Materials and Design 32 (2011) 22282238

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier.com/locate/matdes

Tribological characteristics of electroless NiB coating and optimization

of coating parameters using Taguchi based grey relational analysis
Suman Kalyan Das, Prasanta Sahoo
Department of Mechanical Engineering, Jadavpur University, Kolkata 700 032, India

a r t i c l e

i n f o

Article history:
Received 27 July 2010
Accepted 11 November 2010
Available online 17 November 2010
Keywords:
A. Non-ferrous metals and alloys
C. Coatings
E. Wear

a b s t r a c t
Electroless nickel coating is an autocatalytic coating whose characteristics are very much dependent on
the composition of electroless bath. The present study is an attempt to minimize the friction and wear
characteristics of electroless NiB coatings simultaneously by optimizing the three coating parameters
viz. bath temperature, concentration of reducing agent and concentration of nickel source together with
the annealing temperature. Taguchi based grey relational analysis is employed for the optimization of
this multiple response problem using an L27 orthogonal array. Analysis of variance reveals that concentration of reducing agent has the maximum contribution in controlling the friction and wear behaviour of
NiB coating. The interaction between bath temperature and nickel source concentration is also found to
possess signicant contribution in controlling the friction and wear characteristics. The surface morphology, composition and phase structure analysis are done with the help of scanning electron microscopy
(SEM), energy dispersive X-ray analysis (EDX) and X-ray diffraction analysis (XRD), respectively. Moreover the wear mechanism is studied and found to be in general abrasive in nature.
2010 Elsevier Ltd. All rights reserved.

1. Introduction
The discovery of the electroless method of coating is mainly
accredited to Brenner and Riddell [1] in the middle of the last century. Since then, the potential of the process has been well recognized especially through a series of research works and now the
process is widely used in various industries like electrical, aerospace, automotive, chemical, electronics, etc. [2]. Electroless nickel
coating is the autocatalytic deposition of nickel based alloys (NiP,
NiB, NiWP, etc.) on the substrate dipped in a solution of metal
ions, complexing agents, reducing agents and stabilizers, operating
in a specic metal ion concentration, temperature and pH ranges.
The advantages of this process include uniform deposition and
the ability to coat even non-conducting materials. Hypophosphite
reduced NiP [35] coatings have already been widely accepted
and superior properties (hardness and wear resistance) of NiB
coatings have already created a stir among the researchers [6
16]. NiB coatings are found to possess superior hardness in asdeposited phase compared to NiP coatings [10]. Again the hardness of NiB coating increases even more with heat treatment
[1013]. The increase in hardness of NiB coating with heat treatment is generally attributed to the modication of deposit structure allowing the precipitation of NiB phases according to the
NiB phase diagram [11]. The NiB phases act as barriers for dislocation movement, thereby increasing the hardness further. As
Corresponding author. Tel./fax: +91 33 2414 6890.
E-mail addresses: psahoo@mech.jdvu.ac.in, psjume@gmail.com (P. Sahoo).
0261-3069/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.matdes.2010.11.028

harder materials generally encounter lesser wear, heat-treated

NiB coatings are found to be more wear resistant than the asdeposited ones [9,10,15]. Use of higher heat treatment temperatures and longer times leads to the progressive hardness decrease,
which can be attributed to the nickel grain growth and to the boride coarsening leading to surface brittleness and enhanced dislocation propagation. Moreover, NiB coating possesses a columnar
structure, which is useful in retaining lubricants under conditions
of adhesive wear [16]. Krishnaveni et al. [9] have found that specic wear rate and coefcient of friction (COF) of electroless NiB
coating increases with increase in applied load under pin-on-disc
arrangement. Heat treatment is also found to reduce the COF of
electroless NiB coatings [9] by presenting a virtually incompatible
surface for the hard counter face material. Scratch test by Delaunois and Lienard [11] point towards the fact, that heat treatment
could also increase the adhesion between the Ni and B deposit
and the substrate. Search of improved tribological properties has
led to the formation of duplex coatings of NiP and NiB [10]
and three component coatings of NiBP [17,18]. Composite coatings with improved wear resistance have been formed by incorporating several particles in NiB coatings, viz. diamond, alumina and
silicon carbide [2].
The tribological characteristics such as friction and wear are
highly dependent on the composition of the material. In case of
electroless NiB coatings, although the composition primarily consists of nickel and boron, the characteristics of the coating is found
to vary depending on the bath composition [6]. Thus, the present
study is primarily devoted to the investigation on the dependence

2229

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

of friction and wear characteristics of electroless NiB coating on

various bath compositions. However, a review of the existing literatures revealed that the study of friction and wear characteristics
of NiB coatings have mainly been conned to and around the effect of heat treatment on the coatings. To the best of the authors
knowledge, scientic approaches to obtain an optimum bath formulation for an enhanced (minimum) friction and wear performance have remained unaddressed. Hence, the present
investigation is formulated into an optimization problem based
on grey analysis together with Taguchi method, so that the optimum bath composition for minimum friction and wear can be predicted and also the inuence of the bath parameters on the
tribological behaviour can be better understood. Moreover the
coating is characterized with the help of SEM, EDX and XRD in order to understand the microstructural characteristics of NiB
coatings.

