Scribd
Upload
121 views

Reliability and Maintenance (MANE 4015) : Instructor: Dr. Sayyed Ali Hosseini Winter 2015 Lecture #4

The document discusses probability distribution functions that are commonly used in reliability analysis, including the exponential and Poisson distributions. It provides the equations and properties of these distributions and includes examples demonstrating how to calculate probabilities using the distributions.

Uploaded by

Felipe
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
121 views

Reliability and Maintenance (MANE 4015) : Instructor: Dr. Sayyed Ali Hosseini Winter 2015 Lecture #4

The document discusses probability distribution functions that are commonly used in reliability analysis, including the exponential and Poisson distributions. It provides the equations and properties of these distributions and includes examples demonstrating how to calculate probabilities using the distributions.

Uploaded by

Felipe
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 26

Reliability and Maintenance

(MANE 4015 )
Instructor: Dr. Sayyed Ali Hosseini
Winter 2015
Lecture #4

Some Probability Distribution Functions and Their


Application in Reliability
(Continue From Lecture #3)

MANE4015 Reliability and Maintenance

Lecture #4

Previous Lecture Review (1- Exponential Distribution)


During their useful lifetimes, many components exhibit a
constant hazard rate.
Such a rate implies that the occurrence of failures is purely
random and that there is no deterioration of the strength or
soundness of the components with time.
Constant hazard rate leads to the exponential failure
distribution, most widely used distribution in reliability
analysis.

MANE4015 Reliability and Maintenance

Lecture #4

Previous Lecture Review (1- Exponential Distribution)


Failure rate (hazard rate function):

Failure density function:

Failure distribution function:

=1

Reliability distribution function:

=1

MANE4015 Reliability and Maintenance

Lecture #4

( )

Previous Lecture Review (1- Exponential Distribution)


The parameter is called the failure rate of a given unit time
e.g.

= 5 => implies 5 breakdowns/hr

The Mean Time to Failure (


e.g.

)=

= 1/5 = 0.2 (average 0.2 hour between breakdown)

Expected value = [ ] =

Note: for repairable components, Meantime Between Failures


(
) is used.
= cycle time between failures

MANE4015 Reliability and Maintenance

Lecture #4

Exponential Distribution Model

(a) Failure rate function (hazard rate function) and (b) failure density
function for exponential distribution model
MANE4015 Reliability and Maintenance

Lecture #4

Exponential Distribution Model

(c) Reliability function and (d) failure distribution function for exponential
distribution model
MANE4015 Reliability and Maintenance

Lecture #4

Posteriori Failure Probability


A posteriori failure probability distribution ( ) is defined as
the probability of failure during , given survival up to .
= [ !"#$% &$%"'( (") ' * +,-+' ' *!. .$%)") & $- + ]

MANE4015 Reliability and Maintenance

Lecture #4

Posteriori Failure Probability (Cont.)


/(

Generally, we would like to calculate

Assuming that:
event A = failure during time
event B = survival up to time
0 = survival up to time and failure during ( ,
=

=4

=4

$ &$

(0 )
( )

$ &$

MANE4015 Reliability and Maintenance

85

85

Lecture #4

+ )

$ &$

$ &$

Posteriori Failure Probability (Cont.)


The above formulas are general formulas. If the failure density function is
for example exponentially distributed then the probability of a component
5 and the probability of
surviving up to time
is given by
=
surviving up to time and failing during the interval ( , + ) is :
4

$ &$ = 4

9 &$

= [

9]

=
=

And:
4

$ &$ = 4

9 &$

= [

9]

5
5

As a result:
/

5
85
6
85

$ &$

$ &$

MANE4015 Reliability and Maintenance

=
Lecture #4

=1
10

Posteriori Failure Probability (Cont.)


<= (>) for exponential ?(>) is independent of prior operating
time, in the other words, no matter how long the component
has been in operation, the probability of failure during the
next > is the same memory less property.
Given survival during (0, ), what is the probability of surviving
during ( , + )?
=1

MANE4015 Reliability and Maintenance

( + )
=
( )

Lecture #4

11

Example #1
The lifetime of a certain device is exponentially distributed with
a mean value of 500 hours. What is the probability of this
device operating successfully for at least 600 hours?

MANE4015 Reliability and Maintenance

Lecture #4

12

Example #1 Solution
[ ]=

= 500*%.

1
=
!"#./*%
500
[#"

", 600*%.] = ( ) = (600)

=1

600 =

MANE4015 Reliability and Maintenance

BCC DCC

= 0.3012

Lecture #4

13

Example #2
A batch of 1000 components are believed to have an
exponential failure density function
=
. 50
components fail in the 1st year. Calculate:
(a) The hazard rate
(b) The time for half of the components fail
(c) Number of failures in the 10th year

MANE4015 Reliability and Maintenance

Lecture #4

14

Example #2 Solution
(a) 50 failure out of 1000 components means 5% of the total population
and = 1 year; as a result:
=1

0.05 = 1

1 0.05 =

(b) The time for half of the items to be failed is (


BC

= 0.5 1

C.CB F

= 0.5 1 0.5 =

BC );

= 0.0513

as a result:

C.CB F

= 13.5 G !%.

