Seismic Analysis of Structures: Improvements Due to Probabilistic Concepts - Dynamic Behavior of Concrete and Seismic Engineering - page 310

Figure 8.14. Diagram of a sliding oscillator
Thus, we can obtain the average value of the residual plastic displacement, the
probability for the cumulative plastic displacement to exceed a specific value, etc.
2) Structures presenting a series of adherence and sliding phenomena that can be
represented owing to a Coulomb's law belong to the above category. Thus, the 1-dof
sliding oscillator represented in Figure 8.14 obeys the same behavior law as the
perfect elasto-plastic oscillator when the adherence force equals the friction force.
The sliding displacement plays the same part as the previous plastic displacement.
We can mention studies on the sliding problem carried out with a technique similar
to the reference inelastic spectrum technique (see [NED 99], [NOE 93] and [SAR
98]).
8.8.4. Conventional method of stochastic linearization
The flaws of the inelastic spectrum method result from the fact that the
probabilistic nature of the seismic source is unknown. Thus, a solution may consist
of coming back to the separable probabilistic model defined in the previous sections
and trying to statistically analyze the response of the non-linear system.
The stochastic linearization method tries to solve the problem in a quite
comprehensive way. We are going to present its principles while shedding light on