2. Taguchi method and grey relational theory

Taguchi technique was rst proposed by Genichi Taguchi [19
21]. Taguchi method is a variance reduction technique which can
improve quality of a system at minimum cost. Using an orthogonal array (OA), the number of experiments which may increase
the time and cost can be reduced. A designer must identify the
input, the output, the constraints and the ideal functionality. It
is important to ensure that the resultant functionality resembles
as closely as possible the ideal function. Therefore it is most crucial to develop a means for measuring the deviation between the
actual and the ideal cases. Hence, Taguchi employs signal to
noise (S/N) ratio to measure the performance of the process response. S/N ratio being the ratio of mean to standard deviation
can effectively consider the variation encountered in a set of trials. Again based on the objective of the experiment, S/N ratio
characteristics can be categorized into three criteria: lower-thebetter (LB), higher-the better (HB) and nominal-the best (NB).
The parameter level combination that maximizes the appropriate
S/N ratio is the optimal setting.
The grey system theory was rst proposed by Deng in 1989
[22]. It is similar to fuzzy technique and is an effective mathematical tool to deal with system analysis characterized by imprecise
and incomplete information. The theory is based on the degree of
information known. If the system information is unknown, it is
called a black system; if the information is fully known, it is called
a white system. And a system with information known partially is
called a grey system. Deng [22] had also proposed grey relational
analysis (GRA) in the grey theory that was proved to be an accurate
method for multiple attribute decision making problems. The GRA
method is based on the minimization of maximum distance from
the ideal referential alternative. The aim of GRA is to investigate
the factors that affect the system. The method is based on nding
the relationships of both independent and interrelating data series.
By nding the GRA mathematically, the grey relational grade (GRG)
can be used to evaluate the relational level between referential series and each comparative series. The algorithm of GRA is illustrated
as follows:
(1) Calculation of the grey relational generation in which the set
of experimental results are normalized in between 0 and 1.
(2) Calculation of the grey relational coefcients from the normalized data to represent the correlation between the
desired and actual experimental data.
(3) Calculating the grey relational grade by averaging the grey
relational coefcients. The grey relational grade is treated
as the overall response of the process instead of the multiple
responses of friction and wear.

(4) Performing statistical analysis of variance (ANOVA) for the

input parameters with the grey relational grade and nd
which parameter signicantly affects the process
performance.
(5) Selecting the optimal levels of process parameters.
(6) Conduct conrmation experiment and verify the optimal
process parameters setting.
3. Experimental details
3.1. Preparation of substrate
Steel blocks (AISI 1040) of size 20 mm  20 mm  8 mm are
used as substrates for the deposition of electroless NiB coating.
This particular dimension of the substrate is chosen in accordance
with its counterpart in the multi-tribotester apparatus where the
sample has to be tted for tribological testing.
It is noteworthy here that the tribological characteristics of a
surface may depend on its surface roughness. Now, as electroless
nickel coatings generally follow the surface prole of the substrate,
the prepared substrates in the present study should have similar
surface roughness. Hence, all the substrates before coating are subjected to roughness evaluations (center line average values, Ra) and
the substrates which showed as little as about 0.1% variation in
roughness are selected for electroless NiB coatings. The roughness
measurements are carried out using a surface prolometer (Taylor
Hobson, Surtronic 3+).
3.2. Developing NiB coating
The bath composition is selected after a lot of trials. The selected composition together with the bath condition is shown in
Table 1. Nickel chloride is selected as the nickel source and sodium
borohydride is taken as the reducing agent. Ethylenediamine is the
complexing agent while lead nitrate stabilizes the bath so that
solution does not break down during the coating period. As alkaline NiB baths are generally operated at high pH, the pH of the
present bath is maintained around 12.5. The selected substrates
after proper cleaning are at rst activated by dipping into a palladium chloride solution kept at 55 C. Activated samples are then
submerged into the electroless bath which is maintained at a temperature between 85 and 95 C (according to the OA) with the help
of a hot plate cum stirrer attached with a temperature sensor also
submerged in the solution. The photograph of the electroless setup
is shown in Fig. 1. The coating is carried out in general for a period
of 2 h and the thickness of the coating is found to lie in the range of
2530 lm (Fig. 2). After coating, the samples are subjected to
annealing at different temperatures (250 C, 350 C and 450 C)
as specied in their position in the OA.
There have been several propositions regarding the reaction
mechanism of electroless NiB coatings but one of the mechanisms
proposed by Gorbunova et al. [23] is well supported by experimental evidence. The proposed scheme for the reaction mechanism of

Table 1
Bath constituents and their ranges.
Parameters

Ranges of parameters

Nickel chloride
Sodium borohydride
Ethylenediamine
Lead nitrate
Sodium hydroxide
Bath temperature
pH of solution

