Need a book? Engineering books recommendations...

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

VERTICAL

[Subject Prev][Subject Next][Thread Prev][Thread Next]
 
Thanks, Harold for the text below.
So nice to hear from someone who is on top of the essentials of the issue.
 
By the way, if you are running FEA and want to get a good response,
your mesh along columns must be much denser than usual to get the true effect
of the high-frequency forcing.
 
If anyone wants to see it, I have a simulation of a virtual  building
disintegrating under a vertical component.
 
Gregory
 
 
 

  The vertical component issue is in a state of flux.  But the seismic =
code=20
  developers are making progress.  Previously, we only looked at what I =
will=20
  call inadvertent vertical effects.  These were vertical elemental =
forces=20
  that were caused by the structural response due to lateral ground =
motions. =20
  They occurred in roughly the same time domain as the lateral ground =
motion. =20
  The ground motions were historically fairly easy to quantify.

  The next considerations were direct vertical ground motion effects.  =
This is=20
  where the ground moves in a vertical direction.  The most notable US =
example=20
  was the 1994 Northridge event and Kobe.  Previously, the concern and=20
  concentration was the lateral faults that generated the big lateral =
ground=20
  motions.  After Northridge, we started paying more attention to the =
thrust=20
  faults.  The other type of faulting that can generate large vertical =
ground=20
  motions are the subduction zone faults.

  The big difference is the arrival time of the vertical ground motions. =
 The=20
  vertical ground motions arrive with the P waves and before the large =
lateral=20
  ground motions (mostly S and Love waves).

  Another complicating issue was the mapping.  There are currently no =
maps=20
  that show the vertical ground motions.  You are required to develop a =
site=20
  specific response spectra for the lateral and the vertical ground =
motions in=20
  order to start the procedure.  This is not too much of a problem for ="">nuclear=20
  plants.  But good news is on the way.  There is an effort to develop =
maps=20
  and / or algorithms to quantify vertical ground motions without having =
to=20
  develop site-specific spectra.  There are elements in the DOD and USGS =
that=20
  are working on this and will have it fairly soon.  NEHRP, ASCE 7, and =
IBC=20
  will follow soon.  There are many of the "usual suspects" involved in =
all of=20
  these efforts.

  Generally if you do NOT have a thrust fault or subduction zone fault =
within=20
  10 km of your site, you can ROUGHLY assume about 2/3 of the lateral =
ground=20
  motion will be a vertical ground motion.  If you are within that magic =
10=20
  km, you may have vertical ground motions that may be larger than the =
lateral=20
  ground motions.

  If you perceive that your structure is sensitive to vertical ground =
motion=20
  (i.e. cantilevered structures or elements, liquid containing tanks or=20
  vessels, etc.) and you are within the magic 10 km, you may want to get =
a=20
  vertical response spectra.

  Once you have the input, you then need to decide what to do with it.  =
There=20
  are no R values for vertical structural responses.  You may need to =
consider=20
  amplifying vertical forces or performing a vertical dynamic analysis.  =
I=20
  have done both on various projects.

  The IBC simply disclosed the level of our ignorance that was contained =
in=20
  the UBC.

  Stay tuned, when this is developed further, I will notify the list.  =
Expect=20
  it in early 2005.

  Regards,
  Harold Sprague


Sincerely, Gregory
 
Dr Gregory Szuladzinski
Analytical Service Pty Ltd
5 Warwick St
Killara/Sydney
Australia 2071
T: 612 9418-1842
E: ggg(--nospam--at)bigpond.net.au