International Journal of Advanced Engineering Research and

Science (IJAERS)
Peer-Reviewed Journal
ISSN: 2349-6495(P) | 2456-1908(O)
Vol-12, Issue-2; Feb, 2025
Journal Home Page Available: https://ijaers.com/
Article DOI: https://dx.doi.org/10.22161/ijaers.122.3

Predicting the surface roughness and Tolerance using

regression analysis while performing a boring operation in
AA6061 Alloy
Manohar Pala, Shatabdi Biswal, Karumanchi Viswanatha Sarma, Abdul Hameed H,
Mahesh V
U R Rao Satellite Centre, Bangalore-560017
[email protected]

Received: 28 Dec 2024, Abstract— Aluminium alloys are widely used in aerospace applications
Received in revised form: 01 Feb 2025, and AA6061 is one of the popular alloy which is extensively used in
Accepted: 08 Feb 2025 spacecraft mechanical hardware. Some of the mechanical hardware of
spacecraft mechanisms call for stringent tolerances in larger diameter
Available online: 15 Feb 2025
holes. These holes are achieved through boring operation on CNC
©2025 The Author(s). Published by AI
machining centre by utilizing precision boring head and boring bar.
Publication. This is an open-access article Surface roughness and tolerance of the hole plays an important role in
under the CC BY license functioning of the system. In the current work experiments are carried
(https://creativecommons.org/licenses/by/4.0/). out to study the significant input process parameters which influence the
Keywords— Boring, Taguchi, ANOVA, surface roughness and tolerance of holes. This was done using ANOVA.
AA6061 Also, a model was developed based on linear regression analysis. It was
found that optimum cutting parameters predicted by Taguchi method
improved surface finish and tolerance of holes

I. INTRODUCTION Yahya, E. et.al (2016) performed the machining

Spacecraft consists of several mechanical hardware such experimental work on AA6061 alloy using a vertical milling
as mechanisms, thermal subsystems, and electronics machine to study the effect of tool flutes, cutting speed,
packages. From vast experience and available resources, it depth of cut and feed rate on surface roughness and cutting
is evident that to achieve geometric and dimensional force. Their work established a relationship between input
tolerances, complex shapes and surface finishes required by parameters and output parameters using the response
these spacecraft mechanical hardware, machining is the best surface method. The tests were done using the Taguchi
method of manufacturing. Machining is a manufacturing technique and results were analyzed using ANOVA. Their
process where the material is removed by using different work concluded that surface roughness has only two
cutting tools on different machine tools to obtain close significant parameters (tool flutes and depth of cut) which
tolerances and surface finish. affected surface machining. Gutema, E. M. et.al (2022)
explored the implications of cutting parameters like cutting
Aluminium alloys are widely used in aerospace
speed, feed rate and cutting depth; and nose radius on
applications owing to their good specific stiffness and
surface roughness and temperature in the workpiece while
strength. AA6061 is one such alloy which is commonly
turning the AA6061 work piece. Further, desirability
used in several spacecraft mechanical hardware components
analysis optimization was performed to find the optimum
as it can be heat treated and has good corrosion resistance
values. Their work stressed the importance of cutting speed
properties. Also, it is easily machinable using several
as the most influencing parameter compared to other
machining operations.
parameters. Sivaiah, P., & Chakradhar, D. (2018) discussed
on optimum cutting conditions in turning 17-4

