Existing Steel Joist Assessment

[PROFR]

I am analyzing an existing steel joist. All joist measurement were collected form the field which include joist geometry and member cross section info. I did the analysis considering the design loads on the record drawings but it appears that all top chord members are under-designed by more than 40%. I always experience this whenever I check an exiting joist. Is there a trick that the joist designers know that allows them to push the limits?

 

[SteelPE]

What steel strength are you using? Depending on the age some use 50ksi and there is at least one that uses 55ksi (but doesn't tell anyone).

I assume you took all of your measurements with a micrometer?

Did you submit all of the measurements to the SJI:

https://steeljoist.org/joist-investigation-form/

I gave up a while ago trying to analyze these. Not because I was having issues, but rather because it's quicker... and I have backup in my calculations. It would be interesting to see what the SJI says in regards to the joist size vs your calculations.

 

[PROFR]

Hi SteelPE.
I used micrometer and tried to be very accurate in collecting all the measurements but as I said the joist designers probably know something that nobody else knows. By the way, I used 55 ksi. When you did you calcs, did they make sense to you or you had the same issue like me?

 

[r13]

PROFR,

The joist manufacture has their own method to calculate joist strength, you shall at least find a catalog and find out the differences. The best is as suggested by SteelPE to send an inquiry to SJI to get more exact information.

 

[SteelPE]

Profr,

When we used to calculate the joist capacity they used to make sense. Something like calculating the capacity based upon the joist chord size would come out quite close.

If you are way off, either your calculations are wrong.... or the wrong joists were selected.... or they changed the loading in-between.

Also, I can't say that I have used 55 ksi as I thought SJI capped Fy @ 50ksi.... but I do know some that use 55ksi material.... will design if using 50ksi and then use the extra 5ksi when they are in trouble (my old boss told me this.... he used to work for a mfr).

 

[phamENG]

What do you mean by "under designed by 40%"? Can you elaborate a bit? Are the failures happening in uplift checks or gravity checks? Flexural or axial failure? Or both? Or just when combined? What are your boundary conditions for the analysis? Give a little more of that kind of detail, and somebody may be able to identify a likely input error.

Analyzing these things can be a bit tricky - I've only had to dig that deep into them a few times, and it took some "massaging" of the calculation to get everything to make sense.

Do you know when the building was built? If so, make sure you grab the SJI spec from that time period (currently published in the 75 Year Joist Manual, I think I have a copy of the 60 Year Manual). As retired13 says, they have their own method. There are standardized calculations to determine allowable stresses based on joist geometry. That could help shed some light on how they would have originally designed it.

 

[r13]

Link to Vulcraft design tools. Link

 

[PROFR]

phamENG,
It is the gravity check. Combined Flexural and axial case of members of the top chord. We are adding some extra concentrated loads to an exiting one and I was surprised to discover that the joist is not even safe under the exiting loads before adding any of the new loads. If the joist manufactures have their own methods, shouldn't these methods still abide by the current design codes?

 

[BAretired]

It is possible that the structural steel subcontractor decided to fabricate the joists in his own shop rather than purchasing from a recognized joist manufacturer. In that case, he or his consultant could have erred in the design and SJI would have no way of knowing the capacity of those joists.

There are no special tricks that joist designers know which allows them to push the limits. Joist design is very easy to check if the dimensions and grade of steel are known. An error of 40% in the capacity of the top chord under gravity load should be easily detected.

Check your calculations and also the design loading. If they are correct, then trust them.

BA

 

[Settingsun]

How do the numbers stack up if you leave the flexural component out, ie assume loads applied at the nodes? Could that have been assumed originally?

 

[r13]

The linked file is SJI technical publishing on evaluation and strengthening of existing steel joist. It may not help you for lack of information, but can help you to do the strengthening down the road. Link

 

[oldestguy]

Any chance home construction of joists considered the roof "planking" was to make up some of the compression not available in the joist?
Edit: What role might that decking play when one looks at roof failures under heavy snow loading? Which parts failed? Top chord? My bet is bottom chord.

 

[phamENG]

PROFR: can you post your calculations? Not promising to go through them in a detail, but at this point we have no way of knowing what may or may not be wrong with it. But as BA said - if you have confidence in you calculations and they're telling you the joist won't hold up, then tell the client it wont hold up and recommend a method of reinforcement.

 

[r13]

I guess buckling governs. Need to know the load and assumptions.