En

NCU

a

A

R

R

1

A

K

F

W

B

F

W

1

Eecms

0d

j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144

journa l homepage: www.e lsev ier .com/ locate / jmatprotec

ffect of water spray on friction and wear behaviour ofoncommercial and commercial brake pad materials

.S.M. El-Tayeb ∗, K.W. Liewentre of Computer Aided Design and Knowledge Manufacturing (CCADKM), Faculty of Engineering and Technology, Multimedianiversity, Jalan Ayer Keroh Lama, 75450 Melaka, Malaysia

r t i c l e i n f o

rticle history:

eceived 1 June 2007

eceived in revised form

8 December 2007

ccepted 25 December 2007

eywords:

rictional brake pad

ater spray

ack transfer films

riction

ear

a b s t r a c t

The reduction in the friction coefficient to unacceptable levels for frictional brake pad

applications in a rainy and humidly environment is considered as a serious problem as

it influences the safety of the vehicle. Thus, in the present work, particular emphasize were

given to the effect of water spray on the level and behaviour of friction and wear character-

istics of frictional brake pad (FBP) materials. Four different non-commercial FBP materials

(NF1, NF2, NF4, and NF5) were developed and evaluated along with other two chosen com-

mercial FBP materials (CMA and CMB) using a small-scale tribo-tester of pad-on-disc type.

The results confirmed that in spite of spraying water to the disc, no evidence of HD water

film could be observed. Hence the friction behaviour was influenced by factors other than

HD film. The values of friction coefficient obtained under wet condition were in the range

of dry friction, mixed and boundary lubrication friction. For instant, some values of friction

coefficient for NF1 &NF4 were less than 0.05 which was in the range of friction in mixed

lubrication. Regardless of the type of brake pad materials, the results indicated that spray-

ing water eliminates the establishment of the transfer layer at the friction interface and

suppresses the temperature rise in the disc and so the formation of the char or other reac-

tion products such as oxides. Besides, wear rate of all FBP materials behaved linearly with

contact pressure. Qualitative assessment of the SEM morphologies of brake pad surfaces

confirmed that tribofilms were hardly formed in wet braking. In addition, all brake pad sur-

faces showed formation of contact plateaus “patches” and disintegrations of various sizes

and locations depending on the braking condition. Furthermore, the removal of material was

associated with either mechanical crushing action performed by entrapped wear debris or

due to disintegration of plateaus which were accelerated by spraying the water.

ronmental conditions, dry or wet. Therefore, frictional brake

. Introduction

ffective performance of frictional brake pad at different lev-ls of braking conditions in wet environment is the main

oncern of cars drivers. Frictional brake pads (FBPs) in auto-otive brake system are always the key role of effective and

afe braking performance and blames always go to it when

∗ Corresponding author. Tel.: +60 6 252 3926; fax: +60 6 231 6552.E-mail addresses: nabil.eltayeb@mmu.edu.my, nabil.eltayeb@yahoo

924-0136/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.jmatprotec.2007.12.111

© 2008 Elsevier B.V. All rights reserved.

a brake-related problem arises. This is because brake padsare more vulnerable to various braking parameters includingpedal pressure, sliding velocity, disc temperature, and envi-

.com, mscalec@yahoo.com (N.S.M. El-Tayeb).

pad (FBP) material should be designed to maintain a relativelyhigh and stable friction coefficient irrespective of tempera-ture, age, degree of wear, presence of dirt, humidity, and water

n g t

136 j o u r n a l o f m a t e r i a l s p r o c e s s i

spraying from the road, etc. (Eriksson et al., 2002; Kim and Jang,2000; El-Tayeb et al., 2006).

Even though it is known that brake must operate under avariety of environmental conditions, most standardized anddeveloped tribo tests for brake materials are surprisingly con-ducted under dry sliding conditions. Publications concerningthe effects of wet frictional behaviour are rare in the tribol-ogy literature. Wet and dry braking tests tend to be performedon full-vehicle, on road test (field test) rather than usinglaboratory-scale test (model test) and in this technique bothtire performance and braking system contribute to the results.Additionally, brake manufactures and suppliers often mea-sure the water absorption characteristics of lining materials.Laboratory-scale tribotesting protocols commonly use con-stant load (or constant torque) and constant speed. Wet anddry behaviour is not a part of recommended procedures suchas the Society of Automotive Engineers procedures J 886 (alaboratory-scale, coupon test for determining lining friction),J 2430 (a multi-stage dynamometer test for disc brakes), andSAE J 1802 (a test procedure for drum brake linings) (Blau andMcLaughlin, 2003).

