Flexible Diaphragm & 3 Shear Walls 4

[Once20036]

I've gotten into some discussions with coworkers recently about whether or not it's appropriate to use only 3 shear walls to stabilize a single story building with a flexible diaphragm.
To create an arbitrary example - say you have a 100' x 100' building with a shear wall or lateral frame on the west, north, and east walls.

When the wind blows north/south, I think we can all agree there aren't any issues.

When the wind blows east/west... for a rigid diaphram nobody in the office sees any issues. The north shear wall would take the load and the north/south would act as a pair to eliminate the eccentricity, but is the same true for a flexible diaphragm?

Typically a flexible diaphragm is analyzed as a simply supported beam, and a simply supported beam isn't stable if there's only one support.

A reasonable counter argument is that flexibility only has to do with the difference in stiffness between the supports and the diaphragm. A diaphragm will still have enough stiffness to utilize the east/west walls as torsional restraint.

The first question is simply how other people treat this condition and whether or not it's typical.
The second question is how to quantify that the deck is stiff enough to utilize the torsional restraint.

Thanks in advance!

 

[KootK]

@ DaveAtkins: Thanks for the non-crazy vote. Your last paragraph summarizes my opinions on the matter rather succinctly.

@ MS^2: Can the displacement component that you mentioned not simply be calculated, or at least bounded, by treating it as a rigid body movement of the diaphragm? That's how I've been tackling it.

@ CEL: the SE Asia buildings that I observed directly were mostly frontage buildings. As such, they had open floor plans at the ground floor level to accommodate commercial retail space and/or indoor scooter parking.

I believe that judgement and intuition reign supreme in structural engineering and that calculation should only serve to supplement and inform judgement. I would never encourage anyone to set aside their personal intuition solely based on a technical debate like the one that we're having here. If the argument shifts your needle a bit one way or the other, great. If not, so be it.

It's interesting to note that our collective wisdom regarding torsionally regular buildings will likely turn out to be incorrect. The Christchurch event exposed a bunch of textbook, torsionally regular buildings that experienced the kind of damage one would expect from torsionally sensitive structures.

Examples include low rise buildings with similarly proportioned moment frames on all four sides that wound up having a disproportionate amount of inelastic damage concentrated at one corner. Academia is focusing research attention on this area with an eye towards structural reliability theory. The thinking seems to be that random variation in otherwise regular structures is sufficient to trigger a torsionally significant response.

That mysterious thing that we call"judgement" is in a state of constant evolution, as it should be.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[CELinOttawa]

The "surprise" found in the ChCh EQs has been no surprise at all... Several of Canterbury University's professors have been talking about "true" response modeling for years. I even bought Priestley's "Displacement Based Design" text. Impressive but, to be frank, not practical. Nor does it matter; I paraphrase the eternal words of Wooten: Not knowing doesn't matter if you are still able to produce a reliable structure. In some ways the theory most widely applied in New Zealand (Capacity Engineering) is a school of thought based on not trusting our ability to calculate the real effect of an Earthquake on a structure. You should also read John Scarry's Open Letter to the Structural Engineers of New Zealand. It is a very interesting read, and well worth every practicing Structural Engineer's time.

While it is interesting to see where the "surprises" came from, and the knowing will be of value to every Structural Engineer, this is where the trolley should stop. Not one of the *cough* difficulties *cough* in ChCh was in any way fatal without also involving major issues missed during the (often omitted) site reviews. I for one think that trying to yet again further refine the code would be a mistake. We are already at a level of complexity that is causing errors; Many young engineers are expected to produce results with little supervision, leveraged to death. If we try to make things any more complex, when the existent rules are already proving to be very capable of producing safe and reliable structures, is dumb.

Here I become the nay-sayer that I always oppose when it comes to the ASD versus LSM/LSD debate... As has been stated before in this thread, a robust system with redundancy and secondary load paths is what makes for good Structural Engineering.

P.S. I am confident what will be found in the research being undertaken at

 

[CELinOttawa]

(continues)

Canterbury Uni, et al, will be that the otherwise minor variations normal to a construction site which cause the variation from ideal/model to real world/actual are responsible for changes in stiffness and strength that will prove all but impossible to eliminate. What I mean is that the random variations, ie: bar spacing, specific height, formwork discrepancies, mech/elec openings, cold joints, material strength variations, etc, are going to mean that there is a practical limit to the accuracy of our modelling.

Anyone who expects a structure to behave exactly as modeled is a fool. Practical considerations must be allowed to drive the final solution: Safe and sound, and economically non-crippling, is frankly good enough.

