Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

Rigid Diaphram and shearwall stiffness analysis in residential Construction

[Subject Prev][Subject Next][Thread Prev][Thread Next]

First, allow me to thank John Lawson for his responses. John was the panel speaker at the Wood Fair in the subject related to rigid diaphragm action in wood framed construction. Johns presentation was one of the most in-depth and informative (if not a bit worry-some in content).

So far, here are some of the comments received - I have added some comments to the end of this post:

John Lawson wrote:

"Your prospects of higher fees for a more complex analysis may be too optimistic. Until the building departments enforce uniformity among engineers in the extent of the analysis done, there will always be someone skirting by with limited effort for a cheap fee. Often times, a plan checker is primarily looking for a engineer's stamp indicating someone is taking responsibility than worrying about whether the diaphragm is accurately modeled. Sad but true."

Lynn Howard wrote:

"Ugh, Ugh, Ugh, Ugh!!!!!!!!  I see a great NEED for some slick software to help out on this issue.  Are you listening Enercalc and others!!!

So my question is, do I need to contact all the owners of wood framed buildings I have designed in the past using flexible plywood diaphragms and tell them that their building may not perform properly in an earthquake?   Where did this provision come from?  I have read many accounts of damage to buildings in the Loma Pietra, Northridge, and Kobe earthquakes, and I never remember seeing an engineer decide that the reason for a failure was that the shear forces distributed themselves into the shear walls in a rigid manner.  I would like some history on this one.  WHERE ARE THE BODIES!!!!!!!! "

Bill Nelson commented:

"Doug T. and I presented a seminar on this topic back in february.  personally i'm not sure if the rigid analysis is justifyable for single family resdences but i can tell you that as of now it is the position of the wood siesmology sub-committee that this level of analysis will be required under the 1997 UBC."

Michael Cochran wrote:

"The diaphragm will still probably be rigid even when the aspect ratio is greater than 2-1/2 to 1 when using the definition in the UBC, particularly in the longitudinal direction. I currently do not believe that the diaphragm will be able to transfer loads by rotation in the true sense as is assumed in a rigid concrete diaphragm analysis, particularly if the diaphragm is relatively long.

Example: 150 foot x 50 foot building with shear wall lines in the transverse direction spaced 30 feet on center (total of 6 shear walls lines). Assuming that there are limited number of shear wall elements along each shear wall line due to door openings and other architectural requirements, the diaphragm being 50 feet deep between shear walls at 30 foot on center will most likely calculate out to be rigid between shear wall lines. I do not currently believe that the diaphragm will actually act as a rigid element and transfer forces by torsion to the extreme end walls 150 feet apart. Instead I believe the diaphragm behavior would be more along the lines of a continuous foundation where the shear wall lines are the springs restraining the diaphragm from moving. Since the shear wall lines most likely have different stiffnesses, I would then expect the 6 shear wall lines to have different translational movement. Since the deflections are different, as one wall deflects more than the other wall, the diaphragm will drag load back to the stiffer adjacent wall, causing an increase in deflection of this adjacent wall. The question is then, that when designing by tributary area, some  walls will be underdesigned for shear (the stiffer shear wall lines) and other will have excess capacity (the less stiff shear wall lines which deflect more).

The example above relates more to commercial buildings, but custom homes, etc can have closely spaced shear walls, the diaphragm will most likely behave more rigidly than has been previously assumed.

The impact on residential framing could be minor, depending upon what standards are developed to account for a redistribution of load (true rigid diaphragm rotational analysis or other simplified method). Additional time will be required for calculating the overall deflection of individual shear walls along each shear wall line, but computer programs can be written for this. I believe a simplified method can be developed that would only require one redistribution of seismic loads to adjacent shear wall lines based upon deflection without having to perform a rotational analysis that more help to minimize additional design time. Yes we are going to have to ask for higher fees since we are spending more time, which will be difficult to do unless building officals enforce the requirement for checking shear wall deflections on all designers. Hopefully the owner can be made to understand that this additional effort this will help improve the performance of ths structure during an earhquake, and is worth the additional fees.

In order to protect the architects requirements, we will probably require the use of more braced frames (not moment frames) in custom residential  homes in order to get deflection compatibility between frames and shear walls. Aspect ratios for shear walls will have to be decreased furthur ( 1.5:1, 1:1, etc) depending upon wheather the building is two stories or a single story, or proprietary shear walls such as Simpson Strong-Wall, STS or Hardy Frame are going to need to be used which have been tested for deflection. 

