Professional Documents
Culture Documents
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 336
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
Pw = power requirement (kW) work piece will be turned so that adjacent sections
T = tool life (min.) have different diameters. Or, in its basic form, it can
be defined as the machining of an external surface:
= efficiency of cutting
p = constant (i) With the work piece rotating
n = constant (ii) With a single-point cutting tool, and
Z = constant (iii) With the cutting tool feeding parallel to the
axis of the work piece and at a distance that will
remove the outer surface of the work. Umesh
[22].
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 337
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
important factor in determining finish cutting variables in the model are optimized using the GP
conditions. According to Nithyanandhan et al. technique. We then apply the model to a real life
[10], In machining process, Surface finish is one of data generated from metal machining operations
the most significant technical requirements of the using Cast iron material as the work piece. A
customer. It is pertinent to note that cutting sensitivity analysis is finally carried out to check the
parameters should be selected to optimize the robustness of the model. This approach is intended
economics of machining operations, as assessed by to reduce the time it would take to produce an item
productivity, efficiency or some other suitable of turning process by recommending the optimum
criteria. According to Thakre [20], productivity values of cutting parameters, which form the
could be interpreted in terms of material removal decision variables in the model.
rate in the machining processes. The cutting
conditions that determine the rate of metal removal
are the cutting speed, the feed rate, and the depth of 2 LITERATURE REVIEW
cut. Oberg et al. [11].
Deepak [6] presented an analysis on cutting Attempt is made to review the literature on
speed and feed rate optimization for minimizing optimizing machining parameters in turning
production time of turning process. He performed processes. Various conventional techniques
optimization using only the cutting speed and feed employed for machining optimization include
rate as decision variables in the model, maximum geometric programming, geometric plus linear
cutting speed, maximum feed rate, power and programming, goal programming, sequential
surface roughness as constraints, which he analyzed unconstrained minimization technique, dynamic
independently. It was revealed that the proposed programming, and so on.
method provided a systematic and efficient method
to obtain the minimum production time for turning. In this research work, geometric
Also, the method of GP could be applied programming approach will be used in the modeling
successfully to optimize the production time of of cutting parameters problem in machining
turning process. operations It is a powerful tool for solving some
In this research work, we shall include the special type of non-linear programming problems.
material removal rate so as to reflect the depth of Generally, it has a wide range of applications in
cut in the model, then explore simple, effective and optimization, and particularly, in engineering for
efficient way of optimizing the production time of solving some complex optimization problems.
the turning process within the following operating According to Islam and Roy [9], since late 1960s,
constraints -the maximum cutting speed, maximum GP has been known and used in various fields like
feed rate, maximum depth of cut, power Economics, Physical Sciences, Engineering, et
requirement and surface roughness. GP approach cetera. Non-linear programming problems are
will be used in obtaining the optimal solution. perhaps the most tedious class of optimizing
problems to deal with, because the response
function and sometimes constraints are both non-
linear and there exists no handy transformation to
1.1 Scope and Purpose simplify or reduce them to linear.
This research covers the area of engineering Many applications of GP are on engineering
machining operation that concerns the turning design problems where parameters are estimated.
process. The cutting parameters (cutting speed, feed Following the pioneer work of Taylor (1907) and
rate and depth of cut) form the input (design or his famous tool life equation, different analytical
decision) variables in the developed GP model for and experimental approaches for the optimization of
the minimization of production time. The decision
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 338
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
machining parameters have been investigated, see where large number of responses have to be
Roby et al. [16]. optimized simultaneously.
In 1967 Duffin, Peterson and Zener put a 3) Utility based Taguchi method was found fruitful
foundation stone to solve wide range of engineering for evaluating the optimum parameter setting and
problems by developing basic theories of GP in the solving such a multi-objective optimization
book Geometric Programming, see Das and Roy problem.
[3].
