International Journal of Engineering Materials and Manufacture (2016) 1(1) 11-15
https://doi.org/10.26776/ijemm.01.01.2016.03
M. H. F. Hazza and N. A. Najwa
Department of Manufacturing and Materials Engineering
International Islamic University Malaysia
PO Box 10, 50728 Kuala Lumpur, Malaysia
E-mail: muataz@iium.edu.my
Reference: Hazza, M. H. F and Najwa, N. A. (2016). Optimization of Cutting Parameters to Minimize Tooling Cost in High Speed
Turning of SS304 Using Coated Carbide Tool using Genetic Algorithm Method. International Journal of Engineering Materials and
Manufacture, 1(1), 11-15.
Optimization of Cutting Parameters to Minimize Tooling Cost in High
Speed Turning of SS304 Using Coated Carbide Tool using Genetic
Algorithm Method
Muataz Hazza F. Al Hazza, Nur Amirah Najwa Bt Mohmad Bakhari
Received: 17 August 2016
Accepted: 30 August 2016
Published: 05 September 2016
Publisher: Deer Hill Publications
© 2016 The Author(s)
Creative Commons: CC BY 4.0
ABSTRACT
High speed turning (HST) is an approach that can be used to increase the material removal rate (MRR) by higher
cutting speed. Increasing MRR will lead to shortening time to market. In contrast, increasing the cutting speed will
lead to increasing the flank wear rate and then the tooling cost. However, the main factor that will justify the best
level of cutting speed is the tooling cost which merges all in one understandable measurable factor for manufacturer.
The aim of this paper is to determine experimentally the optimum cutting levels that minimize the tooling cost in
machining AISI 304 as a work piece machined by a coated carbide tool using one of the non-conventional methods:
Genetic Algorithm (GA). The experiments were designed using Box Behnken Design (BBD) with three input factors:
cutting speed, feeding speed and depth of cut and three machining levels.
Keywords: High speed turning, tooling cost, AISI 304, MRR
1 INTRODUCTION
The development of advanced manufacturing technology has been growing up rapidly. One of the advanced
approaches is by increasing the machining speed to increase material removal rate and then shortening time to market,
lowering cost, high accuracy and better quality. One approach for reducing the machining time in machining is by
increasing the speed turning. High speed turning is difficult to define due to the fact of materials are varied for their
hardness. Therefore, high speed turning for one material may still be a low speed for another for example; the high
speed for titanium is a low speed for aluminium [1]. However, these technologies should be justified by economic
study. One of the most effective tools for economic study is by developing a cost model.
In high speed turning the machining zone will be under high temperature and high sliding velocity. Therefore,
the wear progress will be difficult to estimate and predict. However, the wear rate of the cutting tool may give
unacceptable outputs and that will result a low quality of surface roughness [2]. However, estimating the tool wear
is highly valuable to estimate the tooling cost due to the relationship of tool life and material removal during the life
of tool. However, tool insert may reaches its life and should be removed and changed before the tool insert edge
cannot give the desired and accepted roughness. If the cutting tool reaches its life very fast then this will lead to
increase the tooling cost becomes. Therefore, the manufacturer needs to determine the best cutting levels that
minimize the tooling cost. Thus, estimating and then determining the best levels of the independent factors in
machining becomes critical and important.
Determining the input level that can give the optimum values in machining process for one output response is
very useful if the need for that response is important for one application but needs a validate and reliable
mathematical model. In this research a regression empirical model will be developed and then the genetic algorithm
method will be used to determine the minimum tooling cost of high speed turning of AISI 304 using coated carbide
insert.
Optimization of Cutting Parameters to minimize tooling cost in High Speed Turning of SS304 …
12
2 METHODOLOGY
The methodology was three related integrated parts: firstly, the experimental work based on the theoretical study,
then developing the cost model based on the experimental work. Finally, using the genetic algorithm in order to
determine the best cutting levels that will give the minimum tooling cost. Box Behnken design (BBD) has been used
in this research to conduct the experimental work for three independent factors: cutting speed, feeding speed and
depth of cut. Three levels: –1, 0, and 1. However, BBD is easy to conduct in addition to the ability of sequentially.