(c) Number of failures in the 10th year? The amount of failures before year
10 is:
=9 =1

10 = 1

C.CB FK
C.CB F C

0.37% or 370 failed parts so 630 is remaining

= 0.05 +% 5% + * % ,!"'"'( -!% . = 630 0.05 32

MANE4015 Reliability and Maintenance

Lecture #4

15

2- Poisson Distribution
A discrete events occurs in a continuum period of time (e.g. #
of accidents which occur in a given city over given period of
time)
The of occurrence ( ) of the event is assumed constant and the
events are assumed independent of each other (e.g. 5 break
downs/hr or 10 accident/week)
Assume instant replacements or time to replace/repair = 0.

MANE4015 Reliability and Maintenance

Lecture #4

16

2- Poisson Distribution (Cont.)


The Poisson distribution of a discrete random variable
describing the previously-mentioned characteristics is as
follows:
[

[(

) = probability of the event occurring


times in the
interval (0, ) where is Poisson random variable
= Poisson random variable

MANE4015 Reliability and Maintenance

Lecture #4

17

2- Poisson Distribution (Cont.)


If event is a failure of a component, then
components fail in time interval (0, ).

[(

) is probability

The mean value or expected # of failures for Poisson


distribution = \ where the \ can be defined as follows:
\=

=_

[`C

The variance for Poisson distribution is also = \


a!%

=\=

The time between successive failures is a random variable,


what is its probability distribution function?
MANE4015 Reliability and Maintenance

Lecture #4

18

2- Poisson Distribution (Cont.)


Assume event occurred at = 0, what is probability event will
occur again in time interval (0, ). Let random variable = time
between occurrence.

MANE4015 Reliability and Maintenance

Lecture #4

19

2- Poisson Distribution (Cont.)


( ) = ( )

( ", c d ' + $%% '


1

) = 1 ('+ + $%% ' "' (0, ))


=1

is exponentially distributed random variable.


Therefore, if the # of failures occurring in a period of time is
Poisson random variable, the time between failure is
exponentially distributed random variable.

MANE4015 Reliability and Maintenance

Lecture #4

20

Poisson Distribution vs. Exponential Distribution


An interesting feature of these two distributions is that, if the
Poisson provides an appropriate description of the number of
occurrences per interval of time, then the exponential will provide
a description of the length of time between occurrences.
The Poisson describes the number of occurrences of some
event in a fixed period of time.
The exponential describes the time intervals between events
in a Poisson process.

MANE4015 Reliability and Maintenance

Lecture #4

21

Poisson Distribution vs. Exponential Distribution

? f gh ih

Exponential

Poisson

<(>)

MANE4015 Reliability and Maintenance

j(>)

k(>)

_
l`

Lecture #4

22

Example #3
Failure in a production line occurs at an average rate of once
every 3 months.
(a) Compute the probability of having more than 5 failures in a
period of one year.
(b) What is the probability of 2 failures less than one month
apart?

MANE4015 Reliability and Maintenance

Lecture #4

23

Example #3 Solution
(a) Compute the probability of having more than 5 failures in a period of one year.
[,+% *!' 5 !"#$% .] = 1 [# .. *!' +% n$!# + 5 !"#$% .]
Let x = # of failures, 1 failure per 3 months
= 0.333 !"#$% /,+' *

= 1G !% = 12 ,+' *.
> 5 = 1

=
=0 =

5 = 1 ( [ = 0] + [ = 1] + + [ = 5])

0.333 12 C
= 12 =
0!

MANE4015 Reliability and Maintenance

(C.FFF q)

Lecture #4

4C

= 0.01832

24

Example #3 Solution
[
[
[
[
[

= 1] = 0.07326
= 2] = 0. 14682
= 3] = 0.19556
= 4] = 0.19537
= 5] = 0.15629

[ > 5] = 1 [ 5] = 1 ( [ = 0] + [ = 1] + + [ = 5])
> 5 = 1 (0.01832 + 0.07326 + 0. 14682 + 0.19556 + 0.19537 + 0.15629)
[ > 5] = 1 0.785 = 0.215

MANE4015 Reliability and Maintenance

Lecture #4

25

Example #3 Solution
(b) What is the probability of 2 failures less than one month apart?
[2 +'. $ ") !"#$% . 1 ,+' *] = [ < 1]
since

is exponentially distributed random variable

[ < 1] = 1

(C.FFF )

= 1 0.71653 = 0.2837

MANE4015 Reliability and Maintenance

Lecture #4

26

You might also like