International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No.

3
ISSN: 1837-7823

4

RELIABILITY ANALYSIS FOR ROLLING MILLS USING
ALGEBRA OF LOGICS

Dr. Deepankar Sharma
Prof. & Dean-Academics, Divya J yoti College of Engg. & Tech., Modinagar, U.P.
Dr.Anshu Agarwal
Dept. of Maths, Pillai Institute of I.T., Engg., Media Studies & Research, Mumbai
Avneesh Kumar
Research Scholar, Dept. of Maths, Sai Nath University, Ranchi


Abstract
In this paper, authors have considered rolling mills for its reliability assessment. In this present
discussion, two rolls i.e. upper roll and lower roll are depicted as subsystem A and B, and further
categorized as main system (
1 1
,B A ) and its standby redundant (
2 2
,B A ), connected in series
respectively. Redundancy is particularly valuable where it is not possible to do maintenance and of
course one of the method of increasing reliability. There are two standby redundant generators G
1
and
G
2
used to supply power to rolling mills and these can be connected through the change over device
COD. The system has switching devices
12
E and
21
E . To utilize survival subsystem in
i
A and
i
B
) 2 , 1 ( = i a switching device E connecting main and standby system is being used in addition and the
system is called a compound redundant system. Here compound word means to have switching
devices in order to achieve higher systems reliability. Boolean function technique has been used to
formulate mathematical model of the system. For the system presented here, Algebra of logics is being
used to obtain reliability and M.T.T.F. of the considered system. A numerical illustration to detect the
effects of the compound system to the usual standby is examined.

Keywords: Reliability, M.T.T.F., Weibull time distribution, Redundancy, Boolean functions, Algebra
of logics etc.

1. Introduction
In modern industries, systems are designed to be operative for a specified period (called the
mission time), i.e. there should be no failure in any equipment or part of equipment under specified
operating conditions during the total period (operative time, administrative time and repair time).
Behavior analysis of each item of equipment under given operating conditions is helpful to design the
component for minimum failure and to prepare a plan in advance to scheduled preventive maintenance.
Although the basic concepts of many forming processes have remained largely unchanged throughout
history, the details and equipment have evolved considerably. Manual processes were converted to
machine processes during the industrial revolution. The machinery then became bigger, faster and more
powerful. Water wheel power was replaced by steam and then electricity. More recently, computer
controlled, automated operations have immerged. Processes that are normally performed hot, is
presented in this paper.
An obvious reason for the popularity of the hot-working processes is that they often provide an
attractive means of producing a desired shape. At elevated temperature, metals weaken and become
more ductile. With continual re-crystallization, massive deformation can take place without exhausting
material plasticity. In steels, hot forming involves the deformation of weaker, austenite structure, which
then cools to the stronger, room-temperature, ferrite or much stronger non-equilibrium structures.
Among the hot-working processes viz, rolling, forging, extrusion, hot drawing, pipe welding and
piercing, rolling is of major importance in modern manufacturing.
Rolling is usually the first process that is used to convert material into a finished wrought product.
Thick starting stock can be rolled into blooms, billets or slabs or these shapes can be obtained directly
from continuous casting. A bloom has square or rectangular cross section, with a thickness greater than
6 inches and a width no greater than twice the thickness. A billet is usually smaller than a bloom and
has a square or circular cross section. Same form of deformation process, such as rolling or extrusion,
usually produces billets. A slab is a rectangular solid where the width is greater than twice the
thickness. Slabs can be further rolled to produce plate, sheet and strip. These hot-worked products often
form the starting material for subsequent processing using techniques such as cold forming or
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

5

machinery. Sheet and strip can be fabricated into products or further cold rolled into thinner, stronger
material or even into foil. Blooms are billets and can be further rolled into finished product such as
structural shapes or railroad into finished product, such as structural shapes or railroad rail or they can
be processed into semi-finished shapes, such as bar, rod, tube or pipe.
In hot rolling, heated metal is passed between two rolls (taken as subsystem A and B) that rotate in
opposite directions. The gap between the rolls being somewhat less than the thickness of entering
metal. Because the rolls rotate with a surface velocity that exceeds the speed of the incoming metal,
friction along the contact interface acts to propel the metal forward. The metal is squeezed and
elongates to compensate for the decreased in cross-sectional area. The amount of deformation that can
be achieve in a single pass between a given pair of rolls depends on the friction conditions along the
interface, In too much is demanded, the rolls can not advance the material and simply skid over its
surface. Too little deformation per pass, however, results in excessive production cost. Fig-1 shows the
logical block diagram of considered system.

