Response Spectra in Managua, Nicaragua


Seismic Hazard Analysis - part 1 | part 2 | part 3

The parameters of seismic hazard can be separated in three different categories:

  1. Ground movements.
    • acceleracion
    • velocity
    • displacement
  2. Predominant frequency.
  3. Duration.

The seismic phenomenon is fundamentally an aleatory process; uncertainties exist for the distribution of the events both in time and space. This fact in the first place, and the limitations for both quantity and quality of seismological data, in second place, demand to introduce probabilistic concepts in the Seismic Engineering researches.

The methodology to assess the seismic hazard involves the probabilistic evaluation of the following [Larsson & Mattson, 1987]:

  1. When do the earthquakes occur? The probability of occurrence of the earthquakes is calculated.
  2. What is the size of the earthquakes? The maximum magnitude that can be expected over a certain interval of time is defined.
  3. Where do the earthquakes occur? The distance to the earthquake sources that could affect the site under investigation is calculated.
  4. What is the local influence? An attenuation relationship of the seismic waves is set.

The most important concepts to define the seismic action include the return period of an earthquake and the probability of exceedence. These concepts are associated with the useful life (or economic life) of the structures.

Occurrence probability

Occurrence probabilities vs. Useful life, for different Return periods (Carballo, 1993) The return period can be defined as the mean time among seismic events of equal characteristics.

Frequently, seismic phenomena are idealised as Poisson processes. The Poisson distribution is based in that events are independent among each other, also regarding the time between each. The reason to use this distribution is the simplicity and the handling easiness instead of the adaptation of such hypothesis to the reality [Barbat & Canet, 1994].

The return period for events that follow a Poisson process is: T=1/N, where N is the mean number of events with certain characteristics that occur in time.

The exceedence probability (or occurrence probability) is the probability that at least one earthquake of certain characteristics will take place in a given period of time (return period).

The useful lives considered for common urban or industrial constructions are between 50 and 100 years.

The exceedence probability associated characterizes the level of intensity of the design earthquake.

Two exceedence probabilities of interest are considered. The first is associated with the occurrence of moderate magnitude earthquakes, with relatively high probabilities of occurrence. These are called operation earthquakes.

The structure should resist their action without suffering important damages that would put it out of use. Repair of damages should have a reasonable cost.

The other exceedence probability of interest is associated to events of high magnitude, but sporadic; that is the strongest earthquake that would take place only once in the lifetime of a structure, and it is denominated accidental earthquake.

It is economically justified to accept that an earthquake with these characteristics produces important structural damages, whenever at the same time the collapse of the construction; the loss of lives and of material goods is avoided [Barbat & Canet, 1994].

For operation earthquakes exceedence probabilities between 50% and 64% are considered. Smaller probabilities do not correspond to operation conditions [Carballo, 1993]. On the other hand, selecting a reasonable security level to take into account the accidental type actions is a subjective work. The acceptable level of risk is governed by diverse factors, as the economic ability to accept such a risk and the effects that a bad performance of the structures could have on the population [Carballo, 1993]. The National Regulation for the Construction (RNC-83) recommends for the accidental case a 10% probability.

part 2 >
up

Copyright ©2002 - 2007   Rainer M. Parrales   All rights reserved.

Last updated: Sat, 21 Jul 2007