Industrial Lubrication and Tribology

Prediction of specific grinding forces and surface roughness in machining of AL6061-T6 alloy using ANFIS
technique
Yusuf S. Dambatta, Mohd Sayuti, Ahmed A.D. Sarhan, Hamdi Bin Ab Shukor, Nur Aqilah binti Derahman, Sunusi Marwana
Manladan,
Article information:
To cite this document:
Yusuf S. Dambatta, Mohd Sayuti, Ahmed A.D. Sarhan, Hamdi Bin Ab Shukor, Nur Aqilah binti Derahman, Sunusi Marwana
Manladan, (2019) "Prediction of specific grinding forces and surface roughness in machining of AL6061-T6 alloy using ANFIS
technique", Industrial Lubrication and Tribology, Vol. 71 Issue: 2, pp.309-317, https://doi.org/10.1108/ILT-03-2018-0098
Permanent link to this document:
https://doi.org/10.1108/ILT-03-2018-0098
Downloaded on: 15 April 2019, At: 00:52 (PT)
References: this document contains references to 19 other documents.
To copy this document: [email protected]
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

The fulltext of this document has been downloaded 42 times since 2019*
Users who downloaded this article also downloaded:
(2019),"Sustainable machining. Modeling and optimization of temperature and surface roughness in the milling of AISI D2
steel", Industrial Lubrication and Tribology, Vol. 71 Iss 2 pp. 267-277 <a href="https://doi.org/10.1108/ILT-11-2017-0322">https://
doi.org/10.1108/ILT-11-2017-0322</a>
(2019),"Tribological performance of commercial Mobil grade lubricants operating with titanium dioxide nanoparticle additives",
Industrial Lubrication and Tribology, Vol. 71 Iss 2 pp. 188-198 <a href="https://doi.org/10.1108/ILT-04-2018-0147">https://
doi.org/10.1108/ILT-04-2018-0147</a>

Access to this document was granted through an Emerald subscription provided by emerald-srm:376953 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please visit
www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics
(COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

 Prediction of specific grinding forces and
surface roughness in machining of AL6061-T6
alloy using ANFIS technique
Yusuf S. Dambatta
Department of Mechanical Engineering, University of Malaya, Kuala Lumpur, Malaysia
Mohd Sayuti
Department of Engineering Design and Manufacturing, University of Malaya, Kuala Lumpur, Malaysia
Ahmed A.D. Sarhan
King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia
Hamdi Bin Ab Shukor and Nur Aqilah binti Derahman
Department of Mechanical Engineering, University of Malaya, Kuala Lumpur, Malaysia, and
Sunusi Marwana Manladan
Department of Mechanical Engineering, Bayero University, Kano, Nigeria
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

Abstract
Purpose – Optimisation of grinding processes involves enhancing the surface quality and reducing the cost of manufacturing through
reduction of power consumptions. Recent research works have indicated the minimum quantity lubrication (MQL) system is used to achieve
near dry machining of alloys and hard materials. This study aims to provide an experimental analysis of the grinding process during machining
of aluminium alloy (Al6061-T6). MQL nanofluid was used as the lubricant for the grinding operations. The lubricant was formed by suspending
silicon dioxide nanoparticles in canola vegetable oil. The effect of input parameters (i.e. nanoparticle concentration, depth of cut, air pressure
and feed rate) on the grinding forces and surface quality was studied. Adaptive neuro-fuzzy inference system (ANFIS) prediction modelling was
used to predict the specific normal force, specific tangential force and surface quality, the ANFIS models were found to have prediction
accuracies of 97.4, 96.6 and 98.5 per cent, respectively. Further study shows that both the specific grinding forces and surface roughness are
inversely proportional to the nanofluid concentration. Also, the depth of cut and table feed rate were found to have a directly proportional
relationship with both the grinding forces and surface roughness. Moreover, higher MQL air pressure was found to offer better delivery of the
atomised nanofluid into the grinding region.
Design/methodology/approach – Grinding experiments were performed using MQL nanofluid as the lubricant. The lubricant was formed by
suspending silicon dioxide nanoparticles in canola vegetable oil. The effect of input parameters (i.e. nanoparticle concentration, depth of cut, air
pressure and feed rate) on the grinding forces and surface quality has been studied.
Findings – The grinding process parameters were optimised using Taguchi S/N ratio analysis, whereas the prediction of the response
parameters was done using ANFIS modelling technique. The developed ANFIS models for predicting the specific normal force, specific
tangential force and surface quality were found to have prediction accuracies of 97.4, 96.6 and 98.5 per cent, respectively. Further findings
show that both the specific grinding forces and surface roughness are inversely proportional to the percentage of nanoparticle concentration
in the lubricant. Also, the depth of cut and table feed rate were found to exhibit a direct proportional relationship with both the grinding
forces and surface roughness, while high MQL air pressure was observed to offer more efficient delivery of the atomised nanofluid into the
grinding region.
Practical implications – The work can applied into manufacturing industries to prevent unnecessary trials and material wastages.
Originality/value – The purpose of this study is to develop an artificial intelligent model for predicting the outcomes of MQL grinding of the
aluminium alloy material using ANFIS modelling technique.
Keywords Surface roughness, Grinding force, Silicon dioxide nanoparticle, AL6061-T6 alloy, Adaptive neuro fuzzy inference system, (ANFIS)
Paper type Research paper

