Lecture 4 Dr.
Aseel Jameel
1. Reliability Networks
A system can form various configurations in performing reliability
analysis.
1.1 SERIES NETWORK
This is probably the most commonly occurring configuration in
engineering systems, and its block diagram is shown in Figure (1). In this
case n number of units forms a series system. If any one of the units fails,
the system fails. All system units must work normally for successful
operation of the system. A typical example of a series system is four wheels
of a car. If any one of the tires punctures, the car for practical purposes
cannot be driven. Thus, these four tires form a series system. The series
system reliability is :
𝑅𝑠 = 𝑅1 𝑅2 𝑅3 … … 𝑅𝑛
Where:
RS: series system reliability
n: number of units
Figure (1) The Block Diagram of an n-unit Series System
For exponentially distributed times to failure of units, the system
reliability is:
𝑛
𝑅𝑠 (𝑡) = 𝑒− ∑𝑖=1 λi t
where
RS(t): series system reliability at time t.
λi : constant failure rate of unit i, for i = 1, 2, 3,…, n
1
Lecture 4 Dr. Aseel Jameel
1
𝑀𝑇𝑇𝐹𝑠 =
∑𝑛𝑖=1 λi
𝑛
λs(t) = ∑ λi
𝑖=1
where:
MTTFS : series system mean time to failure.
λS(t) :series system hazard (failure) rate.
Example: Assume that a system is composed of five independent and
identical subsystems in series. The constant failure rate of each subsystem
is 0.0025 failures per hour. Calculate the reliability of the system for a 50-
hour mission and the system mean time to failure.
Solution:
R = e−(0.625)
R= 0.5353
1
𝑀𝑇𝑇𝐹 =
5(0.0025)
=80 hours
1.2 Parallel Network
In this case n number of simultaneously operating units form a parallel
system, as shown in Figure (2). Each block in the figure denotes a unit. At
least one of the units must work normally for system success. The parallel
system reliability is given by:
𝑅𝑝𝑠 = 1 − (1 − 𝑅1 )(1 − 𝑅2 )(1 − 𝑅3 ) … … (1 − 𝑅𝑛 )
Where:
2
Lecture 4 Dr. Aseel Jameel
Rps = parallel system reliability.
n = total number of units.
For exponentially distributed times to failure of units, the system reliability
is:
𝑛
𝑅𝑝𝑠 (𝑡) = 1 − ∏ 1 − 𝑒 −𝜆𝑖 𝑡
𝑖=1
where
Rps(t) = parallel system reliability at time t.
Figure (2) An n-unit parallel system
Example: An aircraft has two independent and active engines. At least one
engine must operate normally for the aircraft to fly. Engines 1 and 2
reliabilities are 0.99 and 0.97, respectively. Calculate the probability of the
aircraft flying successfully with respect to engines.
Solution:
Rps = 1 – (1 – 0.99) (1 – 0.97) = 0.9997
3
Lecture 4 Dr. Aseel Jameel
1.3 Standby System
In this case one unit is operating and k units are in standby mode. As soon
as the operating unit fails, it is immediately replaced with one of the
standby units. The system has a total of (k + 1) units. Figure (3) shows a
block diagram of a standby system with one operating and k standby units.
For exponentially distributed unit times to failure, the standby system
reliability is given by:
𝑅𝑠𝑏𝑠 (𝑡) = ∑[𝜆𝑡 𝑖 𝑒 −𝜆𝑡 )]/𝑖!
𝑖=0
𝑘+1
𝑀𝑇𝑇𝐹𝑠𝑏𝑠 =
𝜆
Where:
Rsbs : standby system reliability at time t.
MTTFsbs : standby system mean time to failure.
Figure (3) An (k + 1)-unit standby system
4
Lecture 4 Dr. Aseel Jameel
Example: A system is composed of two independent and identical units —
one working, one standby. The standby switching mechanism is perfect
and the unit failure rate is 0.0005 failures per hour. Calculate the system
mean time to failure and reliability for a l00-h mission. Assume the standby
unit remains as good as new in its standby mode.