In our previous work (El-Tayeb and Liew, 2007) there was aclear and substantial effect of water spray on friction coeffi-cient and wear rate of FBPs when tested against GCI rotor disc.A significant reduction in the friction coefficient to unaccept-able levels for FBP applications was observed which in a rainyand humidly environment is considered as a serious problemas it influences the safety of the vehicle. FBP material shouldhave the capability to produce reasonably high friction evenat high sliding speed in a wet environment.

A limited number of studies investigating the effect ofwater spray on the performance of FBP are available in theliterature and complete analysis of the key role of waterfilm at the interface of brake pad/rotor disc is seldom found.Accordingly, the present work intended to investigate thefriction and wear behaviour of four non-commercial (contain-ing 10 ingredients) and two commercial brake pad materials.The NCFBP materials were developed from non-asbestosorganic-based friction materials for automotive brake sys-tems. A small-scale tribo testing brake pad/rotor disc machinewas implemented to conduct the tribo tests at different

nominal contact pressures (1.11–2.22 MPa) and two differentsliding speeds (1.3 and 2.1 m/s). The effect of continuousbraking under wet conditions on friction and wear character-istics of all commercial and non-commercial FBP materials

Table 1 – Chemical composition (vol.%) of grey cast iron

Element (vol.%) Material: grey cast iron

C 3.0–3.8Si 1.8–2.2Mn 0.2–0.4S 0.04 max.Cr 0.01Cu 0.01P 0.07Mo 0.01Ti 0.15–0.25Fe Bal.

e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144

tested against gray cast iron (GCI) rotor disc was investi-gated. A particular emphasis of this investigation was focusedon gaining some understanding on the role of water filmat the interface from the lubrication theories stand pointand how this affect the process of back transfer tribofilms.Results in this work are based on investigation of currentformulations of non-commercial FBP materials and resultsoutside these formulations and test parameters may bedifferent.

2. Experimental work

2.1. Rotor disc and brake pad materials

Brake materials can be classified as organic, metallic and car-bon (Anderson, 1992). Metallic brake linings mostly of copper,iron, and sintered bronze are used for very high power inputdensity applications such as high-speed railways and racingcars. Carbon–carbon composites are generally used for air-craft and racing car applications where weight is an essentialfactor. Cast iron, particularly gray cast iron materials are com-monly used almost in all automotive frictional brake discs(Blau and McLaughlin, 2003; Anderson, 1992). This is due toits satisfactory wear resistance, excellent damping capacity,high thermal conductivity (stability), and, in particular, rel-atively easy to cast and machine (Eriksson et al., 2002). Thusrotor disc used in the present investigation was chosen as graycast iron (GCI) of grade Flocast 4E (El-Tayeb and Liew, 2007).The Chemical compositions of the GCI rotor disc are given inTable 1.

Although friction materials have been developed for acentury, the formulations used in the most of commercialfrictional brake pad materials are usually obtained by trialand error method (Bijwe, 1997) cited in Yafei (2006). Mostof commercially available automotive friction materials con-tain multiple components, usually consists of 5–25 vol.% offibrous ingredients and the type and relative amounts of thefibres influence many features of brake achievement and wearlife (Mutlu et al., 2006). In the present work, four differentbase matrices of non-commercial FBP materials (NF1, NF2,NF4, and NF5) were manufactured for this work, i.e. non-asbestos semi-metallic type materials containing 10 differentingredients. These ingredients comprise fibre reinforcement,binder, friction modifiers, solid lubricant, abrasive, and filler.Fibres are usually used because of their friction properties,heat resistance, and their thermal conductivity. Besides thatfibres can impart toughness and strength to the binder. Fillermaterials are added to improve or optimize the properties.The abrasive zircon (ZrSiO4) is added to control the build-ing up transfer film on the counter disc and thus increasesthe friction coefficient. Phenolic resin (binder), which is cur-rently used in most commercial friction materials, is addedto hold all ingredients together. The role of natural or syn-thetic rubber is to improve the flexibility of the binder. Therelative amount and type of these compositions are given in

Table 2.

The non-commercial friction materials were manufacturedby dry-mixing, pre-forming, hot press molding at 17.2 MPa and180 ◦C, post-curing, and heat treatment (El-Tayeb and Liew,

j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144 137

Table 2 – The Composition of the non-commercialfriction brake pad materials (vol.%).