1525 g/l
0.61.0 g/l
59 g/l
0.0145 g/l
40 g/l
8595C
12.5

2230

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

3.3. Choice of design parameters

The characteristic of electroless coating is affected by several
factors which include the bath composition as well as the deposited conditions. If all the factors are considered then the experimental design would become very much complicated. Hence, the
three bath parameters viz. bath temperature (A), concentration of
reducing agent (sodium borohydride) (B) and concentration of
nickel source (nickel chloride) (C), which are mainly used by the
researchers to control the properties of electroless nickel coating
are considered as the main design parameters along with their
interactions in the present study. Again, annealing is found to improve the tribological behaviour of electroless nickel coatings.
Hence, the annealing temperature (D) is taken as fourth parameter
in the experimental design to observe its effect on the tribological
behaviour of electroless NiB coatings. All the design factors along
with their levels are shown in Table 2. Three levels are considered
so that nonlinear effects if any can be observed.
3.4. Response variables
Response variables are the outputs of the experimental model. The present study tries to minimize the friction and wear of
electroless NiB coating. Hence, the coefcient of friction (COF)
and wear depth (in microns) are taken as the response variables
in the present case.
3.5. Planning of experiments
Fig. 1. Photograph of electroless setup.

Design of experiments involves the preplanned changes to the

controllable variables (design factors) so that their effects on the
responses can be studied in an organized manner and with the
least number of experimental trials. There are several standard
orthogonal arrays (OA) tabulated by Taguchi which are chosen
based on the number of design factors, their levels and the interactions considered. In the present case, an L27 OA is chosen so that
the individual effects of the chosen coating parameters as well as
their interactions on the process responses (friction and wear)
can be studied. The factors and their interactions are assigned to
the columns of the OA according to the Triangular Table for 3-level
OA [21]. The L27 OA along with the column assignments is shown
in Table 3. The values in each cell of the main parameter columns
(AD) in the array indicate their levels (13). Again in case of interactions, two columns are assigned to a single interaction and the
two cell values in a particular row indicate the levels of each of
the factors involved in the interaction. The unassigned columns
in the OA are kept for the error terms.
Fig. 2. Optical micrograph of cross cut NiB coating.

3.6. Tribological tests

nickel boron plating consists of mainly three steps:
Reduction of nickel:

BH4 4H2 O ! BOH4 4H 4H 4e

2

2Ni

BH4

The tribological tests for the present study are performed in a

multi-tribotester apparatus (TR-25, DUCOM). The tests are carried
out using a block on roller geometry and in dry condition (without
any lubrication). The ambient temperature is about 28 C with a

4e ! 2Ni
2

2Ni

4H2 O ! 2Ni

BOH4

2H2 4H

Table 2
Design parameters and their levels.
Design factors

Unit

Reduction of boron:

BH4

5
H ! BH3 H2 ! B H2
2

Bath temperature (A)

Reducer concentration (B)
Nickel source concentration (C)
Annealing temperature (D)

Hydrolysis of borohydride:

BH4 4H2 O ! BOH4 4H 4H 4e ! BOH4 4H2

Initial condition.

C
(g/l)
(g/l)
C

Levels
1

85
0.6
15
250

90a
0.8a
20a
350a

95
1.0
25
450

2231

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

Table 3
L27 Orthogonal Array with design factors and interactions.
Trial no.

