Multi-criteria optimization and predictive modeling of technological parameters in

turning of Polyoxymethylene (POM C) using desirability function approch

A. Chabbi1, M.A. Yallese1, S. Belhadi1, T. Mabrouki2, F. Girardin3

1
UniversityMay 8, 1954, Structure and Mechanics Laboratory (LMS) Guelma, Algeria.
Email:amel_chabbi@yahoo.fr, yallese.m@gmail.com, belhadi23@yahoo.fr
2
Université de Tunis El Manar, Ecole Nationale d'Ingénieurs de Tunis (ENIT), 1002, Tunis, Tunisie.
Email:Tarek.Mabrouki@yahoo.fr
3
Laboratoire Vibrations Acoustique, INSA-Lyon, 25 bis avenue Jean Capelle, F-69621 Villeurbanne
Cedex, France
Email:francois.girardin@insa-lyon.fr

Abstract

This article focuses on the determination of the optimum cutting conditions leading to
minimum surface roughness, cutting force, cutting power and maximum productivity, in
turning of the POM polymer C, using cemented carbide cutting tool. The optimization is done
by using the response surface methodology technic (desirability function approach), which
resulted in the modeling of considered responses. Furthermore, the analysis of variance
(ANOVA) is used in order to establish the statistical significance of the cutting parameters on
different technological parameters studied. The results revealed that the surface roughness is
strongly influenced by feed rate with a large contribution, followed by cutting depth, whereas,
the cutting speed has no influence. Regarding cutting force, it is found that feed rate and depth
of cut are the most significant terms. The RSM allowed the optimization of the cutting
conditions for minimal surface roughness, cutting force, cutting power and maximal material
removal rate.
Keywords: Polymer POMC, Surface roughness, Cutting forces, MRR, Cutting power,
ANOVA, RSM, Optimization.

Nomenclature

ANOVA Analysis of variance Pc Power (W)

Vc Cutting speed (m/min) DF Degrees of freedom
f Feed rate (mm/rev) MS Adjusted mean squares
ap Depth of cut (mm) α Clearance angle (degree)
MRR Material removal rate γ Rake angle (degree)
SS Sequential sum of squares χr Major cutting edge angle (degree)
Cont.% Contribution ratio (%) RSM Response surface methodology
Fz Tangential force (N) R2 Determination coefficient
Ra Arithmetic mean roughness (μm)

Polymers are increasingly used in industry because of several advantages such as low density,
excellent corrosion resistance, facilitated mass production, low coefficient of friction and the
ability to run quietly and without external lubrication [1, 2]. Among the different types of
polymers, thermoplastics have a difficulty to be cut due to their distinguished characteristics
such as low elastic modulus, the rate of moisture absorption, the high coefficient of thermal
expansion and internal stresses, they present a complex nonlinear mechanical behavior which
is function of structural morphology and external applied factors. Basically, polymers are
classified by taking into account three basic characteristics (density, molecular weight, and
molecular weight distribution) that significantly influence the manufacturing processes and
the end-use properties; this is known as the structure property- relationship [3,4,5,6].

Among the technical thermoplastics we find the polyoxymethylene (POM C) that is known
for fifty years past. It has many applications in the field of mechanics. According to Myriam
Bastard [7], it is usually found in relatively mechanically demanding applications, such as
gears, bodies for pumps, fittings for pneumatic systems, skiing, connection boxes ... etc. The
range of acceptable temperatures a mechanical use of this material spreads to 105 ° C. [8]
The quality of machined surfaces is influenced by various parameters, such as cutting
parameters (cutting speed Vc, feed rate f and depth of cut: ap). [9, 10, 11]
Surface finish is an important parameter in the machining process. In machining of parts,
surface quality is one of the most specified customer requirements. Major indication of
surface quality on machined parts is surface roughness. It has formed an important design
feature in many situations such as parts subjected to fatigue loads, precision fits, fastener
holes, and aesthetic requirements. In addition to tolerance, surface roughness imposes one of
the most critical constraints for the selection of machines and cutting parameters in the
process planning [12]. So, surface roughness is an important measure of the technological
quality of a product and a factor that greatly influences manufacturing cost [13,14]
Several researchers have studied surface roughness, chip formation and tool wear generated
during the machining of polymeric materials. It was concluded that rake angle and transverse
feed rate are significant parameters that determine the surface roughness in polymer
machining. In addition to the above processing variables, the effect of physical and chemical
properties of polymeric materials is much more pronounced during machining operations.
Because of their visco-elastic behavior, it is complicated to reveal the relationship between
machinability and material properties compared to the well established behavior of
polycrystalline steels [15].
M.KADDACHE [10] show, that in general, the surface roughness decreases with the increase
of the hardness of the work piece material for polyethylene, it is much lower than that of the
cutting material, which affects therefore not the life of the cutting tool.
Xiao and Zhang [16] tried to assess the machinability of typical thermoplastic and
thermosetting polymers and understand the effect of their viscous properties on the surface
integrity, chip formation and cutting forces.
V. N. Gait et all [17] made an analysis of the effects of cutting parameters,
cutting speed (v) and the feed rate (f) on aspects of machinability, such that effort
cutting (Fc), surface roughness (Ra), specific cutting force (Ks), and power (p)
during micromachining of polyamide with 30% of fibers (PA66 GF30) with the tool
PCD, the response surface methodology (RSM) was employed, they found that the cutting
force increases with feed rate for a fixed cutting speed.
In addition, the cutting force decreases with increasing cutting speed. The cutting force is
responsive to higher values of advance per turn in relation to lower values of cutting speed.
Secondly, the surface roughness decreases with lower values in advance and higher values of
cutting speed. To reduce the minimum power you have to take lower values possible cutting
speed and feed.
Marin Motoi [18] presented a study on the influence of additions elements
the roughness of the machined surface in the longitudinal stock removal of certain
polyamides. The types of polyamide studied are: PA66, PA66-GF30 and PA66MoS2 (with
PAMoS2). He observed that the polyamide PA66 gives good results in shooting
longitudinal and the further additions of elements that affect polyamide
negatively the surface quality in the case of roughing.
T. MASUZAWA [19] proved that when filming the resulting surface is not perfectly
cylindrical, it includes helical grooves resulting from more or less pointed shape of the cutting
tool and the helical motion tool / work piece. These grooves have a much deeper profile than
the radius of the nose of the tool is low. Furthermore they are all wider than the feed per
revolution is great.
CHEN, Y.K; est all [20] have shown that the geometry of the cutting tool has a significant
impact on the residual stress and can cause vibration during machining. A rounded edge
favors the appearance of compressive stresses. Furthermore the radius of the nose of the tool
has a considerable effect on the integrity of the surface by decreasing the height of the ridges.
Indeed roughness is inversely proportional to the radius. Cutting parameters have significant
effects on the surface quality obtained by shooting. The increase of the cutting speed and
decreasing the advance allow the reduction of the roughness and improve surface quality.
From a theoretical point of view, depth of cut is less important than advance. For a properly
sharpened tool, the roughness is directly related to the amount of feed per rotation.
Finally, the response surface methodology (RSM) continues to be used to develop
optimization models or prediction of the output parameters such as surface roughness, cutting
power, the life of the tool as well as optimizing the cutting regime [21-22].
In the field of machining of metallic materials and alloys, literature is rich and widely
available. By cons very little research is done on the machining of plastics. Compared
literature in this area is less. Therefore, machining plastics need to be studied extensively.
The objective of this study is to correlate the parameters of cutting speed (Vc, ap and f) and
surface roughness (standard criteria) and the tangential cutting force, cutting power and rate
of material removal during the machining of the polyoxymethylene POM C using the RMS.
The development of a second-order model is adopted to predict roughness, cutting force,
cutting power and the rate of abduction of matter. And end modeling which leads to
optimization using the desirability function.

