Upload
lamdieu
View
218
Download
0
Embed Size (px)
Citation preview
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
129
Engineering, Environment
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH1*, Abraham ADELEKE2
1Department of Materials Science and Engineering,
2Obafemi Awolowo University, Ile – Ife, Osun State, Nigeria
E-mail(s): 1 [email protected]; 2 [email protected].
* Corresponding author: phone: 08164027907, 08105862661
Received: July 01, 2017 / Accepted: November 25, 2017 / Published: December 30, 2017
Abstract
The development of asbestos free brake pads using corn husks as alternative filler was
studied with a view to replacing asbestos, which has been known to be carcinogenic.
Corn husks was sourced and milled, before been sieved into sieve grades of 100 and
200 μm. The varying proportions of the as-screened corn husk fibres and silicon carbide
were mixed with fixed proportions of graphite, steel dust and resin to produce brake
pads by using compressional moulding. The hardness, compressive strength, density,
flame resistance, wear rate and porosity of the products were then determined. The result
obtained showed that the brake pad produced with the corn husk passing the finer 100
μm screen gave better compressive strength, higher hardness, lower porosity and lower
rate of wear, consequent on the finer distribution of the corn husks particles in the
matrix. The results obtained for the brake pads were then compared with that of
commercial brake pad (asbestos based and optimum formulation laboratory brake pad,
corn husk based). The results were found to be in close agreement suggesting that corn
husk can be used in the production of asbestos-free brake pads.
Keywords
Corn Husks; Density; Porosity; Hardness; Brake Pads; Wear rate; Flame resistance; Ash
content
Introduction
The purpose of friction brakes is to decelerate a vehicle by transforming the kinetic
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
130
energy of the vehicle to heat, via friction, and dissipating that heat to the surroundings [1].
According to [2], brake pads are steel backing plates with friction material bound to the
surface facing the brake disc. They are components of disks brakes used predominantly in
automobiles. When we apply brakes, the brakepads or shoes, squeeze against the brake drums
or rotors, converting kinectic energy into thermal energy via friction.
Mechanical brake system are divided into the disk and drum brakes. Brake shoes are
located inside a drum for drum brake type so that on application of brakes, the brake shoe is
forced outward and pressed against the drum. Disc brakes operate in similar way except that
while drum brakes are enclosed disc brakes are exposed to environment.
By the arrangement, particulate materials that gradually wear from brake pads are
carried away by breeze into surrounding especially those from disc brakes. Most worn particles
from drum brakes are retained within enclosed drums and inhaled by mechanics when they
open up wheels for maintenance and repairs and gradually poison the human system [3].
According to Elakhame et al., asbestos has some few engineering properties that makes
it suitable for inclusion in brake linings [4]. Some of these include its good sound absorption,
resistance to heat, fire and affordability. Brake pads generally consist of asbestos fibres
embedded in polymeric matrix along with some other ingredients. Use of asbestos fibres is no
longer acceptable because of its carcinogenic nature leading to development of new asbestos-
free friction materials for brake pads.
Studies conducted in 1989 by National Institutes of Health published a report that
showed a high population of brake mechanics were afflicted with pleural and peritoneal
mesothelioma which were linked to chrysotile and asbestos exposure [3].
Consequently, researchers have turned to non-asbestos friction materials as an
alternative. There is hence, a concentration on industrial and agro - waste materials. Curent state
of knowledge have seen researchers replace asbestos with organic materials such as bagasse
[2], palm kernel shells (PKS) [4], and banana peels [10]. According to Elakhame et al., typical
formulations consists of more than 10 ingredients and and more than 300 materials of different
brands [4]. The works show a similar interest in finding a substitute for asbestos as a friction
material in brake linings, hence the purpose of this research is to show that corn husks can be
an alternative friction material in brake linings.
These waste usage will not only be economical but also result in foreign exchange and
environmental control [4]. Olabisi and Ademoh reported that the coefficient of fricion of maize
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
131
husk was determined as 0.5 at density of 45-80Ib/ft3 [3]. In contrast, friction coefficient in the
range of 0.30-0.70 is normally desirable when using brake lining material [4]. Based on these
factors, corn husk was conceived as a possible friction material alternative to asbestos.