4) They asserted that the technique they adopted can
Also in 1976, according to Das and Roy [3], be recommended for continuous quality
Beightler and Phillips gave a full account of entire improvement and off-line quality control of a
modern theory of GP and numerous examples of process/product.
successful applications of GP to real-world
problems in their book Applied Geometric Ojha and Biswal [13] concluded that by
Programming. using weighted method we can solve a multi-
objective geometric programming problem as a
Umesh [22] did analysis on Optimization of vector-minimum problem. A vector-maximum
Surface Roughness, Material Removal Rate and problem can be transformed as a vector-
Cutting Tool Flank Wear in Turning Using minimization problem. If any of the objective
Extended Taguchi Approach. The study dealt with function and/or constraint does not satisfy the
optimization of multiple surface roughness property of a posynomial after the transformation,
parameters along with material removal rate (MRR) then any of the general purpose nonlinear
in search of an optimal parametric combination programming algorithms could be used to solve the
(favourable process environment) capable of problem. This technique could also be used to solve
producing desired surface quality of the turned a multi-objective signomial geometric programming
product in a relatively lesser time (enhancement in problem. However, if a GP problem has either a
productivity). The study proposed an integrated higher degree of difficulty or a negative degree of
optimization approach using Principal Component difficulty, then any of the general purpose nonlinear
Analysis (PCA), utility concept in combination with programming algorithms could be used instead of a
Taguchis robust design of optimization GP algorithm.
methodology. The following conclusions were Ojha and Das [14] also carried out a
drawn from the results of the experiments and research on MultiObjective Geometric
analysis of the experimental data in connection with Programming Problem Being Cost Coefficients as
correlated multi-response optimization in turning: Continuous Function with Weighted Mean Method.
Their conclusion was the same as that of Ojha and
1) Application of PCA was recommended to Biswal [13].
eliminate response correlation by converting
correlated responses into uncorrelated quality Agarwal [1], after thirty-six (36) specimens
indices called principal components which have of Aluminium alloy were machined, noted that the
been as treated as response variables for surface roughness could be efficiently calculated by
optimization. using spindle speed, feed rate and depth of cut as
the input variables.
2) Based on accountability proportion (AP) and
cumulative accountability proportion (CAP), PCA Deepak [6] modeled the cutting speed and
analysis could reduce the number of response feed rate for the minimum production time of a
variables to be taken under consideration for turning operation, using GP approach. The
optimization, which was really helpful in situations maximum cutting speed, the maximum feed rate,
maximum power available and the surface
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 339
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
roughness, were taken as constraints. The results of Pt = Machining Time + Tool Changing Time + Set-
the model showed that the proposed method up Time (1)
provided a systematic and efficient method to
obtain the minimum production time for turning. It We now add the Material Removal Rate (MRR)
was concluded from his study that the obtained
Pt + MRR = Machining Time + Tool Changing
model could be used effectively to determine the
Time + Set-up Time + MRR (2)
optimum values of cutting speed and feed rate that
will result in minimum production time. It was also Let us define:
revealed that the method of GP could be applied
successfully to optimize the production time of PtMRR = production time per piece (min./piece) +
turning process. material removal rate (in.3/min.)
Yograj and Pinkey [23] addressed and solved the Then we have:
problem of parameter optimization in constrained
machining environment using Differential
Evolution (DE), a potential candidate of non-
traditional global optimizers. The mathematical Thus,
simulation turned out as a complex and highly
constrained machining model where the aim was to
minimize the total production cost. The machining
parameters as feed rate, cutting speed and depth of (3)
cut during roughing and finishing passes were the
main process parameters whose optimal values Where,
affect the machining process to greater extent. The
: Oberg et al. [11]
observation of optimal results showed that the
proposed methods (DE), provided promising results The Taylor's tool life denoted by T, used in
on quality and feasibility basis as compared to other Equation (3) above is given by:
existing algorithms. The proposed algorithm found
significantly better optimal solution with less
computational efforts. Thus, DE could be
recommended as a reliable and efficient method for
(4)
solving such complex machining problems and
those with higher degree of complexity. Where, n, p and Z are constants in tool life