The current issue and full text archive of this journal is available on This work was supported by the University of Malaya Postgraduate
Emerald Insight at: www.emeraldinsight.com/0036-8792.htm Research Grant: grant no.: RP039A-15AET.

Received 7 January 2018

Industrial Lubrication and Tribology
Revised 6 March 2018
71/2 (2019) 309–317 17 May 2018
© Emerald Publishing Limited [ISSN 0036-8792] 10 July 2018
[DOI 10.1108/ILT-03-2018-0098] Accepted 14 August 2018

309
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

1. Introduction In this work, the conventional grinding operation was conducted

to investigate the performance of SiO2 nanofluid MQL technique
Recently, there has been a widespread application of
during grinding of tempered grade Aluminium 6061-T6 alloy.
aluminium (Al) alloys in manufacturing industries such as
Silicon carbide (SiC) grinding wheel was used to perform the
aviation and automobile industries. The Al alloys are used in
grinding experiments due to its high rate of self-sharpening
these fields due to their low weight, reasonable cost and
capability. In addition, due to the low hardness of the carbide
desirable mechanical properties (Ooi et al., 2015; Manladan
wheels compared to diamond and CBN wheels, the carbide
et al., 2016). Al alloys also have low density, high corrosion
wheels are less prone to chatter during the grinding. This makes
resistance, high thermal and electrical conductivity, good
the SiC wheel to be a better choice for machining highly ductile
crashworthiness, and highly ductile. Their lightweight makes
materials such as aluminium alloys. Finally, an ANFIS prediction
the aluminium alloys good candidates in achieving reduced fuel
consumption in the automotive industries. This is evident as
shown in previous works that a 10 per cent reduction in the Table II Taguchi array for sixteen experimental conditions
total mass of the auto-vehicle results in about 6 per cent Parameters
reduction of fuel consumptions, without any decrease in the Experiment run A B C D
vehicles’ reliability (Kim and Wallington, 2013). Hence,
significant weight reduction is highly encouraged during 1 1 1 1 1
2 1 2 2 2
manufacturing of auto-vehicles so as to achieve improved
3 1 3 3 3
efficiency, and also to reduce the emission of greenhouse gases
4 1 4 4 4
(Manladan et al., 2016).
5 2 1 2 3
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

AL alloys have better machinability than other similar metallic

6 2 2 1 4
alloys. However, due to their highly ductile nature, there seem to
7 2 3 4 1
be a resultant increase in friction during the machining
8 2 4 3 2
operations, which has been found to cause excessive wheel
9 3 1 3 4
loading and chatter, that are detrimental to the tool (Nouari et al.,
10 3 2 4 3
2003). Previous works have shown that the lubrication system
11 3 3 1 2
used during machining significantly affects the tool life, surface
12 3 4 2 1
quality, and the forces (Ooi et al., 2015; Balan et al., 2013).
13 4 1 4 2
The minimum quantity lubrication (MQL) system has been
14 4 2 3 1
used successfully in grinding of hard materials. The MQL
15 4 3 2 4
system in addition to being economical provides better 16 4 4 1 3
lubrication compared to the conventional flood cooling systems
(Kalita et al., 2012). Furthermore, the MQL nanolubrication
process was found to produce thin continuous tribofilms on the
surface of the workpiece and grinding wheels, thus increasing Figure 1 Experimental set-up
anti-wear abilities (Silva et al., 2013; Peng et al., 2010).
Recently, there have been several studies conducted on applying
artificial intelligence (AI) techniques in modelling and prediction
of different performances in machining processes. The main AI
techniques used include artificial neural network (ANN), fuzzy
logic or their hybrid, i.e. adaptive neuro-fuzzy inference system
(ANFIS) (Sparham et al., 2017). The AI techniques were found to
be highly accurate and adaptive in predicting the machining
responses. In addition, due to the learning and adaptive ability of
the ANFIS, it was confirmed to outperform the predictability of
the regression models, ANN and fuzzy logic [Kar et al., 2014;
Yilmaz and Kaynar, 2011). In addition, the review of works of
literature have shown that there are no works conducted on the
prediction of the performance of MQL grinding of the aluminium
alloy material using ANFIS modeling.