The main objective are to develop asbestos free brake pads by using of corn husks as a
friction material. Performance evaluation of the produced samples include density, hardness,
microstructural analysis, wear rate, ash content tests, porosity and compressive strength. The
significance of this work lies in the advantage that it provides to our local automotive industry
while minimizing the danger caused by asbestos to health and environment.
Material and Methods
The experimental brake pads were developed from carefully selected materials [1] and
following the procedure as described by [3], [5], [6] and the products were evaluated with the
test methods elaborated by Elakhame et al [4].
The materials used were: Phenolic resin (phenol formaldehyde), corn husks particles,
steel dust, graphite and silicon carbide shown in Figure 1.
Silicon Carbide
Corn Husks
Graphite
Steel Dust
Figure 1. Research materials
The role of each materials indicated in Table 1.
Table 1. Materials used and their functions
S/N Material Function
1 Phenolic Resin (Phenol Formaldehyde) Binder / Matrix
2 Silicon Carbide Heat Resistant Material
3 Steel Dust Abrasive
4 Graphite Lubricating Material
5 Corn Husk Filler Material
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
132
The base raw material, corn husk was collected from a farm centre, properly sun-dried
and cleaned to remove impurities. Thereafter, it was crushed with hammer and milled into
powder using ball milling machine (Model 87002 Limoges-France, A50, …, 43). After this, it
was sieved to pass 100 and 200 μm screen apertures.
Specimen brake pad production
The technique of powder metallurgy as described by [5], [6] was adopted.
The production of brake pad consists of a series of unit operations including mixing,
cold and hot pressing, cooling, post-curing and finishing [7].
For each formulation, quantities expressed in percentage, weights presented for fillers,
abrasives, friction modifier and reinforcement were measured into the mixing vessel and
thoroughly mixed for 15 minutes to ensure homogeneity.
The desired amounts of phenolic resin (phenol formaldehyde) was poured into a
separate container and appropriate quantity of hardener added; with resin to hardener
percentage of 64.3% to 35.7% to form the binder, that was thoroughly stirred for about 10
minutes to obtain a uniform mixture.
Thereafter, the binder mixture was poured into the powdered friction material mixture
and stirred further to obtain a paste-like homogenous mixture.
The formed paste was poured into mould cavities that already had powdered talc applied
for ease of component removal after casting, cold pressed and allowed to cure for 90 minutes
[3]. The general methodological phases are shown in Figure 2.
Figure 2. General methodological phases
Raw Materials
Sourcing
Materials
Processing
Fabrication of Mould/
Plate Metal
Mixing Compacting
Surface Finish Curing
Optimization of Product
Formulation
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
133
The formulations used to produce the brake pads are shown in Table 2.
Table 2. Specimen formulations in (g*)
S/N Ingredients A B C D E
1 Corn Husks 40 45 50 55 60
2 Silicon Carbide (SiC) 26 21 16 11 6
3 Graphite 5 5 5 5 5
4 Resin 14 14 14 14 14
5 Steel Dust 15 15 15 15 15
g* - represents grams (a unit of mass)
The mixture of resin, graphite and steel dust that weighed 34 g was blended with varying
proportions of corn husks and silicon carbide that weighed 66 g, to make brake pads that
weighed 100 g [4].
Figure 3 shows the picture of the samples of organic brake pads produced using corn
husks as filler.
Figure 3. Produced samples
Specimen brake pad analysis. Brinell hardness Test
The resistance of the composites to indentation was carried out through the Brinell
hardness testing equipment to BS240, using a Tensometer (M500-25KN, Gunt Hamburg
Hardness Tester and WP300) pressing hardened steel ball with diameter D into a test specimen.
Based on ASTM specification, a 10 mm diameter steel ball was used and the load applied P
was kept stable at 3000 kg/f.
The diameter of the indentation d was measured along two perpendicular directions,
using an optical micrometer screw gauge.