www.ijaers.com Page | 23
Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

precipitation-hardened stainless steel using the Taguchi improved cutting conditions and enhanced quality in
optimization method. Input parameters like cooling predicting surface roughness, are mentioned in the work.
environment (cryogenic, MQL and dry) were considered Deshpande, A. A., & Rehman, M. A. A. (2022) reviewed
apart from cutting parameters, to study their effects on the machining process modelling literature related to
surface roughness and flank wear. Aggarwal, A et.al (2008) surface roughness. The effectiveness of different statistical
investigated the effects of cutting parameters, nose radius and mathematical models like RSM, Fuzzy Logic, Artificial
and cutting environment on power consumption of a CNC Neural Networks (ANN) and Support Vector Machines
turning machine while machining AISI P-20 tool steel. (SVM) are discussed in the work.
Their work utilized Response Surface Methodology (RSM) Spacecraft mechanisms like solar array deployment
and Taguchi methodology. Their work discussed that mechanisms, solar array drive mechanisms, dual gimbal
cryogenic working conditions contributed more in reducing antenna, antenna pointing mechanisms etc., play a crucial
the power consumption other than the cutting speed, which role in the success of a spacecraft mission. These
remained the highest influencing parameter in power mechanisms comprise complex shapes, stringent tolerances
consumption. Karkalos, N. E., Efkolidis, N., Kyratsis, P., & and high surface finishes, critical for mechanism
Markopoulos, A. P. (2019) conducted a comparison study functioning. Some of the mechanisms call for the close
for performance of various neural network models like tolerated holes whose dimensional accuracies are in the
Multi-Layer Perceptron (MLP), the Radial Basis Function range of a few microns and surface finish in the range of 0.4
Neural Network (RBF-NN), and the Adaptive Neuro-Fuzzy to 1.6 microns. For larger diameter holes and holes for
Inference System (ANFIS) models with the performance of which standard reamers are unavailable, a boring operation
multiple regression model for drilling experiments on an is known to be the best alternative to achieve these stringent
AA6082-T6 workpiece. The experiment was conducted for requirements. Boring is a subtractive manufacturing
different cutting parameters and also with three cutting tools technique used to enlarge a previously produced hole yet
(solid carbide drilling tools) diameters of 8mm, 10mm and enhance its dimensional accuracy and surface finish. The
12mm. The depth of holes drilled was 30mm. The work process uses a single-point cutting tool to remove material
concluded that the MLP method performed better in all parts from the interior of a workpiece.
cases compared to other methods. However other than
Apart from achieving the stringent tolerances, high
multiple regression models, MLP was observed to be surface finishes and complex shapes of the components,
competitive for smaller data sets. Sastry, M. N. P et.al realizing the hardware in the short lead times to meet the
(2012) investigated the effect of process parameters on project schedules is also a challenging requirement.
MRR using RSM while machining Aluminium alloy and Realizing the hardware with short lead times without
resin using an HSS cutting tool. A close relationship
between observed and predicted values was observed. Do
Duc, T et.al (2020) presented a method of predicting the
surface roughness in the hole-turning operation of 3X13
steel. An experimental matrix was prepared using Central
Composite Design (CCD) and RSM was used to develop a
quadratic polynomial model to predict surface roughness.
Apart from this SVM algorithm was also used and their
study showed SVM to be a better process for predicting
surface roughness. Aamir, M et.al (2021) investigated the
effect of the multi-spindle drilling process on dimensional
hole tolerances, such as hole size, circularity, cylindricity,
and perpendicularity. In addition to this, defects during
drilling operation was also studied. The materials
considered for the study were AA2024, AA6061 and
AA5083. Their work used an uncoated carbide twist drill of
6mm diameter. AA2024 was found to have more
dimensional stability compared to other materials. Spindle
speed is found to influence the most in affecting the hole
size and cylindricity errors. Trinh, V. L. (2024) reviewed
Figure.1 : Boring operation on 3-Axis CNC Vertical
the methods followed in predicting the surface roughness of
machining Centre, DMG MORI D650V
the machining processes. The benefits like reduced cost,

www.ijaers.com Page | 24
Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

compromising the quality requires the maximum utilization as this parameter is the widely used parameter for measuring
of the machine and machining parameters. To achieve the surface roughness in our organization.
correct balance among these a study has to be carried out for 1 𝐿
fixing the ideal machining parameters. Taguchi method was 𝑅𝑎 = ∫ |𝑧(𝑥)|𝑑𝑥 (1)
𝐿 0
found to be the most popular method in Design Of
Experiments and regression analysis to be most suitable if Where
data set is small. To the best of the authors’ knowledge no L = Evaluation Length
work related to boring of AA6061 is done. z(x) = Profile height function
Average of three measurements of Ra (in µm) in each hole,
Hence, in this work prediction of surface roughness
measured by using portable surface roughness tester (as
using regression analysis is carried out while boring the
shown in Fig 2a) was considered. A cutoff length of 0.4 mm
AA6061 material.
for each measurement was considered while taking Ra
II. EXPERIMENTAL METHODOLOGY readings.
Taguchi L9 Design Of Experiments (DOEs) are used to The diameter of the tolerated hole of ∅22+0.021
+0.00 𝑚𝑚 was
optimize parameters for the surface roughness and tolerance measured using 3-point digital bore micrometer by
of the hole on AA 6061 alloy. Depth of cut, Feed and Speed
were the parameters taken in to consideration. Since three
levels and three factors considered, L9 orthogonal array
(OA) is used in this study. Design of experimental (DOE)
has been used for reducing the number of experiments. The
experimental plan having values with units, symbols and
levels are listed in the Table 1.
AA6061 material with dimensions 250 mm × 150 mm ×
25 mm was used for the experimental study for boring of
holes with diameter of 22 mm as per experimental plan and
for the confirmation tests and evaluation of the obtained
models. Nine holes of each three with the diameter 21.6
mm, 21.4 mm and 21.2 mm were machined at feed of 2000
mm/min, speed of 6000 rpm and depth of cut 0.5 mm by
circular pocket milling operation on DMG MORI D650V 3-
Axis CNC vertical machining centre. As per the
experimental plan, each hole was enlarged to the diameter
of 22+0.021
+0.00 𝑚𝑚 by boring operation in two passes with a
precision digital boring head of make Microkom and (b)
(a)
carbide tool insert as shown in Fig.1 and Fig.2. The hole
Figure 2: a). Precision Digital Boring head MicroKom –
diameter of 22 mm with H7 tolerance was selected as the
BluFlex 2 b). Boring bar with Carbide Insert
same dimension is found to repeat in several of spacecraft
mechanical hardware.

