Civil Engineering Reference
In-Depth Information
be selected arbitrarily. However, it is suggested that E [ B i ]
k *)/ e g ( T ) ;
where e g ( T ) is the number of shocks by time T (i.e., time-window for the
study) according to the trend function g ( t ).
=
( v 0
15.4.5 Comparison of graceful deterioration models
Consider a structure that deteriorates continuously with time; for example
a concrete sewage pipe subjected to bacterial attack. The deterioration
process can be approximated by a discrete shock-based approach according
to the models presented in previous sections. Consider that the pipe is put
into service at time t
=
0 with an initial capacity v 0
=
100 (capacity units);
the minimum acceptable capacity is k *
25. Three models for graceful
deterioration that follow a Gamma process are presented in Fig. 15.5. The
cases shown were constructed for a scale parameter and the following time-
dependent shape parameter functions: v ( t )
=
0.125 t 2 and v ( t )
=
5.25 t , v ( t )
=
=
exp(0.11 t ).
Consider now the case in which shocks are iid and occur at fi xed time
according to a function g . In Fig. 15.6 three sample paths of the deteriora-
tion history process are shown. Every process corresponds to a different
trend function g . It can be clearly observed that the times at which the
random increments are evaluated depend on the trend function. Thus, they
80
Δ t
70
60
Fix time intervals
50
Δ t
Δ t
40
30
v ( t ) = 5.25 t
v ( t ) = 0.125 t 2
20
10
v ( t ) = exp 0.11 t
0
0
5 0 520
25
Time
30
35
40
45
50
15.5 Gamma model for graceful deterioration (adapted from Sánchez-
Silva and Riascos-Ochoa, 2012a,b).
 
Search WWH ::




Custom Search