 

[TXStructural]

I don't see this linked here, a 1999 discussion on the issue:

http://www.seaint.org/wood.htm

In my book, it is generally unacceptable to design with a truly 3-sided, c-shape any longer. Too many bad things happen under lateral loads when this layout is used. Implementing some kind of 4th side reduces the risk, even if it is very short, stiff wall segments or a moment frame.

We have also had success by stiffening the diaphragm (such as by adding diagonal ties) to create a stiff lateral load path.

As far as distribution based on diaphragm and wall stiffness for 4-sided systems, I like Dr. Richard Klingner's (U Texas, retired) approach:
compute distribution based on 1) rigid (ratio of wall length) AND 2) flexible (proximity/equidistance) and then design based on the higher value for each wall. For example, if you have two lines of wall, one twice the length of the other, the flexible scenario puts half on each wall, while the rigid scenario puts twice as much into the longer wall (2/3 & 1/3). Thus, you design the long wall to take 2/3 of the force and the short wall to take 1/2 the force (the greater of the two scenarios for each wall.) This approach could easily be used for a "C plus moment frame", even though doing so by stiffness would result in about 100% of lateral for the wall and 50% of lateral into the frame (infinitely stiff wall compared to a moment frame.)

 

[Archie264]

>>>Here I become the nay-sayer that I always oppose when it comes to the ASD versus LSM/LSD debate... As has been stated before in this thread, a robust system with redundancy and secondary load paths is what makes for good Structural Engineering.<<<

Sorry, can't have it both ways.

(Just funning with you.)

 

[KootK]

I must have a copy of that Scarry Letter. Do you have a copy that you can post CEL? I did some serious googling but was only able to find links to documents discussing the letter.

In my opinion, structural reliability theory is of great practical value. Few things exemplify dispassionate rationality better than statistics, at least until they wind up in the hands of some politician.

One of the best features of structural reliability theory is that the results generally do not lead to a bunch of extra code complexity. Rather, they lead to a simple factor being adjusted one way or another. That is my expectation for the fallout from Christchurch. Currently, torsionally regular buildings are heavily favoured by seismic codes. The data suggest that a reduction of that favouritism may be in order.

I agree completely when it comes to avoiding additional complexity in our design work. I find the following examples particularly salient:

1) There is code committee talk of eliminating the need for accidental eccentricity when non-linear RSA/Time History analysis methods are used.

2) In the US, some buildings are required to have dual lateral systems, usually shear walls in combination with moment frames, UNLESS performance based design methods are employed.

In both of these instances, safety margins are being relaxed in exchange for employing a more complex analysis method and, ostensibly, "knowing more". This, while I find that I often have to correct even licensed junior engineers with regard to which direction their reinforcing hooks should be pointed.

It has long been accepted dogma in our profession that results can be "skinnied" when more sophisticated analyses are performed. Modern analysis methods have become so complex, however, that I almost feel that it's time for na additional safety factor to be instituted to cover the increased odds of error when wildly complicated methods are used.

phi_material; gamma_load; phi_method too frickin' complicated for humans.




The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Archie264]

>>>Modern analysis methods have become so complex, however, that I almost feel that it's time for na additional safety factor to be instituted to cover the increased odds of error when wildly complicated methods are used.<<<

Assuming that's tongue-in-cheek then wow, that was a good one!

 

[Hokie93]

@KB4894: To the best of my knowledge, ASCE has not published a solution to the "3-Sided Diaphragm" Trial Design Problem mentioned above. I just checked the ASCE website and did not see a solution. If I recall correctly, for certain problems in the past, it has taken 12 months or longer for a solution to be posted.

 

[CELinOttawa]

KootK: I did not know that the current "vogue" in the code committees was a tendency toward relaxing some of the duplications. Nice to hear, to be frank.

As for the Scarry letter, I do indeed have copy which I could post. I will not, however, do so.

The letter was and remains an open letter to STRUCTURAL ENGINEERS. John Scarry is a personal friend, and a very sensible man. He knew when he penned the letter that it would shock and possibly unnecessarily frighten the layman, or be used to embarrass the profession. He is wise enough to know that unintended consequences are the main outcome of every act and omission. As such it was never made publicly available and is *not* available anywhere on the web. Hence your dry search despite earnest googling.

I will distribute a copy of the Open Letter to any member who's concentration is listed as "Structural" or "Civil". Email me on m (no space here) quinn <at> celottawa (dot) ca.