Table I Experimental parameters (factors and levels)

Levels
S/n Parameters Symbol 1 2 3 4
1 MQL Nanoparticle concentration (Wt.%) A 0 2 4 6
2 Depth of cut ap ( m m) B 3 6 9 12
3 Feed rate (m/min) C 10 15 18 23
4 Air pressure (Bar) D 2.45 4.9 7.36 9.7

310
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

model was developed to predict and analyse the variation of the 4. Experimental results
input parameters with each output parameters.
The results and the deviations for the specific normal and
tangential grinding forces obtained after each experimental run
2. Design of experiments are given in Figure 2(a)-(c). In addition, the surface roughness,
In this research, an experimental investigation was conducted to which was a function of the surface quality was acquired by
find the relationship between the MQL nanolubrication
parameters (i.e. air pressure and percentage concentration of
Figure 2 Experimental results
nanoparticles), and the grinding process parameters (i.e. depth of
cut and feed rate). A summary of the parameters and levels
of each variable are provided in Table I. The concentration of
nanoparticle in the fluid was selected according to the optimised
range given by (Ooi et al., 2015; Molaie et al., 2016). Also, the
feed rate was chosen according to the allowable range of the
surface grinding machine. The combination of process
parameters based on Taguchi L16 experimental design for each
run is given in Table II.

3. Methodology
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

The experiments were performed using the Naga Ichi model

(NI-450AV2) surface grinding machine at a constant rotational
speed of 3000 rpm. As the recommended cutting speed range
for grinding such high-grade alloys is between 20 and 30 m/s,
the cutting speed as obtained by equation (1) was found
suitable for machining the aluminium alloy. The surface
grinding operation was conducted on aerospace AL tempered
grade Al6061-T6 workpiece of ultimate tensile strength
between 260-310MPa and hardness of 105HV on the Vickers
scale. The tool used for the grinding experiment is the Saint-
Gobain Norton’s silicon carbide grinding wheel (25A100-
JVBE). The diameter, thickness and hole-diameter of grinding
wheel are 150 mm, 13 mm and 31.75mm respectively. The
setup for the experiment is shown in Figure 1.

p Dn
Cutting speed Vs ¼ (1)
60000

Where D is the wheel’s diameter and n is its rotational speed

(rpm).
The settings for the MQL equipment was done according to
(Sayuti et al., 2014). The properties of the SiO2 are listed in
Table III. Also, the grinding forces were obtained using the
Kistler three-axis dynamometer (type 9272). The average of
the measured cutting forces in X-axis (tangential force) and Z-
axis (normal force) was obtained after five grinding passes. The
procedure was repeated three times, and the average of
the grinding forces was recorded. This was done to ensure the
accuracy of the acquired forces.

Table III Properties of SiO2

Properties SiO2
Specific heat 1.0J/g-K
Thermal expansion coefficient 5.6  107/K
Melting point 1,600°C
Density 2.2 g/cm3
Dielectric constant 3.9
Dielectric strength 107
Conductivity at 330K (W/cm-K) 0.014