2. Assumptions
1. The system consists of two main units
1 1
,B A and two standby units
2 2
,B A connected in
series, respectively.
2. At 0 = t ,
1
A and
1
B start operating while
2
A and
2
B are kept as standby.
3. None of the standby units degrade in unused condition.
4. The system ceases to function when both the units of any subsystem are non-operative.
5. Units are non-identical and statistically independent.
6. The system has two modes of states, viz; normal and failed.
7. Switching devices used are imperfect.
8. All failures are exponentially distributed.

3. List of Notations Used
3 1
, x x : States of generator
2
x : State of change over device.
5 4
, x x : States of main rolls A
1
and B
1
.
8 7
, x x : States of standby rolls A
2
and B
2
.
10 9 6
, , x x x : States of switching devices E, E
12
, E
21
, respectively.
' xi
: Negation of 10 , 2 , 1 , = i
xi
.
/ : Conjunction / Disjunction
Ri
: Reliability of i
th
state, 10 , 2 , 1 = i .
xi
:

= . 10 , 2 , 1 , 0
, 1
i state bad
state good

i
: Failure rate of i
th
component of the system.

4. Formulation of Mathematical Model
Using Boolean function technique, the conditions of capability of the successful operation of complex
system in terms of logical matrix are expressed as:
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

6

(
(
(
(
(
(
(
(
(
(
(

=
8 7 6 3 2
10 7 5 3 2
9 8 4 3 2
5 4 3 2
8 7 6 2 1
10 7 5 2 1
9 8 4 2 1
5 4 2 1
10 2 1








) , , (
x x x x x
x x x x x
x x x x x
x x x x
x x x x x
x x x x x
x x x x x
x x x x
x x x F (1)

5. Solution of the Model
By application of algebra of logical, equation (1) may be written as
| | f x x x x F ) , , (
2 10 2 1
= (2)
where,
(
(
(
(
(
(
(
(
(
(
(

=
8 7 6 3
10 7 5 3
9 8 4 3
5 4 3
8 7 6 1
10 7 5 1
9 8 4 1
5 4 1








x x x x
x x x x
x x x x
x x x
x x x x
x x x x
x x x x
x x x
f
(3)
Now substituting the following in equation (3):
| |
5 4 1 1
x x x A = (4)
| |
9 8 4 1 2
x x x x A = (5)
| |
10 7 5 1 3
x x x x A = (6)
| |
8 7 6 1 4
x x x x A = (7)
| |
5 4 3 5
x x x A = (8)
| |
9 8 4 3 6
x x x x A = (9)
| |
10 7 5 3 7
x x x x A = (10)
| |
8 7 6 3 8
x x x x A = (11)
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

7

Input x
6


x
4
x
7

x
1 COD


x
10
x
9

x
2


x
3
x
5
x
8



Output

Fig-1: Logical Block Diagram of the System

G
1

G
2

B
1

A
1
A
2

B
2

E
21

E
12

E
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

8

We obtain

(
(
(
(
(
(
(
(
(
(
(

=
8
7
6
5
4
3
2
1
A
A
A
A
A
A
A
A
f (12)
Using orthogonalisation algorithm, equation (12) may be written as:
(
(
(
(
(
(
(
(
(
(
(

=
8 7 6 5 4 3 2 1
7 6 5 4 3 2 1
6 5 4 3 2 1
5 4 3 2 1
4 3 2 1
3 2 1
2 1
1







A A A A A A A A
A A A A A A A
A A A A A A
A A A A A
A A A A
A A A
A A
A
f (13)
By algebra of logics, we compute the following
(
(
(

=
5 4 1
4 1
1
1


x x x
x x
x
A
| |
9 8 4 1
5 4 1
4 1
1
2 1


x x x x
x x x
x x
x
A A
(
(
(

=
| |
9 8 5 4 1
x x x x x = (14)
Similarly, we obtain
| |
10 7 5 4 1 3 2 1
x x x x x A A A = (15)
| |
8 7 6 5 4 1 4 3 2 1
x x x x x x A A A A = (16)
| |
5 4 3 1 5 4 3 2 1
x x x x A A A A A = (17)
| |
9 8 5 4 3 1 6 5 4 3 2 1
x x x x x x A A A A A A = (18)
| |
10 7 5 4 3 1 7 6 5 4 3 2 1
x x x x x x A A A A A A A = (19)
| |
8 7 6 5 4 3 1 8 7 6 5 4 3 2 1
x x x x x x x A A A A A A A A = (20)

