Need a book? Engineering books recommendations...

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

RE: A seismic question with no (easy) answer?!

[Subject Prev][Subject Next][Thread Prev][Thread Next]
I read the code much more conservatively.  1617.3 provides the general
requirement that buildings be separated by a sufficient distance to avoid
"damaging contact."  What is damaging contact?  I don't know.  Maybe it's
contact that destroys some portion of diaphragm of a building, maybe it's
the knifing effect you refer to, maybe it's just damage to the veneer that
poses a falling hazard to people below.  Frankly, I don't know so I assume
the latter and keep the building separated a distance that they
(theoretically) won't touch during an earthquake.

To do this I take the amplified (inelastic) deflections for each building
and add them together to determine my expansion joint size.  This is more
conservative than using the SRSS method allowed for higher design
categories, and I know that seems like too much, but I can't justify any
other way. I know that I'm saying that two buildings will displace out of
phase, during a once in 2500 year event and the likelihood of that is less
than winning lotto - but I don't know enough about the research behind the
code that led to the requirements.  Is equivalent lateral force less
conservative than modal analysis?  Are the period calculations so inexact
that this is some sort of factor of safety?  Is the code covering for the
fact that in SDC A,B,C you don't account for all of the structural

To get back to your original post - I'm not in California, I'm also in the
northeast, but if I was upgrading a building to meet current seismic code
under your scenario I guess I'd take the easy way out and ask for a code
modification.  As backup, I'd provide information showing that the
non-compliance of the neighboring building makes it impossible for the
project to comply.  I would expect to be required to upgrade the project I
was working on to a reasonably stiff building with shear walls or
concentrically braced frames.  If you've shown that you've done everything
in your power to make the building work I can't imagine that a reasonable
code official would deny the request.  

You say this may become a problem when IBC kicks in.  I believe it's already
a problem.  BOCA '96 has the same requirement.  The biggest problem is no
one attempts to address it.  I've heard people say "c'mon, when was the last
earthquake that caused damage?" or "if the slabs are at the same level, it
doesn't matter" or "I just double for the number that RAM spits out (without
increasing by the def amp factor)." If you follow the letter of the code, or
even if you follow the SRSS method, you'll have big (like 4") expansion
joints at the first story of two moment frames buildings separated by a
joint.  And you'll be one of the very few in the NE to actually do it - and
you'll become known as the guy that's too conservative in your designs.
Then who's going to want to hire you?

Please reply with any comments.  I've also been looking to hash this out
with someone.  I'd love to find a reason to not have to be Mr. Conservative.

Peter Griem, P.E.

-----Original Message-----
From: Cliff Schwinger [mailto:clifford234(--nospam--at)]
Sent: Saturday, September 28, 2002 11:49 AM
To: seaint(--nospam--at)
Subject: RE: A seismic question with no (easy) answer?!

I haven't had any responses to my post, but I think I came up with a
satisfactory answer to my question about seismic pounding at expansion
joints.  Here it is:

Section 1620.3.6 requires that buildings that are SCD "D" and higher be
designed so that they are separated from adjoining structures by the
square root of the sum of the squares of the deflections of each of the
adjoining structures.  This provision is a "detailing" requirement in
Section 1620.  If I "read between the lines" I guess I could infer from
this requirement that pounding for SDC "A", "B" and "C" buildings can be
ignored. This seems to make sense. Just because buildings pound during
an earthquake does not mean that they are going to have severe damage.
The level of damage that might occur on SDC "A", "B" and "C" buildings
would probably be acceptable. This is because these buildings are either
buildings in lesser restrictive Seismic Use Groups in regions of
moderate or high seismicity or are buildings in the more "important"
SUG's in areas of low seismicity. It would not make sense to require an
office building in Philadelphia (where there has never been seismic
damage to any building structure) to meet the same prohibitions against
pounding that are required of an office building in San Francisco.

I looked at a number of photos of pounding damage that occurred on some
buildings in California during the Loma Prieta earthquake and these
photos reinforce my opinion that I am on the right track with my
conclusion. "Pounding" does not equal "collapse". Of course if I was
designing a building in an area of low seismicity (for instance,
Philadelphia) I would still want to keep building separations
sufficiently wide to prevent slab edges from "knifing" columns if I had
conditions where adjacent floors did not align. I am still not sure what
the answer is to expansion joint widths in building facade elements. The
expansion joint gap should probably be wide for facade components such
as precast panels, but I would think light-weight facades such as EIFS
panels could have smaller expansion (seismic joint widths). The
consequences of pounding precast panels would be more severe than the
consequences of lightweight facade elements pounding.

Sound reasonable?