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
134
The mean value was taken and incorporated into Eq. (1) to obtain the Brinell Hardness
Number (BHN) [4].
𝐵𝐻𝑁 = 2𝑃 ÷ 𝜋𝐷 (𝐷 − √(𝐷2 – 𝑑2) ) (1)
Where: P - the load applied; D - the diameter of hardened steel ball into a test specimen;d - the
diameter of indentation.
Compressive strength Test
The compressive strength test was carried out using the Tensometric Machine. The
samples of diameter 29.40 mm were subjected to compressive force, loaded continuously until
failure occurred. The load at which failure occurred was then recorded.
Ash content Test
About 1.20 g ± 0.1 g of the sample was weighed in a cooled crucible previously oven
dried by heating in a furnace at 5500C for 1 hour.
Then the samples were charred by heating on a hot plate thereafter. The charred samples
were thereafter taken into the furnace and heated at 5500C for 1 hour, then cooled in a desiccator
and weighed. This cycle of heating, cooling and weighing was repeated until a constant weight
is obtained. The % ash was obtained as, Eq. (2):
% 𝑎𝑠ℎ = (𝑊2 − 𝑊0) /(𝑊1 − 𝑊0) × 100 (2)
Where: W0 - weight of empty crucible; W1 - weight of crucible + sample; W2 - weight of crucible
and residue i.e. after cooling.
Density Test
The true density of the samples will be determined by weighing the samples mass on a
digital weighing machine, which would divide the volume. The volume was obtained via liquid
displacement method.
The formula is show in Eq. (3), below:
𝐷𝑒𝑛𝑠𝑖𝑡𝑦 (𝜌) = ( 𝑀 ÷ 𝑉) × 10 (3)
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
135
Where: M - the mass of test piece (g); V - the measuring volume of test piece (cm3) by liquid
displacement method.
Wear Rate Test
The wear rate for the samples was measured using pin on disc machine by sliding it over
a cast iron surface at a load of 1 kg, sliding speed of 125 and 250 rev/min, and sliding distance
of 2000 and 4000 m. All tests were conducted at room temperature. The initial weight of the
samples was measured using a single pan electronic weighing machine with an accuracy of
0.01 g.
During the test, the pin will be pressed against the counterpart rotating against a cast
iron disc (hardness 65 HRC) of counter surface roughness of 0.3 μm by applying the load.
A friction detecting arm connected to a strain gauge held and load the pin samples
vertically into the rotating hardened cast iron disc. After running through a fixed sliding
distance, the samples were removed, cleaned with acetone, dried, and weighed to determine the
weight loss due to wear. The differences in weight measured before and after tests give the wear
of the samples. The formula used to convert the weight loss into wear rate is in Eq. (4) [8]:
𝑊𝑒𝑎𝑟 𝑟𝑎𝑡𝑒 = 𝑊𝑒𝑎𝑟 𝑟𝑎𝑡𝑒 = 𝑆/𝑊 (4)
Where: ΔW is the weight difference of the sample before and after the test (mg); S - total sliding
distance (m).
Porosity
A sample of diameter 29.40 mm with a different height thickness of as thick as possible
was used. The specimens was weighed to the nearest mg, and then soaked in oil and water
container at 90-1000C for 8 hours.
The samples were left for 24 hours and then taken out from the oil container. Finally,
the test samples were weighed to the nearest mg.
The formula to calculate porosity is given as Eq. (5) [4]:
𝑃𝑜𝑟𝑜𝑠𝑖𝑡𝑦 (𝜌) = (𝑀2– 𝑀1) ÷ 𝐷 × 100 ÷ 𝑉 (5)
Where: D - the density of test oil and water; M2 - the mass of test piece after absorbing oil and
water (g); M1 - the mass of test piece (g); V - the volume of test piece (cm3).
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
136
Results and discussions
Brinell hardness Test
Figure 4, shows that variation of hardness with composition with sieve grades of 100
μm and 200 μm.