Raw materials Sample code

NF1 NF2 NF4 NF5

Metal fibre: steel fibre 15 20 15 20Friction modifiers: brass 6 6 6 6Cashew dust 10 10 10 10Solid lubricant: graphite (C) 8 8 8 8Abrasive: zircon (ZrSiO4) 3 3 3 3Binder (matrix): phenolic resin 20 15 20 15Rubber (SBR) – – 10 10Organic fibre: aramid pulp 10 10 – –Fillers, reinforcements: CaCO3 8 8 8 8

2tMCtT

2

I9pMcb

Table 3 – Chemical composition of two commercialfriction brake pad materials (wt.%)

Element Material (wt.%)

CMA CMB

Fe 62.5539 40.2219Ba 3.4754 6.2559Al 0.0506 1.0377Mg 2.1001 0.6904Ag 0.5116 0.2149Na 0.0885 0.1614K 0.2968 0.4185Cu 0.0025 0.3152Mn 0.1057 0.2299Ga 0.0965 0.1664

The compression tests of brake pad specimens werecarried out on a Gunt Hamburg Universal Material Tester

Fb

BaSO4 20 20 20 20

007). All non-commercial friction materials were manufac-ured at CL Industry Sdn. Bhd. of brake pad manufacturing,alaysia. Two other commercial FBP materials (CMA andMB) were chosen for this work. The chemical composi-

ions of these two commercial FBP materials are listed inable 3.

.2. Preparation of specimens

n the present work, specimens of size 9.5 mm ×.5 mm × 20 mm were machined from non-commercial brakead plates of size 250 mm × 250 mm × 20 mm by MAZAK CNCilling machine, Fig. 1. The microstructures of the polished

ross-section of non-commercial and commercial frictionalrake pad materials are shown in Fig. 2.

ig. 1 – Frictional brake pad materials (FBPs). (a) Non-commerciarake pad specimens; and (d) specimen’s dimensions.

Sr 0.0496 0.1464Zn – 0.3751Others 30.6688 49.7663

2.3. Hardness and compression tests

Surface hardness of both non-commercial and commercialmaterials was measured using a Shore D hardness tester (TimeGroup Inc. Shore D Durometer TH210) and Brinell hardnesstesting machine (Gunt Hamburg Universal Material TesterWP300). Before measuring the hardness, surfaces of the spec-imens were polished using fine abrasive paper. At least fivereplications of hardness test for each specimen were madeand the average value was reported in Table 4. Experimentalscatter was about ±2 HB and ±2 SHORE D.

WP300. Each specimen of initial cross-sectional area of190 mm2 was placed between the lower cross-member and

l brake pad material; (b) commercial brake pad; (c) prepared

138 j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144

Fig. 2 – Microstructure of non-commercial and commercial FBP materials: (a) NF1; (b) NF2; (c) NF4; (d) NF5; (e) CMA; and (f)

CMB.

the cross-head and the load was applied up to 20–50% of thebreaking load. Three replications of compression tests weremade for each material and the average results are givenin Table 4.

2.4. Friction and wear test procedure

The friction and wear test procedure was described in ourprevious work (El-Tayeb and Liew, 2007). However a briefdescription is given here. A small-scale-tribo tester (SSTT) wasimplemented to conduct continuous sliding friction and weartests at low and moderate braking conditions, i.e. nominalpressure (up to 2.22 MPa) and speed (up to 2.1 m/s) under wetconditions. A schematic diagram and photo of the SSTT areshown in Fig. 3, respectively. The GCI rotor disc of 135 mmdiameter and 10 mm thickness has an initial average sur-face roughness (R ) of 1.263 �m and surface hardness of

a

185.52 ± 6.20 HV30kgf (Vickers Hardness Tester 430/450 SVA,Wilson Wolpert).

A single brake pad was tested against the GCI rotor disc attwo different sliding velocities (1.3 and 2.1 m/s) and different

Table 4 – The physical properties of the Noncommercial and Co

Raw materials

NF1 NF2

Brinell hardness 23.84 ± 1.91 21.82 ± 1.60 20.Shore D hardness 88.30 ± 1.39 87.86 ± 1.80 86.Density (gm/cm3) 2.6821 2.6464 2.Ult. comp. stress (MPa) 85.53 ± 16.59 68.77 ± 18.92 56.

nominal contact pressures from 1.11 to 2.22 MPa for 5 min slid-ing time. Before starting each test, the specimen was rubbedagainst an abrasive paper of grade 180 to ensure better uni-form contact between the specimen and the rotor disc duringtribo-test. Frictional force at the sliding interface was mea-sured using strain gauge (previously calibrated) mounted onthe load level that hold the specimen and friction coefficientfor each applied nominal contact pressure was obtained overthe steady state range. Weight loss was determined by mea-suring the weight difference using 0.1 mg electronic balance(SHIMADZU AW-220).