Before doing shear wall deflection analysis, we are going to need manufacturers of holdowns to provide deflection of their holdowns under  loads (I believe only Simpson publishes this information), wood crushing loads and corresponding deformation on sill plates, and design shrinkage factors for wood. Without this information being standardized (or methods) it will be hard to justify this design criteria."

Bill Allen Responds:

" First, I believe you make quite an assumption that our fees can automatically go up because suddenly the Code is more complex. There is evidence that, on several occasions, code writers are totally insensitive to the business impact of their efforts. From my point of view, it has been a monumental task convincing my clients that my outrageous fees are justified for the prior code changes. At this point in time, I feel that I have discuss this issue with my clients and give him/her a choice: Simplified Static w/rediculously high design forces and "normal" feel or Static with higher fee.

Second, I believe that probably most SFRs perform as if they had rigid diaphragms anyway based on the aspect ratios of the diaphragms and the relative stiffnesses of the shear resisting elements. Analysis by flexible diaphram methods have just been more convenient. Looking at the code specification for flexible vs. rigid, I doubt if many SFRs would have diaphragm deflections twice that of the shear wall deflections particularly when considering the contribution due to the hold downs."
Bob Kazanjy wrote (his own opinions:>):
"The rigid diaphragm method is indeed correct (although a PITA), It is an issue of relative stiffness of the various structural elements (we Me's think a lot about springs; dynamics is our thing). Like a system of springs (some in parallel & some in series). Depending on the relative stiffness of each spring the behavior (load sharing & displacement) will different. Most residential wood floor dias are very rigid when compared to the shear resisting elements. I think most residential systems will respond with ETABS type behavior; X, Y, Theta "
Dennis Wish comments:
My comment about raised fee's was meant to be more cynical than an actual prediction. In this case, I happen to agree more with Bill Allens comments since I have too have faced some very stiff competition as well as resistance from owners who do not understand our purpose.
The presentation for rigid diaphragm analysis used fairly rectangular models with consistency in diaphram levels, depths between shear resisting elements and with no mention of discontinuities created by openings (stairs and skylights). An 8000 square foot one-story custom home - typical of the work I do - would calculate as a rigid diaphragm from the position that the walls are close enough to reduce the aspect ratio to 1  or less.
Learning to distirbute load for a rigid diaphram will require not simply a new methology to me, but will take time for me to build up a "feeling" of how the materials will perform as a system. I think this might be the case with others whose practices have developed around the principles of flexible diaphragms.
I can understand the deflection stiffness for a floor diaphragm, but isn't a rigid diaphragm analysis for an unblock plywood horizontal panel a bit overkill? Has any consideration been given to diaphragm openings, discontinuities caused by diaphragm elevations, diaphragms which are not flat - ie, vaulted ceilings, gables and hip roofs, saw-tooths, shear transfers by "california framing" etc?
One suggestion, although not a simple solution, would be to consider the approach used for the design of URM buildings. As Michael Cochran pointed out, shearwalls of different stiffness along various lines of shear will or should force the diaphragm to distribute load based upon the stiffness (deflection) of the resisting wall. This is essentially (if I understand Michaels opinion) the same opinion held in the design of cross-walls rather than shearwalls. The wall is designed as a damper and does not proportionally draw shear from the diaphragm.
I think the difference, in one respect, is that a rigid diaphragm is expected to distribute shear through rotation - a concept dispelled in wood diaphragms by the amount of damage incurred in open-front structures.  However, if the diaphragm - at least at the roof level - is considered somewhat flexible, than a more rational approach seems logical.
As Lynn Howard suggests, "where are the bodies". John Lawsons examples justifying diaphragm rigidity (if I recall correctly) came from examples of tilt-up diaphragm failures rather than smaller residential models. This is disturbing only because we are complicating a design that has historically performed well as long as the coordination between Architect, Engineer and Builder was clearly defined AND where the Engineer provided appropriate Observation in the field to help educate the builder and reduce field errors.
Any other comments? From our perspective, it seems easy for those creating the code to forget the "hard sell" that design engineers must do to convince the public of the necessity of this. The transition period creats a great deal of mistrust and anger among other building professionals which translates to our clients seeking other professionals to perform the work. By the time they find out that the other professional is bound by the same rules, it's too late for those of us who have lost the work. In addition, if we can not realistically raise our fee's then we end up working for less. Finally, I think that the rational to financially penalize the home-buyer for the inadequacy of the building profession (architects, engineers, builders - et al) is a questionable move. I also think that including residential construction into the code compliance is a bit premature without looking at the performance of more realistic models without the uniformity found in commercial structures.
Dennis Wish PE