equation; depend on the many factors like tool
geometry, tool material, work piece material, etc.,
3 MATHEMATICAL MODELING FOR according to Deepak [5].
OPTIMIZATION In Tool-life testing, very high standards of
systematic tool testing were set by F.W. Taylor in
According to Arua et al. [2], Modeling is the work which culminated in the development of
very important in operations research. Though high speed steel. The variable of cutting speed, feed
model has different shades of meaning Here we rate, depth of cut, tool geometry and lubricants, as
have a GP model formulation, which is theoretical well as tool material and heat treatment were
and has a mathematical expression. studied and the results presented as mathematical
The Production time to produce a part by relationships for tool life as a function of all these
turning operation is denoted by Pt, and expressed as parameters. These tests were all carried out by lathe
follows: turning of very large steel billets using single point
tools. Such elaborate tests have been too expensive
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 340
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
(5)
It is necessary to state that a function of the
form, as presented in equation (5) above can also be
written as:
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 341
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
optimization of cutting parameters in this research In the same manner, when we apply
work, it is pertinent to define such problems that GP nonlinear programming to engineering design
deals with. We then consider problems in which the problems, the objective function and the constraint
objective and the constraint functions are of the functions frequently take the form of GP (as above),
following type: Hillier and Lieberman [7]. We now justify the use
of GP approach for optimization of cutting
parameters by comparing our model with equations
(6) and (7), which vividly, we can discover that the
approach is satisfied.
where In relation to the above definition of GP,
therefore we proceed as follows:
(6) (A)
Subject to the following constraints:
Maximum Cutting parameters (Maximum
It is assumed that: Cutting speed, Maximum Feed rate,
Maximum Depth of cut), Power and Surface
i. All Cj > 0, Cj, of course, are constants. roughness constraints
ii. N < , that is, N is finite.
The constraints above imply the following
iii. The exponents, aij are unrestricted in sign.
condition:
iv. The function, f(x) takes the form of a
polynomial except that the exponents, aij vfDc Pw Ra vmax fmax Dc max Pw max Ra max
may be negative. For this reason, and
because all Cj > 0, f(x) is called a where,
posynomial, Taha [19]
Similarly, when we apply nonlinear
programming design problems, as well as certain
economics and statistics problems, the objective
function and the constraint functions frequently take
the form:
vmax fmax Dc max are the maximum cutting parameters
allowable on the lathe, Pw max and Ra max are the
maximum power required and maximum surface
roughness respectively, allowable for the turning
where, operation.
(7)
Substituting, inequality above implies that:
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 342
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
(8)
4.2 Constrained Minimization
Now, to enable us perform optimization, we
When an objective function is to be
apply the conclusion from the above derivations.
minimized, it is workable if either the minimization
Thus we have: problem is constrained (or unconstrained)
depending on the problem formulation and its
environment. We discover that most engineering
optimization problems are subject to constraints. If
(B) the objective function and all the constraints are
expressible in the form of posynomials, GP can be
used most conveniently to solve the optimization
v 0, f 0, Dc 0 (the non-negativity constraints); problem: see Rao [15]. GP problem whose
while 1, 2, 3 and 4 are machining constants. parameters, except for exponents, are all positive
are called posynomial problems, whereas GP
We now wish to formulate this as a Zero- problems with some negative parameters are
degree-of-difficulty problem, considering the referred to as signomial problems: see Ojha and
number of functions in the objective function and Biswal [13].
the constraints, and also the number of decision
variables in the objective function. The degree of Let us consider the constrained
difficulty is defined as the number of terms minus minimization problem whereby we are required to
the number of variables minus one, and is equal to find the design variables, X = xi, i = 1, 2, . . ., n,
the dimension of the dual problem, Ojha and which minimize the objective function:
Biswal [13].
Simply, we have the degree of difficulty
as follows:
(9)
Degree of difficulty = (Nn1)
Subject to:
Where,
N = the number of functions in the objective
function and the constraints (that is, the total
number of posynomial terms in the problem), which
(10)
is equal to four (4).
Where,
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 343
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
(17)
We then have the corresponding dual The simultaneous equations (14), (15), (16)
functions as: and (17) above, which are the Normality and the
Orthogonality conditions, can be put in a matrix
form:
(18)
(11)
Where,
A is a (4 x 4) square matrix, W is a (4 x 1) column
vector and b is a column vector, or a (4 x 1) identity
matrix corresponding to A and W: Taha [19].