Using equation (14) through (20), equation (13) becomes
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

9

(
(
(
(
(
(
(
(
(
(
(

=
8 7 6 5 4 3 1
10 7 5 4 3 1
9 8 5 4 3 1
5 4 3 1
8 7 6 5 4 1
10 7 5 4 1
9 8 5 4 1
5 4 1








x x x x x x x
x x x x x x
x x x x x x
x x x x
x x x x x x
x x x x x
x x x x x
x x x
f (21)

Putting the value of f from equation (21) in equation (2), we obtain:
(
(
(
(
(
(
(
(
(
(
(

=
8 7 6 5 4 3 2 1
10 7 5 4 3 2 1
9 8 5 4 3 2 1
5 4 3 2 1
8 7 6 5 4 2 1
10 7 5 4 2 1
9 8 5 4 2 1
5 4 2 1
10 2 1








) , (
x x x x x x x x
x x x x x x x
x x x x x x x
x x x x x
x x x x x x x
x x x x x x
x x x x x x
x x x x
x x x F (22)
Since, R.H.S. of equation (22) is disjunction of pair-wise disjoint conjunctions, therefore the reliability
of considered system as a whole is given by:
{ } 1 ) , , ( Pr
10 2 1
= = x x x F R
S

|
8 7 6 1 5 4 10 7 5 1 4 9 8 4 1 5 5 4 1 2
R R R R S S R R R R S R R R R S R R R R + + + =
|
8 7 6 3 5 4 1 10 7 5 3 4 1 9 8 4 3 5 1 5 4 3 1
R R R R S S S R R R R S S R R R R S S R R R S + + + +
where,
i
R is the reliability corresponding to system state
i
x while 10 , 2 , 1 , 1 = = i R S
i i
.
Thus, we have
|
9 8 4 3 8 7 6 1 10 7 5 1 9 8 4 1 5 4 3 5 4 1 2
R R R R R R R R R R R R R R R R R R R R R R R R
S
+ + + + + =

9 8 5 4 3 1 8 7 6 5 4 1 8 7 6 3 10 7 5 3
R R R R R R R R R R R R R R R R R R R R + + + +

8 7 6 5 3 1 8 7 6 4 3 1 10 7 5 4 3 1
R R R R R R R R R R R R R R R R R R + + +

10 7 5 4 1 9 8 5 4 1 5 4 3 1 8 7 6 5 4 3
R R R R R R R R R R R R R R R R R R R R +

10 7 5 3 1 9 8 5 4 3 9 8 4 3 1 8 7 6 5 1 8 7 6 4 1
R R R R R R R R R R R R R R R R R R R R R R R R R

8 7 6 5 3 8 7 6 4 3 8 7 6 3 1 10 7 5 4 3
R R R R R R R R R R R R R R R R R R R R
|
8 7 6 5 4 3 1
R R R R R R R (23)

6. Some Particular Cases
CASE I: When reliability of each component is R:
In this case, setting R i R
i
= = ) 10 2 , 1 ( in equation (23), we obtain
8 7 6 5 4
6 11 5 2 R R R R R R
S
+ + = (24)


CASE II: When failure rates follow Weibull time distribution:
Let
i
be the failure rate corresponding to system state
i
x , then in this case, reliability of that
component is given by
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

10

-10 - - - 1,2 all for , ) ( = =

i e t R
t
i
i

where is some positive parameter.
Using above results in equation (23), we obtain the reliability of considered system at time instant t as:
{ } { }

t b t a t R
j
j
i
i
SW
=

= =
exp exp ) (
13
1
14
1
(25)
where,
2 5 4 1 1
+ + + = a
5 4 3 2 2
+ + + = a
9 8 4 2 1 3
+ + + + = a
10 7 5 2 1 4
+ + + + = a
8 7 6 2 1 5
+ + + + = a
9 8 4 3 2 6
+ + + + = a
10 7 5 3 2 7
+ + + + = a
8 7 6 3 2 8
+ + + + = a
8 7 6 5 4 2 1 9
+ + + + + + = a
9 8 5 4 3 2 1 10
+ + + + + + = a
10 7 5 4 3 2 1 11
+ + + + + + = a
8 7 6 4 3 2 1 12
+ + + + + + = a
8 7 6 5 3 2 1 13
+ + + + + + = a
8 7 6 5 4 3 2 14
+ + + + + + = a
5 4 3 2 1 1
+ + + + = b
9 8 5 4 2 1 2
+ + + + + = b
10 7 5 4 2 1 3
+ + + + + = b
8 7 6 4 2 1 4
+ + + + + = b
8 7 6 5 2 1 5
+ + + + + = b
9 8 4 3 2 1 6
+ + + + + = b
9 8 5 4 3 2 7
+ + + + + = b
10 7 5 3 2 1 8
+ + + + + = b
10 7 5 4 3 2 9
+ + + + + = b
8 7 6 3 2 1 10
+ + + + + = b
8 7 6 4 3 2 11
+ + + + + = b
8 7 6 5 3 2 12
+ + + + + = b
8 7 6 5 4 3 2 1 13
+ + + + + + + = b