During wet tests, a metal container and a plastic capsurrounded the counterface disc and the specimen (Fig. 3b)were used to collect and drain the water during the experi-ments. Fresh water was continually sprayed into the interfacebetween the brake pad specimen and the rotor disc.

The morphologies of frictional surfaces (worn surfaces)

were examined by scanning electron microscopy (SEM) (JEOL,JSM 840). Before taking micrographs, the worn surfaces of allthe specimens were coated with a thin layer of gold using ionsputtering (model JEOL, JFC-1600).

mmercial FBP materials (vol.%)

Sample code

NF4 NF5 CMA CMB

98 ± 1.62 19.61 ± 1.05 24.62 ± 1.98 14.75 ± 1.4026 ± 1.10 83.90 ± 0.99 87.52 ± 2.20 83.28 ± 0.854655 2.5649 3.3704 2.072457 ± 10.50 57.83 ± 6.38 65.11 ± 12.05 67.20 ± 14.82

j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144 139

F stert

3

ThaorarraicmwBnc(an

3

Ipw2cd

ig. 3 – Schematic diagram and photo of small-scale tribo-teest set-up.

. Results and discussion

he NF4 and NF5 showed lower compression strength andardness compared to NF1 and NF2. The composition of NF1nd NF2 contained 10% aramid pulp and 0% rubber. On thether hand NF4 and NF5 contained 0% aramid pulp and 10%ubber (SBR). It is known from the literature (Yafei, 2006) thatramid pulp enhances the mechanical strength of FBP mate-ials and reduces the thermal conductivity. This may be theeason why NF1 and NF2 gave higher compression strengthnd hardness. Phenolic resin is invariably used as bindern friction materials due to good combination of mechani-al properties such as high hardness, compressive strength,oderate thermal resistance, creep resistance and very goodetting capability with most of the ingredients (Gurunath andijwe, 2007; Kim et al., 2008). The high hardness of the phe-olic resin is attributed to the increase in the hardness duringuring process (Kim et al., 2008). It is clear that from the resultsFig. 4) higher properties of hardness and compression aressociated with FBP materials possessing higher vol.% of phe-olic resin.

.1. Friction results

n literature, general trends show that the friction and wearerformance do not go together. If friction behaviour is good,

ear performance is poor and vice versa (Gurunath and Bijwe,

007). Dry friction and wear rate of non-commercial andommercial FBP materials tested against GCI rotor disc atifferent nominal contact pressure and different sliding veloc-

Fig. 4 – Hardness and compression

(SSTT). (a) Continuous contact position and (b) wet braking

ities were reported in our previous work (El-Tayeb and Liew,2007). All non-commercial FBP materials, as well as, CMBshowed insignificant difference in the coefficient of friction(COF) under dry sliding condition, indicating that the COF wasinsensitive to the type of FBP materials and this was attributedto the formation of back transfer films. Only CMA has sub-stantially lower COF than the other. Furthermore, the frictioncoefficient showed a slight increase with increasing eithernominal contact pressure or sliding velocity. Fig. 5 shows acomparison between dry and wet friction coefficients at 1.11and 2.22 MPa for two different sliding speeds (1.3 and 2.1 m/s)for both non-commercial and commercial FBP materials. Com-paring the results of wet braking tests for friction coefficient(Fig. 5), one can see clearly that a significant reduction in thefriction coefficient at high-speed (2.1 m/s) to unacceptable lev-els for FBP applications. According to Shorowordi et al. (2004)the industry standard range of friction coefficient fall withinthe range of 0.30–0.45 for automotive brake system. Under wetcondition the range of COF (shown in Fig. 5) is 0.05–0.16. In arainy and humidly environment this is considered as a seriousproblem as it influences the safety of the vehicle. FBP materialshould have the capability to produce reasonably high frictioneven at high sliding speed in a wet environment. It should bealso noted that the wet results (Fig. 5) showed large variationin the COF between the different types of FBP materials testedunder identical conditions.

Apparently, water film did not only cool the interface and

suppresses the back transfer film but it also brings anothereffect to the picture which is the type of FBP materials.To explain the effect of the water, three modes of lubrica-tion should be defined in connection with the wet results

properties of FBP materials.

