015 0713

015-0713
Application of Taguchi Method in process improvement of turning of a Superalloy
NIMONIC 80A
Ricardo Batista Penteado - Univ. Estadual Paulista – UNESP – Department of Production
Av.Ariberto Pereira da Cunha, 333 Pedregulho, 12515-410 Guaratingueta , SP, Brazil
email. taiuva@gmail.com, phone + 55 – 12 – 31232855 fax: +55 – 12 - 31232855
Thiago Garcia Hagui - Univ. Estadual Paulista – UNESP – Department of Production
email. thiago_hagui@hotmail.com, phone + 55 – 12 – 31232855 fax: +55 – 12 - 31232855
José Carlos Faria - Univ. Estadual Paulista – UNESP – Department of Materials and
Technology Av.Ariberto Pereira da Cunha, 333 Pedregulho, 12515-410 Guaratingueta , SP,
Brazil email faria207@bol.com.br , phone + 55 – 12 – 31232855 fax: +55 – 12 - 31232855
Messias Borges Silva – Univ. Estadual Paulista – UNESP – Department of Production
email. messias.silva@feg.unesp.br, phone + 55 – 12 – 31232855 fax: +55 – 12 - 31232855
Marcos Valério Ribeiro - Univ. Estadual Paulista – UNESP – Department of Materials and
Technology Av.Ariberto Pereira da Cunha, 333 Pedregulho, 12515-410 Guaratingueta , SP,
Brazil email. mvalerio@feg.unesp.br, phone + 55 – 12 – 31232853 fax: +55–12-31232852
Abstract
In this article was used the Taguchi Method as a tool to improve a turning of superalloy
NIMONIC 80A, using a cemented carbide TNMG 160408-23 C standard H10A (ISO S10)
and a holder PTGNR K16, in order to optimize the cutting length (Lc). Was used the L8
orthogonal array with three factors: cutting speed, feed rate and cutting depth, operating in
two levels and three interactions, setting yet the value of flank wear in 5 mm as final goal for
each experimental condition. The analysis of variance (ANOVA) of the response variable
cutting length showed a significant influence of the process of variable: cutting speed and
depth of cut, as well as their interactions. The Taguchi Method showed to be very useful in
process improvement, because provides confident information about influence of factors on a
response variable and require less number of experiments than traditional methods to improve
the process.
Keywords: Taguchi Method, Machining, Design of Experiment
Create PDF files without this message by purchasing novaPDF printer (http://www.novapdf.com)

1. Introduction
Economic globalization and the rapid and continuous appearing of new technologies
mobilized organizations to obtain the maximum degree of competitiveness, high quality
products in short time in order to ensure their survival and growth in the market
(Benardos,&Vosniakos, 2002; Besseris, 2009) .
In this scenario, companies started to look for quality not only in their products, but
also in its production process, thus increasing considerably their production, and launching on
the market at a higher quality product to meet consumer is increasingly demanding.
For this purpose, this study proposes the use of the technique Design of Experiments
(DOE), more specifically, the Taguchi Method to optimize the process of turning on CNC
(Computer Numerical Control) in this case for the super alloy Nimonic 80A.
2. DOE - Design of Experiment
The concept of Design of Experiment was first introduced in early 1920 in a small
station of Agricultural Research in England, by a scientist called Sir. Ronald Fisher. He
showed as a valid experiment could be conducted in the presence of many natural conditions
variables, such as: temperature, soil conditions and rainfall. The principles of experimental
design first used in agriculture have been adapted successfully in industry and in military
applications since 1940 (Franceschini & Macchietto, 2008 ; Kim & Kalb, 1996).
The DOE is a technique used to define what data, in what quantity and conditions
should be collected during an experiment, trying to satisfy two major goals: the statistical
accuracy of the response and lower cost (Gunasegaram et al., 2009).
Benardos,&Vosniakos (2002) The most efficient method of experimental planning is
DOE, offer even greater opportunities for improving the quality and increase productivity.

