> On May 31, 2000, Mr. Turk wrote an email with the following
> information. "On July 14, 1998, I received a packet of maps
> with the cover sheet stating: 'NEHRP MAPS 1994 EDITION.'
> Maps 6, 8, 10 and 12 show spectral accelerations for parts of
> California in excess of 300%g.
E.V. Leyendecker responded:
> Maps 6 and 8 are maps of CA for 10% probability of exceedance
> in 50 years of spectral accelerations for 0.3 sec and 1.0 sec
> respectively. Maps 10 and 12 are maps of CA for 10% probability
> of exceedance in 250 years of spectral accelerations for 0.3 sec
> and 1.0 sec respectively... On the short period 10% in 250
> year maps the spectral values reach 3.0 g, but what is the
> point? These are spectral accelerations so numbers of this
> magnitude are not unusual. I can not speculate on what
> numbers Mr. Turk is comparing to arrive at his factor of 4.
> It is absolutely incorrect to say that the USGS maps were
> divided by 4 or any other number to account for ductility.
> The code equations account for ductility (with R factors
> in its current form) and these factors are not
> always 4, nothing new here. I repeat - No USGS maps of any
> ground motion parameter on a map we have published at any
> probability level have ever been divided by a factor to
> account for ductility.
I can sympathize with Roger Turk's confusion with the NEHRP maps,
since seeing 300%g contours on a map implies much greater accelerations
than used in design. But the problem is that we are used to seeing
acceleration contours on maps which represent ground motion. As
E.V. Leyendecker notes, these maps with high acceleration values are
"spectral acceleration" maps. In reading the FAQ's on the web site for
the USGS maps (see "How do I use these maps?" under "What is probabilistic
ground motion?"), there is an explanation that the spectral acceleration
is based on "a mass on top of a rod having a particular natural vibration
period". As I understand this, the spectral acceleration map thus
represents the maximum acceleration of a single degree of freedom (SDOF)
system with a natural period of the value noted on the map, and thus
the acceleration is magnified above ground acceleration.
Now comparing this to the worst case acceleration from the 1997 UBC:
V = 2.5CaIW/R ; if we set R=1.0 to neglect ductility, I= 1.0, and assume
Soil Type D (default soil type), for Zone 4 and near fault conditions:
Ca = 0.44Na = 0.44x1.5(max) = 0.66 and V(max) = 2.5*0.66*W = 1.65W.
This is still much less than 300%g. To E.V. Leyendecker's attention: It
would be helpful if you could clarify why there is still a noticeable
difference in apparent spectral values.