Need a book? Engineering books recommendations...

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

Vertical Earthquake Forces

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Ben & Harold both provided insight, but here is the real story.
As many of you know, the 1994 and earlier UBC provisions were generally prepared by SEAOC's Seismology Committee.  It never had provisions for vertical earthquake effects, except for a few cases, e.g. cantilver elements, as Harold notes.  Basically SEAOC members did not believe that vertical earthquake was a significant factor for most structures, even though vertical shaking clearly existed.  There are several reasons for this:
1- most structures have substantial reserve capacity for vertical loads.  e.g. we design them for live load that is not likely to be present when the earthquake occurs and also use an effective factor of safety for vertical on the order of 1.6 or higher.
2- Vertical earthquake shaking tends to be very high frequency.  As a result it does not generate much displacement and can not do much damage to most elements.
The IBC provisions are based on the 1997 NEHRP Recommended Provisions, developed by the Building Seismic Safety Council.  The NEHRP Provisions are themselves based on the ATC3.06 document developed in the mid-1970s.  The ATC3.06 document and all editions of the NEHRP Provisions included requirements to design for vertical earthquake shaking. For high seismic regions, the requirement amounted to designing for 20% of the weight as a vertical force.  In regions of lower seismicity, this was reduced.
In preparing the 1997 UBC, the SEAOC Seismology Commiitte wanted to make a transition for users, from the older style provisions to the NEHRP format so that when the IBC was published, engineers would be familiar with it.  One of the things the 1997 UBC did was adopt the ASCE-7 load combination equations.  Under the 1994 and earlier UBC's, the strength design equation for seismic was:
1.4 (D+E) + other loads
Under the new ASCE-7 load combinations, the strength equation was:
1.2D + E + other loads
SEAOC did not believe that vertical eathquake effects were important for most building structures.  However, they did note that the new strength load combinations were less conservative than those traditionally used in the UBC. This made us uncomfortable, since earthquake provisions are largely based on the observation of building performance in earthquakes, and most of the buildings we were observing used the UBC approach.
Therefore, we adopted the NEHRP vertical earthquake effect - which amounts to 0.2D in Zone 4.  The result  is that you end up with the old-style load combination (1.4D+E).  This was required only for the strength load combination because the above problem did not exist in Allowable Stress Design.
When the IBC was put together, SEAOC tried to make it compatible with 1997 UBC, but many folks from the NEHRP process believed that the vertical earthquake was a real effect that needed to be accounted for. - so, as occurs in a political process, the IBC came out different from the UBC.
What should it really be?  As Harold suggests we don;t know the answer.  The fact that vertical earthquake shaking is real is undeniable.  Does it cause failure of building structures?  Who knows- This will clearly be an area of further development in future codes.
Ron Hamburger
Chair, Building Seismic Safety Council, Provisions Update Committee
& Past Active Seismology Committee member