2 Experimental Procedure
2.1 Work piece material, cutting insert and tool holder
The goal of this experimental work was to investigate the effects of cutting parameters on
surface roughness, and to establish a correlation between them.

In order for this cutting speed, feed rate and depth of cut were chosen as process parameters.
In this study, a society of control conventionnal tour "TOS TRENCIN" SN40 model with a
power of 6.6 kilowatts pin was used to perform turning operations on round bar
Polyoxymethylene POM C with 80 mm diameter 400 mm long.
The material of the workpiece used was a polyoxymethylene polymer (POM C) .It has a good
recovery yield; it is insensitive to cracks and has excellent workability properties. It has low
moisture absorption, high hardness, rigidity, good toughness and impact resistance, even at
low temperatures and high resistance to fatigue. All these important characteristics add to the
cost that gives impressive performance. This material is recommended for the production of
mechanical parts (gears, bearings heavily loaded cam wheels etc ...).

The physical properties and the mechanical POMC are shown in Table 1.

All turning operations were made by Carbide SPUN120408 (r=0.8 mm), manufactured by
Seco society, the cutting insert is mounted on a tool holder with the designation PMOCSDPN
2525M12.
 2.1 Measurement setup
The criterion measures of the surface roughness (arithmetic mean roughness Ra) are obtained
instantly after each pass roughing by means of a Mitutoyo Surftest SJ-201 roughness meter. It
consists of a diamond point, with a radius of 5 µm moving linearly on the machined surface.
The length examined is 4 mm with a cutoff of 0.8 mm and the measured values of Ra are
within the range 0.05–40 µm. To prevent errors and recovery for more precision, roughness
measurement was performed directly on the work-piece without dismounting it from the lathe.
The measurements were repeated three times on three generatrices also placed at 120°. The
cutting forces were measured in real time with a Kistler three-component dynamometer model
9257B linked via a multichannel charge amplifier (type 5011B) to high-impedance cable.
Illustration of measured forces and surface roughness is given in Fig. 1.

2.3 Planning of experiments

To study the impact of various cutting parameters (VC, f, ap) on different output parameters
such as Surface Roughness, Cutting Force, material remove rate and power of cut, we have
compiled a factorial design of 3 factors, each factor has 3 levels. The levels of the parameters,
shown in Table 2 are selected in the ranges recommended by the manufacturer of the cutting
tools.

Cutting speeds were chosen less than a maximum one given by the supplier of materials to
prevent the increase of the temperature in the cutting area to keep the state of the workpiece
material and avoid loss of material properties.
The advances were selected supperior at 0.05 to avoid the discharge of matter, and then had
worked in the interval advances generally used for machining polymers. [10]

The statistical treatment of the data was made in three phases. The first phase was concerned
with the ANOVA and the effect of the factors and of their interactions. The second phase, a
(quadratic regression), has allowed to obtain the correlations between the parameters.
Afterwards, the results were optimized.

2.4 Response surface methodology (RSM)

Box and Wilson (1951) have proposed response surface methodology for the optimization of
experiments. According to, Myers and Montgomery (1995) the RSM is an empirical modeling
approach for the determination of relationship between various process parameters and
responses with the various desired criteria and searching the significance of these process
parameters on the coupled response It is a sequential experimentation strategy for building
and optimizing the model [23].

The invoked procedure includes six steps; the first is to define the independent input variables
and the desired output responses; then adopt an experimental design plan.

The third step is to perform a regression analysis with the quadratic model of RSM; also
perform a statical ANOVA for the independent input parameters in order to identify witch
parameters significantly affect the response.

The determination of the situation of the quadratic model of RSM and decide whether the
model of RSM needs screening variables or not; and finally the optimization of conducting
confirmation experiment and verifying the predicted performance characteristics [24, 25].

In this investigation, the relationship between the INPUT consisting of the cutting conditions
(cutting speed, Vc; feed rate, f and depth of cut, ap) and the dependent variable that
corresponds to the resulting effects of interest in the experiment (OUTPUT), designated by W,
and indicative of machinability aspects (surface roughness, (Ra), tangentiel composante of
cutting force (Fz), power cutting (Pc) and material remove rate (MRR), is presented as:

W= ϕ ( Vc , f , ap )

(1)
where W is the desired machinability aspect and φ is the response function. The
approximation of W is proposed by using a quadratic mathematical model, which is suitable
for studying the interaction effects of process parameters on machinability characteristics. In
the present work, the RSM-based second-order mathematical model is given by the following:

k k k
W =x 0 + ∑ y i X i + ∑ y i , j X i X j + ∑ y ii X 2i
i=1 i, j i=1

(2)
Where x0 is the free term of the regression equation, the coefficients y1, y2,..,yk and y11, y22,..ykk
are the linear and the quadratic terms, respectively, while y12, y13,..yk−1 are the interacting
terms.

Statistical data treatments are carried out in two steps. In the first one, the ANOVA and the
effect of each factor and its interactions are determined. To achieve this goal, the response
surface plots are generated considering two parameters at time while the third one is kept
constant. The second step focused on the optimization and modeling aspects using RMS
outputs.

To determine the energy consumed during the machining operation, the cutting power Pc (W)
related to the cutting force Fz is often measured. Another commonly used scale to quantify
the work done is to calculate the MRR. These aspects of the machining such as cutting power
(Pc) and material removal rate (MRR) were calculated with the results obtained by the
tangential force using the following equations:

Fz×Vc
Pc=
60
(3)

MRR = 1000 × Vc × f ×ap (4)

Where Pc is cutting power (W), Fz is the tangential force (N), f is feed rate (mm/rev), ap is
depth of cut (mm), Vc is cutting speed (m/min), and MRR is material removal rate (mm3/min).

3 Results and discussion

Table 3 shows all the values of the response factors, surface roughness (Ra), cutting force
(Fz), power (Pc), and material remove rate (MRR), and was made with the objective of
analyzing the influence of the cutting speed (Vc), feed rate (f), and depth of cut (ap) on the
total variance of the results. These results (Ra, Fz) were obtained as a result of various
combinations of levels of mixing function in cutting parameters the full factorial design L27.
The results of (MRR, Pc) were calculated with the results obtained by the tangential force
using Equations 3 and 4.

