Need a book? Engineering books recommendations...

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

Re: x brace question

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Tarek Mokhtar wrote:

> For an ordinary x braced steel frame with tube steel braces
> can one assume the brace buckling length for in and out of plane
> to be 1/2 the actual brace length ?
> 

I have read several other responses that indicate that out-of-plane 
buckling of the compression diagonal is not braced by the tension 
diagonal.  That is generally NOT the case.  You might refer to the 
following articles (and their references) that discuss theoretical 
and experimental work on this topic:

Picard and Beaulieu in AISC Engineering Journal, 3rd Quarter 1987, 
pp. 122-126.

Picard and Beaulieu in AISC Engineering Journal, 4th Quarter 1988, 
pp. 156-160.

Stoman's discussion of Picard and Beaulieu papers.  Discussion was 
in AISC Engineering Journal, 4th Quarter 1989, pp. 155-159.

To summarize for common cases (braces pinned at ends and connected to 
each other at the crossing, same section and material properties for 
both braces):

When C/T < 1.6, K=0.5 for full length of brace.  (Where C is the 
force in the compression brace and T is the force in the tension 
brace.)

When T = 0 in one brace, use K = 0.72 for full length of compression 
brace. (Note that this is NOT both braces in compression, but is one 
brace in compression and the other unstressed.)

Others mentioned using K=0.65.  The AISC codes (ASD and LRFD) use 
K=0.65 instead of K=0.5 for a fixed-fixed column condition because 
infinite fixity is impractical.  The conditions leading to K=0.5 (of 
full length) for X-braces is different because it is based on bracing 
due to axial stiffness and not flexural rigidity.  Using K=0.5 for 
the full length is conservative as long as C is not greater than 
1.6T.  Using K=0.65 for balanced conditions (C=|T|) would be 
analogous to using 1.3 times the distance between braces for typical 
compression members--that is, unnecessarily conservative.

Many designers (and authors of papers) use K<1 where they can 
identify partial end restraint (for instance, in trusses with all 
welded connections).  Similarly, some use K<0.5 for full length of X 
braces where the design provides end restraint in the connection.

As a practical note, ductile detailing of X-braced systems assures 
that the tension brace really can brace the compression brace because 
the tension connections are strong enough to guarantee that 
compression brace buckling controls (and the braces are usually 
balanced).  Unless your braces pick up a considerable amount of 
gravity load compression, one brace is much larger than the other, or 
your braces do not cross at midlength, you should not end up with 
conditions that would require using K>0.5 of the full length.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Michael Valley                                   E-mail: mtv(--nospam--at)skilling.com
Skilling Ward Magnusson Barkshire Inc.                  Tel:(206)292-1200
1301 Fifth Ave, #3200,  Seattle  WA 98101-2699          Fax:        -1201