The benefits of DOE include the possibility of improving performance in the process,
avoiding the trial and error to find solutions (Kleijnen et al., 2005). For Antony (2006) DOE
emphasizes the development and use of regression models to predict process behavior under
different process conditions to a certain level of confidence.
Montgomery (1997) states that before starting the experimentation is important to
establish the planning of experiments. The author emphasizes the importance of the problem
domain by all those involved in the experiment, and recommends that during the
implementation process is carefully monitored to ensure that everything be done according to
plan.
The application of DOE has gained acceptance around world as an essential tool for
improving the quality of goods and services. This recognition is partly due to the work of
Genichi Taguchi, a Japanese expert in Quality, which promoted the use of DOE in designing
robust products - relatively insensitive to environmental variations (Antony 2006; Padke
1989; Rosa et al. 2009).
3. Taguchi Experimental Design
“The experimental design is widely used to optimize process parameter values in order
to improve the quality properties of a product for example” (Rosa et al. 2009). Conventional
experimental design methods are generally complex and not always reach the desired
objectives. Moreover, these methods require a large number of experiments when the number
of process parameters increases.
Taguchi developed the foundations of Robust Design introduced in the 1950s and
1960s and the application of his method in electronics, automotive, photographic and many
others industries has been an important factor in the rapid industrial growth of Japanese
industries (Padke, 1989). Taguchi method uses a special design of orthogonal arrays that

allows studying the whole parameter space with a limited number of experiments (Ross,
1996). Besides, this method provides other advantages: it reduces economically the variability
of the response variable, shows the best way to find out the optimum process conditions
during laboratory experiments, it is an important tool for improving the productivity of the
R&D activity and it can be applied to any process.
The steps of the Taguchi experimental design (Barrado et al., 1996) are: (a) to select
the output variable(s) (response(s)) to be optimized; (b) to identify the factors (input
variables) affecting output variable(s) and to choose the levels of these factors; (c) to select
the appropriate orthogonal array; the arrays are found in literature (Taguchi & Konishi, 1987);
(d) to assign factors and interactions to the columns of the array; (e) to perform experiments;
at this step it is important to randomize the trials in order to minimize the systematic error; (f)
to analyze the results using analysis of variance (ANOVA); (g) to determine the optimal
process parameters; (h) to perform confirmatory experiments, if it is necessary.
The analysis of variance (ANOVA) is applied in order to test the equality of several
means, resulting in what process parameters (factors) are statistically significant. The results
of ANOVA are presented in a table that displays for each factor (or interaction) the values of:
• SS: sum of squared deviations from the mean. For n values of yi and the mean value .
SS = (1)
• d.f.: degree of freedom which is the number of levels for each factor minus 1.
• MS: mean of squares.
MS = (2)
• F: F is the ratio between the mean of squares effect and the mean of squares error.
F = (3)

F-test is used to see the significance of each factor (or interaction) on the response variable or
signal-to-noise ratio.
4. Machining
Kitagawa et al (1997) relates that due to the difficulty in the cut of superalloy the
demand for machining in higher speeds is increasing in order that reaches high productivity
and lower cost of machining.
Aeronautical industries manufacture components of motors with alloys of titanium,
nickel alloys and also special stainless steels. These industries are characterized by a great
cost in the production of these workpieces machining and one of the great reasons of this
elevated cost is the cost hour/machine. For this reason, it is interesting to reduce the times of
workpieces machining and to increase the tool life, because the costs hour in the industry
aeronautics they are higher than in the conventional industries (RIBEIRO, 2003; López De
Lacalle et al. 1998).
In research of good achievements, the machine CNC (Computer Numerical Control)
had as initial objective to solve the machining of the components of the high complexity, and
afterwards came support the reduction of unproductive time (Tzeng et al., 2009). However the
utilization of these machines only become economically feasible when guaranteed the proper
employment of these machines and tools during the machining process, that is, the efficiency
of machining process depends of the utilization of optimized cutting data, updated and
adequate to the real work conditions.
Good results in machining operation are related with an achievement of the component
wished measurements, selection of cutting tools and cutting conditions is an essential element
in process planning for machining, because each material would be influenced by different
machining parameters (Wang et al. 2007; Yih-fong, 2006).