3.1 Analysis of variance (ANOVA)

Analysis of variance (ANOVA) is the statistical method used to interpret experimental data.
ANOVA is used for detecting differences in average performance of groups of items tested. It
breaks total variation down into accountable sources. The main objective of ANOVA is to
extract from the results how much variation each factor causes relative to the total variation
observed in the result [26].

The ANOVA table consists of sum of squares and degrees of freedom. The mean square is the
ratio of sum of squares to degrees of freedom and F-ratio is the ratio of mean square to the
mean square of the experimental error. As per ANOVA, the calculated value of Test-F of
developed model should be more than F-table for the model to be adequate for a specified
confidence interval [16].

In ANOVA table, P-value is the probability (ranging from 0 to 1) that the results observed in
a study (or results more extreme) could have occurred by chance.

 If p value > 0.05, the parameter is insignificant;

 If p value < 0.05, the parameter is significant.
The sum of squarens is performed into contributions from the polynomial model and the
experimental value and was calculated by the following equation:

Nn f
N
SS f = ∑ ( y i− y ) 2
Nnf i=1 (5)

N
1
y= ∑ y i
Where N i =1 i is the average of responses, yi is the average response observed in

experiments where the factor f takes its ith level, N is the total number of experiments and

N nf is the level of each factor f.

The mean square is the ratio of sum of squares to degrees of freedom was calculated by the
following equation:

SS i
MS=
DF i (6)

The F-value is the ratio of mean square of regression model to the mean square of the
experimental error was calculated by the following equation:
 MS i
F−value=
MS error (7)

The model is adequate at 95% confidence level whenever the F-calculated value is greater
than the F-tabulated value. When P-values are less than 0.05 (or 95% confidence), the
obtained models are considered to be statistically significant. Thus, it is demonstrated that the
terms adopted in the model have significant effects on the responses.

The determination coefficient (R2), which is defined as the ratio of the explained variation to
the total variation, is a measure of the goodness of fit. In other words, the more R 2 approaches
unity, the better the response model fits the actual data [27].

∑ ( y i− y ) 2
R ²=
∑ ( y i− y )2 (8)

The last but one column of ANOVA tables shows the factor contribution (percentage; Cont
%) on the total variation, indicating the degree of influence on the result

SS f
Cont %= ×100
SST
(9)

The obtained results are analyzed using the statistical analysis software Design-Expert 9.

Tables 4, 5, 6, and 7 illustrate ANOVA results for surface roughness (Ra), cutting force (Fz),
cutting power (Pc), and material remove rate (MRR) respectively, for the 95% confidence
level.. The main purpose was to analyze the influence of the cutting speed (Vc), the feed rate
(f), and the depth of cut (ap) on the total variance of the results.

Surface finish is an important parameter in machining process. Surface roughness has

received serious attentions for many years. It has formulated an important design feature in
many situations such as parts subject to fatigue loads, precision fits, fastener holes and
aesthetic requirements. In addition to tolerances, surface roughness imposes one of the most
critical constraints for selection of machines and cutting parameters in process planning [28].

The value of 'P' to the model of the surface roughness Ra is less than 0.05 (Table 4),
indicating that the model is adequate and that the terms in the model have a significant effect
on the response, which is desirable. The main effect of feed rate factor (f), depth of cut (ap)
are significant model terms with contribution of (67.00% and 15.53%) respectivelly. It can be
seen that the feed rate (f) is the most important factor affecting Ra, this may be explained by
the fact that the increase thereof leads to helical grooves resulting from the shape of the tool
and tool-workpiece helical movement. These grooves are even deeper and wider than the
advance is high. Indeed, the thicknesses of the layer removed by each turn increases with
increasing the speed (f), which increases the amount of heat generated. The plastic
deformation of the workpiece is generally proportional to the amount of heat produced and
promotes the degradation of the surface roughness. It was concluded by [11] that the advance
is an important parameter which determines the surface roughness in the machining of
polymer. Other model terms can be said to be not significant.These furrows are deeper and
broader as the feed rate increases. In the same sense, Leonardo R. Silva et al. [29] found that
the surface roughness increased as feed rate was elevated for both work materials in precision
turning PA66 polyamide, Also M. Vijaya Kini [30] discovred that the feed rate is the main
influencing factor on the roughness, followed by the depth of cut. Feed rate tool nose radius
interaction and cutting speed-tool nose radius interaction have the highest influence on the
surface roughnes in finish turning of polymer pipe.

From the analysis of Table 5 it can be seen that the model is adequate and that the terms in the
model have a significant effect on the response which is desirable. It can be apparent seen that
the feed rate (f), depth of cut (ap) two level interaction effect of feed rate and depth of cut
(f×ap) and product (ap×ap) all have significant effect on the cutting force (Fz). But, the effect
of feed rate is the most significant factor associated for cutting force with 33.11 %. The next
largest factor influencing Fz is the depth of cut. Its contribution is 26,40% to the model. The
product (ap×ap) is less significant, Other model terms can be said to be not significant.

Mohd. Suhail Ansari et al. [31], developed that cutting force is significantly influenced (at a
95% confidence level) by feed rate, depth of cut, while cutting speed and insert radius have a
small influence. The cutting force increases with the increase of feed rate and depth of cut,
respectively, in Turning of Polytetrafluoroethylene polymer.

From Table 6, the value of 'P' for the model of cutting the power is less than 0.05, indicating
that the model is adequate and the terms in the model have a significant effect on the
response, which is desirable, from analysis of the influence, it can be apparently seen that the
feed rate is significant factor (Cont=31,87%) followed by the interaction (f×ap) with 19.29%
of contribution, then depth of cut with contribution of 17.83%. Cutting speed and interactions
(Vc×ap), (Vc×f), Vc², ap² and f² does not have meaning on the cutting power Pc.

The material removal rate, MRR, can be defined as the volume of material removed divided
by the machining time. Another way, MRR is to imagine an "instantaneous" material removal
rate as the the rate at which the cross-section area of material being removed moves through
the workpiece. Table 7 presents ANOVA results for the material removal rate. The model is
adequate and that the terms in the model have a significant effect on the response, which is
desirable. It can be stated that all the imput parameters such as feed rate, depth of cut and
cutting speed are significant, depth of cut and feed rate have almost the same statistical
significance (37.01%) followed by cutting speed (16.55 %). The interactions (Vc×f, Vc×ap
and f×ap) were found to be less significant, but ap², Vc² and f² are not significant.

The same results found by M. Vijaya Kini et al. [30], they concluded that For material
removal rate, the depth of the cut is the main influencing factor on the roughness, followed by
the tool nose radius. Among the interaction effects, cutting speed-tool nose radius has the
highest influence.

