Chapter 7 - 02

Reliability and Maintainability

2015/16

1
Failure Rate & Mean Time Between Failures

Failure Rate
 Assume a batch of N items out of which a number k have failed

at time t. The total cumulative time T can be evaluated in one

of the following ways.

i. If it is assumed that each failure is replaced as it occurs, the

cumulative time is
T  Nt
ii. If items are not replaced as they fail, for non-replacement
condition the cumulative time is given by

T  t1  t2   tk   N  k  t 2
Definition:
 For a stated period in the life of an item, the ratio of the total

number of failures to the total cumulative observed time is

defined as the observed failure rate.
total number of failures
observed failure rate 
total cumulative observed time
k
  , ˆ is the failure rate of N items observed
ˆ
T
Units of λ:
- percentage failures per 1000 h,
- failures /h
- failures /106 h 3
Observed Mean Time Between Failures (MTBF)
Definition
For a stated period in the life of an item, the mean value

of the length of time between consecutive failures,

computed as the ratio of the total cumulative observed
time to the total number of failures is defined as the
MTBF.

ˆ T ˆ 1
MTBF    , or ,  
k ˆ
4
Observed Mean Time to Fail (MTTF)
Definition
For a stated period in the life of an item, the ratio of

cumulative time to the total number of failures is defined

as MTTF. T
MTTF 
k
The difference between MTBF and MTTF
 MTTF is applied to items that are not repairable (e.g. Bearings)
 MTBF is applied to items that are repairable.
 MTBF excludes downtime. Therefore it is the mean up-time
between failures.

5
Mean Life
Definition
 Mean life is defined as the mean of the times to failure

where each item is allowed to fail.

 MTBF and MTTF can be calculated over any period of time.

Mean life must include the failure of every item.

6
7
8
9
Reliability Function
Consider the probability of an item failing in the
interval between t and t+dt.
Given the failure rate λ(t), the probability that the item
may fail in the interval t to t+dt, provided it has
survived until time t, is given by the conditional
probability
PE2    t  dt
E1

 where E1 is survival up to time t with the survival probability

given by the reliability

R t   PE1
 And, E2 is item failing between time t and t+dt.
10
The probability of failure in the interval t to t+dt
unconditionally is f(t)dt
 where f(t) is the failure probability density function.
This probability is obtained by the multiplication
theorem which states that

f  t  dt  PE1 and E2
Noting that

PE1 and E2  PE1  PE1

E2

f  t  dt  PE1 and E2  R t     t  dt
11
Thus, the failure rate is obtained to be

f t
t  
R t 
The probability that an item may fail between running
times 0 and t is

F  t    f  t dt  1  R t 
t

Differentiating this equation,

 dR t 
f t 
dt
12
Substituting for f(t) in the equation for λ (t),

1 dR t 
t    
R t  dt
Integrating both sides gives

R t  dR t 
    t dt  
t

0 1 R t 
Which gives the reliability function

R t   exp    t  dt 
t

 0 
13
Assuming constant failure rate ,

R t   exp  t   e  t

14
MTBF
Consider the (N-k) items that survived at t. Let (N-k) be
Ns(t). Then
Ns t 
R t  
N
In each time interval dt, the time accumulated is

N s  t  dt
At time t  the total time accumulated is


 N s  t  dt
0
15
MTBF is then given by

1   N s  t  dt
MTBF   N s  t  dt  
N 0 0 N

  R  t  dt
0

For constant λ,

  1
MTBF   R t dt    t
e dt 
0 0 

16
1.

17
2..

18
 System Reliability with Weibull Failure Probability
Distribution
 In Weibull probability distribution of failure, the failure rate

varies with time. The failure rate and reliability function are
given by, respectively.
Where :

  t     t  t0 
 1  β = shape parameter (slope)

   = Scale parameter (characteristic

life)
 t t 


R t   exp  0
 
    
For   1 failure rate increases with time resulting in
decreasing reliability.
19
1.

20
21
22
Maintainability

 Maintainability is a characteristic of design and installation which is expressed as

the probability that an item will conform to specified conditions within a given
period of time, when maintenance action is performed in accordance with
prescribed procedures and resources.

 The objective of a maintainability program should be to influence equipment

design to assure that maintenance of the equipment can be accomplished

efficiently and safely.

 Maintainability requirements are usually contractual, and in such cases it is

essential that the test method and the conditions under which it is to be carried
out
23 are carefully defined. Achievement of specified Mean Time to Repair is rather
 Maintainability design features include:

 access to replaceable assemblies

 clear markings for convenience and safety

 maximum use of interchangeability

 elimination of unnecessary adjustments

 clearly written maintenance instructions

 adequate use of test points for fault isolation

 These allow rapid restoration of service by minimizing

downtime.
24
 Maintainability Function (Exponentially disritibuted)
 The time-to-restore probability density function g(t) for the

exponential times-to-restore distribution is

g  t   e  t

 μ is equipment repair, replacement or restoration rate.

 Given μ, the mean time to correctively repair, replace or restore

the equipment to satisfactory operation is given by

1
MTTR 

t
1 
gt   e MTTR
MTTR
25
 The maintainability function for the exponential time to restore

distribution is
M  t1   P t  t1    g  t dt
t1

0
t1
  e dt  t
0
 t1
 1 e

 M  t1  is the probability that a repair will be completed

successfully in time t1.

26
 If the time t1=MTTR, an equipment with an exponential times-to-

restore function has probability of being restored in the given time

M  t1  MTTR   1  e  0.6321

 Maintainability probability of 90% is obtained for

t1  2.3026 MTTR
 The 50% probability of accomplishing any repair is obtained in
time

t  0.69315 MTTR
1


t1 is called the median time to restore
27
 Determination of Mean Time To Repair MTTR

 In many practical applications, determination of MTTR is not

easy.
 MTTR is the mean of the distribution of equipment repair time
n
and can be estimated from T i i
MTTR  i 1
n


i 1
i

Where :
 T is the time needed to repair the equipment when the ith part
i
fails
 λi is the constant failure rate of the ith repairable part of the
equipment
 n is the number of repairable parts in the equipment 28
29
 A-priori and a-posteriori maintainability determination

a) The a-priori maintainability function M(ti) is predicted

from

M  t1   P t  t1    g  t dt
t1

 t1 is the restoration time for which M(ti) is to be predicted, and

 g(t) has to be determined ahead of time.

30
b) The a-posteriori maintainability M(ti) is determined
from actual data collected and is given by

N  t  t1 
M  t1  
NT

Where:
 N  t  t1  number of maintenance actions which were successfully
completed in time t1 or less

 NT is the total number of maintenance actions undertaken out of

which N  t  t1  were completed successfully in time t1 or less.
31
Availability:
a) Steady State Availability (inherent availability)
Availability is the available up-time of an
equipment. This is the probability that an
equipment, when used under stated conditions and
ideal support environments, will operate
satisfactorily at any given time.

Ass 

where Ass = steady state availability
= system constant repair rate
= system constant failure rate
32
Substituting for repair rate and failure rate, steady state
availability is
MTBF
Ass 
MTBF  MTTR

 In the calculations of Ass , preventive maintenance down

time, supply down time, queuing downtime,
administrative down time are excluded.
Ass is useful to designers.

33
b) Operational Availability:

This can be defined as the probability that an equipment,

when used under stated conditions in an actual
environment, will operate satisfactorily at any time.
MTBM
Ao 
MTBM  MDT
where
 MTBM = Mean Time Between Maintenance actions

 MDT = Mean Down Time; sum of the mean corrective and

preventive maintenance time intervals including supply down
time, administrative down time, etc.
34
 The End!!

35