The optimization of machining process have been one of the options for
manufacturing industries, looking for better achievements, represented mainly by the cost
reduction or manufacturing time (Diniz et al, 2001).
5. Experimental Procedures
Based on the analysis of the experiments and their results, it intends to guide the
choice of production parameters, which will permit a greater cutting length possible,
evaluating the best arrangement between the chosen parameters and the effect of the
interaction between them.
5.1. Machining tests
The external cylindrical turning tests were carried out considering machining
parameters (semi-rough).
The tests were performed on a CNC lathe MACH-9-CENTUR 30S, 25 to 3500 rpm,
with power of 10 hp, brand ROMI from the Department of Materials and Technology, Faculty
of Engineering, Guaratinguetá Campus, UNESP. Was used a chip carbide uncoated TNMG
160408-23 Class H10A (ISO S10) and a support, PTGNR 2020 K16 (Sandvik).
After each step of turning, to the measures of the diameters of the components was
used a Mitutoyo caliper from 0 to 150 mm with a sensitivity of 0.05 mm.
In the test was used a cutting fluid designed to meet the machining operations for
ferrous metals and aluminum alloys, the Tapmatic - type ME-1. This fluid consists of a
concentrated microemulsion of 10% in water. It contains in its composition a mixture of
mineral oils, esters, amides boric, surfactants, biocides and defoamers. The fluid was applied
to conventional way (abundant).
To calculate the cutting length was used the Equation 3 (Diniz et al., 2001):

)(
1000.
..
m
f
DLf
Lc

 (3)
where Lf is the length of advance of the operation, equal to 185 mm, D is the diameter
obtained in the pass in question (in mm) and f is the feed rate used (mm/rev).
5.2. Determination of the factors to be investigated
The aim is to evaluate the influence of the following factors on the characteristics of
the machined components, that is, the best arrangement of factors and levels that provide a
greater cutting length (Lc). The levels and the factors are showed on Table 1.
Table 1: Factors and their levels
Factors of Variation
Level A: Cutting Speed (Cs) B: Feed (f) C: Depth of Cut (ap)
1 75 (m/min) 15 (mm/rev) 0,8 (mm)
2 90 (m/min) 18 (mm/rev) 1,6 (mm)
5.2. Selection of the Orthogonal Array
The sequence of experiments with the combination of factors and levels is
determined by an orthogonal array. This array will determine the number of experiments to be
performed, ensuring that all levels of all factors will be tested an equal quantity. The
appropriate array is selected according to the number of factors, levels and resolution
intended.
In this case, for 3 factors at two levels was selected L8 orthogonal array. This array
includes the implementation of eight runs showed on the Table 2.

Table 2: Orthogonal Array L8
Factors
Runs 1 2 3 4 5 6 7
1 1 1 1 1 1 1 1
2 1 1 1 2 2 2 2
3 1 2 2 1 1 2 2
4 1 2 2 2 2 1 1
5 2 1 2 1 2 1 2
6 2 1 2 2 1 2 1
7 2 2 1 1 2 2 1
8 2 2 1 2 1 1 2
Factor column
Interaction column
Figure 2: Linear graphs of the orthogonal array L8 (a) and (b)
Source: Adapted from (Ross, 1996)
6 Results and Discussion
6.1. Taguchi Array and Results
With the array planned and the experiments made, was constructed the Table 3 where
A, B and C are the factors to be studied and Lc is the result of the cutting length for each
configuration of input factors in their respective array for the flank wear set at 0.5 mm.
1
6
4
2
7
5
3
3 5
6
42
1
7
(a) (b)

Table 3: Set of factors for a Taguchi Array L8
Number of columns, position of the factors, their interactions and their levels
1 2 3 4 5 6 7
Runs A B AxB C AxC BxC empty Lc(m)
1 1 1 1 1 1 1 1 1811,60
2 1 1 1 2 2 2 2 227,50
3 1 2 2 1 1 2 2 914,42
4 1 2 2 2 2 1 1 253,11
5 2 1 2 1 2 1 2 255,72
6 2 1 2 2 1 2 1 193,72
7 2 2 1 1 2 2 1 134,32
8 2 2 1 2 1 1 2 124,63
6.2. Calculating the Effects
From the results obtained in the experiment the effects of factors on the response was
calculated with the aid of statistical software Minitab®, as showed on Table 3.
Table 4: Factors Effect on the Response
Effects of factors and their interactions
Level A B C AxB AxC BxC
1 801,7 622,1 779,0 574,5 761,1 611,3
2 177,1 356,6 199,7 404,2 217,7 367,5
6.3. Analysis of Variance
On the Table 5 presents the analysis of variance (ANOVA) of the effects of factors on
the response variable cutting length (Lc).