To better view the results of the analysis of variance, a Pareto graph is built (Fig. 2). This
figure ranks the cutting parameters and their interactions of their growing influence on the
surface roughness (Ra), cutting force (Fz), power (Pc) and material removal rate (MRR).
Effects are standardized (F-value) for a better comparison. Standardized values in this figure
are obtained by dividing the effect of each factor by the error on the estimated value of the
corresponding factor. The more standardized the effect, the higher the factor considered
influence. If the F-table values are greater than 4.45; the effects are significant. By cons, if the
values of F-table are less than 4.45; the effects are not significant. The confidence interval
chosen is 95%.

3.2 Regression analysis

The relationship between the factors and the performance measures was modeled by quadratic
regression. The regression equations obtained are given below by equations (10), (11), (12)
and (13) with respective coefficients of determination R2 of 90.25%; 95.95% 89.13% and
99.53%. These regression models are useful in predicting the response parameters with
respect to the input control parameters.
Ra= 0,413 + 0,003 Vc - 7,126 f - 0,369 ap - 4,0252e - 006 Vc×Vc+
0,001 Vc×f - 0,0005 Vc×ap + 44,878 f×f + 2,270 f×ap +0,192
(10)
ap×ap

Fz = -3,992 + 0,013 Vc - 436,166 f + 55,625 ap - 2,810e-005

Vc×Vc+0,077 Vc×f - 0,0150 Vc×ap + 553,299 f×f + 222,792 f×ap
– 17, 065 ap×ap
(11)

Pc = 439,799 - 1,217 Vc - 6132,51 f + 327,319 ap + 0,0005 Vc×Vc +

4,436 Vc×f + 0,178 Vc×ap + 7647,31 f×f + 1684,71 f×ap- 143,732
ap×ap (12)

MRR = 147.413-0.320×Vc- 921.333×f- 73.706× ap+2.000× Vc (13)

×f+0.160× Vc × ap+460.666× f × ap+4.005E-019× Vc×Vc
-2.15292E-012× f×f+1.595E-014× ap× ap

The above
models can be used to predict surface roughness Ra, cutting force Fz, cutting power Pc and
material removal rate MRR at the particular design points. The differences between measured
and predicted responses are illustrated in figs 3 (a, b, c and d). These figures indicate that the
quadratic models are capable of representing the system under the given experimental
domain. It can be seen that, the results of the comparison proves that predicted values of
different technological studied parameters are very close to those readings recorded
experimentally.

In order to better understand the interaction effect of variables on response factors, three-
dimensional (3D) plots for the measured responses were created based on the model equations

(Eqs (10) to (13)). Since each model had three variables, one variable was held constant at the
center level for each plot; therefore, a total of four response surface plots were produced for
the responses (figs 4 to 7).
 3.3 Interactions effect on responses (3D plots and contours)

The 3D surface graphs for surface roughness, cutting force, power and material remove rate
are shown in Figs. 4, 5, 6, and 7.

Figures 5a and 5b show that the effect of cutting speed on surface roughness is negligible
compared to the effects of feed rate and depth of cut. Figure 5c shows that the increase of both
feed rate and depth of cut leads to the increase of surface roughness. However, it is noted that
the effect of feed rate is more important, it’s simillary by V. N. Gaitonde et al. [32].

Figures 5a and 5b reveal that the effect of the cutting speed on cutting force Fz is not
significant. Figure 5c illustrates the interaction effect between feed rate and the depth of cut.
In fact, at low depth of cut the influence of feed rate is not significant. Conversely, when large
values of depth of cut are used, the effect of feed rate becomes important. It can be concluded
that the feed rate exhibits maximum influence on cutting force (Table 4), because of the
sheared chip cross section that is getting larger together with the volume of the deformed
polymer. Therefore, the workpiece material becomes more resistant to shearing and there
should be much effort applied to remove the chip. The increase in the feed rate induces a
larger volume of the cut material in the same unit of time, besides establishing a dynamic
effect on the cutting forces. It also leads to corresponding increase in the normal contact stress
at the tool chip interface and in the tool chip contact area. But, the cutting force decreased
with increase in cutting speed, because the temperature increased in the shear plane area,
which resulted in a reduction in the shear strength of the material.

Fig 6 shows the estimated response surface for power (Pc) in relation to the cutting
parameters of depth of cut (ap) and feed rate (f), while the cutting speed (Vc) is kept at the
middle level, as can be seen from this figure, the power tends to increase, considerably with
increase in the feed rate and depth of cut. Also, it can be observed that at lower cutting
parameters, there is a small resistance to cutting tool, and while at the higher cutting
parameters, the work material offers more resistance to cutting tool, thus increasing the
friction. Hence, the cutting force increases due to increase in friction, which in turn increases
the power. When the feed rate or depth of cut are increased with the increase of cutting
velocities, high power was required to deform the material within a short period of time. The
same result was found by V. N. Gaitonde et al. [32].
From the interaction plot of Fig 7, it can be observed that the MRR sharply increases with
cutting speed, feed rate and depth of cut. It should be noted that the maximal material removal
rate occurred for the combination of the three highest values of the parameters (Vc, f and ap).

The diagnostic checking of the model has been carried out using residual analysis and the
results are presented in Fig. 8. In the figures, the color point indicates the value of surface
roughness.

The normal probability plot of Ra (a), Fz (b), Pc (c) and MRR (d) is presented in Fig 9. The
figure revealed that the residuals fall on a straight line implying that the errors are distributed
normally. Figure 9 shows the residuals with respect to the sixty four experimental runs of Ra
(a), Fz (b), Pc (c) and MRR (d). The residuals do not show any obvious pattern and are
distributed in both positive and negative directions. This implies that the model is adequate
and there is no reason to suspect any violation of the independence or constant variance
assumption.

The plots below illustrate a random pattern of residuals on both sides of 0.00 and do not
expose any recognizable patterns. Thus, it implies that there is nothing awesome about the
residuals in Fig 9

4. Optimization of response

In the present study, desirability function optimization of the RSM has been employed for
surface roughness, cutting force, cutting power and material remove rate optimizations. The
optimization module searches for a combination of factor levels that simultaneously satisfy
the requirements placed on each of the responses and factors in an attempt to establish the
appropriate model. During the optimization process, the aim was to find the optimal values of
cutting parameters to minimize the values of surface roughness, and maximize the value of
material remove rate. To resolve this type of parameter design problem, an objective function,
F(x), is defined as follows:

n 1
1/n
( )
D=( d 1×d 2×. . .×d n ) = ∏ d i
i=1
n

(14)

F( x)=−D (15)
Where di is the desirability defined for the ith targeted output and wi is the weighting of di and
n is the number of responses in the measure. For simultaneous optimization each response
must have a low and high value assigned to each goal. On the worksheet, the "Goal" field for
responses must be one of five choices: "none", "maximum", "minimum", "target", or "in
range". Factors will always be included in the optimization, at their design range by default,
or as a maximum, minimum of target goal. The meanings of the goal parameters are:

In the case of searching for a maximum, the desirabilit is rewritten as follows:

di = 0 if response < low value

0  di  1 as response varies from low to high

di = 1 if response > high value

In the case of searching for a minimum, the desirability can be defined by the following
equations:

di = 1 if response < low value

1  di  0 as response varies from low to high

di = 0 if response > high value

We are interested in three optimization approaches that we called the "quality optimization",
"Productivity optimization", and the last is the combination between the two preceeding
optimizations.