Figure 4. Variation of hardness with composition
The 100 μm and 200 μm formulations are however too closely related in sizing. The
samples 100 μm sieve grade of different formulation A-E, have the highest hardness value of
254, 235, 228, 222 and 211 HB, respectively.
The higher hardness value for the 100 μm compared to the 200 μm is as a result of the
reduced particle size of the corn husk fibre. This results in the increase in the surface area,
leading to an increasing bonding ability to the resin.
The more homogenous distribution of the filler and resin, as the particle size gets finers
is due to proper bonding and closer packing distance This is in agreement with results from
other research [4].
Compressive strength Test
Figure 5, shows that the compressive strength of the samples follows similar trends to
that of the hardness value.
254235 228 222 211 212 218 208 211 201
0
50
100
150
200
250
300
100μm
100μm
100μm
100μm
100μm
200μm
200μm
200μm
200μm
200μm
A B C D E A B C D E
Har
dn
ess
(H
B)
Composition (g)
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
137
Figure 5. Variation of compressive strength with composition
Ideally, the higher the density; which is the ratio of the amount of matter to volume, the
higher the compressive strength.
Density Test
Figure 6, shows the results of the density.
Figure 6. Variation of density with composition
Generally, density decreased as the corn husk particles increased in composition and
particle size. The trend in the density can be attributed to the increase in the volume of the
particle size, leading to a less close packing density.
This leads also leads to a decrease in homogeneity of the friction lining composites. The
111 109 108 110101
109 10599 98
89
0
20
40
60
80
100
120
100μm
100μm
100μm
100μm
100μm
200μm
200μm
200μm
200μm
200μm
A B C D E A B C D E
Co
mp
ress
ive
Str
en
gth
(K
pa)
Composition (g)
1,7411,621
1,504
1,786
1,3011,446 1,421
1,31 1,251 1,206
00,20,40,60,81
1,21,41,61,82
100μm
100μm
100μm
100μm
100μm
200μm
200μm
200μm
200μm
200μm
A B C D E A B C D E
De
nsi
ty (
g/cm
3)
Composition (g)
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
138
100 μm particle sizes has a higher density. This is due to the close packing of the corn husk
particles, creating more homogeneity in the entire composite phase.
Figure 7, shows the variation between the compressive strength and the density.
Figure 7. Variation of density with compressive strength
The results shows that the higher the density, the higher the compressive strength, due
to close packimg of the particles leading to reduced interatomic spaces between the particles.
Lower porosity of the specimen was observed for composition with lower sieve grade.
Compositions of 100 μm, generally show lower porosity compared to compositions 200 μm.
This result can be attributed to better dispersion of particles and good interfacial bonding
with the resin. Theoretically, a high porosity will lead to higher water and oil absorption, which
will lower the friction coefficient and lead to an increase in the wear rate of the brake pads [4].
Ash content Test
From Figure 8, it can be seen that the % value of the ash content increases as the particle
size increased.
85
90
95
100
105
110
115
1 2 3 4 5 6 7 8 9 10
Co
mp
ress
ive
Str
en
gth
(K
Pa)
Density (g/cm3)
Series1 Series2
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
139
Figure 8. Variation of ash content with composition
The 100 μm brake pad specimens had a lower ash content compared to the values
obtained for the 200 μm. This increase in ash content as sieve grade increases can be attributed
to an increase in pores sizes as the sieve grade increases. With increasing pore sizes, less close
packing, the charring of the specimens will yield a higher ash content. This is due to less proper
bonding between the resin and the filler – Corn Husk Particles [4].
The lower the particle size at optimum composition, the better the dispersion and lower
the ash content value.
The results of the wear rate showed an increase in the wear rate as particle size and
composition increases. This can be easily attributed to a lower bonding ability of the resin to
the filler – Corn Husk Particles. The increase in wear is also as a result of less closing packing
density. The wear, which is the loss of material or the erosion of a material from its derivative
position as a result of mechanical action of the opposite surface.
Another characteristic factor that increases the wear rate is the porosity of the specimen.
The higher the porosity, the higher was the wear rate.