Process Levels
Unit Symbol
Parameters 1 2 3
Depth of
mm d 0.1 0.15 0.2
Cut
Feed mm/min f 100 250 500
Speed rpm s 4000 6000 8000
Table 1 Process parameters and their levels
A portable surface roughness tester Surtronic S-128 of
make Taylor-Hobson was utilized to measure Ra (arithmetic
mean surface roughness) value of holes produced by boring
(a) (b)
operation on the basis of the ISO 4287-1997 norms. Eq (1)
defines Ra value. Ra value was considered for observation, Figure 3: a) Portable surface roughness tester Surtronic S-128
b.) 3-Point Digital Bore Micrometers by Mitutoyo

www.ijaers.com Page | 25
 Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

Process parameters Experimental results S/N ratio results

Run d f s Surface roughness Tolerance Tolerance
Ra (dB)
mm mm/min rpm Ra (µm) of the hole (mm) (dB)
1. 0.1 100 4000 0.6 0.037 4.4370 28.6360
2. 0.1 250 6000 0.8 0.092 1.9382 20.7242
3. 0.1 500 8000 1.2 0.094 -1.5836 20.5374
4. 0.15 100 6000 0.4 0.089 7.9588 21.0122
5. 0.15 250 8000 0.3 0.099 10.4576 20.0873
6. 0.15 500 4000 0.8 0.046 1.9382 26.7448
7. 0.2 100 8000 0.3 0.107 10.4576 19.4123
8. 0.2 250 4000 0.4 0.056 7.9588 25.0362
9. 0.2 300 6000 1.0 0.092 0.0000 20.7242

Table 2 Experimental plan, Experimental results and S/N ratios

Mitutoyo (as shown in Fig 2b). The average of three values a. Effect of Process parameters on Ra and
was taken while measuring the tolerance of the hole. tolerance of hole

Fig. 4 depicts the effect of process parameter viz. depth

III. RESULTS AND DISCUSSIONS of cut (DOC) (d) in mm, Feed (f) in mm/min and Speed (s)
in RPM, on Ra. Results show that Ra value decreased with
The S/N ratio is defined as the ratio of mean of readings
increasing depth of cut but slightly raised with further
to the standard deviation of the same and is used to measure
increasing DOC. Ra value showed jump in the value with
the quality characteristic deviating from the desired value
by the Taguchi technique. Taguchi used the term signal for
wanted value i.e mean and noise for unwanted value i.e
standard deviation, which are determined for a response.
Taguchi divided S/N ratio into three categories namely,
Mean of Means

higher-the-better, nominal-the-better and smaller-the-

better. In the present work smaller-the-better Eqn. (2) is
used for Ra and tolerance achieved. The achieved results
and S/N values for Ra and tolerance, are listed in Table 2.
1
S/N ratio for smaller the better = −10 log ∑(𝑅)2 (2)
𝑛
Where, n = No. of observations
R = Observed data for each response
Figure 4: Effect of process parameters on Ra
increasing feed rate. With increasing the cutting speed Ra is
observed to increase but reduces with further rise in the
cutting speed.
Mean of Means

Fig. 5 depicts the effect of process parameters on

tolerance of holes. It is observed that increasing the DOC is
affecting the tolerance of holes. Same but more significant
Effect of speed on tolerance of the holes is noticed, with
sharp decline in hole quality with increase in speed.
However, feed rate seems to affect the hole quality but tend
to improve it with increase in its values.