The object of the first is to find the minimum of the surface roughness (Ra) using a
combination of different input parameters such as cutting speed (Vc), the depth of cut (ap)
and feed rate f. In the second, we seek the maximum of material remove rate MRR.
And finally the optimization is to do a combination of both productivity and quality, the
purpose of the combination of two types of quality and productivity, is to have a best case
surface of the workpiece machined with a minimum of energy supplied. The factor ranges
defined for each optimization are summarized in Table 8.

As shown in Table 8, three configurations were studied; optimization of quality that is

recommended for better surface quality but with low productivity with desirability of
(0.964).The second is the optimization of productivity, this optimization is to increase
productivity but against one loses part surface quality with desirability of (0.969).
The last optimization is a compromise between surface quality and productivity that we are
interested because it assembles the best surface quality and maximum productivity, we took
the importance degre for each factor =1.

The solutions are presented in Table 9 which is having maximum value of desirability
i.e.0.712, is tabbed. The optimum values of cutting speed, feed rate and depth of cut to
maximize the MRR (177.063 mm3/min) and minimize the Ra (0.969µm) are (Vc=628 m/min,
f=0.08 mm/rev and ap=3 mm). Table 9 summarizes the results for each type of optimization.

In our work are we interested in optimizing combination to win the good quality and better
productivity, and that is the purpose for industry, table 10 shows the results obtained for this
case.

Graphic ramp function for MRR and Ra overall desirability is shown in fig. 10. In this figure
the points in red on the cutting velocity curves, feed rate and cutting depth are defining the
optimal values. The optimal value corresponding response namely MRR and Ra is also
exposed by blue dot on the curves of the above.

The contour graphe is presented in the figure.11, It Presents Values Optimal cutting speed
and feed rate whether that plot shows the increase after the number of revolutions and feed
spindle in increasing the Opportunity Value MRR and Ra.

Graph of desirability is presented in the fig 12, It Presents Values Optimal cutting speed and
feed rate.This parcelle manifeste Que l'augmentation de la suite du Nombre de tours et
animale broche Dans augmentation de la Valeur de l'Opportunité de MRR et Ra.

Figure 13 presents the bar graph of desirability for the cutting conditions and the responses
together with a combined desirability =0.727.

Conclusion :

This study focuses on the determination of the optimum cutting conditions leading to
minimum surface roughness, cutting force, cutting power and maximum productivity, in
turning of the polymer POM C.

From the results discussed above, the flowed conclusions could be drawn:

- Based on the ANOVA analyse for surface roughness Ra, it’s found that the feed rate
(f) is the most important factor affecting Ra.
- Tangential force was highly affected by feed rate (f), depth of cut (ap), two level
interaction effect of feed rate and depth of cut (f×ap) and product (ap×ap) all have
significant effect. But, the effect of feed rate is the most significant factor associated
for cutting force followed by the depth of cut, their contributions are 33,11% and
26,40% respectivelly.
- Feed rate has highest influence on cutting power to perform the machining operation
with a contribution of 31, 87% followed by the interaction (f×ap) with 19.29% and the
depth of cut with contribution of 17.83%.
- Material remove rate is affected by all the imput parameters such as feed rate, depth of
cut and cutting speed, depth of cut and feed rate have almost the same statistical
significance (37.01%) followed by cutting speed (16.55 %).
- Comparison between predicted and measured values of surface roughness Ra, cutting
force Fz, cutting power Pc and material removal rate MRR proves that predicted
values of different technological studied parameters are very close to those readings
recorded experimentally.
- The effect of cutting speed on surface roughness is negligible compared to the effects
of feed rate and depth of cut. The increase of both feed rate and depth of cut leads to
the increase of surface roughness. However, it is noted that the effect of feed rate is
more important.
- The effect of the cutting speed on cutting force Fz is not significant. In fact, at low
depth of cut the influence of feed rate is not significant. Conversely, when large values
of depth of cut are used, the effect of feed rate becomes important.
- Power increases with feed rate for any given value of cutting speed. The lower values
of cutting speed and feed rate is necessary for minimizing the power.
- MRR sharply increases with cutting speed, feed rate and depth of cut. It should be
noted that the maximal material removal rate occurred for the combination of the three
highest values of the parameters (Vc, f and ap).
- In productivity optimization, the best combination of cutting parameters is (Vc=628
m/min, f=0.24 mm/rev and ap=3 mm) leads to the best values of the material removal
rate MRR=438.773 (mm3/min). This corresponds to a value of desirability equals
0.969.
 - To optimizing the quality, the best results for the surface roughness and the tangential
cutting force after the 27 tests carried out are: Ra= 0.440µm ; from the following input
values (Vc=628 m/min, f=0.08 mm/rev and ap=1.44 mm) with a desirability of 0.964.
- In combined optimization ; the values of MRR and Ra found using the imput
parameters (Vc=628 m/min, f=0.08 mm/rev and ap=3 mm) are : (MRR=177.063
mm3/min ; Ra= (0.969µm) with a desirability of 0.712.
- The mathematical models of Ra, Fz, Pc and MRR are very reliable and they represent
an important industrial interest, since they help to make predictions within the range of
the actual experimentation.

References:

1. Q. Jiang, L.C. Zhang, M. Pittolo, (2000), The dependence of surface finish of a

spectacle polymer upon machining conditions, Progress of MachiningTechnology.
Beijing: Aviation IndustryPress, 7–12.