311
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

taking the roughness measurements along the axis of the analysing the Taguchi design are respectively shown in
grinding wheel using the Mitutoyo SJ-210 profilometer with Figure 4(a)-(c).
0.8 mm cut-off length. The surface roughness was taken from From Figure 4(a), it was found that the optimum settings of
three different points on the work material, and the average the grinding process parameters for the specific normal force are
value was recorded. A typical example of the roughness 6 per cent nanoparticle concentration (Level 4), depth of cut of 3
measurement is shown in Figure 3. m m (Level 1), a feed rate of 15 mm/min (Level 2) and air
The Taguchi experiment design process is an important tool pressure of 4.9 Bar (Level 2). Similarly, in Figure 4(b), it could be
that eases the difficulty of experimental work by significantly seen that the optimum parameter settings for the specific
reducing the experimental runs. In other to optimize the tangential grinding force are 6 per cent nanoparticle
results, the objective function of the smaller-is-better [given in concentration (Level 4), depth of cut of 3 m m (Level 1), a feed
equation (2)] was selected. Moreover, the S/N ratio method rate of 10 mm/min (Level 1) and air pressure of 9.7 Bar (Level 4).
was used to obtain the optimised results of the experimental Finally, the optimum process parameters for the surface
runs as done by (Emami et al., 2014; Dambatta et al., 2018), roughness as obtained in Figure 4(c) was 6 per cent nanoparticle
the calculated values of S/N ratios for the specific tangential concentration (Level 4), depth of cut of 3 m m (Level 1), a feed
force (F’t), specific normal grinding force (F’n) and surface rate of 10 mm/min (Level 1) and air pressure of 2.5 Bar (Level 1).
quality are tabulated in Table IV.
   5. Analysis and discussion
1 Xn
S ¼ log X
i¼1 i
2
(2) ANFIS modelling represents a mapped relation between input
n
and output parameters through a hybrid learning process to
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

Where n is the number of runs and Xi is the ith result obtained obtain an optimum distribution of defined membership functions
in an experimental run. (Jang, 1993). The ANFIS model in this work was built on
The plots of S/N ratios for specific normal force, specific MATLAB R2015a software. The developed fuzzy inference
tangential grinding forces and surface roughness obtained by system consists of five different layers of neurons. The factors and

Figure 3 Example of surface roughness measurement

Evaluation Profile
2.5
䣝䣛䢢䢪䤖䣯䢫䣟

2.0
1.5
1.0
0.5
0.0
– 0.5
– 1.0
– 1.5
– 2.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
䣝䣚䢢䢪䣯䣯䢫䣟

Table IV Experimental results and S/N ratio

Measured specific
Exp. Measured specific normal grinding force F 0 N (N/ tangential Measured surface roughness Ra
run mm) S/N Fn grinding force F 0 t (N/mm) S/N Ft ( m m) S/N Ra
1 4.9231 13.8448 3.6923 11.3459 0.444 7.05234
2 6.4616 16.2068 4.8462 13.7080 0.6115 4.27207
3 10.7692 20.6437 8.3077 18.3896 0.658 3.63548
4 12.9231 22.2273 8.6154 18.7055 0.9219 0.70632
5 3.38462 10.5902 3.0769 9.7623 0.453 6.87804
6 6.1539 15.7830 3.3846 10.5901 0.5158 5.75037
7 8.3077 18.3896 7.3846 17.3665 0.902 0.89587
8 7.5385 17.5457 6.6154 16.4111 0.752 2.47564
9 3.2308 -10.1862 2.6923 8.6025 0.4794 6.38604
10 4.3077 12.6849 5.0000 13.9794 0.6846 3.29126
11 4.9231 13.8448 3.2308 10.1862 0.462 6.70716
12 7.0769 16.9969 5.2308 14.3714 0.5258 5.58359
13 2.4616 7.8243 2.6923 8.6025 0.4013 7.93062
14 3.8462 11.7006 2.3077 7.2636 0.3698 8.64066
15 5.2308 14.3714 2.1539 6.6645 0.412 7.70206
16 6.4616 16.2068 2.0000 6.0206 0.5131 5.79596

312
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

Figure 4 Main effects plot of S/N ratios for (a) specific normal force-Fn; (b) specific tangential force-Ft and (c) surface roughness-Ra

Main Effects Plot for SN ratios

Data Means-Specific normal force

Nanofluid concentration Grinding depth

–10

–12

–14

Mean of SN ratios
–16

–18
1 2 3 4 1 2 3 4
Feed rate Air pressure
–10

–12

–14

–16

–18
1 2 3 4 1 2 3 4

Signal-to-noise: Smaller is better

(a)
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

Main Effects Plot for SN ratios

Data Means - Specific Tangential Force
Nanofluid concentration Grinding depth

–8
–10
–12
Mean of SN ratios

–14
–16
1 2 3 4 1 2 3 4
Feed rate Air pressure

–8
–10
–12
–14
–16
1 2 3 4 1 2 3 4
Signal-to-noise: Smaller is better

(b)