Figure 5: Effect of process parameters on tolerance of hole

www.ijaers.com Page | 26
Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

Mean S/N ratio

Process Parameters Symbol Rank
Level 1 Level 2 Level 3
Depth of Cut (mm) d 1.5972 6.7849 6.1388 2
Feed (mm/min) f 7.6178 6.7849 0.1182 1
Speed (rpm) s 4.7780 3.2990 6.4438 3
Table 3 Mean S/N ratio response table for Ra

Mean S/N ratio

Process Parameters Symbol Rank
Level 1 Level 2 Level 3
Depth of Cut (mm) d 23.30 22.61 21.72 2
Feed (mm/min) f 23.02 21.95 22.67 3
Speed (rpm) s 26.81 20.82 20.01 1
Table 4 Mean S/N ratio response table for tolerance of holes

b. Selection of optimum cutting conditions for Ra The conformation experiments were conducted at the
and tolerance of hole Taguchi predicted optimum cutting conditions, and the
The obtained S/N ratio response table for Ra is shown in results are tabulated in Table 5 and Table 6.
Table 3 and Fig. 6 depicts the mean S/N ratio graph obtained
in R software. Higher S/N means there is minimum
variation difference between required output and measured
output. It can be seen from Fig 4 that the highest mean value
Mean of S/N ratio

of S/N ratio for Ra is obtained for DOC value of 0.15mm,

federate of 100mm/min and speed value of 8000 RPM. Thus
the predicted optimum cutting parameters for obtaining the
best surface finish i,e least Ra value are DOC = 0.15 mm,
f = 100 mm/min and N = 8000 RPM.
It can be seen from Fig. 7 that the highest mean value of
S/N ratio for tolerance of hole is obtained for DOC value of
0.1 mm, federate of 100 mm/min and speed value of 4000
Figure 6: Mean S/N ratio of Ra
RPM. Thus the predicted optimum cutting parameters for
obtaining the best tolerance value for hole i.e least tolerance
value are DOC = 0.1 mm, f = 100 mm/min and
N = 4000 RPM.
Mean of S/N ratio

c. Conformation Test
To validate the Taguchi predicted optimum conditions
conformation test needs to be conducted.
The predicted S/N ratio is calculated based on the formula
given in Eqn. (3) (Sivaiah, P., & Chakradhar, D. (2018))

𝜀𝑝𝑟𝑒𝑑𝑖𝑐𝑡𝑒𝑑 = 𝜀𝑖 + ∑(𝜀0 − 𝜀𝑖 ) (3) Figure 7: Mean S/N ratio of tolerance of holes

𝑖=1

Where Improvement in S/N ratio for Ra and tolerance of holes

𝜀𝑖 = Total mean S/N ratio is observed at the optimum cutting conditions (d2-f1-s3)
𝜀0 = Mean S/N ratio at an optimal level when compared to the S/N ratio of initial process parameters
n= No. of input process parameters d2-f2-s2. The S/N ratio for Ra at the optimum

www.ijaers.com Page | 27
Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

Initial process Optimal process parameters

parameters Prediction Experiment
Level d2-f2-s2 d2-f1-s3 d2-f1-s3
Surface roughness (µm) 0.59 0.30
S/N ratio (dB) 7.188 11.165 10.45
Improvement in S/N ratio (dB) 45.38 %
% improvement in Ra 49.15 %
Table 5 Conformation test results for Ra

