Kalpa Publications in Engineering

Volume 1, 2017, Pages 239–245

ICRISET2017. International Conference on Re-
search and Innovations in Science, Engineering
&Technology. Selected Papers in Engineering

Optimum Selection of Mechanical Drives Using

TOPSIS
Ketan Tamboli1, P M George 2
1
Mechatronics Engg. Deptt.,G H Patel College of Engg. & Tech.,V V Nagar,Gujarat,India,388120
2
Mechanical Engg. Deptt.,B V M Engg College,V V Nagar,Gujarat,India,388120

Abstract— New product development demands the analysis of each of the component from various
engineering concepts.If a machine is expected to have a mechanical power transmission system, obviously
number of options can be thought of.The power transmission system involves many hardware and many
options are available.These mechanical drives are expected to be light,energy efficient and maintenance
free. In the present work,a data matrix comprising of 10 attributes and 10 alternatives is presented .It is
then operated for optimum selection in form of ramking using one of the welknown Multi Attribute
Decision Making( MADM) method called Technique for Order Preference by Similarity to Ideal Solution
(TOPSIS).The results are given as ranking for the purpose of selectionwhich sets initiative for preliminary
design.However the approach presented is also useful for existing designs which needs to be changed to
enhance performance.

Keywords— Mechanical drives, TOPSIS,MADM

I. INTRODUCTION

The present scenario in industrial environment is highly concerned the conservation of energy in
one form or other. With the increasing demand for energy due to rapid industrialization, it has
become necessary to relook into energy consumptions for any industrial unit. The generation and
transmission of power is highly costly and hence the end product becomes uneconomical. The
power transmission especially for the mechanical drives need to be attended consciously for the
purpose of its effectiveness. In the present times, due to developments in computer hardware and
software technology, it is possible to calculate and decide the best suitable choice for the
mechanical drives. Number of researchers have used MADM methods such as SAW(Simple
Additive Weight )method[1],AHP(Analytical Heirarchy Process)method[2,3],Modified TOPSIS
method[4] and VIKOR(VIsekriterijumsko KOmpromisno Rangirange ) [3]. In the present work,
TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method is used for
optimally selecting the best drive.Many researchers
[5,6,7,8,9,10,11]have used MADM methods for various applications such as for optimal selection
of materials etc.

A. Shukla, J.M. Patel, P.D. Solanki, K.B. Judal, R.K. Shukla, R.A. Thakkar, N.P. Gajjar, N.J. Kothari, S.
Saha, S.K. Joshi, S.R. Joshi, P. Darji, S. Dambhare, B.R. Parekh, P.M. George, A.M. Trivedi, T.D. Pawar,
M.B. Shah, V.J. Patel, M.S. Holia, R.P. Mehta, J.M. Rathod, B.C. Goradiya and D.K. Patel (eds.),
ICRISET2017 (Kalpa Publications in Engineering, vol. 1), pp. 239–245
Optimum Selection of Mechanical Drives Using TOPSIS Dr Ketan Tamboli and Dr. P.M. George

II. JUSTIFICATION FOR SELECTION OF DRIVE

In mechanical drives, number of factors (attributes) are associated with each drive such as center
distance, power transmitting capacity, speed ratio, working conditions (temperature range),
corrosive atmosphere, etc.), maintenance (lubrication, vibration, noise, etc.). the conflicting
requirements of different attributes when number of choices (options/alternatives) are available
emphasizes the need of a scientific approach to select the optimum choice.The attributes
considered are important from performance ,life and maintenance point of view.As an example
the minimum centre distance is offerd by gear drive with more power transmitting capacity but
has more cost .The contradictory matters related to all attributes compels the use of scientific
approach for optimum selection.

III. DEVELOPMENT OF DATA MATRIX

A large number of sources [12,13,14,15,16,17,18] are referred to compile the information for
various mechanical drives and their attributes, which is represented in Table 1 of Appendix as a
10×10 matrix, known as data matrix. It is assumed that all the alternatives are possible to be used
in a perticular application.The fuzzy attributes are categorised on 1-5 scale and its consideration
is taken in the method.The beneficial and nonbeneficial attributes are to be desired for higher and
lower values respectively.

IV. MADM METHODS AND TOPSIS

Following description summarizes about TOPSIS, a wellknown MADM method used for the
present work. The mathematical procedural steps are detailed here.The relative weights as
required in TOPSIS are decided based on the importance of a particular over other attributes
which will give adequate performance of the system with minimum maintenance and repair.
However, depending upon the requirement of particular application, these weights could vary and
group of experienced designers can decide upon them. Also, in the literature [1] the five point and
eleven point scale is suggested which can be used as subjective measure of each attribute. In the
present case, the relative weight vector used for all the methods is kept same i.e. is 0.15, 0.20,
0.20, 0.10, 0.05, 0.10, 0.05, 0.05, 0.05, 0.05. The computer codes in MATLAB are developed
for all of these methods. The codes are verified with the existing results of other application in
the literature and then applied for the present case.
A. TOPSIS Algorithm
The TOPSIS method was developed by Hwang and Yoon [19].The steps are shown below.

