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

RE: Rigid Diaph. Analysis

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Good question and the basis of most of my arguments against the new code philosophy. I first discovered this when substituting Hardy Frames for plywood shearwalls where wall aspect ratio's were high and stiffer elements were needed. The stiffness of the steel braced frame is so much greater than wood that more force was absorbed in the analysis and no matter how many braced frames I used, unless the entire structure was converted to a brace framed lateral system, the system would almost always fail until enough rigid elements were added to balance the loads.
This points to the difference of mixing materials with significantly different stiffness values. The difference in stiffness between wood and steel braced frames are significant enough to throw the balance off in a rigid diaphragm analysis.
The next problem is that the code does not offer any guidance as to the recommended method of design. It does test for a Flexible Diaphragm, but does not clearly indicate a "rigid" diaphragm. It does, however, consider the diaphragm to be "Non-Flexible" and thus leaves the decision as to how to balance the results of either OR both to the engineer. I have not, admittedly, kept current on this issue, but I don't recall any changes in the last 20 months or so that would suggest that the Seismology Committee is more definitive on their advice as to how best this is to be addressed in the field.
Once you combine the Flexible results with the Rigid, you "unbalance" the Rigid analysis by creating stiffer elements. In reality, this is probably a moot issue as you are now working far below the capacity level of the resisting materials - in other words, just increasing their reserve - but theoretically, the rigid analysis is intended to be accurate as to how the elements must be arranged so as to balance the rotation and distribution of shear to each element. As I noted before, the stiffer the element, the more shear it takes initially and the greater the risk that it will be the first to fail. In reality, this is probably not the case as the redundant members come into play.
The envelope solution is, in my opinion, the most conservative and the most costly to the owner as it tends to significantly over-design the structure. While I would continue to argue that the flexible analysis is and was sufficient to distribute shear, complianance to the flexible design alone violates conformance to the code and is not recommended by SEAOC. Designs using Flexible OR rigid alone, leaves the possibility that distribution may be higher on one and not the other, and therefore, implies that the engineer may have missed something if liability issues are a concern.
Finally, with the passage of the resent California Supreme Court issue regarding construction defects or non-compliance with the codes (builders can not be sued unless there is reported damage or injury), the engineer will have some protection if he or she decides (by professional judgment) to use the less restrictive results.
I think most of the community has forgotten that these issues still exist and are, as yet, unresolved as to a suggested method supported by SEAOC.

Dennis S. Wish, PE
The Structuralist Administrator for:
AEC-Residential Listservice
(208) 361-5447 E-Fax
ICQ # 95561393

 -----Original Message-----
From: Larry Hauer [mailto:lrhauer(--nospam--at)]
Sent: Wednesday, January 10, 2001 7:33 AM
To: seaint
Cc: Larry Hauer
Subject: Rigid Diaph. Analysis

I am presently designing a two story Type V, (steel/wood), commercial building which has blocked plywood roof and floor diaphragms and a combination of lateral resistive elements; plywood shear walls, OMRF's, and at the center of the building in one direction, a two story chevron braced frame, (2 bays). I have used an R of 4.5 due to the mixed use of lateral resistive elements, and of course, the requirements of the '97 UBC.
My problem is, in using an "envelope" solution for the lateral forces, the braced frame due to it's stiffness, takes almost the entire lateral load in one direction when the diaphragms are assumed rigid. I know, from a practical stand point, that due to the rigidity of the braced frame in comparison to the OMRF's and plywood shear walls it will take more load, but taking almost the entire load causes all sorts of problems such as increasing the rho, overturning, collector forces, etc. In relationship to the stiffness of the braced frame, the diaphragm should be considered flexible, but in relationship to the other elements, rigid. Is there some rationalization, code exception, or modeling, that would allow me to be more practical in how much load this braced frame will take?
Thanks in advance,
Larry Hauer SE