2. N. Dusunceli, O.U. Colak, (2008), The effects of manufacturing techniques on visco-

elastic and viscoplastic behavior of high densitypolyethylene (HDPE), Mater. Design
29 1117–1124
3. M. Alauddin, Plastics and their machining: A review, J.Mater. Proc. Technol. 54
(1995) 40–46
4. P. Gombette, I. Ernoult, Physique des polym`eres: II.Propri´et´es m´ecaniques,
Herman Editeurs, Paris, 2005
5. N. Dusunceli, B. Aydemir, The effects of loading history and manufacturing methods
on the mechanical behavior of high density polyethylene, J. Elastomers Plastics 43
(2011) 451–468
6. K.O. Xiao, L.C. Zhang, The role of viscous deformation in the machining of
polymers, Int. J. Mech. Sci. (Pergamon), 44 (2002) 2317–2336,
www.elsevier.com/locate/ijmecsci
7. Myriam Bestared, (2006), Etude de la durabilité de pièces thermoplastiques.
Application au polyoxyméthylène, thèse de doctorat `a l’E.N.S.A.M.
8. JW.Carr, C. Feger, Ultraprecision machining of polymers, Precision Engineering,
Precision Engineering, Volume 15, Issue 4,(1993) 221–237.
9. A.K. Kobayashi, Machining of plastics, Mc Graw-Hill, New York, 1967.
10. Kaddeche, M., Chaoui, K., &Yallese, M. A. (2012). Cutting parameters effects on the
machining of two high density polyethylene pipes resins. Mechanics &
Industry, 13(05), 307-316
11. Q. Jiang, L.C. Zhang, M. Pittolo, (2000),The dependence of surface finish of a
spectacle polymer upon machining conditions, In: D. Chen et al. Editors, Progress of
Machining Technology, Aviation Industry Press, Beijing, pp. 7–12
12. Minitab User Manual Release 13.2 Making data analysis easier. MINITAB Inc., State
College, PA, (2001).
13. Keresztes, Kalacska, Zsidai, Dobrocsi, Machinability of engineering polymers‖,
Sustainable Construction and Design (2011), pp. 106-114.
14. A.E. Young, J.H. Wilson, The critical rake angle in the machining of plastics, Plastics
& Rubber Processing 39 (1978) 77–84.
15. A.E. Young, J.H. Wilson, The critical rake angle in the machining of plastics, Plastics
& Rubber Processing 39 (1978) 77–84.
16. R. Suresh, S. Basavarajappa, G.L. Samuel, Some studies on hard turning of AISI 4340
steel using multilayer coated carbide tool, Department of Mechanical Engineering,
Canara Engineering College, Mangalore 574 219, Karnataka, India, 2012
17. V.N.GAITONDE, J. P.DAVIM & ALL; Machinability study in microturning of PA66
GF30 polyamide with a PCD tool, journal of composite materials, 2009; 43; P725
18. M.MOŢOI; Studies regarding the influence of the improving elements upon the
roughness of the surface processed of polyamides, technologies in machine building,
2008, ISSN, PP1635-1636

19. T. MASUZAWA; State of the art of micromachining, annals of CIRP, 2000, 49,
PP473–488.
20. Y.K. CHEN, O.P.MODI, A.S. MHAY, A. CHRYSANTHOU, J.M. O’SULLIVAN;
The effect of different metallic counter face materials and different surface treatments
on the wear and friction of polyamide 66 and it’s composite in rolling–sliding contact,
Wear, 2003, 255, PP714–721
21. Syed Altaf Hussain, V. Pandurangadu, K. Palani Kumar, “Cutting power prediction
model for turning of GFRP composites using response surface methodology”,
International Journal of Engineering, Science and Technology, Vol 3, pp.161-171,
(2011).
22. A.J Makadia, J.I. Nanavati, “Optimization of machining parameters for turning
operations based on response surface methodology”, Measurement, 46(4), 2013, pp.
1521-1529.
23. Rajesh Kumar Bhushan. Optimization of cutting parameters for minimizing power
consumption and maximizing tool life during machining of Al alloy SiC particle
composites, Shri Mata Vaishno Devi University, Katra, Jammu, J&K, India. 2012.
24. Gaitonde VN, Karnik SR, Faustino M, Davim JP (2009) Machinability analysis in
turning tungsten-copper composite for application in EDM electrodes. J Refract Met
Hard Mater 27:754–763
25. Aouici, H., Yallese, M.A., Chaoui, K., Mabrouki, T. and Rigal, J.F. (2012) ‘Analysis
of surface roughness and cutting force components in hard turning with CBN tool:
 prediction model and cutting conditions optimization’, Measurement, Vol. 45, No. 3,
pp.344–353.
26. Rutherford, A, 2001. Introducing ANOVA and ANCOVA. A GLM Approach. SAGE
Publications, London, England.
27. Süleyman N, Süleyman Y and Erol T. Optimization of tool geometry parameters for
turning operations based on the response surface methodology. Measurement 2011;
44: 580–587.
28. Feng CX, Xian Feng W. Development of empherical models for surface roughness
prediction in finish turning. Int J Adv Manuf Technol 2002;20(5):348–56.
29. LEONARDO R. SILVA, A. M. ABRA˜ O, J. PAULO DAVIM, The Effect of Tool
Geometry on the Machinability of Polyamide During Precision Turning, Department
of Mechanics, Federal Center for Technological Education of Minas Gerais
-CEFET/MG, Av. Amazonas 5253 Nova Suı´c¸ a, Belo Horizonte MG, CEP: 30.480-
000, Brazil
30. M. Vijaya Kini, A.M. Chincholkar, Effect of machining parameters on surface
roughness and material removal rate in finish turning of +-30 glass fibre reinforced
polymer pipes, Department of Mechanical and Manufacturing Engineering, Manipal
Institute of Technology, Manipal University, Manipal 576 104, India.
31. Mohd. Suhail Ansari, Deepak Sharma and Sagar Nikam, Study of Cutting Forces and
Surface Roughness in Turning of Bronze Filled Polytetrafluoroethylene, Student at
Dept of Material Science and Nano Technology, University of Petroleum and Energy
Studies, Dehradun, Uttrakhand, International Journal of Advanced Mechanical
Engineering. ISSN 2250-3234 Volume 4, Number 2 (2014), pp. 151-160
32. V. N. Gaitonde, S. R. Karnik, L. R. Silva, A. M. Abr, and J. P. Davim, Machinability
Study in MicroTurning of PA66 GF30 Polyamide with a PCD Tool, 1Department of
Industrial and Production Engineering, B. V. B. College of Engineering and
Technology, Hubli, Karnataka, India, Materials and Manufacturing Processes, 24:
1290–1296, 2009

Tables list
Table 1 Physical and mechanical characteristics of POM C
Table 2 Factors Levels
Table 3 Experimental results for Ra, Fz, MRR and Pc.
Table 4 ANOVA result for surface roughness Ra
Table 5 ANOVA result for tangential force Fz
Table 6 ANOVA result for the power Pc
Table 7 Analysis of variance for MRR
Table 8 Goals and factors ranges for optimization of quality, productivity and combined
Table 9 Optimization results.
Table 10 Optimization solutions for combined optimization.
Figures list
Fig. 1. Schematic diagram of the experimental set-up.
Fig. 2. Graphs of Pareto, for effect cutting parameters on: a) surface roughness, b) cutting
force, c) power and d) material removal rate.
Fig. 3. Measured vs. Predicted values of technological parameters: (a) surface roughness, (b)
tangential force, (c) cutting power and c) material remove rate.
Fig. 4. Estimated Response Surface for surface roughness Ra, depending on Vc, f, ap.
Fig. 5. Estimated Response Surface for tangential cutting force Fz depending on Vc, f, ap.
Fig. 6. Estimated Response Surface for power Pc depending on Vc, f, ap
Fig. 7. Estimated Response Surface for MRR, depending on Vc, f, ap.
Fig. 8. Normal probability of residuals for, (a) Ra, (b) Fz, (c) Pc and (d) MRR data.
Fig. 9. Plot of residuals vs. Run numbers for, (a) Ra, (b) Fz, (c) Pc and (d) MRR data.
Fig. 10. The contour graphe of optimization combined.
Fig. 11. Ramp function plot for combined.
Fig. 12. Graph of desirability.
Table. 1 Physical and mechanical characteristics of POM C

Properties Values
Density 1.41 g/cm3
Absorption of moisture 0.2%
Resistense tensile 67 MPa
Elongation at break 30%
Module of elasticity tensile 2800 MPa
Shock resistance 6 KJ/m²
The hardness BALL 150 MPa
melting temperature 165°C
thermal conductivity 0.31 W/ (m*k)
Specific heat capacity 1.5 KJ/ (Kg*K)
linear thermal expansion coefficient 110*10-6 K-1
Deflection temperature under load 110 °C
dielectric constant 3.8

Table 2 Factors Levels

Leve
Vc, m/min f, mm/rev ap, mm
l
 1 310 0.08 1

2 440 0.16 2

3 628 0.24 3

Table 3 Experimental results for Ra, Fz, MRR and Pc.

