Need a book? Engineering books recommendations...

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

[SEAOC] More on Steel Frame Connections - Response to DWish

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

Sorry it has taken so long for me to get back to you on this.  Your note a
month or so ago (10-23-96) brings up even more good questions that are not
truly being addressed by the current research programs.  I write this back to
you with unfortunately more questions and  fewer answers.  

Your project specific question raises yet another issue that appears to be
just outside of the current scope of the SAC investigations.  That is, What
do we do for Ordinary Moment Frame Connections when the expected loading
should not require plastic behavior??

In my reading and discussions on the subject I have found two main opinions: 

1)  Increase the design force to insure elastic performance.

This is actually the approach that the City of Los Angeles outlined in the
Interdepartmental Correspondence that I quoted from last time.  I did not
include the OMRF section, sorry.  It reads:

Ordinary Moment Resisting Frames (OMRF): 
Cyclic testing is not required.  However, all components of the structure and
their connections shall have adequate strength to resist the design
earthquake elastically.  The connections for OMRF structures shall be
designed using a lateral force reduction coefficient (Rw) of unity while the
beams and columns shall be designed using the standard lateral force
reduction coefficients (Rw) specified by the code.

OMRF?s with a Flexible Diaphragm:
Supervisors may approve a less stringent connection design on a case by case

2)  The ORMF connection still has to be capable of developing a measure of
ductiliy.  The current recognized measure of ductility is the ability to
develop the capacity of the beam.

Within the realm of my discussions with a number of different researchers
this second opinion seems to stem from the recognition that a number of the
pre-Northridge joints failed at stress levels well below the elastic limit of
the material.  This fact appears to be getting ?lost in the shuffle?.  

Although most of the published information focuses on large scale projects
with concrete floor diaphragms, we need to remember that a number of small,
2-3 story buildings with plywood diaphragms had the same type of damaged
connections as the bigger buildings.  I had the misfortune of having to crawl
through the open web joists in the second floor level of one such building
that had approximately 30% of it ORDINARY MOMENT FRAME connections damaged.
 The damage was very consistent with that discussed for large buildings
including, HAZ tearing, column flange divot pullout, and beam flange
cracking.  One significant item was the percentage of TOP FLANGE damage.
 Also consistent with larger buildings, no significant evidence of plastic
yielding of beam or column flanges was observed.  There were however, a
number of connections that had evidence of panel zone yielding, (by evidence,
I refer to the visible flaking of the primer.)

I?m not sure that your statement suggesting SAC Guidelines recommends the
avoidance of bolted connections is entirely accurate.  My thoughts:

It is my understanding that SAC/FEMA is recommending that any connection
proposed for a project either be tested or based on a connection which has a
design procedure founded on the results of a test program.  Obviously there
needs to be considerable caution when extrapolating test results.  A number
of bolted moment connections have been tested through various organizations
including SAC/FEMA, NIST, UC Berkely and IFC/Kaiser; the last one being a
proprietary connection.  The test results for bolted connections have varied.
 Most are able to achieve significant rotational performance, however they do
not as readily achieve the strength requirement for the given beam.  (It
should be noted that, in many opinions this still satisfies the needs.)  One
of the restrictions to using bolted connections is the limitation to smaller
beam sizes or significant bolting requirements.  Although, I believe the
IFC/Kaiser connection has been able to achieve the flexural strength of a
W36x150; I?m not sure of the actual details.  A common failure mode is the
deterioration of strength with sometimes sudden failure at the bolt holes due
to the reduction of net area and potential for stress risers at the bolt
holes.  A typical hysteric loop has signs of pinching, due to slip of the
bolts.  The question becomes; does the slip create the potential for
increased drift of the overall frame system?  I personally feel that the
slippage is not as great an issue as the potential for failure at the reduced
net section.  

I do agree with you whole heartedly that the cost of a testing program is
prohibitive for a small job.  MNH-SMRF has spent close to 1.5 million
unsubsidized dollars in testing, developing, and qualifying their connection.
 Even a scaled back testing program for a building that uses only one
beam-to-column size range and therefore needs only to test one configuration
can expect to spend over $50,000.  I?m willing to bet this exceeds your
design fee for a small job by a factor of 3-4.

The final item in your note appears to describe a connection which eliminates
the backer bar concerns.  From your description it sounds like you are
discussing the repair condition and not the concepts for the new
construction.  Regardless, the main focus at this point is to get the hinge
to form away from the face of column.

Your frustration is noted in your closing paragraphs.  As a suggestion you
recommend resolving the design issues as soon as possible.  As I see it, the
main issue holding back most currently available connection designs is a
validated design value for the through thickness strength of the column
flange material.  To my knowledge, with the exception of the MNH-SMRF and
IFC/Kaiser connections, ALL other proposed connections are adaptations of the
pre-Northridge connection which rely on this through thickness property.
 Interestingly enough, rumor has it that the 8.3 million dollar budget for
SAC Phase II allocates less than $100,000 to the through thickness issues
while it allocates $400,000 to review the socio-economic impact issues.
 Seems to me that providing a good design basis and connections which will
not fracture suddenly, will reduce the socio-economic impact without spending
so much to study it.

One of the most important things we, as the design community, need to do is
make sure that SAC addresses all of our questions. At the SEAOC Convention in
Hawaii, SAC presented the scope of work completed and proposed.
 Unfortunately they did not allow questions to be raised from the floor.
 These questions include the big, as well as, the little jobs.  Some of the
time it appears that the committees is charge become so focused on the
specifics that the broader picture gets lost.  I suggest that we need to
express our concerns to the SAC oversight committee and make sure reasonable
and rational answers are provided.

In response to your frustration that it seems we are faced with paying SEAOC,
SAC or AISC fees for seminars, we have to remember that most of the fees go
to renting the space where we can all meet in order to discuss the issues.  I
am aware of the upcoming SEAOC Short Design Seminars in December to present
the miscellaneous proprietary connections (see yesterdays mail).  AISC is
also going to have a series presenting the current options and design
requirements, look for this in the near future.

Ken O?Dell
email: mnhkodell(--nospam--at)