Main Effects Plot for SN ratios

Data Means - Surface roughness
Nanofluid concentration Grinding depth

7
6
5
Mean of SN ratios

4
3
1 2 3 4 1 2 3 4
Feed rate Air pressure

7
6
5
4
3
1 2 3 4 1 2 3 4
Signal-to-noise: Smaller is better

(c)

313
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

levels of the Taguchi experimental design were used to train the depth and feed rate. This observation corroborates the finding of
Sugeno-type, fuzzy interface system (FIS) using the results authors’ (Emami et al., 2014).
obtained from the grinding experiment (given in Table IV). The
5.1.2 Effect of minimum quantity lubrication and grinding
steps are taken to train the ANFIS model were similar to that of
parameters on the specific tangential grinding force
author’s (Dambatta et al., 2018; Maher et al., 2014).
Finally, the ANFIS model for the specific grinding forces and Specific tangential force is the force which is purely involved
surface quality were created using 16 rules based on the input with grinding power (plowing, rubbing and chipping actions).
parameter settings and results obtained. The ANFIS model Hence, the lubrication technique used during the machining
was trained using 100 iterations, and the average epoch errors operations is expected to significantly affect the magnitude of
for the specific normal force, specific tangential force, and the tangential force (Rowe, 2013). Previous researchers found
surface roughness are 6.6961  106, 4.9432  106 and that the nanofluid MQL system proffers about 58 per cent
5.9290  107 respectively. reduction to the specific tangential force as compared to the
flood cooling process (Molaie et al., 2016; Dambatta et al.,
5.1 Discussion of adaptive neuro-fuzzy inference system 2017). The relationship between the specific tangential
result grinding force and the nanofluid concentration is shown in
5.1.1 Effect of minimum quantity lubrication and grinding Figure 6(a). It could be seen that the specific tangential force is
parameters on the normal grinding force inversely proportional to the concentration nanofluid. As the
Studies have shown that the normal forces obtained during the specific tangential grinding force contributes significantly to the
grinding process occur due to the asperities on the carbide tool. power consumed during grinding operations, then a higher
From Figure 5(a), it could be seen that an increase in the nanoparticle concentration would produce higher efficiency in
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

concentration of the SiO2 nanoparticle greatly helps to reduce to the grinding process. Moreover, the air pressure was found to
normal grinding forces. This shows that he MQL oil provides have a negligible effect on the specific tangential force as shown
suitable lubricity in the grinding operations. Furthermore, the air in Figure 6(a). At the lower feed rates (10m/min), the specific
pressure of the MQL was found to have little effect on the specific tangential grinding force is between 2-4 N/mm [see
normal grinding forces. In addition, it was found that the specific
normal grinding forces are directly proportional to the grinding Figure 6 Effect of process parameters on the specific tangential
grinding force (N/mm)
Figure 5 Effect of process parameters on the specific normal grinding
force (N/mm)

314
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

Figure 6(b)]. Moreover, an increase in the grinding feed rate microscopy. From the microscopic images obtained, the
produces higher specific tangential grinding force. Similarly, images from experimental run 2 and 14 were compared due to
from Figure 6(b), it could be seen that the grinding depth is the similarity in the process parameters. The average surface
directly proportional to the specific tangential force. roughness obtained from the sample machined in experimental
run 2 was found to be 0.6115 m m. The surface of the
5.1.3 Effect of minimum quantity lubrication and grinding workpiece was found to contain discontinuities and severe
parameters on surface roughness surface damages as shown in Figure 8(a). Whereas in
Figure 7(a) shows the variation of surface roughness with the Figure 8(b), it could be seen that the surface profile of the
air pressure and nanofluid concentration. The results show that sample from experimental run 14 has a smoother machining
an increase in the nanofluid concentration results in improved condition.
surface quality. Also, it can be seen that the suspended SiO2 Generally, as the concentration of the silicon dioxide
nanoparticle have high lubricity. This is because the results nanoparticle in the oil was increased from 0 per cent to 6 per cent,
indicate that when the pure vegetable oil was used as the the surface roughness was found to decrease by about 60 per cent.
nanolubricant, the surface roughness was very poor. However,
with the addition of nanoparticles in the MQL oil, the surface 6. Validation of the adaptive neuro-fuzzy
roughness decreases significantly. Furthermore, the surface inference system model
roughness was found to have a directly proportional
relationship with the feed rate. The higher feed rate was found To validate the accuracy of the ANFIS model, eight
to deteriorate the surface roughness of both the work material confirmation tests were conducted using different input
and grinding wheel. Finally, the surface roughness at a higher parameters within the range of values of the developed model.
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

depth of cut was found to be very poor. The experimental conditions and results are shown in Table V.
The positive lubrication effect provided by the nanofluid was The values of measured and ANFIS predicted responses
also analysed using field emission scanning electron from the verification experiment are presented in Figures 9-11
respectively.
The overall accuracy of the models was calculated using
Figure 7 Effect of process parameters on the surface roughness ( m m) equation (3). In general, the average accuracy of the ANFIS