Case III: When failure rates follow exponential time distribution:
Exponential distribution is a particular case of Weibull distribution for = 1 and is much useful in
numerous practical problems. The reliability of the system in this case at any instant t, can be
obtained by putting = 1 in equation (25), and is

International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

11

{ } { }

= =
=
13
1
14
1
. exp . exp ) (
j
j
i
i SE
t b t a t R (26)
Also, an important reliability parameter viz; M.T.T.F., in this case, is given by
dt t R F T T M
SE
) ( . . .
0

=


= =
|
|
.
|

\
|

|
|
.
|

\
|
=
13
1
14
1
1 1
j j i i
b a
(27)
where, s b s a
j i
and have been mentioned earlier.

7. Numerical Illustration
Setting
(i)
i
=0.3 ( ) 9 , 2 , 1 = i and =2 in equation (25);
(ii)
i
=0.3 ( ) 9 , 2 , 1 = i in equation (26);
and (iii)
i
( ) 9 , 2 , 1 = i = = 0, 0.1, 0.2 ..1.0, in equation (27);
One can draw the graphs shown in fig-2 and 3, respectively.


Fig-2: Reliability Vs Time Fig-3: Mean time to failure Vs

8. Conclusion and Future Research
In this paper, the author has analyzed rolling mills system for its reliability assessment with the aid of
Boolean function technique and algebra of logics. Reliability of the system as a whole has obtained in
three different cases, M.T.T.F. of the system has also obtained. Graphical illustration followed a
numerical example has appended in the last to highlight important results of the study.
A critical examination of fig - 2 reveals that reliability of the system decreases rapidly in case, failures
follow Weibull time distribution but it decreases in a smooth and better way when failures follow
exponential time distribution.
Study of fig-3 yields that the M.T.T.F. of the system decreases catastrophically in the beginning but
thereafter it decreases approximately in a constant manner.

9. References
[1]. Adachi, K. and M. Kodama (1980); Availability analysis of two unit warm standby system with
inspection time, Microelectron Reliab., 20, 449-455.
[2]. Agnihotri, R.K. and S.K. Satsangi (1996); Two unit identical system with priority based on repair
and inspection, Microelectron Reliab., 36, 279-282.
International J ournal of Computational Intelligence and Information Security, J une 2014 Vol. 5, No. 3
ISSN: 1837-7823

12

[3]. Dhillon, B.S. and N-Yang (1992); Reliability and availability analysis of warm standby with
common cause failure and human error, Microelectron Reliab., 32, 561-576.
[4]. Goel, L.R., P. Srivastava and R. Gupta (1992); Two unit cold standby system with correlated
failures and repairs, Int. Jr. of System Science, 23(3), 379-391.
[5]. Goel, L.R. and P. Srivastava (1991); Profit analysis of a two unit redundant system with provision
for rest and correlated failures and repairs, Microelectron Reliab., 31(5), 827-833.
[6]. Gopalan, M.N., R. RadhaKrishna and A. Vijay Kumar (1984); Cost benefit analysis of a two unit
cold standby system subject to slow switch, Microelectron Reliab., 24, 1019-1021.
[7]. Kumar, A. and R. Lal (1979); Stochastic behaviour of a two unit standby system with constant
failure and intermittently repair facility Int. Jr. of System Science, 10(6), 589-603.
[8]. Murari, K. and Vibha Goel (1984) Comparison of two-unit cold standby reliability models with
three types of repair facilities Microelectron Reliab., 24(1), 35-39.
[9]. Osaki, S. (1972); Reliability of a two-unit standby redundant system with preventive maintenance,
IEEE Trans. Reliab., R-21, 24-29.
[10]. Singh, S.K. and A.K. Mishra (1994); Profit evaluating of a two unit cold standby redundant
system with two operating systems; Microelectron Reliab., 34(4), 747-750.