Step 1. Formulation of the Data Matrix

The data matrix comprises of alternatives and attributes for each of them. Thus an element a[i,j] of the
data matrix A gives the value of the jth attribute in a non-normalized form/units, for the ith alternative.
Therefore for m alternatives with n pertinent attributes, the data matrix is an m×n matrix.

Step 2. Relative Importance Matrix Formulation

In case an expert system is being used, this may also be supplied by a panel of experts. This
information is stored in an n×n matrix, say R, with all its diagonal elements as unity.

Step 3. Formulation of Normalized Data Matrix

Since all the attributes are with different units it is not possible to compare them.Hence a
normalization is done as per below.

240
Optimum Selection of Mechanical Drives Using TOPSIS Dr Ketan Tamboli and Dr. P.M. George

( )
( ) ( )
∑ ( ( ))

Step 4. Eigenvalue Formulation

In step 2, the information is stored in a pair-wise basis. Thus if

(2)

then ideally
( )

To overcome the human inconsistency, a drawback, the eigenvalue method is employed. This method
seeks to find a weight matrix vector, W, where:
(4)
RW=λW
and
W=[w_1 w_2 w_3……]T. (5)
From above equation we have:
(R-λI)W=0 (6)
To avoid the trivial solution,
|R-λI|=0 (7)
The solution of equation (4)gives the complete eigen spectrum which is the invariant of ‘Attribute
Relative Importance Matrix System’. The solution of the system equation (4) for the largest
eigenvalue, 𝜆 gives the weight vector W.

Step 5. Formation of the Weighted Normalized Data Matrix(V)

This matrix gives a true comparable value of each attribute. Thus:

v(i,j)=w_j*n(i,j) (8)
The resulting matrix is an m×n matrix.

Step 6. Generating the hypothetical best solution (HBS) and the hypothetical worst solution (HWS)
Figure 1 shows graphically the concept of HBS and HWS for two sample attributes. The hypothetical
worst solution is also an imaginary solution, but with least optimal favorable values.In present case
since 10 attribute s are to be compared with 10 alternatives(options) it is inevitable to use computer as
calculations will be many and will be out of reach of manual calculations.Also graphically the
solution is not possible to achieve.
Thus:
(9)

[ ( ) ( ) ( )] ( )
[ ]

241
Optimum Selection of Mechanical Drives Using TOPSIS Dr Ketan Tamboli and Dr. P.M. George

[ ( ) ( ) ( )] ( )
[ ]
(12)
In some cases, minimum of certain attributes, for example cost, weight etc. may give HBS and vice
versa for HWS.

Fig.1 Graphical representation of HWS and HBS Fig.2 Radar Graph for best 2 alternatives

Step 7. Determining the Euclidean Distances

The distances are calculated as per below:
⁄
{∑ [ ( ) ] } (13)
⁄
{∑ [ ( ) ] } (14)

Step 8. Determination of the Metric Index

The metric index is the relative closeness to the ideal solution, and is given by:

( )
It is a design evaluation index and simultaneously considers distances from the worst and the best
solutions. Thus, resting on a much larger platform for comparison, the merit index is a true measure
for attribute alternative design.

Step 9. Establishing a Preference Order

Referring to the Figure 1, it is apparent that if and On the other hand,
if and This implies that the solution with the highest value be

242
Optimum Selection of Mechanical Drives Using TOPSIS Dr Ketan Tamboli and Dr. P.M. George

given the highest rank, and so on. Thus, arranging in descending order of values, a preference
order is obtained.

Step 10. Selection of the Optimum Design

From the preference order, the design solution with the highest value is the optimal alternative
design for the given application.
This procedure will be of immense help to the manufacturer in resolving design conflicts and
in making tradeoffs. Being scientific and objective, the algorithm has in-built application
effectiveness and is thus helpful for making customized products resulting in enhanced customer
satisfaction.