Initial process Optimal process parameters

parameters Prediction Experiment
Level d2-f2-s2 d1-f1-s1 d1-f1-s1
Hole tolerance (mm) 0.14 0.040
S/N ratio (dB) 20.292 28.032 27.958
Improvement in S/N ratio (dB) 34.64 %
% improvement in hole tolerance 71.42 %
Table 6 Conformation test results for tolerance of holes
cutting conditions is found to be 10.45 as against value From the Table 3 it is evident that Ra was mostly
at initial process parameters which is 7.188. Percentage influenced in the order of feed rate, DOC and speed
improvement in Ra while using optimum cutting conditions respectively. Whereas from the Table 4, tolerance of hole
when compared to initial process parameters, is found to be was mostly influenced in the order of speed, DOC and feed
49.15%. Also, for Ra the variation between predicted S/N respectively. From ANOVA analysis it is observed that both
ratio and S/N ratio obtained after conducting the experiment Ra and tolerance of holes are affected greatly by DOC which
at optimum conditions, is found to be 6.40% with respect to ranks 2 for both the parameters as seen in Table 3 and Table
the predicted value. 4 respectively.
The S/N ratio for tolerance of holes at the optimum
cutting conditions(d1-f1-s1) is found to be 27.958 as against V. MODELING
value at initial conditions which is 20.292. Percentage
Since predicting the Ra and tolerance holes for different
improvement in hole tolerance while using optimum cutting
diameters through experiments is practically difficult in
conditions when compared to initial process parameters is
nature, need for a reliable mathematical model for
observed to be 71.42%. Also, for the tolerance of holes, the
predicting Ra and tolerance for holes is required. From
variation between predicted S/N ratio and S/N ratio
literature survey it was noticed that mathematical model
obtained after conducting the experiment at optimum
developed using regression method is found to be reliable
conditions, is found to be 0.26% with respect to the
in predicting the output parameters pertaining to machining
predicted value.
processes and in particular conventional machining
processes.
IV. ANOVA FOR Ra AND TOLERANCE OF HOLES Hence, in the present study most widely used statistical
One of the aims of the experiment conducted is to tool, regression analysis is used to develop a mathematical
determine the significant input parameters which affect the model for output parameters Ra and tolerance of holes,
Ra and tolerance of holes. ANOVA is most widely used based on the dependent input parameters DOC, feed and
statistical tool which determines the significant input speed.
parameters that affect output parameters. Table 3 and Table The predictive equations for Ra and tolerance of holes,
4 present the significance of input process parameters on Ra based on regression analysis are shown in Eqn. (3) and (4)
and tolerance of holes, respectively. The rank column shows respectively
the significance of an input parameter on the output
𝑅𝑎 = 0.98 − 5.47 × 𝑑 + 0.00104 × 𝑓 − 0.00031 × 𝑠
parameter.
+ 0.0033 × 𝑑 × 𝑓 + 0.00026 × 𝑑 × 𝑠
(𝑅2 = 80.25%) (3)

www.ijaers.com Page | 28
Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

𝑇𝑜𝑙𝑒𝑟𝑎𝑛𝑐𝑒 = −0.0462 + 0.261 × 𝑑 − 0.000129 × 𝑓 the tolerance of the holes. Percentage of variation between
+ 0.000024 × 𝑠 + 0.000747 × 𝑑 × 𝑓 predicted and experiment values for Ra was found to be
− 0.000061 × 𝑑 × 𝑠 7.5% and 6.67% for first and second set respectively.
(𝑅2 = 88.72%) (4) Whereas for tolerance of hole, percentage of variation was
The co-efficient of determination (Sivaiah, P., & found to be 10% and 5% for first set and second set
Chakradhar, D. (2018)) R2 was used to check the capability respectively.
of the developed models. The co-efficient of determination
value varies from zero to one, closer to one better is the VI. CONCLUSION
model in predicting the values. In the present study, the
Following are the conclusions drawn from the
developed regression models for Ra and tolerance of holes
experiments conducted
have R2 of 80.25% and 88.72% respectively. To validate the
developed models, conformation tests were conducted and • The predicted optimum cutting parameters for hole
results are mentioned in Table 7 and Table 8. diameter 22+0.021
+0.00 𝑚𝑚, obtaining the best surface
finish i.e least Ra value are d = 0.15 mm, f = 100
For Ra from Table 7 conformation tests for performed
mm/min and s = 8000RPM. The optimum conditions
for run numbers 2, 4 and 8 and error between predicted
are determined using Taguchi method and is
model and experimental values are varying from -7.5% to represented as d2-f1-s3. The Ra value at these
11%. Similarly, for tolerance of holes mentioned Table 8, optimum cutting conditions was found to improve by
the error is varying from -2.17% to -7.86%. 49.15% compared to d2-f2-s2 values. the predicted
optimum cutting parameters for obtaining the best
Ra (µm) tolerance value for hole i.e least tolerance value are
Run Error % d = 0.1 mm, f = 100 mm/min and s = 4000 RPM. The
Experimental Predicted optimum conditions determined is represented as d1-
2 0.8 0.74 -7.5 f1-s1. The tolerance of holes at these optimum
4 0.4 0.37 -7.5 conditions was found to improve by 71.42%
8 0.4 0.44 11 compared to d2-f2-s2 values.
• ANOVA has shown that Ra was mostly influenced
Table 7 Conformation results for the developed
Tolerance of the hole (mm) by feed rate, DOC and speed respectively. Whereas
Run models for surface roughness Error %
tolerance of hole was mostly influenced by speed,
Experimental Predicted
DOC and feed respectively.
4 0.089 0.082 -7.86 • Experiment was further conducted for the two set of
6 0.046 0.045 -2.17 process parameters viz Set 1: d = 0.15 mm, f = 150
mm/min, s = 2000 RPM and Set 2: d = 0.15 mm, f =
9 0.092 0.088 -4.34 280 mm/min, s = 2200 RPM, to conform the
Table 8 Conformation results for the developed equations modelled using linear regression analysis.
models for tolerance of holes It was observed that percentage of variation between
predicted and experiment values for Ra was found to
Further to test the modelled equations Eqn. (3) and (4), be 7.5% and 6.67% for first and second set
experiments were conducted for two set of process respectively. Whereas for tolerance of hole,
parameters as mentioned in Table 9. From the table it is percentage of variation was found to be 10% and 5%
observed that variation is less for Ra and slightly more for for first set and second set respectively.

