tables, but SJI told me there is no such provision.

Thank you dik,

I appreciate your recommendation, yet as mentioned above to r13, I can't access a fair portion of the joists due to the secure lab space.

The unit we're adding doesn't require snow drift considerations per ASCE 7-16 due to it's narrows sides.

Thank you JoshPlumSE,

1) Agreed

2) That makes sense, I'm just hesitant to bank on the table value still being reliable since the load diagram produces a smaller load and since SJI told me that there isn't any requirement to design joists for the tabular load as a minimum value for strength.

3) That makes sense. Thank you for the reference!

4) Agreed; unfortunately, the joist Designer, CANAM, doesn't have records for these joists since the project is from 1996.

I wish SJI would mandate that standard joists have no less strength than what corresponds to the load tables. Additionally, I wish they would come up with standard joist configurations and dimensions; it's painful how we only know the joist depth from their designation...having to take field measurements of the members and welds is such a pain. I get giving manufacturers flexibility, but it would be nice to standardize.

Thanks again!

 

[r13]

I think the owner's rejection is irrational. Without verification, you are forced to use the capacity based on whichever information provided is the lowest. It means the joist likely need to be strengthened, or even be replaced, while in fact/reality it might need not to. Then what, will the owner give up the planned additions?!

 

[JoshPlumSE]

Well, you could always have instructions on your design drawings that require the contractor to confirm that specific joists consists of members that are the size and depth associated with a joist of that callout.

I worked on a roof project where we did something similar with our construction drawings. One set of instructions for the contractor to replace any "damaged" trusses that were encountered. And, a different set of instructions for them to reinforce any existing undamaged trusses.

 

[JedClampett]

There's several ways of specifying joists.

Giving the actual joist size. That way you know what it's good for.
Giving a joist size asking to design for certain loads. You might get a joist slightly larger than the loads due to round off.
Giving the joist depth and dead load and total load 26K200/340.

I don't like joists at all, but if I have to, I prefer specifying them the first way. That way I know how much design margin I have. If you put it in the joist suppliers hands, that's it. They'll tell you the joist is good for the load specified and not an ounce more. goodbye, click.
And don't get me started on their calculations. Hey guys! No one uses DOS anymore.
This didn't help the original poster, but it allowed me to vent.

 

[jochav52802]

Thank you r13,

I agree, and am planning to present the option of them allowing us to measure the joist components/welds in order to check it's actual capacity. Our alternative design calls for an independent structure that will cost an additional $100k, so hopefully that will inspire the client to allow us to do the inspection.

Thank you JoshPlumSE,

I like the idea, it's just at this point our client won't allow us to even look at the joists, so that situation won't improve during construction. Additionally, since joist components are not standardized and vary depending on the manufacturer, it's difficult to say what the components/welds associated with that joist should be without looking.

Again, I really wish SJI would standardize joist components, welds and geometry as the non-standardized route makes it tough on the Engineer, especially down the road when the joist manufacturer and building owner no longer have design documents available. Hopefully someday our codes will require that design documents be kept on a database so that they can be found later down the road to assist with additions/alternations.

 

[jochav52802]

Thank you JedClampett,

I agree with you, the first option is best.

There are frequently cases where we have a combination of uniform, partially uniform and concentrated loads to deal with, so it makes sense to provide diagrams in those cases where the tabulated uniform load isn't enough to summarize the situation. In those cases, I think we need a naming convention to indicate when the standard tabulated uniform load capacities don't apply, as in the case when the joist manufacturer designs for an Engineer's load envelope that yields lower design moment and shear capacities. I'd suggest 24K7M instead of 24K7 for instance. I made this suggestion to SJI today.

I bought SJI's Technical Digest 12, which is a bit misleading as their example 1.1 indicates that all one needs to consider when checking an existing joist for new loads is the joist designation on the engineering drawing, yet as this post has shown, if the original EOR specifies a lower load envelope than the tabulated values, SJI's approach would be unconservative. To assist Engineers that may modify my designs in the future, I don't think it would hurt to indicate whether the tabulated SJI values were used as a lower bound on the moment and shear design capacities.

 

[dik]

Contractually that's bad if you are looking to tender this out... better to select a single contractor and have him do the work. Choose wisely... If any uncertainties that cannot be determined before the work commences, then consider a matter of reconstructing a new structure over the existing. In that manner the lab is secure during construction... lab workers look good in hardhats, anyway.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik

 

[jdcollins]

I had a job once that was very similar to what you're describing. Typically, we will gain access to the full length of the joists to be loaded and inspect for general damage, as well as record all the member size and geometry. For this project, half the length of the joist was in a secured space that the owner WOULD NOT allow access to. So, we ended up proposing an exterior steel frame that spanned from column to column and beam to beam. It was probably much more expensive, but they went with it.

Another option would be to put the new equipment somewhere on the roof that you can access all of the joist framing, and then routing whatever power/duct work, etc. you need to the used space.