The wear rate is a very important factor in the determining the friction coefficient.
Ideally, the lower the wear rate, the higher the friction coefficient.
41
43 4344
46
4243 43
48 48
36
38
40
42
44
46
48
50
100μm
100μm
100μm
100μm
100μm
200μm
200μm
200μm
200μm
200μm
A B C D E A B C D E
Ash
Co
nte
nt
(%)
Composition (g)
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
140
Porosity
Figure 9, shows the variation of porosity, hardness with composition. Lower porosity
of the specimen was observed for composition with lower sieve grade. Compositions of 100
μm, generally show lower porosity compared to compositions 200 μm.
Although, the values obtained for the porosity show infintesimal difference, the effect
on the hardness of the composite is pronounced.
Comparing compositions of 100 μm and 200 μm for formulations A, a 0.06% change in
porosity, results in a 42 HB value of hardness. This result can be attributed to better dispersion
of particles and good interfacial bonding with the resin.
Theoretically, a high porosity will lead to higher water and oil absorption, which will
lower the friction coefficient and lead to an increase in the wear rate of the brake pads [4].
Figure 9. Variation of porosity with hardness
Wear rate Test
Figure 10, relates the values of the wear at different speeds (30 min, 1 kg, 12 5rev/min
and 250 rev/min (g/km)).
0,91 0,91 0,92 0,92 0,98 0,97 0,93 0,89 0,93 0,91
254235 228 222 211 212 218 208 211 201
0
50
100
150
200
250
300
100μm
100μm
100μm
100μm
100μm
200μm
200μm
200μm
200μm
200μm
A B C D E A B C D E
Har
dn
ess
, Po
rosi
ty (
HB
, %)
Composition (g)
Series1 Series2
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
141
Figure 10. Variation of wear rate with composition
It shows the increase in wear rate as particle size. The graph shows a higher elevation,
which is a higher wear rate for formulations of the 200 μm formulations.
The results of the wear rate showed an increase in the wear rate as particle size and
composition increases. The increase in wear, which is a loss of material due to mechanical
action can be easily attributed to a lower bonding ability of the resin to the filler – Corn Husk
Fibre- as percentage composition of filler increases. The increase results in less closing packing
density, which leads to an increase in rate of wear.
Figure 11, relates the wear to the Brinell hardness and it shows that the lower the wear
rate, the higher the hardness value. This inverse realtionship between the wear rate and the
hardness can be attributed to the reduced porosity as wear rate reduces.
A lower porosity, as shown in Figure 9 is directly related to a reduction in particle size
which results in better interfacial bonding.
Improved interfacial bonding, due to better dispersion of the fibres in the matrix will
lead to an improved hardness value as shown in Figure 11. This can also be attributed to an
increased coeffieicnt of friction as particle size and porosity decreases.
11
11,5
12
12,5
13
13,5
14
14,5
100μm100μm100μm100μm100μm200μm200μm200μm200μm200μm
A B C D E A B C D E
We
ar R
ate
(g/
km)
Composition (g)
Series1 Series2
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
142
Figure 11. Variation of hardness, wear rate with composition
The results of this research work indicates that samples containing 100 μm of
formulation (A-E) gave better properties than those of 200 μm formulation.
Hence, the lower the sieve grades of corn husk particles, the beter the properties. The
100 μm sieve size results were compared with that of commercial brake pads (asbestos based)
as shown in Table 3.
Table 3. Comparison of the results obtained with standard
Property @ 100 (μm) Optimum formulation
laboratory brake pad (Corn
Husk based)
Commercial brake pad
(asbestos based)
Hardness, Brinell (at 3000 Kgf) 254 – 211 101
Porosity (%) 0.91 - 0.98 0.52
Ash Content (%) 41 – 46 54
Density 1.741 – 1.301 1.320
Compressive Strength ( KPa) 111 – 101 110
Wear Rate (g/km) 12.191- 13.782 3.800
The results are in close agreement. However, the higher hardness value is due to the
presence of SiC, which helps to increase the hardness of the brake pad composite.