Table 5: ANOVA to the Mean
Source of Variation SS d.f MS F
A 780150 1 780150 8,24
B 140996 1 140996 1,49
C 671119 1 671119 7,09
A*B 57984 1 57984 0,61
A*C 590632 1 590632 6,24
B*C 118853 1 118853 1,25
Residual Error 94717 1 94717
Total 2454451 7
SS: sum of squares; d.f.: degree of freedom. MS: mean of squares (equal to SS/d.f.); F:
MSeffect/MSerror
In order to obtain the most significant variables, was used the judgment proposed by
Phadke (1989), which considers influent those variables that present F ≥ 2.0. Performed this
analysis, it is observed that the factors A, C and the interaction between AxC are the most
significant variables in the process.
6.4. Influence of factors on the cutting length
With the results on Table 4 and Figure 3, taking into account that the objective is
reaching the largest possible value of cutting length, can be observed the great importance of
the factors A and C and their interaction AxC, being confirmed by the values F on Table 5.

21
800
600
400
200
21 21
21
800
600
400
200
21 21
A
MeanofMeans
B AB
C AC BC
Main Effects Plot for Means
Data Means
Figure 3: Effects of levels of factors
Thus, the optimum condition of the adjust of control factors, deduced from analysis of
the response of cutting length, in order to achieve longer life of the tool is to put all the factors
in its lower levels.
Analyzing the best fit for the cutting speed (factor A) at level 1, that is, working with
lower speeds there is a gain in tool life thus increasing the response variable cutting length.
This best setting is due to the studied components (Nimonic 80A), they have a high heat
resistance as a characteristic, making the increase of cutting speed, increases the friction
between tool and material, and much of that heat is absorbed by the tool making its life
decreased.
For the Factor B is observed on Table 4 which has no great influence on the process,
because there is a good distribution of heat for both levels of feed rate used, so that there isn't
great significance in the variation, what is not observed with Factor C, that with the increasing
of the depth there is no time enough to dissipate heat through the material, causing heating of
the tool and reducing its hardness and consequently the tool’s life.

Analyzing the Figure 4 (a) can be observed that the cutting length decreases while that
there is a increasing in the depth of cut (ap). This behavior is presented in the same way with
the cutting speed (Cs).
Figure 4: Response surface (a), Contour (b) to the cutting length (Lc) depending on
the depth machined and cutting speed.
The Figure 4 (b) presents on a similarly graphs with the contours of that same
arrangement, being the part in dark shade of blue the lowest values of cutting length and the
darker green the highest values. This make an evident conclusion that the higher cutting speed
and Lc, shorter the length of material being machined, thus proving the explanations of the
behavior of factors.
It can be pointed out from the Figure 4 (a) the presence of interaction, not by the
flatness of the response surface, which does not happen on the Figures 5 (a) and 6 (a) where
there is no evidence that there is a presence of interaction between the factors.

Figure 5: Response surface (a), Contour (b) the cutting length (Lc) depending on the
feed rate and cutting speed.
Figure 6: Response surface (a), Contour (b) the cutting length (Lc) depending on the
feed and depth of cut
Comparing the figures 4 (b), 5 (b) and 6 (b) can be proved that the parameters Cs and
ap produce a much greater effect on the Lc once it reaches the greatest variations in Lc and
therefore on the life of the tool, ensuring the analysis of the Table 4.

7. CONCLUSIONS
The use of Taguchi orthogonal array has helped in the choice of factors and
parameters of the test application, defined the sequence of experiments performed and the
properties measured. After the experiments, it was possible to identify the influence of factors
on the characteristics of measures and identify the parameters that gave the best results about
the Lc.
Before these conditions, the maximum response in the cutting length according the
parameters of forward cutting speed, feed rate and depth of cut, is obtained when these three
parameters have their values adjusted in the minimum experimental levels, meaning bigger
productivity for the process, by a greater rate of material removal.
In the present study can be seen that there was a strong influence of the parameters
cutting speed and depth of cut as well as their interaction, so that there was a correlation
between these parameters and the life of the tool, that is, increasing either of these two
parameters, Vc, or ap, there will be a decrease in life of the tool, possibly due to the low
abrasion of the alloy that increases the temperature of the tool, because the alloy has low
thermal conductivity.
Taguchi Method showed to be efficient and suitable for the process improvement,
proposing adjustments that will provide an improvement in the productivity.