Process % of variation w.r.t

Diameter of the Ra (µm) Tolerance(mm)
Parameters from predicted
hole (mm)
models Predicted Experiment Predicted Experiment Ra Tol
d = 0.15,
f = 150, 0.4 0.43 0.02 0.018 7.5 10
s = 2000
22
d = 0.15,
f = 280, 0.6 0.56 0.02 0.019 6.67 5
s = 2200
Table 9 Conformation results for the developed model for surface roughness of hole

www.ijaers.com Page | 29
Pala et al. International Journal of Advanced Engineering Research and Science, 12(2)-2025

ACKNOWLEDGEMENTS
Authors would like to acknowledge Director, URSC and
CGM, Facilities for their valuable support and
encouragement in conducting this experiment. Authors
would like to thank Mr. Ajin Surendran S and Mr. Dilip
Kumar, IFF-1/MFF and OLQC for assisting in the
experiment. Authors would also like to thank all the
technical and non-technical staff of URSC and everyone
who were directly or indirectly involved in this endeavor.

REFERENCES
[1] Aamir, M., Tolouei-Rad, M., Giasin, K., Vafadar, A., Koklu,
U., & Keeble, W. (2021). Evaluation of the surface defects and
dimensional tolerances in multi-hole drilling of AA5083,
AA6061, and AA2024. Applied Sciences, 11(9), 4285.
[2] Aggarwal, A., Singh, H., Kumar, P., & Singh, M. (2008).
Optimizing power consumption for CNC turned parts using
response surface methodology and Taguchi's technique—a
comparative analysis. Journal of materials processing
technology, 200(1-3), 373-384.
[3] Deshpande, A. A., & Rehman, M. A. A. (2022). Surface
Roughness Prediction using Empirical Modelling
Techniques:-A Review. SAMRIDDHI: A Journal of Physical
Sciences, Engineering and Technology, 14(01 SPL), 108-117.
[4] Do Duc, T., Ba, N. N., Van, C. N., Nhu, T. N., & Tien, D. H.
(2020). Surface roughness prediction in CNC hole turning of
3X13 steel using support vector machine algorithm, Tribol.
[5] Gutema, E. M., Gopal, M., & Lemu, H. G. (2022).
Minimization of surface roughness and temperature during
turning of aluminum 6061 using response surface
methodology and desirability function analysis. Materials,
15(21), 7638.
[6] Karkalos, N. E., Efkolidis, N., Kyratsis, P., & Markopoulos, A.
P. (2019). A comparative study between regression and neural
networks for modeling Al6082-T6 alloy drilling. Machines,
7(1), 13.
[7] Sastry, M. N. P., Devi, K. D., & Reddy, K. M. (2012). Analysis
and optimization of machining process parameters using
design of experiments. Industrial Engineering Letters, 2(9),
23-32.
[8] Sivaiah, P., & Chakradhar, D. (2018). Modeling and
optimization of sustainable manufacturing process in
machining of 17-4 PH stainless steel. Measurement, 134, 142-
152.
[9] Trinh, V. L. (2024). A Review of the Surface Roughness
Prediction Methods in Finishing Machining. Engineering,
Technology & Applied Science Research, 14(4).
[10] Yahya, E., Ding, G., & Qin, S. (2016). Prediction of
cutting force and surface roughness using Taguchi technique
for aluminum alloy AA6061. Australian Journal of Mechanical
Engineering, 14(3), 151-160.

www.ijaers.com Page | 30