Flexible Diaphragm Deflection

[Once20036]

Along a similar vein as my previous thread is the question of diaphragm deflection between lateral supports.
SDI has a procedure and tons of tables to determine the deflection of a metal roof deck between supports.

In the past, I've considered this carefully in the design and spaced lateral frames to keep this deflection reasonable (2" +/-, depending on the specifics)
Often, this means some sort of lateral restraint at every 120' or so (typically 40'x40' bays in industrial structures).

Recently, I've seen several buildings that utilize braced frames at exterior bays only. For a 300' wide building, this means that there would be significant (8" +/-) deflection between the bays, which strikes me a grossly unacceptable.

Am I mis-interpreting SJI's deflection calculations? It seems like this is significant when you think about the exterior beams/columns deflection with the deck this much and all the second order effects that it would create.

As always, thanks in advance.

 

[JoshPlumSE]

As long as those columns are designed to be able to handle that deflection (and the moments that go with them) then I would think it would be okay. The question is whether folks are really designing those columns to take these deflections?

 

[Once20036]

That leads to another interesting question - one that may have been discussed last month in a different thread -

Does the code weigh in on how much deflection is permissible before the eccentricity must be specifically addressed? I don't image that ever 1/4" deflection needs to get 5 pages of calculations, but is this a judgement call or is it specifically addressed somewhere?

 

[JAE]

Whenever I've done large diaphragms like that I always add additional PDelta lateral forces to the system to deal with it.

 

[TDIengineer]

For a diaphragm deflection of very long span, I would think that flexure is governing the deflection more than shear deflection of the deck itself.

Maybe the designers of these buildings you have seen are performing their chord calculation in a different manner. Not sure what your exterior wall is in this situation, but there are several elements that can be considered for chord action and stiffness that many designers overlook. Such as the ledger, the wall itself, all of the rebar in the wall and parapet etc... Maybe these buildings you have seen have a large chord element you're not seeing.

Just a thought.

And yes, I agree, as long as you have checked deflection compatibility of your gravity carrying elements which includes both forced deflections and accompanying p-delta effects, I don't see an issue with it.

 

[RFreund]

@Josh and TDI -

When you say 'columns' are you referring to all columns or only columns that participate in the LFRS?


EIT

http://www.HowToEngineer.com

 

[steellion]

I consider the diaphragm deflection along with the drift of the LFRS to determine the total deflection at each point and make sure I feel comfortable that deflection is within acceptable limits. Check your numbers, I've gone beyond 300' for a warehouse and the deflection was less than 1". It all depends on the width of the diaphragm. If you have a building that is 300' in one direction and 100' in the other, you may have deflection issues. If the building is 300'x300', the diaphragm deflection may be minimal.

Also, I typically use 0.70 x the service load for wind for deflection calculations to get to a 10-year recurrence interval. Make sure you're using service loads, not strength loads.

 

[JAE]

Not sure why you use the 0.7. Can you elaborate?

Is that allowed by code somewhere to reduce the interval? (I'm aware of the IBC allowing this for components and cladding but not for mwfrs)

 

[Once20036]

The most recent time I dealt with this on a new design, the diaphragm was roughly 150' x 300'. I was comfortable with lateral frames on the exterior only but needed to add moment frames for the 300' direction due to this diaphragm deflection. The final bays ended up at 150'x100'. In the 100' direction, I was getting roughly 1" diaphragm deflection of which 0.95" was due to shear deflections.

Going hand in hand with my previous post, this question was raised by the review of drawings by others. "What they're using as chords" would need to be another thread all together and is an excellent question.

The general concensus above seems to be that anything is ok (from a strength standpoint) and long as the p-delta effects are included in the design. How does this typically work it's way into the work flow for a typical warehouse design? I`m not aware of a way to include it in a ram Model, and checking each column individually could be cumbersome on a large building. Is there an efficient way to check a number of columns simultaneously?

 

[JAE]

Take the deflection of each column line across the span of the diaphragm. Then calculate a lateral force required to resist the PDelta effect using the sustained loads (Dead plus sus. LL) as the vertical loading. Apply that to the building as an additional lateral force. We usually factor this up a bit to account for the iterative nature of PDelta forces (diaphragm deflects - PDelta occurs - diaphragm deflects more - more PDelta - etc.)

 

[RFreund]

My question to Josh and TDI is somewhat a leading question.
But yes as JAE describes is what I would expect. If you are designing with steel, then the Direct Analysis Method already deals with this using Notional loads. The idea of the DAM could be applied to other situations as well.

EIT

http://www.HowToEngineer.com

 

[steellion]

JAE, I take it back, I do not take a 0.70 reduction for diaphragm deflection. Half of my jobs are in ASCE 7-05 and the other half are in ASCE 7-10 right now. For 7-10, you have to use a coefficient to bring the wind loads back to serviceability load cases. That's what I was thinking of. Thanks for catching that.

 

[structSU10]

ASCE 7 has a table in the commentary that has conversion factors for MRI, and for 10 years, if V=85-100 that number is 0.84.

 

[Hokie93]

@ JAE: when checking building drift due to wind loads, do you use the same wind pressures as were used to size the members for strength/capacity? I thought it was fairly common practice to use a lesser mean recurrence interval (say 10 or 25 years) when checking drift due to wind. The ASCE 7-05 Commentary (yes, I realize it is the Commentary and not in the main body of the standard) has the conversion table structSU10 mentioned and ASCE 7-10 goes a step further in the Appendix C Commentary and provides wind speed maps for various mean recurrence intervals. Being that IBC 2009 does not specify the load to be used in the drift analysis ("Structural Systems and members thereof shall be designed to have adequate stiffness to limit deflections and lateral drift", from section 1604.3) and historically the building codes and referenced standards are primarily concerned with strength, I think it is reasonable to use a lesser mean recurrence interval for drift purposes.

 

[TDIengineer]

Regarding the 0.7 multiplier... The 2010 CBC which was based on ASCE7-05 design wind loads allowed you to take 0.7W when checking serviceability for component and cladding wind design pressures. There is not a reduction for the MWFRS.

Table 1604A.3 attached, footnote f.

 

[wannabeSE]

The CBC and IBC don't provide any wind drift loading requirements and do not specify a maximum drift limit. For design, I usually refer to the ASCE 7 commentary.

 

[steellion]

For diaphragm deflection, I do not take a reduction multiple because the Steel Deck Institute examples do not take a reduction.

For MWFRS drift, our company standard is to limit story drift at the center or rigidity to H/400 and at any point to H/300 under 0.70 x ASCE7-05 Wind (50-year recurrence service-level wind).