Net Uplift & LRFD on Steel Joists

[structuraldan]

With the help of a senior engineer, I designed a large open bay steel warehouse building, 105' wide by 400' long by 45' high, and used steel joists and girders for framing. The joists are 105' long, 8' deep, moment resisting and have restrained top and bottom chords. Standard joist tables are not applicable due to roof slope and joist depth.

Due to these conditions, and use of LRFD design, we decided the best approach was to provide shear and moment envelopes for each joist design, instead of unfactored wind loads per joist tributary area. This would allow us more control as to what would be provided and provide best interpretation of design requirements. To assist, I also indicated envelopes are based on factored loads and sign convention for uplift on the drawings.

The contractor's fabricator insists on requesting net uplift load psf, "per SJI specification requirements" in the latest RFI. Coworkers in the office agree that not only the information provided on the drawings was more than sufficient for design, but required due to OUR design interpretation of the building system, vs. the joist designer's need for his component design. (Although an important one, presumably using ASD with stress increase allowance for wind forces.)

Is there a viable reason why the joist designer cannot use the information provided to size the joist members? (Also why is LRFD not adequately addressed in SJI?)

I'd like to get as many comments as soon as possible to give due consideration to the RFI. Thanks

 

[JAE]

What do your V/M envelopes pertain to? Are they one combination or do you provide multiple combinations for design? Are the dead, live and wind V/M values given in separate envelopes?

Or by the word envelope, do you mean that the worst case shear and moment values are shown on one diagram?

You may be limiting the joist engineer's ability (in their own mind) to design the way they want to, the way they are used to, or the way their computer program demands input.

Sometimes in RFI's I ask for clarification as to the reason/intent of the request...

 

[TFL]

thye need uplift for negative bending/compression in bottom chord/ and bracing of botom chord.
they would need this unless you specified a negative moment to design for - or there is never compression in botom chord.

 

[structuraldan]

The "envelopes" are worst case load combinations on one diagram, so + and - values and max/min locations are shown for V & M. I understand that they need uplift values to design members and bridging, and that with this value, again presumably unfactored, they can plug into their black box to design members and use a 1/3 increase in stress to minimize sizes.

This is what we were trying to avoid since this does not incorporate interaction with other load combinations or the use of load factors.

I guess it comes down to information they could derive or I could back-calculate for them...

Thanks for the posts!

 

[haynewp]

Is the 1/3 increase allowed in the Code you are using?
Are you requiring the joist designs to be signed and sealed?

I think that they can use whatever method they want, ASD or LRFD, if they are going to take full responsibility of the design. You may should have put the requirement in your specifications that the joist manufacturer has to use LRFD if it is that important to you. Unfortunately, to get them to use LRFD and your diagrams instead of what their software is set up for may require an additional fee at this point.

 

[connect2]

Well i think its fair enough to supply your load criteria as 'specified' loads, ie no load factors applied, and then let the 'pre-engineered' component manufacturer use either LFRD or WSD as the case may be. It is in the end though your responsibility as EOR to review. It helps all parties to see whats Live Load, Dead load, Wind Load, and Sesimic Load seperately. It's an accounting procedure for sure keeping track of it all. I would never assume that that you are supplying the ultimate or necessary 'design' criteria for an element that your not actually designing and are infact asking others to design. Give them the 'real' loads and let them figure out how to handle the various load combinations, load factors and factors of safety as the case may be for THEIR propietory product. Then review it or design it yourself.