V. RESULTS AND DISCUSSION

The methodology is detailed in section IV. Tables 1,2,3 and 4 are intermediate calculated stepswhich
are shown in Appendix. The highlighted values in the Table 4 are considered for radar graph as shown
in Figure 2. The respective relative closeness of alternatives 1 to 10 to ideal solution is 0.3389,
0.3333, 0.4665, 0.4566, 0.4514, 0.3411, 0.3374, 0.3718, 0.4689, 0.4449. The relative closeness sorted
is 0.4689, 0.4665, 0.4566, 0.4514, 0.4449, 0.3718, 0.3411, 0.3389, 0.3374, 0.3333. The order of
preference of different alternatives is 9-3-4-5-10-8-6-1-7-2. The best alternative is 9. Figure 2 shows
the comparative variations of different attributes for best two alternatives.

VI. CONCLUSION AND WAY FORWARD

The application of TOPSIS method for the selection of optimum mechanical drives is presented.
Upon application of the TOPSIS, the ranking of alternatives is obtained. The friction drive is ranked
as first followed by flat belt drive,v-belt drive etc.Since the drive selection case is not attempted by
any of the earlier researcher the validation of the present work is not possible here.However the
approach presented would aid to existing method of selction as the TOPSIS is a scientific approach
for optimum selection. Though the results are satisfactory, the designer can still have freedom to
analyze the real life application and may increase the size of the data matrix. Additionally, many more
MADM methods can also be tried.

ACKNOWLEDGEMENT
The encouragement and support provided by Sophisticated Instrumentation Center for Applied
Research and Testing (SICART), an institute sponsored by Department of Science and Technology,
Govt.of India, New Delhi jointly by Charutar Vidya Mandal, Vallabh Vidyanagar, Gujarat is highly
appreciated and acknowledged.

REFERENCES
[1] R.V.Rao, Decision making in the manufacturing environment, Springer, 2007.
[2] Saaty T.L., The analytic hierarchy process, McGrow Hill, New York, 1980.
[3] Saaty T.L., Fundamentals of decision making and priority theory with the AHP, RWS publications, Pittsburg, 2000.
[4] Deng H., Yeh C.H., Willis R.J., Inter-company Comparison using Modified TOPSIS with objective weights,
Computers and Operations
Research, 27, 963-973, 2002.
[5] Fishburn P C, Additive utilities with incomplete product set; applications to priorities and assignments, Operations
research society of America, Baltimore, 1967.
[6] Ananthasuresh G.K., Ashby N.F., Concurrent Design in material selection for trusses, Notes of Workshop on optimal
design, France, 26-28 Nov., 2003.

243
Optimum Selection of Mechanical Drives Using TOPSIS Dr Ketan Tamboli and Dr. P.M. George

[7] Abdalla S., Kizil M.S., Canbulat I., Development of a method for layout selection using analytical hierarchy process, 13th
coal operators conference, University of Wollongong, Australia, pp.27-37, 14-15 Feb.2013
[8] Thirumalai R., Senthilkumar J.S., Intelligent selection of optimum machining parameters in turning of Inconel 718,
International journal of advanced engineering technology, Vol.2, Issue 4, pp.167-173, Oct-Dec.2011.
[9] Zoran D., Sasa M., Dragi P., Application of the AHP method for selection of a transportation system in mine planning,
Professional paper, Belgrade, Underground mining engineering, 19, 93-99, 2011.
[10] Dagdeviren M., Yavuz S., Kilinc N., Weapon selection using the AHP and TOPSIS methods under fuzzy environment,
Journal of expert systems with applications, Elsevier, 36, pp.8143-8151, 2009.
[11] Chauhan A., Vaish R., Magnetic material selection using multiple attribute decision making approach, Journal of
materials and design, Elsevier, 36, pp.1-5, 2012.
[12] S.C.Sharma, Agrawal, Machine Design. Kataria Publications, New Delhi 2010.
[13] R.Karwa, Machine Design. Laxmi Publications, New delhi, 2011
[14] T.G.Hicks, Machine Design Calculations Reference Guide, McGrawHill Publications,1987.
[15] B.J.Hamrock, Bo Jacobson, S.R.Schmid, Fundamentals of Machine Elements, McGraw Hill Publications, 1999.
[16] J.E.Shigley, Co Mischke, R.G.Budinas, K.J.Nisbett, Mechanical Engineering Design, Tata McGraw Hill Publications,
New Delhi, 1996.
[17] G.M.Maitra, L.V.Prasad, Handbook of Mechanical Design, Tata McGraw Hill Publications, New Delhi, 1997.
[18] www.reidsupply.com
[19] Yoon K P, Hwang C L, Multi-attribute decision making, SAGE publication, Beverley Hills, CA.

