Need a book? Engineering books recommendations...

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

Re: DOG BONE CONNECTION - HELP

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Most girders in moment frames are sized for drift, so the actual forces 
and moments due to dead, live and wind loads may not be anywhere near the 
strength of the beam, even with a reduced beam section (RBS). Under a 
major seismic event, however, we expect hinging in the RBS and the rest 
of the frame (end connections included!) must be capable of developing 
this. So think of the RBS as a seismic fuse. At the same time, it sure 
has to be strong enough to resist the other load combinations without 
yielding (for instance, under gravity load). In the end, dogbones have 
more of an impact on the drift, although that impact is typically only on 
the order of a 5 percent increase in required stiffness.

While the dogbone approach is usually a viable option for high-seismic 
moment frames, you will find that it cannot be made to work in every case 
because there are practical limits on how much flange area you can cut 
away. If you can't cut enough away to protect the end connection (with 
nominal elasticity) as the RBS yields, you'll have to go another route, 
like a haunch, coverplates, or ribs. Note that very modest amounts of end 
connection reinforcement have been used in combination with the reduced 
beam section to extend the useful range of this approach. This is 
discussed in the paper to which you referred.

Charlie

MarthaAlic wrote:
> 
> FEMA gives you a design procedure to determine the increased forces at the
> connection assuming a plastic hinge is formed. The paper 'Ultimate Strength
> Considerations for Seismic Design of the Reduced Beam Section, Engineering
> Journal/ First Quarter/1997' gives you the guidelines to determine the
> properties of the reduced section. I understand how to do the moment
> connection, but how can you just reduce a beam section and assume it is
> adequate for dead+live+seismic or any load combination that I might have? If
> my beam properties are reduced, how can it still be adequate? Is there a check
> that I need to perform at the reduced section to find out if the beam is
> adequate for my load combinations. Do you know about a design example to
> follow?
> Thank you
>