Figure 1 concluded the activities and tasks need to be done in order to achieve the objective of the research in
developing tooling cost and then determine the optimum cutting parameters to minimize the tooling cost in high
speed turning for SS304 Using Coated Carbide Tool.
Figure 1: Research methodology.
3 EXPERIMENT PROCEDURE
Experimental works was conducted on CNC turning machine type Power Path 15 HS – High Speed Version (spindle
ASA A 2-5”) and the insert chosen for this study was a coated cemented carbide type (TNGA 16 04 08 T1020) to
turn work piece of AISI 304. Under dry cutting conditions with cutting speed from 500 up to 700 m/ min, feed
speed of 1000 to 2000 mm/min and depth of cut 0.1 to 0.3 mm. Based on BBD with three centre points, fifteen run
have been conducted. Table 1 shows the machining levels.
Al Hazza et al., (2016): International Journal of Engineering Materials and Manufacture, 1(1), 11-15
13
Table 1: Machining levels.
Max Min
Cutting speed (m/min) 500 700
Feeding speed (mm/min) 1000 2000
Depth of cut (mm) 0.1 0.3
4 DEVELOPMENT OF TOOLING COST MODEL
Tooling cost model developed based and is limited in this research to the cost of tool holder and tool insert. The
model was developed based on calculating the cost of removal one cubic centimetre. Actually, tooling cost is inversely
proportional with tool life and proportional with the machining time Therefore, the tooling cost is calculated based
on the work of [4-7]. The machining time components used in this research was based on machining time model
developed by [6]. However, the final tooling cost was calculated based on the following equation [3]: However, all
the experimental results is concluded in Table 2. The results was analysed using the DoE 6.0.8. The analysis of variance
(ANOVA) was conducted to develop an empirical model that can be used to estimate the tooling cost. Analysis of
variance (ANOVA) is concluded in Table 3 which shows significant with value of 0.002 and lacks of fit of 0.9574.
Therefore, the model can be used to navigate the design space. The following model has been used to determine the
optimum cutting levels in the boundary of the design that can minimize the tooling cost.
5 GENETIC ALGORITHM
The genetic algorithm has been implemented using the developed model in the previous section in order to minimize
the total tooling cost using three different independent decision variables: cutting speed, feed speed and depth of
cut. Table 4 concluded the objective functions decision variables, constrains. In the implementation, the chromosome
values of each individual are generated randomly from the ranges of these values. However, the simulation has been
repeated for four different runs, each for 1000 iterations. The optimum results was concluded in Table 5. Finally
Figure 2 shows the results for different runs that give the minimum tooling cost. The results show for different runs
that after 100 iterations the values of the objective function become stable and only few points are giving results far
from the final one.
3
3
( )
( / )
( )
Tc
C Rm
Cost Rm Cm
MRV cm
Table 2: Result of experiment.
No. of Run
Cutting Speed [Vc]
(m/min)
Feeding Speed
[Vf] (mm/min)
Depth of Cut
[d] (mm)
Tooling Cost
(RM/cmᵌ )
1 700.00 1500.00 0.30 0.051136
2 700.00 1500.00 0.10 0.273389
3 500.00 2000.00 0.20 0.157189
4 600.00 1500.00 0.20 0.053346
5 500.00 1500.00 0.10 0.193171
6 700.00 2000.00 0.20 0.267183
7 600.00 2000.00 0.10 0.100728
9 600.00 1000.00 0.10 0.096816
10 600.00 1500.00 0.20 0.177876
11 600.00 1500.00 0.20 0.592315
12 600.00 2000.00 0.30 0.078657
13 600.00 1000.00 0.30 0.588838
14 700.00 1000.00 0.20 0.176317
15 500.00 1000.00 0.20 0.160914
17 500.00 1500.00 0.30 0.173868
Optimization of Cutting Parameters to minimize tooling cost in High Speed Turning of SS304 …
14
Table 3: ANOVA table for tooling cost.