APPENDIX
Table 1 Data matrix for mechanical drives (10 drives, 10 attributes)

Drives Lead Variable Gear

Ball Flat V Roller Silent Rope Friction
screw speed drive
screw belt belt chain chain drive drive
drive
Attributes
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Centre Distance (mm) (CD)
1000 1000 5000 500 2000 600 600 10000 300 800
(NB)
Power Trans. Capacity (kw)
300 150 125 190 100 250 250 750 200 500
(PTC) (B)

Speed Ratio (SR) (B) 0.3 2.5 80 150 150 25 25 25 15 600

Efficiency (%) (EFF) (B) 45 98 96 95 96 99 99 96 95 80

Lower Opg. Temp. (0c)

20 20 -40 -55 -40 -20 -20 55 20 60
(LOT) (NB)

Power/Weight Ratio
0.6 0.65 0.8 0.7 0.75 0.95 0.9 1.2 0.80 0.6
(PWR) (B)

Maint.(poor-excellent:1-5)
3 4 5 5 3 3 3 4 4 3
(MAI) (B)

Vibrations.(high-
2 2 4 4 4 3 3 5 5 2
negligible:1-5) (VIB) (B)

Slip.(high-negligible:1-5)
5 5 2 3 3 5 5 4 1 5
(SLP) (B)

Working condition(worst-
3 2 4 4 4 5 5 4 2 2
good:5-1) (WC) (NB)

Table 2: Normalized Matrix

244
Optimum Selection of Mechanical Drives Using TOPSIS Dr Ketan Tamboli and Dr. P.M. George

Alternative CD PTC SR EFF LOT PWR MAI VIB SLP WC

1 0.10 0.40 0.00 0.45 -2.75 0.50 1.00 1.00 0.20 0.75
2 0.10 0.20 0.00 0.99 -2.75 0.54 0.75 1.00 0.20 0.75
3 0.50 0.17 0.13 0.97 1.38 0.67 0.60 0.50 0.50 0.50
4 0.05 0.25 0.25 0.96 1.00 0.58 0.60 0.50 0.33 0.50
5 0.20 0.13 0.25 0.97 1.38 0.63 1.00 0.50 0.33 0.50
6 0.06 0.33 0.04 1.00 2.75 0.79 1.00 0.67 0.20 0.25
7 0.06 0.33 0.04 1.00 2.75 0.75 1.00 0.67 0.20 0.25
8 1.00 1.00 0.04 0.97 -1.00 1.00 0.75 0.40 0.25 0.50
9 0.03 0.27 0.03 0.96 -2.75 0.67 0.75 0.40 1.00 1.00
10 0.08 0.67 1.00 0.81 -0.92 0.50 1.00 1.00 0.20 0.50

Table 3: Relative Importance Matrix

Alternative CD PTC SR EFF LOT PWR MAI VIB SLP WC
1 1 0.34 0.5 0.23 0.75 0.5 0.34 0.5 0.14 0.34
2 2.99 1 0.41 0.34 0.67 0.41 0.59 0.59 0.5 0.75
3 2 2.44 1 0.5 0.75 0.67 0.5 0.67 0.26 0.95
4 4.44 4.26 0.5 1 0.59 0.41 0.67 0.59 0.5 0.86
5 1.34 1.5 1.34 1.69 1 0.14 0.26 0.67 0.41 0.5
6 2 2.44 1.5 2.44 7.41 1 0.67 0.59 0.5 0.67
7 2.99 1.69 2 1.5 3.92 1.5 1 0.5 0.5 0.67
8 2 1.69 1.5 1.69 1.5 1.69 2 1 0.5 0.59
9 7.41 2 3.92 2 2.44 2 2 2 1 0.67
10 2.99 1.34 1.05 1.16 2 1.5 1.5 1.69 1.5 1

Table 4: Normalized Decision Matrix

Alternative CD PTC SR EFF LOT PWR MAI VIB SLP WC
1 0.09 0.28 0.00 0.16 0.16 0.23 0.25 0.18 0.39 0.4
2 0.09 0.14 0.00 0.34 0.16 0.25 0.33 0.18 0.39 0.4
3 0.43 0.12 0.12 0.33 -0.33 0.31 0.42 0.35 0.16 0.27
4 0.04 0.18 0.23 0.33 -0.45 0.27 0.42 0.35 0.23 0.27
5 0.17 0.09 0.23 0.33 -0.33 0.29 0.25 0.35 0.23 0.27
6 0.05 0.23 0.04 0.34 -0.16 0.37 0.25 0.27 0.39 0.13
7 0.05 0.23 0.04 0.34 -0.16 0.35 0.25 0.27 0.39 0.13
8 0.87 0.7 0.04 0.33 0.45 0.47 0.33 0.44 0.31 0.27
9 0.03 0.19 0.02 0.33 0.16 0.31 0.33 0.44 0.08 0.53
10 0.07 0.47 0.93 0.28 0.49 0.23 0.25 0.18 0.39 0.27

245