Column numbers
1A

2B

3 AB

4 AB

5C

6 AC

7 AC

8 BC

9D

10

11 B  C

12

13

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

1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3

1
1
1
2
2
2
3
3
3
1
1
1
2
2
2
3
3
3
1
1
1
2
2
2
3
3
3

1
1
1
2
2
2
3
3
3
2
2
2
3
3
3
1
1
1
3
3
3
1
1
1
2
2
2

1
1
1
2
2
2
3
3
3
3
3
3
1
1
1
2
2
2
2
2
2
3
3
3
1
1
1

1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3

1
2
3
1
2
3
1
2
3
2
3
1
2
3
1
2
3
1
3
1
2
3
1
2
3
1
2

1
2
3
1
2
3
1
2
3
3
1
2
3
1
2
3
1
2
2
3
1
2
3
1
2
3
1

1
2
3
2
3
1
3
1
2
1
2
3
2
3
1
3
1
2
1
2
3
2
3
1
3
1
2

1
2
3
2
3
1
3
1
2
2
3
1
3
1
2
1
2
3
3
1
2
1
2
3
2
3
1

1
2
3
2
3
1
3
1
2
3
1
2
1
2
3
2
3
1
2
3
1
3
1
2
1
2
3

1
2
3
3
1
2
2
3
1
1
2
3
3
1
2
2
3
1
1
2
3
3
1
2
2
3
1

1
2
3
3
1
2
2
3
1
2
3
1
1
2
3
3
1
2
3
1
2
2
3
1
1
2
3

1
2
3
3
1
2
2
3
1
3
1
2
2
3
1
1
2
3
2
3
1
1
2
3
3
1
2

relative humidity at 85%. Table 4 lists the parameters with their

values for the tribological test, which are kept constant in case of
all specimens. A schematic diagram of the tribotester apparatus
is given in Fig. 3. The NiB coated specimens are held stationary
with the help of an attachment and made to slide against a rotating
steel roller coated with titanium nitride of hardness 85 HRc. The
speed of the roller and the duration of tests can be controlled via
a computer attached to the tribotester. Loads can be applied by
placing dead weights on the loading pan which is attached to a
loading lever. The loading lever is pivoted near the normal load
sensor which helps in the measurement of effective normal load
applied on the specimen. A beam type load cell (1000 N capacity)
is used by the frictional force sensor to measure the frictional force
which is plotted in real time on the computer screen. Wear is measured in terms of displacement (in microns) with the help of linear
voltage resistance transducer. Wear displacement measured is
actually the sum of the wear at specimen surface and that at the
counter face surface. But as the hardness of the coatings is found
to be around 7072 HRc which is quite lower than the counter face
material (85 HRc), the counter face material encounters negligible
wear compared to the coated specimen. Hence, the measured displacement can be taken as a representative of the actual wear
depth encountered by the coating surface. It is worth noting that,
in general wear is measured in terms of wear volume or mass loss.
But in the present case, wear is expressed in terms of displacement
or wear depth. Hence, to ensure that the wear measurements are
accurate, the displacement results for wear are compared with
the weight loss of the specimens and almost linear relationship

Table 4
Tribological testing parameters and their values.
Test parameter

Value

Normal load (N)

Roller speed (RPM)
Duration (s)

25
60
300

is observed between the two for the range of test parameters considered in the present study.
3.7. Coating characterization
The characterization of the coating is necessary so that it can be
made sure that the coating is properly developed. In the present
study, surface morphology of the coating is observed through scanning electron microscope (SEM) (JEOL, JSM 6360 and FEI Quanta
200) in order to analyse the effect of heat treatment on the microstructure of the deposits. SEM is also done after the wear testing to
see the pattern of the sliding tracks in order to predict the wear
mechanism. Energy dispersive X-ray analysis (EDAX Corporation)
is used to determine the composition of the coating in terms of
the weight percentages of nickel and boron. The different precipitated phases both before and after heat treatment are identied
with the help of X-ray diffraction (XRD) analysis (Rigaku, Ultima
III).
4. Results and discussion
4.1. Grey analysis
The experimental results obtained from friction and wear tests
are included in Table 5. The nal response needed for processing
with Taguchi analysis is the grey relational grade which is obtained
through the following set of calculations:
4.1.1. Grey relational generation
Before calculating the grey relational coefcients (GRCs), these
series data can be dealt with depending on three types of pre-processing including: (1) normalization by maximum value, (2) normalization by minimum value, and (3) normalization by
objective value. This stage is known as grey relational generation.
In the present study as friction and wear are to be minimized,
the pre-processing of normalization by minimum value is adopted.

2232

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

Fig. 3. Sketch of the multi-tribotester.

The processed data after grey relational generation is given in Table 6. Larger normalized results correspond to the better performance and the best normalized result should be equal to 1.

Table 5
Experimental results for friction coefcient and wear
depth.
Exp. no.

COF

Wear (lm)

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

0.184
0.107
0.132
0.314
0.411
0.108
0.300
0.131
0.571
0.093
0.282
0.209
0.083
0.246
0.311
0.368
0.326
0.31
0.207
0.449
0.274
0.409
0.124
0.267
0.364
0.255
0.477

1.87
2.98
12.66
10.97
3.91
2.52
9.1
10.84
5.77
6.29
11.29
7.34
6.614
14.61
11.26
9.627
8.54
6.34
7.74
5.88
4.19
11.46
2.09
8.25
8.57
2.95
2.86

4.1.2. Grey relational coefcient

Grey relational coefcients are calculated to express the relationship between the ideal (best = 1) and the actual experimental
results. The grey relational coefcient ni(k) can be calculated as:

ni k

max yi k  yi k
max yi k  min yi k

where D0i = ||x0(k)  xi(k)|| = difference of the absolute value between x0(k) and xi(k) , Dmin and Dmax are respectively the minimum
and maximum values of the absolute differences (D0i) of all comparing sequences and r is the distinguishing coefcient which is
used to adjust the difference of the relational coefcient, usually
r e {0, 1} [22]. The distinguishing coefcient weakens the effect of
Dmax when it gets too big, enlarging the different signicance of
the relational coefcient. The suggested value of the distinguishing
coefcient, r, is 0.5, due to the moderate distinguishing effects and
good stability of outcomes. Therefore, r is adopted as 0.5 for further
analysis in the present case.
4.1.3. Grey relational grade
In the grey relational analysis, the grey relational grade is used
to show the relationship among the series. The overall multiple response characteristics evaluation is based on grey relational grade
which is calculated as follows:

A linear data pre-processing corresponding to lower-the-better

(LB) criterion can be calculated as:

xi k

Dmin r Dmax
D0i k r Dmax

where xi(k) is the value after grey relational generation while min
yi(k) and max yi(k) are respectively the smallest and largest values
of yi(k) for the kth response; k being 1 (COF) and 2 (wear depth).

ai

n
1X
n k
n k1 i

where n is the number of performance characteristics (2 in present

case). The grey relational grade results are given in Table 7. Higher
the grey relational grade, the closer is the experimental value to the
ideal normalized value. Thus, higher grey relational grade indicates
that the corresponding parameter combination is closer to the
optimal.

