Structural steel Transfer Trusses : Deflection limit

[boobathims]

Dear All
We are designing a steel Transfer truss which supports a floor above. We have limited the Live load deflection as L/600 and Total deflection as L/400.
During the discussion the client representative is saying L/1000 is required that for Total load. According to me , Total load deflection =L/1000 is too
stringent. I would like to know more about the best practice in the steel world.

Regards
Siddha

 

[271828]

I don't know what the best practice is, but I would do the following in general:

1. Camber to counteract almost all of the dead load deflection. Only the structure weight in this calculation. If this deflection is very small, then forget the camber.

2. Limit the deflection due to superimposed dead loads plus live loads to avoid violating floor levelness requirements.

3. Limit the live load deflection. By the IBC, this must not exceed L/360, but that seems too lenient for a transfer truss. I don't have a solid reason for it, but your L/600 seems like a good limit.

You might check out AISC Design Guide 3 to see if it includes help for this.

The client's L/1000 limit might be your best friend on this. Transfer trusses and girders can keep an engineer up at night. It's very nice when the truss or girder is controlled - by a wide margin - by something that wouldn't cause collapse. LOL.

Also, imagine if they're saying you need L/1000 and you use a more lenient limit, and something goes wrong. Something is out of level, a finish cracks, you name it. Even if the problem is not caused by your design, that might be hard to prove. As the design and construction teams grope around searching for the cause, your more flexible (than they originally wanted) truss would be the star of the show.

 

[phamENG]

There are some instances where high deflection requirements make sense, others less so. No idea what this space is or what will be in it, so I don't know if L/1000 is just a 'feel good' number for a layman or a legitimate technical specification.

I agree with 271828, though, that if the client is asking for a more stringent value than is the norm it's best to give it to them. If they can't give you a reason, then you might offer a second concept design of your L/600 and show them how much money it could save so they can decide if their value is really worth it. But if they say they want it, then give it to them if it's possible.

Sheave beams for traction elevators are limited to a deflection of L/1666.

 

[MIStructE_IRE]

Give them the difference in steel tonnage for L/600 vs L/1000.

L/1000 is OTT, but regardless, the “I told you so” factor is already in place now! As such, I’d tend to run with it.

 

[dik]

If he wants L/1000, give it to him... it's his nickel. If you fail to do so, might bring on a lot of liability that you don't want... I've always wanted to do a truss with W21x201 chord sections... but, never had the need.

MIS... you don't even want to do that; it looks bad in court if you've influenced the client.

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

-Dik

 

[Tomfh]

L/1000 is stringent but not that uncommon.

If a client wanted a heavier solution I’d happily give it to them.

 

[rapt]

The problem is with the differential settlement on the supported structure. At less than span/1000, the Australian concrete code requires that you check for the effects of differential settlement of supports.

A good designer should check for that differential settlement in all cases.