Conclusions
The formulation of non-asbestos brake pads using corn husks as a filler, presents an
industry alternative to the carcinogenic asbestos used in commercial production. Physical and
wear tests were carried out and the results obtained were found to compare well with those of
150
170
190
210
230
250
270
290
1 2 3 4 5 6 7 8 9 10
Har
dn
ess
, We
ar r
ate
(H
B, g
/km
)
Composition (g)
Series1 Series2 Series3
Leonardo Electronic Journal of Practices and Technologies
ISSN 1583-1078
Issue 31, July-December 2017
p. 129-144
143
asbestos based friction lining material. From the results, the conclusion reached showed that
the samples 100 μm sieve grade of corn husk fibre gave the best properties of all.
Compressive strength, hardness, densities and porosity of the samples produced were
observed to decrease with increase in sieve aperture size while wear rate and percentage
charrred increasd as the sieve aperture size increased. The higher the percentage volume of the
corn husk particles from A –E, the higher the porosity, density, wear rate and percentge charred.
Hence, based on the above tests properties of the brake pads composite, corn husks as filler can
be used as an alternative to existing fillers, such as asbestos, in brake pad composites.
Non asbestos brake pads made from corn husks are therefore very suitable ecofriendly
replacement for asbestos and many agro-biomass friction materials in automotive brake pads.
From Table 3, the deviation of the some of the properties values such as the hardness
from the asbestos based brake pads can be attributed to the individual properties of the
component elements.
The presence of silicon carbide and steel dust, materials of high hardness will lead to an
increase in the overall hardness value compared with pads made from asbestos. Another factor
is the process technology and manufacturing parameters such as the powder mixing duration,
compaction pressure, compaction duration, post curing temperature and time [9]. This variation
in process parameters would lead to a variation in the properties obtained.
References
1. Blau P.J., Compositions, functions and testing of friction brake materials and their
additives, Being a report by Oak Ridge National Laboratory for U.S Dept. of Energy, 2001.
2. Aigbodion V.S., Akadike U., Hassan S.B., Asuke F., Agunsoye J.O., Development of
asbestos free brake pad using bagasse, Tribology in Industry, 2010, 32, p. 12-18.
3. Olabisi A.I, Ademoh N.A., Development and evaluation of maize husks (asbestos - free)
based brake pad, Industrial Engineering letters, 2015, 5 (2), p. 67-80.
4. Elakhame Z.U., Alhassan O.A., Samuel A.E., Development and production of brake pads
from Palm Kernel composites, International Journal of Scientific and Engineering Research,
2014, 5, p. 735-744.
Development of asbestos free brake pads using corn husks
Wisdom ASOTAH, Abraham ADELEKE
144
5. Edokpia R.O., Aigbodion V.S., Obiorah O.B., Atuanya C. U., Evaluation of the properties
of eco-friendly brake pad using egg shell particles–gum, Arabic ScienceDirectR, Elsevier
B.V. DOI: 10.1016/j.rinp.2014.06.003, 2014.
6. Bashar D., Peter, Madakson B., Joseph M., Material selection and production of a cold-
worked composite brake pad, World Journal of Engineering and Pure and Applied Science
(WJEPAS), 2012, 2 (3):96.
7. Gurunath PV and Bijwe J., Friction and wear studies on brake-pad materials based on
newly developed resin, Journal of engineering tribology, 2007, 263 (7), p. 1212-1219.
8. Osterle W., Griepentrog M., Gross T., Urban I., Chemical and microstructural changes
induced by friction and wear of brakes, Wear, 2001, 251, p. 1469-1476.
9. Talib R.J., Mohmad S.S., Ramlan K., Selection of best formulation for semi-metallic brake
friction materials development, Powder Metallurgy, Dr. Katsuyoshi Kondoh (Ed.), ISBN:
978-953-51-0071-3, 2012.
10. Masrat B., Sheikh S.S., Owais B., Friction and wear beahviour of disc brake pad material
using banana peel powder, International Journal of Research Engineering and Technology,
2015, 4, p. 650-659.