Need a book? Engineering books recommendations...

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

Re: Redundancy Factor

[Subject Prev][Subject Next][Thread Prev][Thread Next]
    I've been looking at the ICBO vol. 1 Seismic Design Manual regarding the
redundancy issue for shearwalls.  As I read the UBC, it says to determine what
percentage of  the story shear is carried by each wall and then apply the
10'/wall length adjustment factor.  The example on page 50 does just that (but
since their example walls are all 10' or more there was no penalty).
    However, on page 54 there is a different calculation which looks completely
wrong.  On page 54, rho supposedly is related to the number of 10' walls, which
is senseless.  In my design work, a 9.5' long wall is a handsome shearwall but
using the "method" on page 54, N would be zero and rho becomes 2. That says add
6" to my walls and halve the design load--dumb idea.
    It looks to me like the "wordsmithing" in the UBC got it right, although
the standard wall should probably be 8' not 10' since 99% of our test data is
on 8' walls.
    Also, the load distribution examples show the "flexible" diaphragm
distribution.  For the diaphragms dimensions shown, rigid distribution is more
likely.  Using narrow shearwalls and flexible diaphragms I expect rho would get
pretty high.
Chuck Utzman

"Ron O. Hamburger" wrote:

> A recent thread of discussion on the list server has called my attention to
> what I believe is an uintentional and also unfortunate problem with this
> factor in the 97 UBC.
> When the committee first developed this factor, the intent  was that the
> rmax represent the % of the story shear carried by the most heavily loaded
> element.  We then proceeded to define what an "element" is.  For example,
> each brace is an element, etc.  When we got to shear wall structures, the
> intent was that each individual wall pier across a horizontal plane cut
> through the building would be an "element".  Then, someone on the committee
> noted that if you had a 100' x 100' tltup type structure, with a number of
> 20' wide panels, this would be considered to have high redudance (because
> each 20' panel would be an element) however, if you had the same structure
> with cast-in-place walls, then it would be non-redundant, as the whole side
> of the strucure would be only one element.  In order to solve this problem,
> for shear walls, we introduced the rule that when a shear wall exceeded 10'
> in length, each 10; segement (or part thereof) could be considered an
> element.  The intent was as follows  - If you have a wall line with 10 - 4'
> piers between windows, each pier would be an element.  If you had a wall
> line with a 40' wall, you would have 4 elements.
> Somehow, in the word smithing that went into the actual code language, this
> logic got badly messed up.  Now each wall segement is multiplied by 10/lw.
> This has the desired effect for long walls, but has a penalty effect for
> short wall segments.  This was, in my opinion, never intended.
> This has some serious negative impacts on wood frame construction.
> Please consider this matter, at your next Seismology Committee meeting.