Source
Sum of
Squares DF
Mean
Square
F
Value Prob > F
Model 0.30711 9 0.034123 20.18702 0.0020 significant
A 0.060565 1 0.060565 35.82974 0.0019
B 0.038284 1 0.038284 22.64872 0.0051
C 0.088878 1 0.088878 52.57925 0.0008
A2 0.016824 1 0.016824 9.952853 0.0252
B2 0.062186 1 0.062186 36.78888 0.0018
C2 0.008493 1 0.008493 5.024655 0.0751
AB 0.033702 1 0.033702 19.93782 0.0066
AC 0.138012 1 0.138012 81.64661 0.0003
BC 0.04324 1 0.04324 25.58052 0.0039
Residual 0.008452 5 0.00169
Lack of Fit 6.82E-06 1 6.82E-06 0.003231 0.9574 not significant
Pure Error 0.008445 4 0.002111
Cor Total 0.315562 14
Tooling cost = +1.47020-9.44817E-003* Cutting speed+4.16756E-003 * feed speed-9.82922* depth of
cut+7.34688E-006* Cutting speed2-6.96458E-007 * feed speed2 +5.22014* depth of cut2-
2.80425E-006* Cutting speed * feed speed+0.018575* Cutting speed * depth of cut-3.17638E-
003* feed speed * depth of cut
Table 4: GA objectives, decision variables, constrains and parameters.
Objective function Decision variables constrains
Minimize Tooling cost Cutting speed
Feeding speed
Depth of cut
500≥Vc≥700
1000≥Vc≥2000
0.1≥Vc≥0.3
Number of Individuals in Population= 20
Number of generation= 1000
Crossover Rate = 35%
Mutation Rate = 8%
Table 5: GA optimization results.
Run Cutting speed Feeding speed Depth of cut Tooling cost
1 500 2000 0.3 0.35779
2 500 1989 0.3 0.035988
3 500 1467 0.3 0.046178
4 500 1276 0.3 0.049914
Al Hazza et al., (2016): International Journal of Engineering Materials and Manufacture, 1(1), 11-15
15
Run 1 Run 2
Run3 Run4
Figure 2: four different runs for 1000 iteration.
6 CONCLUSIONS
Results concluded in table 5, show that the main flexible factor is the feeding speed which varied from 1276 to 2000
mm/min. In contrast, the cutting speed and depth of cut is constant with the values of 500 m/min and 0.3 mm. The
results show that to minimize tooling cost, the cutting speed should be in the lowest level while the depth of cut
should be in the maximum level. In addition the lowest tooling speed will be achieved in the highest feeding speed.
REFERENCES
1. Muataz Hazza F. Al Hazza, H., Nabilah, M. Yusof, B. M., Najwa, N. A., Eikhwan, M., Khazairul, M., & Taha,
A. H. (2013). Experimental Study of flank wear in High Speed Turning of Stainless Steel AISI 304. Journal of
Advanced Science and Engineering Research, 3(2), 96-103.
2. Ozel, T. and Nadgir, A. (2002). Predictions of flank wear by using back propagation neural network modeling
when cutting hardened H-13 steel with chamfered and honed CBN tools. International Journal of Machine Tools
& Manufacture, 42(2), 287–297.
3. Adesta, E. Y. T., Al Hazza, M., Agusman, D., & Sutjipto, A. G. E. (2012). Development of Tooling Cost Model
for High Speed Hard Turning. Advanced Materials Research, 418, 1482-1485.
4. Gara, S., Bouzid, W., Hbaieb, M., & Amar, M. B. (2007). Cutting speed optimization in high speed turning.
International Review in Mechanical Engineering, 1, 237-244.
5. Gara, S., Bouzid, W., Ben Amar, M. and Hbaieb M. (2008). Cost and time calculation in rough NC turning.
International Journal Advanced Manufacturing Technology, 40(9-10), 971-981.
6. Adesta, E. Y. T., & Al Hazza, M. H. (2011). Machining Time Simulation in High Speed Hard Turning. Advanced
Materials Research, 264, 1102-1106.
7. Al Hazza, M. H. F., Konneh, M., Iqbal, M., Taha, A. H., & Hasan, M. H. (2015). Using the Desirability Function
as an Effective Tool in Target Costing Model. Advanced Materials Research, 1115, 126-129).
8. Kalpakjian Serope and Steven R. Schmid, 2010. Manufacturing engineering technology. Machining centers,
machine tool structures and machining economics (pp 208). Prentice Hall.