Machining parameters Response parameters

Vc, f, ap Ra, Fz, MRR, Pc,
N° m/min mm/rev mm µm N mm3/min W
1 314 0.08 1 0,67 22,98 25,12 120,26
2 314 0.08 2 1,22 34,99 50,24 183,11
3 314 0.08 3 1,61 22,38 75,36 117,12
4 314 0.16 1 1,04 14,15 50,24 74,05
5 314 0.16 2 1,3 41,63 100,48 217,86
6 314 0.16 3 2,31 52,45 150,72 274,44
7 314 0.24 1 2,58 9,67 75,36 50,60
8 314 0.24 2 2,47 86,92 150,72 454,88
9 314 0.24 3 4,17 88,5 226,08 463,15
10 440 0.08 1 0,56 19,7 35,2 144,46
11 440 0.08 2 1,13 29,12 70,4 213,54
12 440 0.08 3 0,68 19,92 105,6 146,08
13 440 0.16 1 1,3 20,17 70,4 147,91
14 440 0.16 2 1,69 45,93 140,8 336,82
15 440 0.16 3 2,79 41,16 211,2 30,84
16 440 0.24 1 2,69 16,25 105,6 119,16
17 440 0.24 2 2,51 70,39 211,2 516,19
18 440 0.24 3 3,57 82,09 316,8 601,99
19 628 0.08 1 0,3 10,99 50,24 115,02
20 628 0.08 2 1,19 18,46 100,48 193,21
21 628 0.08 3 0,54 7,09 150,72 74,20
22 628 0.16 1 0,72 14,98 100,48 156,79
23 628 0.16 2 1,2 43,36 200,96 453,83
24 628 0.16 3 2,06 39,93 301,44 417,93
25 628 0.24 1 2,41 10,14 150,72 106,13
26 628 0.24 2 2,27 66,6 301,44 697,08
27 628 0.24 3 3,42 75,07 452,16 785,73
Table 4 ANOVA result for surface roughness Ra.

Source SS DF F-value P-value Cont% Remarks

Model 24.68 9 17.49 < 0.0001 Significant 

A-Vc 0.59 1 3.77 0.0691 2,16% Not singnifcant 

B-f 18.32 1 116.88 < 0.0001 67,00% Significant 

C-ap 4.25 1 27.09 < 0.0001 15,53% Significant 

AB 6.613E-003 1 0.042 0.8397 0.02% Not singnifcant 

AC 0.10 1 0.66 0.4295 0,38% Not singnifcant 

BC 0.40 1 2.53 0.1304 1,45%  Not singnifcant

A^2 0.054 1 0.34 0.5655 0,20% Not singnifcant 

B^2 0.49 1 3.16 0.0935 1,81% Not singnifcant 

C^2 0.22 1 1.41 0.2507 0,81% Not singnifcant 

Residual 2.66 17      

Total 27.34 26      

Table 5 ANOVA result for tangential force Fz

Source SS DF F-value P-value Cont% Remarks

Model 16497.59 9 44.74 < 0.0001 Significant

A-Vc 420.98 1 10.27 0.0052 2,45% Not singnifcant

B-f 5693.50 1 138.96 < 0.0001 33,11% Significant

C-ap 4539.44 1 110.79 < 0.0001 26,40% Significant

AB 11.55 1 0.28 0.6024 0,07% Not singnifcant

AC 67.67 1 1.65 0.2160 0,39% Not singnifcant

BC 3812.05 1 93.04 < 0.0001 22,17% Significant

A^2 2.62 1 0.064 0.8032 0,02% Not singnifcant

B^2 75.24 1 1.84 0.1931 0,44% Not singnifcant

C^2 1747.40 1 42.65 < 0.0001 10,16% Significant

 Residual 696.55 17

Total 17194.14 26

Table 6 ANOVA result for the power Pc

Source SS DF F-value P-value Cont% Remarks

Model 1.007E+006 9 15.49 < 0.0001 Significant

A-Vc 60604.21 1 8.39 0.0100 5,36% Not significant 

3.601E+00
B-f 5 1 49.82 < 0.0001 31,85% Significant 

2.015E+00
C-ap 5 1 27.88 < 0.0001 17,82% Significant 

AB 37741.30 1 5.22 0.0354 3,34% Not significant  

AC 9572.83 1 1.32 0.2657 0,85% Not significant  

2.180E+00
BC 5 1 30.16 < 0.0001 19,28% Significant 

A^2 1011.12 1 0.14 0.7130 0,09% Not significant  

B^2 14372.37 1 1.99 0.1765 1,27% Not significant  

1.240E+00
C^2 5 1 17.15 0.0007 10,97% Not significant  

Residual 1.229E+005 17      

Total 1.130E+006 26      

Table 7 Analysis of variance for MRR

Source SS DF F-value P-value Cont% Remarks

Model 2.732E+005 9 403.71 < 0.0001 Significant

A-Vc 45433.04 1 604.15 < 0.0001 16,55% Significant

1.016E+00
B-f 5 1 1350.67 < 0.0001 37,00% Significant

1.016E+00
C-ap 5 1 1350.67 < 0.0001 37,00% Significant

AB 7670.58 1 102.00 < 0.0001 2,79% Significant

AC 7670.58 1 102.00 < 0.0001 2,79% Significant

 BC 16298.02 1 216.72 < 0.0001 5,94% Significant

A^2 0.000 1 0.000 1.0000 0,00% Not significant

B^2 0.000 1 0.000 1.0000 0,00% Not significant

C^2 0.000 1 0.000 1.0000 0,00% Not significant

Residual 1278.43 17

Total 2.745E+005 26

Table 8 Goals and factors ranges for optimization of quality, productivety and combined

Constraints
Lower Upper Importance
Lower Upper
Name Goal Weigh Weigh Quality Productivity Combined
Limit Limit
t t
Vc (m/min) in range 314 628 1 1 3 3 3
f (mm/rev) in range 0.08 0.24 1 1 3 3 3
ap (mm) in range 1 3 1 1 3 3 3
Ra (µm) Minimize 0.3 4.17 1 1 5 - 1
Fz (N) Minimize 7.09 88.5 1 1 - - 1
MRR (mm3/mn) Maximize 25.12 452.16 1 1 - 5 1
Pc (W) Minimize 30.84 785.73 1 1 - - 1

Table 9. Optimization results.