Equation (11) is transformed to:
Then, we have the arrangement as follows:
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 344
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
A W b
1 1 1 0 w1 1 1 (26)
0 0 1 1 w4 0 0
(28)
(31)
(21)
(32)
(22)
Equating equations (20) and (21), then (33)
solving simultaneously, we have:
Substituting the values of the normalizing
weights appropriately, in equation (13) above we
have:
(23)
(24) (34)
Now, we optimize the cutting parameters;
the cutting speed (v), the feed rate (f) and Depth of
(25) cut (Dc), using the following conditions:
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 345
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
(35)
(42)
Substituting equations (40) and (41) in
(36)
equation (42), therefore:
(37)
From equation (34),
(43)
, which implies that: For convenience in application, however,
equation (42) can be used once equations (40) and
(41) are derived. This also implies that from
equation (42), any of the cutting parameters can be
(38) made the subject which is dependent on the others,
or simply put, the dependent variable.
From equation (35),
In summary, the cutting parameters, the
cutting speed (v), the feed rate (f) and the depth of
, which implies cut (Dc) have been optimized for a set of constraints
that: combined together, to yield the desired positive
result. Thus, when different values of the machining
constants and other relevant parameters for turning
operation are provided (depending on the metallic
material used as work piece, and its dimension), the
(39) equations above will enable us to obtain the values
of the cutting parameters; the values of the cutting
Equating equations (38) and (39), then parameters are tested for minimum production time
solving for v, therefore: and material removal rate, using the model in either
equation (11) or equation (12) above.
Also, for convenience, if data for corresponding
estimate values of the cutting parameters are made
(40)
available via engineering machining handbooks or
Substituting equation (40) in either equation by mere estimation during turning process (that is,
(38) or equation (39), and solving for f, therefore: varying the cutting parameters), equation (11) or
equation (12) above is the solution for deriving the
minimum production time with material removal
rate of any type of metallic material used as work-
piece. In this instance, it is now easy to single out
(41)
the cutting parameters that correspond to the
Also, from equation (37), minimum production time (the optimum values).
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 346
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 347
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 348
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
16 235 0.0179 0.245 22.99 Figure 1: Variation of cutting speed (v) against
17 250 0.0189 0.26 24.2 PtMRR(v,f,Dc)
18 265 0.0199 0.275 25.87
19 280 0.0209 0.29 28
20 295 0.0219 0.305 30.56
21 310 0.0229 0.32 33.55
22 325 0.0239 0.335 36.98
23 340 0.0249 0.35 40.83
24 355 0.0259 0.365 45.13
25 370 0.0269 0.38 49.88
26 385 0.0279 0.395 55.08
27 400 0.0289 0.41 60.74
28 415 0.0299 0.425 66.89
29 430 0.0309 0.44 73.52
30 445 0.0319 0.455 80.66
31 460 0.0329 0.47 88.32
32 475 0.0339 0.485 96.5
33 490 0.0349 0.5 105.23
34 505 0.0359 0.515 114.52
35 520 0.0369 0.53 124.39 The curve obtained between production time
36 535 0.0379 0.545 134.84 with material removal rate and cutting speed, Fig.
37 550 0.0389 0.56 145.9 1 above reveals that a smaller value of cutting
speed results in a high production time. It is due to
38 565 0.0399 0.575 157.57
the fact that a smaller cutting speed increases the
39 580 0.0409 0.59 169.88 production time of parts. Also, it will decrease the
40 595 0.0419 0.605 182.85 profit rate due to the production of a lesser number
41 610 0.0429 0.62 196.48 of parts. In the same premise, if the cutting speed is
42 625 0.0439 0.635 210.79 too high, it will also lead to a high production time
43 640 0.0449 0.65 225.8 due to excessive tool wear and increased machine
downtime (time during which work or production is
44 655 0.0459 0.665 241.53 stopped). The optimum cutting speed is somewhere
45 670 0.0469 0.68 257.99 in-between too slow and too fast which will
46 685 0.0479 0.695 275.19 yield the minimum production time and maximum
47 700 0.0489 0.71 293.16 production rate at the same cutting speed.