Figure 8 Surface quality of MQL grinding with machine settings

315
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

Table V Parameter settings for confirmation test and results Figure 10 Comparison of validation results between measured and
Parameters Measured ANFIS predicted ANFIS predicted specific tangential grinding force
S/n A B C D F’N F’T Ra ( m m) F’N F’T Ra ( m m)
1 0 7 10 4.90 6.38 4.9 0.611 6.44 4.83 0.609
2 2 10 15 7.36 8.96 6.86 0.622 9.02 7.1 0.631
3 4 8 23 4.90 4.85 5.1 0.714 4.72 5.2 0.701
4 6 11 23 9.70 8.08 5.2 0.600 8.12 4.59 0.606
5 0 6 10 7.36 5.98 4.1 0.542 6.29 4.05 0.555
6 2 3 18 9.70 3.36 2.70 0.472 3.24 2.72 0.478
7 4 9 15 2.45 6.88 5.31 0.533 7.04 5.2 0.525
8 6 3 10 4.90 3.23 2.92 0.424 3.07 2.79 0.416

Figure 9 Comparison of validation results between measured and

ANFIS predicted specific normal grinding force
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

Figure 11 Comparison of validation results between measured and

ANFIS predicted surface roughness

models for the specific normal grinding force, specific

tangential grinding force and surface roughness was found to be
97.4, 96.6 and 98.5 per cent, respectively.
 
1 XN Vexp:  Vpred:
Accuracy A ¼ 1   100%
n i¼1 Vexp:
(3)
grinding forces and surface roughness. Thus, it could be
n = number of experimental runs. concluded that a higher amount of nanoparticles in the
lubricant produces a more efficient machining operation.
7. Conclusion Furthermore, the ANFIS model developed for predicting
This research involves using statistical analysis and ANFIS specific normal force, specific tangential force, and surface
modelling to analyse, predict and optimize the grinding roughness were found to have prediction accuracy of 97.4
process parameters during grinding of Al6061-T6 alloy. per cent, 96.6 per cent, and 98.5 per cent respectively.
The work was done using MQL nanolubrication with SiO 2
nanoparticles suspended in vegetable oil as the lubricant.
References
The results obtained from the experiments show that the
use of SiO 2 nanoparticles during the grinding operation Balan, A., Vijayaraghavan, L. and Krishnamurthy, R. (2013),
helps to significantly lower the specific grinding forces, and “Minimum quantity lubricated grinding of inconel 751
subsequently, reduce the surface roughness of the alloy”, Materials and Manufacturing Processes, Vol. 28 No. 4,
machined aluminium alloy. The 6 per cent by weight of pp. 430-435.
nanoparticles suspended in oil was found to have the Dambatta, Y.S., Sarhan, A.D.D., Sayuti, M. and Hamdi, M.
optimal performance, having lowest values of specific (2017), “Ultrasonic assisted grinding of advanced materials

316
 Specific grinding forces and surface roughness Industrial Lubrication and Tribology
Yusuf S. Dambatta et al. Volume 71 · Number 2 · 2019 · 309–317