Cliff Schwinger 

> -----Original Message-----
> From: Cliff Schwinger [mailto:clifford234(--nospam--at)]
> Sent: Thursday, September 12, 2002 11:37 PM
> To: seaint(--nospam--at)
> Subject: A seismic question with no (easy) answer?!
> Your following message has been delivered to the list
>   seaint(--nospam--at) at 20:47:22 on 12 Sep 2002.
> How do engineers working on projects in earthquake prone areas
> (California) handle expansion joints in existing buildings when
> additions or other modifications are made to the structures and the
> building code requires the existing structure to conform to current
> seismic requirements for new structures?
> Here is a fictitious example (but one that will be very common in the
> years ahead under the IBC Code in the northeastern U.S.):
> -	A 20 year old eight story structural steel building was designed
> for future vertical expansion for 4 additional floors.
> -	The building has a 1 ½" expansion joint down the middle and the
> two halves are separated by double columns.
> -	The 4 additional floors are now being added - but only on the
> north side of the expansion joint.  No work is being done to the
> existing structure on the south side of the expansion joint.
> -	This fictitious building is located in an area of low
> seismicity.  When the original structure was designed the building
> required no consideration of seismic forces.  The present code is IBC
> 2000 and there are sizable seismic loads that must be considered
> the building is an "essential facility" and is founded on poor soil.
> -	The buildings will "bang" during the design earthquake.  The
> "banging" was not an issue prior adding the new floors, but now that
> new floors are being added, IBC 2000 requires the lateral load
> system in the half of the building with the new floors to be upgraded
> meet IBC 2000 requirements.  The building code does not require the
> structurally independent framing on the other side of the expansion
> joint to be upgraded because no work is being done to the existing
> structure on that side of the expansion joint.  Even if we did upgrade
> the lateral load resisting system in this half of the building, the
> existing expansion joint is so small that the halves would still bang
> during the design earthquake. (All floor slabs align at the expansion
> joint, so there is no potential for floor slabs to "knife through"
> columns.)
> The questions I have concern banging of the building halves at the
> expansion joint:
> 1.	What do we do about the "banging building" situation? Even if we
> were to cut the slab back and put in a big expansion joint, the double
> columns are close together and they'll hit each other. In fact maybe
> "banging (steel) columns" are not so bad due to the ductility of
> 2.	The code doesn't prohibit buildings from "banging" - it just
> says that there can't be any "damaging contact". If building segments
> bang together there may be lots of small damage, but I would think
> if both buildings are of the same construction type, if all of the
> floors align on either side of the expansion joint, and if the
> structural framing is ductile, then banging action would not cause
> collapse.  What is the definition of "damaging contact"?  If damaging
> contact means damage that causes a threat to public safety, then I
> say that slab edges hitting each other are not a concern.  On the
> hand if facade components "bang" at the expansion joint (same 1 ½" gap
> at expansion joint in facade), then that would be a threat to public
> safety and that condition would have to be corrected.
> 3.	What is done in California when renovations or additions are
> made to old buildings with expansion joints and those joints are too
> small to prevent contact during a big earthquake?
> I heard that there is software that can actually model the "banging
> building" phenomenon.  Is this type of analysis common in California?
> Can anyone shed some light regarding practical solutions to this
> problem?  As I said at the beginning of this post, the specific
> that I outlined in this post all relate to a fictitious building, but
> this problem will have to be addressed in many renovation projects
> in the northeast U.S. where I live when IBC 2000 kicks in.
> Thanks.
> Cliff Schwinger
> ******* ****** ******* ******** ******* ******* ******* ***
> *   Read list FAQ at:
> *
> *   This email was sent to you via Structural Engineers
> *   Association of Southern California (SEAOSC) server. To
> *   subscribe (no fee) or UnSubscribe, please go to:
> *
> *
> *
> *   Questions to seaint-ad(--nospam--at) Remember, any email you
> *   send to the list is public domain and may be re-posted
> *   without your permission. Make sure you visit our web
> *   site at:
> ******* ****** ****** ****** ******* ****** ****** ********

******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at:
*   This email was sent to you via Structural Engineers 
*   Association of Southern California (SEAOSC) server. To 
*   subscribe (no fee) or UnSubscribe, please go to:
*   Questions to seaint-ad(--nospam--at) Remember, any email you 
*   send to the list is public domain and may be re-posted 
*   without your permission. Make sure you visit our web 
*   site at: 
******* ****** ****** ****** ******* ****** ****** ******** 

******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at:
*   This email was sent to you via Structural Engineers 
*   Association of Southern California (SEAOSC) server. To 
*   subscribe (no fee) or UnSubscribe, please go to:
*   Questions to seaint-ad(--nospam--at) Remember, any email you 
*   send to the list is public domain and may be re-posted 
*   without your permission. Make sure you visit our web 
*   site at: 
******* ****** ****** ****** ******* ****** ****** ********