48 715 0.0499 0.725 311.91
49 730 0.0509 0.74 331.45
50 745 0.0519 0.755 351.81
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 349
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
Figure 2: Variation of feed rate (f) against Figure 3: Variation of depth of cut (Dc) against
PtMRR(v,f,Dc) PtMRR(v,f,Dc)
Also, the curve between the production time Again, the curve obtained between production time
with material removal rate and the feed rate, Fig. with material removal rate and depth of cut, Fig.
2 above indicates that a smaller feed rate will 3 above depicts that a smaller value of depth of cut
result in high production time. A smaller feed rate results in a high production time. It is due to the fact
means the number of revolutions should be that a smaller depth of cut increases the production
increased. The more the number of revolutions, the time of parts. Also, it will decrease the profit rate
less will be the production time. However, a very due to the production of a lesser number of parts.
high feed rate is not advisable as it will increase the Similarly, if the depth of cut is too high, it will also
tool wear and surface roughness leading to lead to a high production time due to excessive tool
increased machining time and machine downtime, wear and increased machine downtime. The
which will result to high production time. So, the optimum depth of cut is somewhere in-between
optimum feed rate is somewhere in-between too too low and too high which will yield the
small and too high which will result in the minimum production time, maximum production
minimum production time and maximum rate and better quality of machined part.
production rate at the same feed rate.
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 350
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
www.ijaert.org
International Journal of Advanced Engineering Research and Technology (IJAERT) 351
Volume 4 Issue 11, November 2016, ISSN No.: 2348 8190
Steel Material Using Taguchis Method, [22] Umesh, K., (2009). Optimization of Surface
IJISET - International Journal of Innovative Roughness, Material Removal Rate and
Science, Engineering & Technology, Vol. 1(4) Cutting Tool Flank Wear in Turning Using
[11] Oberg, E., Jones, F. D. and Horton, H. L. Extended Taguchi Approach, A thesis
(1985). Machinerys Hand-Book, 22nd Revised submitted in partial fulfillment of the
Edition, Industrial Press Inc. New York. requirements for the degree of Master of
[12] Oberg, E., Jones, F. D., Horton, H. L. and Technology in Production Engineering,
Ryffel, H. H. (2012). Machinerys Handbook, National Institute of Technology Rourkela
29th Edition, Industrial Press Inc. New York. 769008, India (Unpublished).
[13] Ojha, A. K. and Biswal, K. K. (2010). Multi- [23] Yograj, S. and Pinkey, C. (2012). Analysing
Objective Geometric Programming Problem Constrained Machining Conditions in Turning
with Weighted Mean Method, (IJCSIS) Operations by Differential Evolution, Advances
International Journal of Computer Science and in Mechanical Engineering and its
Information Security, Vol. 7, No. 2 Applications (AMEA) Vol. 2, No. 3, 201-206
[14] Ojha, A. K. and Das, A.K. (2010).
MultiObjective Geometric Programming
Problem Being Cost Coefficients as
Continuous Function with Weighted Mean
Method, Journal of Computing, Volume 2(2).
[15] Rao, S. S. (2009). Engineering Optimization:
Theory and practice, Fourth Edition, John
Wiley & Sons, Inc. Hoboken, New Jersey.
[16] Roby, J., Josephkunju, P. and Roy, N. M.
(2013). Effect of Work Material, Tool
Material on Surface Finish in Turning
Operations, International Journal of
Engineering and Innovative Technology
Volume 2(7)
[17] Shirpurkar, P. P., Bobde, S.R., Patil V.V. and
Kale B.N. (2012). Optimization of Turning
Process Parameters by Using Tool Inserts- A
Review, International Journal of Engineering
and Innovative Technology (IJEIT) Volume
2(6)
[18] Taha, H.A. (2002). Operations Research: An
Introduction, Pearson Education, Seventh
Edition, ISBN 81-7808-757-X.
[19] Taha, H.A. (2007). Operations Research: An
Introduction, Pearson Education, Eighth
Edition, ISBN 0-13-188923-0.
[20] Thakre, A. A. (2013). Optimization of Milling
Parameters for Minimizing Surface Roughness
Using Taguchis Approach, International
Journal of Emerging Technology and
Advanced Engineering, Volume 3(6)
[21] Trent, E.M. (1987). Metal Cutting, Butterworth
and Co. (Publishers) Ltd, ISBN 0-408-10603-4.
www.ijaert.org