REFERENCES
Antony, J. 2006. Taguchi or classical design of experiments: a perspective from a practitioner.
Sensor Review Volume 26 · Number 3, 227–230.
Barrado, E., M. Vega., P. Grande., J. L. Del Valle. 1996. Optimization of a purification
method for metal-containing wastewater by use of a Taguchi experimental design. Water Res.
30, 2309–2314.
Benardos, P. G., G. C. Vosniakos. 2002. Prediction of surface roughness in CNC face milling
using neural networks and Taguchi’s design of experiments. Robotics and Computer
Integrated Manufacturing 18, 343–354.
Besseris, G. J. 2009. Multi-response unreplicated-saturated Taguchi designs and super-
ranking in food formulation improvement. International Journal of Quality, Statistics, and
Reliability, 26 (4), 341-368.
Diniz, A.E., F.C. Marcondes., N.L. Coppini. 2001. Tecnologia da Usinagem dos Materiais.
São Paulo: Artliber (in Portuguese).
Franceschini, G., Macchietto, S. 2008. Model-based design of experiments for parameter
precision: State of the art. Chemical Engineering Science, 63, 4846 – 4872.

Gunasegaram, D. R., D. J. Farnsworth., T. T. Nguyena. 2009. Identification of critical factors
affecting shrinkage porosity in permanent mold casting using numerical simulations based on
design of experiments journal of materials processing technology, 209, 1209–1219.
Kim. J. S., J. W. Kalb. 1996, “Design of Experiment, an Overview and Application
Example,” Medical Device and Diagnostic Industry, Vol. 18, Part 3, pp. 78-88.
Kleijner, J.P.C., S.M. Sanchez, T.W. Lucas, T.M. Cioppa. 2005. State-of-the-Art Review: A
User’s Guide to the Brave New World of Designing Simulation Experiments. Journal on
Computing, v.17, n.3, p.263–289.
Kitagawa, T., A. Kubo., K. Maekawa. 1997. Temperature and wear of cutting tools in high-
speed machining of Inconel 718 and Ti-6Al-6V-2Sn. Wear, v. 202, p. 142 - 148.
López de Lacalle, L.N., J.I. Llorent., J.A. Sánchez. 1998. Improving the cutting Parameters
the machining of Nickel and Titanium alloys, in Annals of the CIRP, v. 47.
Montgomery, D. C. 1997. Design and analysis of experiments. 4 ed. USA: John Wiley&Sons.
Phadke, M.S. 1989. Quality Engineering using Robust Design, Prent ice Hall, New Jersey.
Ribeiro, M. V., M. R. V. Moreira., J. R. Ferreira. 2003. Optimization of titanium alloy (6Al-
4V) machining. Journal of Materials Processing Technology, v 143/144C, p. 453-458.

ROSA , J. L., A. Robin., M. B. Silva., C. A. Baldan., M. P. Peres. 2009. Electrodeposition of
copper on titanium wires: Taguchi experimental design approach. journal of materials
processing technology. 209, 1181–1188.
Ross, P.J., 1996. Taguchi Techniques for Quality Engineering. McGraw-Hill, USA.
Taguchi, G., Konishi, S., 1987. Taguchi Methods: Orthogonal Arrays and Linear Graphs.
American Supplier Institute, USA.
Tzeng, C. J., Y. H. Lin., Y. K. Yang., M. C. Jeng. 2009. Optimization of turning operations
with multiple performance characteristics using the Taguchi method and Grey relational
analysis. journal of materials processing technology, 209, 2753–2759.
Wang, X., Z. J. Da, A. K. Balaji, I. S. Jawahir. 2007. Performance-Based Predictive Models
and Optimization Methods for Turning Operations and Applications: Part 3—Optimum
Cutting Conditions and Selection of Cutting Tools. Journal of Manufacturing Processes. Vol.
9, n 1. 61-74.
Yih-fong, T. 2006. Parameter design optimisation of computerised numerical control turning
tool steels for high dimensional precision and accuracy Materials and Design 27, 665–675.

015 0713

More Related Content

What's hot (18)

Viewers also liked (14)

Similar to 015 0713 (20)

Recently uploaded (20)

015 0713