2233

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

Table 6
Grey relational analysis for friction coefcient and wear depth.
Exp. no.

Normalized data

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

Values of D0i

Grey relational coefcient

COF

Wear depth

COF

Wear depth

COF

Wear depth

0.7940
0.9508
0.8995
0.5266
0.3278
0.9487
0.5553
0.9016
0.0000
0.9795
0.5922
0.7418
1.0000
0.6659
0.5327
0.4159
0.5020
0.5348
0.7459
0.2500
0.6086
0.3319
0.9159
0.6229
0.4241
0.6475
0.1926

1.0000
0.9128
0.1530
0.2857
0.8398
0.9489
0.4324
0.2959
0.6938
0.6530
0.2605
0.5706
0.6276
0.0000
0.2629
0.3911
0.4764
0.6491
0.5392
0.6852
0.8178
0.2472
0.9827
0.4992
0.4740
0.9152
0.9222

0.2059
0.0491
0.1004
0.4733
0.6721
0.0512
0.4446
0.0983
1.0000
0.0204
0.4077
0.2581
0.0000
0.3340
0.4672
0.5840
0.4979
0.4651
0.2540
0.7500
0.3913
0.6680
0.0840
0.3770
0.5758
0.3524
0.8073

0.0000
0.0871
0.8469
0.7142
0.1601
0.0510
0.5675
0.7040
0.3061
0.3469
0.7394
0.4293
0.3723
1.0000
0.7370
0.6088
0.5235
0.3508
0.4607
0.3147
0.1821
0.7527
0.0172
0.5007
0.5259
0.0847
0.0777

0.7082
0.9104
0.8327
0.5136
0.4265
0.9070
0.5292
0.8356
0.3333
0.9606
0.5507
0.6594
1.0000
0.5995
0.5169
0.4612
0.5010
0.5180
0.6630
0.4000
0.5609
0.4280
0.8561
0.5700
0.4647
0.5865
0.3824

1.000
0.8516
0.3712
0.4117
0.7574
0.9074
0.4683
0.4152
0.6202
0.5903
0.4034
0.5380
0.5731
0.3333
0.4041
0.4509
0.4884
0.5876
0.5204
0.6136
0.7330
0.3991
0.9666
0.4996
0.4873
0.8550
0.8654

4.1.4. Grey relational ordering

In relational analysis, the practical meaning of the numerical
values of grey relational grades between elements is not absolutely
important, while the grey relational ordering between them yields
more subtle information. The combination yielding the highest
grey relational grade is assigned an order of 1 while the combination yielding the minimum grade is assigned the lowest order. The
ordering of the present grey grades is shown in Table 7.

4.2. Analysis of signal to noise ratio

Taguchi method uses S/N ratio to convert the experimental results into a value for the evaluation characteristic in the optimum
parameter analysis. In the present work, S/N ratio analysis is done
with grey relational grade as the performance index. As grey relational grade is to be maximized, the S/N ratio is calculated using
higher-the better (HB) criterion and is given by:

S=N 10 log

Table 7
Grey relational grade and its order.
Exp. no.

Grey relational grade

Order

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

0.8541
0.8810
0.6019
0.4627
0.5920
0.9072
0.4988
0.6254
0.4768
0.7754
0.4771
0.5987
0.7866
0.4664
0.4606
0.4561
0.4948
0.5528
0.5917
0.5068
0.6469
0.4136
0.9114
0.5348
0.4761
0.7208
0.6239

4
3
11
24
13
2
18
9
21
6
20
12
5
23
25
26
19
15
14
17
8
27
1
16
22
7
10

 X 
1
1
n
y2

where y is the observed data and n is the number of observations.

The S/N ratio is preferred to the traditional means as the former
can capture variability within a trial condition.
4.3. Factor and interaction effects
As the experimental design is orthogonal, the separation of each
coating parameters at different levels is possible. For example, the
mean of grey relational grade for factor A at levels 1, 2 and 3 can be
calculated by taking the average of the grey relational grade for the
experiments 19, 1018 and 1927, respectively. The mean of the
grey relational grade for each level of other coating parameters can
be computed in the similar manner. The mean of the relational
grade for each level of the combining parameters is summarized

Table 8
Response table for grey relational grade.
Level

1
2
3
Delta
Rank

0.6556
0.5632
0.6029
0.0924
2

0.6593
0.6150
0.5473
0.1121
1

0.5906
0.6306
0.6004
0.0401
3

0.6058
0.6206
0.5952
0.0255
4

Total mean grey relational grade = 0.6072.