Optimizatio Desirability
Vc f ap Ra MRR
n

Productivity 628 0.24 3 - 438.773 0.969

Quality 628 0.080 1.445 0.441 - 0.964
Combined 628 0.08 3 0.969 177.063 0.712

Table 10 Optimization solutions for combined optimization.

Solutions
Number Vc f ap Ra Fz MRR Pc Desirability
1 627.999 0.088 3.000 0.969 7.090 177.061 126.095 0.712 Selected
2 628.000 0.088 3.000 0.977 7.423 178.571 128.781 0.712
3 627.999 0.089 3.000 0.986 7.814 180.340 131.943 0.712
4 626.992 0.091 3.000 1.003 8.454 182.607 136.221 0.711
5 627.999 0.087 2.992 0.958 7.091 175.285 125.942 0.711
6 625.058 0.087 3.000 0.974 7.090 175.270 123.497 0.710
7 627.999 0.094 3.000 1.029 9.631 188.490 146.726 0.710
8 627.999 0.084 3.000 0.940 5.813 171.228 115.833 0.710
9 627.998 0.097 3.000 1.052 10.561 192.619 154.340 0.709
10 622.675 0.087 3.000 0.978 7.095 173.850 121.457 0.709

Fig 1 Schematic diagram of the experimental set-up.

 140 12.00%
120 10.00%
100
8.00%
80
6.00%
60
4.00%
40
20 2.00%

0 0.00%
B-f C-ap Model A-Vc B² BC C² AC A² AB
F-value

14.00% F-Value 160

12.00% 140

120
10.00%
100
8.00%
80
6.00%
60
4.00%
40
2.00% 20

0.00% 0
B-f C-ap BC Model C² A-Vc AC B² A² AB

35 14.00%
30 12.00%
25 10.00%
20 8.00%
15 6.00%
10 4.00%
5 2.00%
0 0.00%
BC C² Régression B-f AB C-ap B² AC A-Vc A²
F-Value
 250 14.00%

12.00%
200
10.00%
150 8.00%

100 6.00%

4.00%
50
2.00%

0 0.00%
BC AB AC C-ap B-f A-Vc A² B² C²
F-Value

Fig 3 Graphs of Pareto, for effect cutting parameters on: a) surface roughness, b) cutting
force, c) power and d) material removal rate.

4.5 Ra mesured
4 Ra Predicted
3.5
3
2.5
Ra, µm

2
1.5
1
0.5
0
1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627
Experimental run order

100
90
80
70
60
50
Fz, N

40
30
20
10
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27
Experimental run order
 900 Pc
800 Mesure
700 d
600
500
Pc, W
400
300
200
100
0
1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627
Experimental run order

500
450 MRR
Mesured
400
350
MRR, mm3/min

300
250
200
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Experimental run order

Fig 3 Measured vs. Predicted values of technological parameters: (a) surface roughness, (b)
tangential force, (c) cutting power and c) material remove rate.
 a b c

Fig 4 Estimated Response Surface for surface roughness Ra, depending on Vc, f, ap.

a b c

Fig 5 Estimated Response Surface for tangential cutting force Fz depending on Vc, f, ap.

a b c
Fig 6 Estimated Response Surface for power Pc depending on Vc, f, ap

a b
c

Fig 7 Estimated Response Surface for MRR, depending on Vc, f, ap.

De sign-E xpert® Softw are
Ra
Co lorp intsby value of
Ra :4
.17
0.3

Desig n-Exp ert®S oftwa re

Fz
Fz:Colr point sbyv alueo f
8 .5
7. 09

Normal Plot of Residuals Normal Plot of Residuals

99
99

95
95
90
90
80
80
70
Normal %Probability

Normal %Probability
70

50
50

30 30
20 20
10 10
5 5

1 1

-3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00

Externally Studentized Residuals Externally Studentized Residuals

a b
MDesigR n-Exp ert®S oftwa re

MRColr :points byv alueo f

45 2.16
25 .12

PDec sign- Expert ®Sof tware

Co lorp intsby valu eof
Pc :785.7 3

30.84
Normal Plot of Residuals
Normal Plot of Residuals
99

99 95
90
95
90 80
70

Normal %Probability
80
Normal %Probability

70 50

50
30
30 20
20
10
10 5
5

1
1

-3.00 -2.00 -1.00 0.00 1.00 2.00 3.00

-6.00 -4.00 -2.00 0.00 2.00

Externally Studentized Residuals Externally Studentized Residuals

c d

Fig 9 Normal probability of residuals for, (a) Ra, (b) Fz, (c) Pc and (d) MRR data.

De sign- Expert ®Sof tware

Ra
Co lorp intsb yvalu eof
Ra :4.17
0.3

FzDe sign- Expert ®Sof tware

Co lorp intsb yvalu eof
Fz :88.5

7.09

Residuals vs. Run

Residuals vs. Run
4.00
6.00

2.00 4.00
Externally Studentized Residuals

Externally Studentized Residuals

2.00

0.00

0.00

-2.00

-2.00

-4.00 -4.00

1 6 11 16 21 26 1 6 11 16 21 26

Run Number Run Number

a b
 De sign- Expert ®Sof tware
PDec sign- Expert ®Sof tware MR R
Co lorp intsb yvalu eof
PCoc :7lorp intsb yvalu eof MR R:
85.7 3 452.1 6
30.84 25.1

Residuals vs. Run Residuals vs. Run

4.00 4.00

2.00
2.00

Externally Studentized Residuals

Externally Studentized Residuals

0.00

0.00

-2.00

-2.00

-4.00

-4.00
-6.00

1 6 11 16 21 26
1 6 11 16 21 26

Run Number
Run Number

c
d

Fig 10 Plot of residuals vs. Run numbers for, (a) Ra, (b) Fz, (c) Pc and (d) MRR data.
Des
ig n-Exp ert®So
f tware
Fa
cto rCoding: Actual
Des
i rabil ty
DesignPoints
1
0

1 =A:Vc
X
X=B:f
2

ActualFactor
C:ap=3

Desirability
0.24
0.2
B : f (m m /re v )

0.4
0.2

0.16

0.6

0.12

Prediction 0.712111

0.08
314 392.5 471 549.5 628

A: Vc (m/min)

Fig. 11 The contour graphe of optimization combined.

 Fig 12 Ramp function plot for combined

Desirability

A:Vc 1

B:f 1

C:ap 1

Ra 0.827118

Fz 1

Pc 0.873821

MRR 0.355796

Com bined 0.727505

0.000 0. 250 0. 500 0. 750 1. 000

Fig 13 Graph of desirability.