for biomedical and aerospace applications–a review”, The quantity lubrication using oil-based nanofluids”, Journal of
International Journal of Advanced Manufacturing Technology, Cleaner Production, Vol. 129, pp. 212-222.
Vol. 92 Nos 9/12, pp. 1-34. Nouari, M., List, G., Girot, F. and Coupard, D. (2003),
Dambatta, Y.S., Sarhan, A.D.D., Sayuti, M. and Hamdi, M. “Experimental analysis and optimisation of tool wear in dry
(2018), “Comparative study on the performance of the machining of aluminium alloys”, Wear, Vol. 255 Nos 7/12,
MQL nanolubricant and conventional flood lubrication pp. 1359-1368.
techniques during grinding of Si 3 N 4 ceramic”, The Ooi, M.E., Sayuti, M. and Sarhan, A.A. (2015), “Fuzzy logic-
International Journal of Advanced Manufacturing Technology, based approach to investigate the novel uses of nano
Vol. 96 Nos 9/12, pp. 1-18. suspended lubrication in precise machining of aerospace AL
Emami, M., Sadeghi, M.H., Sarhan, A.A.D. and Hasani, F. tempered grade 6061”, Journal of Cleaner Production, Vol. 89,
(2014), “Investigating the minimum quantity lubrication in pp. 286-295.
grinding of Al 2 O 3 engineering ceramic”, Journal of Cleaner Peng, D.X., Chen, C.H., Chang, Y.P. and Chang, S.Y.
Production, Vol. 66, pp. 632-643. (2010), “Size effects of SiO2 nanoparticles as oil additives on
Jang, J.S. (1993), “ANFIS: adaptive-network-based fuzzy tribology of lubricant”, Industrial Lubrication and Tribology,
inference system”, IEEE Transactions on Systems, Man, and
Vol. 62 No. 2, pp. 111-120.
Cybernetics, Vol. 23 No. 3, pp. 665-685.
Rowe, W.B. (2013), Principles of Modern Grinding Technology,
Kalita, P., Malshe, A.P. and Rajurkar, K.P. (2012), “Study of
William Andrew, Elsevier.
tribo-chemical lubricant film formation during application of
Sayuti, M., Erh, O.M., Sarhan, A.A.D. and Hamdi, M.
nanolubricants in minimum quantity lubrication (MQL)
(2014), “Investigation on the morphology of the machined
grinding”, CIRP Annals-Manufacturing Technology, Vol. 61
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)

surface in end milling of aerospace AL6061-T6 for novel

No. 1, pp. 327-330.
Kar, S., Das, S. and Ghosh, P.K. (2014), “Applications of uses of SiO 2 nanolubrication system”, Journal of Cleaner
neuro fuzzy systems: a brief review and future outline”, Production, Vol. 66, pp. 655-663.
Applied Soft Computing, Vol. 15, pp. 243-259. Silva, L.R., Corrêa, E.C.S., Brandão, J.R. and de Ávila, R.F.
Kim, H.C. and Wallington, T.J. (2013), “Life-cycle energy and (2013), “Environmentally friendly manufacturing: behavior
greenhouse gas emission benefits of lightweighting in analysis of minimum quantity of lubricant-MQL in grinding
automobiles: review and harmonization”, Environmental process”, Journal of Cleaner Production, in press.
Science & Technology, Vol. 47 No. 12, pp. 6089-6097. Sparham, M., Sarhan, A.A.D., Mardi, N.A., Hamdi, M. and
Maher, I., Eltaib, M.E.H., Sarhan, A.A.D. and El-Zahry, R.M. Dahari, M. (2017), “ANFIS modeling to predict the friction
(2014), “Investigation of the effect of machining parameters forces in CNC guideways and servomotor currents in the
on the surface quality of machined brass (60/40) in CNC end feed drive system to be employed in lubrication control
milling–ANFIS modeling”, The International Journal of system”, Journal of Manufacturing Processes, Vol. 28 (Part 1),
Advanced Manufacturing Technology, Vol. 74 Nos 1/4, pp. 168-185.
pp. 531-537. Yilmaz, I. and Kaynar, O. (2011), “Multiple regression, ANN
Manladan, S.M., Yusof, R., Ramesh, S., Fadzil, M., Luo, Z. (RBF, MLP) and ANFIS models for prediction of swell
and Ao, S. (2016), “A review on resistance spot welding of potential of clayey soils”, Expert Systems with Applications,
aluminum alloys”, The International Journal of Advanced Vol. 38 No. 5, pp. 5958-5966.
Manufacturing Technology, Vol. 90 Nos 1/4 pp. 1-30.
Molaie, M., Akbari, J. and Movahhedy, M. (2016), Corresponding author
“Ultrasonic assisted grinding process with minimum Mohd Sayuti can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]

317
 This article has been cited by:

1. Athanasios Bogiatzis, Basil Papadopoulos. 2019. Global Image Thresholding Adaptive Neuro-Fuzzy Inference System Trained
with Fuzzy Inclusion and Entropy Measures. Symmetry 11:2, 286. [Crossref]
Downloaded by University of Malaya At 00:52 15 April 2019 (PT)