2234

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

Fig. 4. Main effects plot for grey relational grade.

in the multi-response performance index table (Table 8). In addition, the total mean of the grey relational grade of the twenty-seven experiments is also calculated, as shown in Table 8. The
response table also contains ranks based on the delta values. The
delta value is calculated by subtracting the largest value from the
lowest from among the values in each column. Basically, a design
factor with a large difference in the grey relational grade from
one factor setting to another indicates that the factor or design
parameter is a signicant contributor to the achievement of the
performance characteristic. From the response table it is found that
parameter B is the most signicant factor in controlling the friction
and wear characteristics of NiB coatings. Fig. 4 shows the main effect plot of grey relational grade. The main effect plot gives the
optimal combination of coating parameters for minimum friction
and wear. As the larger the grey relation grade is, the closer will
be the product quality to the ideal value. Hence, the optimal combination of parameters is found to be A1B1C2D2. It may be noteworthy here that characteristics of NiB deposits are very much
affected by the ratio of concentrations of nickel and borohydride
ions in the solution. An improper balance between the concentrations of nickel and borohydride can lead to poor and rough deposits unsuitable for tribological applications [2]. The optimal levels of
nickel source (C2) and reducing agent (B1) obtained from the present study may be helping in striking a proper balance between the
two (nickel and borohydride ions) for achieving a smooth deposit
with good wear resistance. In general, the hardness of the coating
is quite responsible for providing the coating with the necessary
resistance against wear. Now, Ziyuan et al. [24] have observed that
the highest hardness achieved by NiB coatings is after a heat
treatment at 350 C. Again, Krishnaveni et al. [9] have observed a
peak at 350 C in the plot of hardness vs annealing temperature.
Hence, the increased wear resistance observed at D2 (350 C) in
the present study is quite consistent with the ndings of others
[9,24]. Also 350 C is a critical temperature of microstructure transformation as Ni2B forms during this process, which is a key contributor to the hardness of NiB coating [25]. Heat treatment
beyond a temperature of 450 C leads to a decrease in coating

Table 9
Results of ANOVA for grey relational grade.

Fig. 5. Interaction effects plot for mean of grey relational grade: (a) A vs B, (b) A vs C
and (c) B vs C.

Source

DF

SS

MS

% Contribution

A
B
C
D
AB
AC
BC
Error
Total

2
2
2
2
4
4
4
6
26

0.03867
0.05733
0.00784
0.00294
0.05567
0.13678
0.06726
0.24446
0.61095

0.01934
0.02866
0.00392
0.00147
0.01392
0.03420
0.01681
0.04074

0.47
0.70
0.10
0.04
0.34
0.84
0.41

6.33
9.38
1.28
0.48
9.11
22.38
11.01

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

Table 10
Results of conrmation test.
Initial parameter

Level
Wear
COF
Grade

A2B2C2D2
14.329
0.396
0.3882

Optimal parameter
Prediction

Experimental

A1B1C2D2

A1B1C2D2
4.405
0.129
0.7784

0.6072

Improvement of grey relational grade = 0.3902.

hardness and hence the wear resistance due to grain coarsening.

Heat treatment in general results in a reduction of the friction coef-

2235

cient of various electroless coatings compared to the as-deposited

coatings. The heat-treated NiB deposits are found to present a virtually incompatible surface for the hard counter face material and
hence results in a reduction of coefcient of friction [9]. Hence, the
optimal combination of parameters is justied. The main effect
plot also gives a rough idea about the relative signicance of the
parameters on the system response. If the plot for a particular
parameter has the highest inclination, then that parameter has
the most signicance. Whereas the plot which is near horizontal
has no signicance. From Fig. 4 it can be observed that parameter
B has the most signicance while parameter A is also quite significant. In the interaction effect plots (Fig. 5), the nonparallelism of

Fig. 6. EDX plots of NiB coatings: (a) 0.6 g/l NaBH4, and (b) 1.0 g/l NaBH4.

2236

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

the plots indicates that some amount of interaction exists between

the two factors, whereas intersecting lines are an indication of
strong interaction. From Fig. 5, it can be seen that quite strong
interaction exists between all the factors as far as the friction
and wear characteristics of electroless NiB coatings are
concerned.

4.4. Analysis of variance

The main purpose of the analysis of variance (ANOVA) is the
application of a statistical method to identify the effect of individual factors on the process response. The Taguchi experimental
method could not judge the effect of individual parameters on
the entire process, thus the percentage of contribution using ANOVA is used to compensate for this effect. Results from ANOVA (Table 9) can determine very clearly the impact of each factor on the
process results. In the present study, ANOVA is performed using
Minitab [26]. It is seen that parameter B i.e. concentration of reducing agent has the maximum contribution in controlling the friction
and wear characteristics of electroless NiB coating. Moreover, it is
found that parameter A i.e. bath temperature also has quite a contribution in controlling the tribological characteristics. Among the
interactions, the interaction between A and C (A  C) has the most
contribution in controlling the tribological behaviour of electroless
NiB coating followed by interaction between B and C (B  C) and
that between A and B (A  B). The concentration of borohydride in
the bath determines the content of boron in the coating. The boron
content of NiB coating again determines the properties inherited
by the coating [6]. Hence, the concentration of reducing agent hav-

ing the signicant contribution over controlling the tribological

behaviour is quite expected.
4.5. Validation test
A validation test is performed to see whether any improvement
in results is obtained through the condition suggested by the optimum parameter analysis compared to the initial condition. Also a
^ at the optimal condition is
prediction of the grey relational grade, c
calculated as follows:

c^ cm

o
X
ci  cm

i1

where cm is the total mean grey relational grade, ci is the mean grey
relational grade at the optimal level, and o is the number of the
main design parameters that signicantly affect the friction and
wear of electroless NiB coating. The comparison of the predicted
grey relational grade, experimental grey relational grade and the
grey relational grade at the initial condition is shown in Table 10.
The mid-level combination of coating parameters is assumed as
the initial condition. From the table, it is found that the improvement of grey relational grade at the optimal condition is about
0.3902 which is about 64% of the mean grey relational grade. This
is considered to be a signicant improvement.
4.6. Microstructure study
The chemical composition of electroless coatings are analysed
using one of the latest EDX detectors without any Beryllium win-

Fig. 7. SEM micrographs of the coating surfaces: (a) as-deposited, (b) annealed at 250 C (c) annealed at 350 C and (d) annealed at 450 C.

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

dow, enabling the detection of light elements like boron with considerable accuracy. The beryllium window if present absorbs all
the soft X-rays emanating from the lighter elements thereby preventing their detection. The EDX plots are shown in Fig. 6 and boron content in terms of weight percentages is found to be in the
range of 5.727.46 while the remaining is mostly nickel.
The SEM micrographs of the coating surfaces in as-deposited
and heat treated (at 250 C, 350 C and 450 C for 1 h) conditions
are shown in Fig. 7. The surface exhibits a cauliower like structure
which strongly points towards the coating possessing a lubricious
behaviour [11] and hence low friction coefcient. Surface of the
NiB coatings appears to be dense and matte grey in colour with
low porosity. Also by careful observation, it can be noted that the
NiB nodules are quite deated and at in as-deposited condition
but gradually grow in size with increase in heat treatment temperature giving rise to coarse grained structure.
The XRD analysis (Fig. 8) shows that the NiB lm is almost
amorphous in as-deposited phase but turns crystalline with heat
treatment. This is evident from the presence of microcrystalline
peaks in as-deposited phase whereas broad peaks of Ni, Ni2B and
Ni3B are found in samples heat treated at 350 C. Hence, the increased wear resistance and higher hardness of NiB coatings annealed at 350 C may be attributed to the precipitation hardening

2237

Fig. 9. SEM micrograph of the worn surface.

phenomenon. The solid solubility of the coating changes with temperature to produce various nickel boride phases (as observed in
XRD plots) which may be impeding the movement of dislocations
(defects in crystal lattice) thereby contributing to the strength of
the coating. This may be the reason for the NiB coating to display
superior tribological behaviour after being heat treated at the concerned temperature.
4.7. Wear mechanism
The worn surface of some of the coated samples is observed under SEM and they seem to suffer similar microstructural changes.
One of the SEM micrographs is displayed in Fig. 9 and it can be observed that load is taken by some of the nodules while the others
remain almost undeformed. The presence of longitudinal grooves
along the sliding direction with high degree of plasticity can be
clearly observed. This is indicative of the occurrence of micro-cutting and micro-ploughing effect and characterized as ductile failure. Almost no pits or prows are observed on the worn surface.
Hence it can be concluded that the abrasive wear is the predominant phenomenon. The same trend is observed for other combinations of deposition parameters within the experimental regime
considered in this study.
5. Conclusion

Fig. 8. XRD patterns of electroless NiB deposit in: (a) as-deposited and (b)
annealed at 350 C.

The coating process parameters (bath temperature, reducing

agent concentration, nickel source concentration) together with
the annealing temperature are optimized in order to minimize
the friction and wear of electroless NiB coatings. Grey relational
analysis is successfully employed in conjunction with Taguchi design of experiments to optimize this multiple response problem.
The optimal combination of parameters is found to be A1B1C2D2
(lowest level of bath temperature, lowest level of reducing agent
concentration, middle level of nickel source concentration and
middle level of annealing temperature). Also through ANOVA it is
revealed that concentration of reducing agent has the maximum
contribution in controlling the friction and wear behaviour of electroless NiB coating. Among the interactions, interaction between
A and C has the maximum contribution towards controlling the tribological behaviour of NiB coating. The coating surface resembles
that of a cauliower surface under SEM. The coating also appears to
be dense and light grey in colour. The XRD plots showed that the
electroless NiB coating is a mixture of amorphous and crystalline
phase in as-deposited condition. But with heat treatment, the coating turns crystalline. This is ascertained by the presence of Ni2B

2238

S.K. Das, P. Sahoo / Materials and Design 32 (2011) 22282238

and Ni3B peaks in the XRD plot of NiB coating heat treated at
450 C. The pattern of the sliding tracks showed that abrasive failure is the predominant wear mechanism.

Acknowledgement
One of the authors (Suman Kalyan Das) would like to acknowledge the research support provided by Council of Scientic and
Industrial Research, India: File No. 9/96(0621)2K10-EMR-I dated
05/03/2010.

References
[1] Brenner A, Riddell GE. Nickel plating by chemical reduction. US Patent:
US2532282; 1950.
[2] Mallory
GO,
Hadju
JB.
Electroless
plating:
fundamentals
and
applications. AESF: Orlando; 1991.
[3] Agarwala RC, Agarwala V. Electroless alloy/composite coatings: a review.
Sadhana 2003;28(3&4):47593.
[4] Hari Krishnan K, John S, Srinivasan KN, Praveen J, Ganesan M, Kavimani PM. An
overall aspect of electroless NiP depositions a review article. Metall Mater
Trans A 2006;37A:191726.
[5] Sahoo P. Wear behaviour of electroless NiP coatings and optimization of
process parameters using Taguchi method. Mater Des 2009;30:13419.
[6] Anik M, Krpe E, Sen E. Effect of coating bath composition on the properties of
electroless nickelboron lms. Surf Coat Technol 2008;202:171827.
[7] Delaunois F, Petitjean JP, Lienard P, Jacob-Duliere M. Autocatalytic electroless
nickelboron plating on light alloys. Surf Coat Technol 2000;124(23):2019.
[8] Dervos CT, Novakovic J, Vassiliou P. Vacuum heat treatment of electroless NiB
coatings. Mater Lett 2004;58:61923.
[9] Krishnaveni K, Sankara Narayanan TSN, Seshadri SK. Electroless NiB coatings:
preparation and evaluation of hardness and wear resistance. Surf Coat Technol
2005;190:11521.

[10] Sankara Narayanan TSN, Krishnaveni K, Seshadri SK. Electroless NiP/NiB

duplex coatings: preparation and evaluation of microhardness, wear and
corrosion resistance. Mater Chem Phys 2003;82:7719.
[11] Delaunois F, Lienard P. Heat treatments for electroless nickelboron plating on
aluminium alloys. Surf Coat Technol 2002;160:23948.
[12] Kanta AF, Vitry V, Delaunois F. Effect of thermochemical and heat treatments
on electroless nickelboron. Mater Lett 2009;63:26625.
[13] Oraon B, Majumdar G, Ghosh B. Improving hardness of electroless NiB
coatings using optimized deposition conditions and annealing. Mater Des
2008;29:14128.
[14] Vitry V, Delaunois F, Dumortoier C. Mechanical properties and scratch test
resistance of nickelboron coated aluminium alloy after heat treatments. Surf
Coat Technol 2008;202:331624.
[15] Srinivasan KN, Meenakshi R, Santhi A, Thangavelu PR, John S. Studies on
development of electroless NiB bath for corrosion resistance and wear
resistance applications. Surf Eng 2010;26(3):1538.
[16] Gawrilov GG. Chemical (electroless) nickel plating. Surrey: Portcullis Press
Ltd.; 1979.
[17] Ivanov MV. Electroless nickelboronphosphorus coatings: protective and
functional properties. Prot Met 2001;37(6):5926.
[18] Jiang B, Xiao L, Hu S, Peng J, Jhang H, Wang M. Optimization and kinetics of
electroless NiPB plating of quartz optical ber. Opt Mater 2009;31:15329.
[19] Taguchi G. Introduction to quality engineering. Tokyo: Asian Productivity
Organization; 1990.
[20] Ross PJ. Taguchi techniques for quality engineering. 2nd ed. New
York: McGraw-Hill; 1996.
[21] Roy RK. A primer on the Taguchi method. Mich: Society of Manufacturing
Engineers: Dearborn; 1990.
[22] Deng J. Introduction to grey system. J Grey Syst 1989;1(1):124.
[23] Gorbunova K, Ivanov M, Moiseev V. Electroless deposition of nickelboron
alloys mechanism of process, structure, and some properties of deposits. J
Electrochem Soc 1973;120:613.
[24] Ziyuan S, Deqing W, Zhimin D. Surface strengthening pure copper by NiB
coating. Appl. Surf. Sci. 2004;221(14):628.
[25] Rao QL, Wang HW, Zhou RH. Structure transformations and property of
electroless NiB coating. Cailiao Gongcheng/J Mater Eng 2000;6:303.
[26] Minitab User Manual (Release 13.2). Making data analysis easier. MINITAB Inc.
State College, PA: USA; 2001.