140 j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144

ts of

Fig. 5 – Comparison between dry and wet friction coefficien(b) 1.11 MPa and 2.22 MPa at 2.1 m/s.

including hydrodynamic (HDL), mixed lubrication (ML), andboundary lubrication (BL). In hydrodynamic lubrication (HDL),a thick film which is many times thicker (5–500 �m) than theheight of the irregularities on the sliding surfaces preventssolid–solid contact and provides a very low friction coefficient(� = 0.001–0.003) and negligible wear (Bhushan, 2002). In addi-tion, this COF increases slightly with sliding speed because ofviscous drag. The thick film is maintained between the tworubbing surfaces and creates a sufficient pressure to supportthe external applied load. Thus by introducing water to therubbing surfaces, the behaviour of the contact should be gov-erned by the bulk physical properties of the water, notablyviscosity, and the frictional characteristics should arise purelyfrom the viscous shearing of the water film. In mixed lubrica-tion mode, there may be more frequent solid contacts but atleast a portion of the sliding surface remains supported par-tially by HD film. The solid contacts could lead to a cycle ofadhesion, metal transfer; wear partial formation, and even-tual seizure. Mixed lubrication is also referred to as quasi-HD,partial fluid or thin-film (typically 0.025–2.5 �m) and the fric-tion coefficient in this mode lies in the range of (0.003 < � < 0.1).The third mode is the boundary lubrication, in which the solidsurfaces are so close to each other that surfaces interactionbetween mono or multi molecular films of lubricants and thesolid asperities dominate the sliding behaviour and the fric-tion coefficient in this mode of lubrication is about 0.1.

When the water spray was introduced to the sliding

interface, very low friction coefficient was observed at bothsliding speeds, i.e. the friction coefficient was in the range of(0.12–0.27) at 1.3 m/s and (0.02–0.2) at 2.1 m/s for brake padsamples (Fig. 5a and b). In particular, NF1 and NF5 showed

Fig. 6 – SEM micrographs showing evidences of direct contact of

FBP materials tested at (a) 1.11 MPa and 2.22 MPa at 1.3 m/s

friction coefficient of (0.02–0.05) at 2.1 m/s which was closeto the range of friction in mixed lubrication. Initially, it wasexpected that introducing water during the tests not onlyremove most of the heat that generated during dry tests (El-Tayeb and Liew, 2007) but also develop HD water film andprevent solid-solid contact. However, general observation ofthe friction coefficient values obtained for brake pads underwet conditions reveals that the friction coefficients are stillwithin the range of dry sliding friction (0.12–0.27 at 1.3 m/s and0.02–0.2 at 2.1 m/s) and not in the range of hydrodynamic fric-tion (0.001–0.003). This is an interesting result since it suggeststhat the hydrodynamic speculation has to be reconsidered andan explanation why friction coefficient under wet braking con-ditions still in the order of dry sliding. It appears that withinthe speed range tested, the pressure generated in the waterfilm was not enough to build up a thick water film in order tosupport the applied normal load and prevent solid–solid con-tact. As a consequence, the rubbing process was dominatedby the physical properties of the direct contact of solid asperi-ties which occurred frequently and partially by very thin waterfilm (mixed lubrication). The SEM micrographs of brake padsurfaces tested under wet braking conditions (Fig. 6a–c) depictsome grooves produced by entrapped wear particles betweenrotor disc and brake pad surface which evidences the directcontact of solid asperities, marked with c. Also, it should benoted that in the present results (Fig. 5a and b), lower frictioncoefficient is associated with higher speed. This is in con-

trast to hydrodynamic theory in which higher speed meansmore fluid is dragged into the interface leading to relativelyhigher friction. Actually, lower speed may bring some asperi-ties into contact to support external load especially at low fluid

solid asperities producing grooves during wet braking tests.

j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144 141

F g su

pgtfawwoatsi

ssdfiapTqeoat

Fa

ig. 7 – SEM micrographs showing clean and smooth rubbin

ressure and these are the trapped asperities that caused therooves in the brake pad surface shown in Fig. 6a–c. In addi-ion to this, the observed low values of friction coefficientsor NF1, NF4, and NF5 suggest the involvement of both mixednd boundary lubrication modes. In summary, the results ofet friction coefficients confirmed that the friction behaviourere influenced by factors other than HD film, and the valuesf friction coefficient were in the range of dry friction, mixednd boundary lubrication friction. If our hypothesis is true,hen more research should be dedicated to further the under-tanding of the mixed and boundary lubrication mechanismsn wet braking applications.

The SEM micrographs of rubbing surfaces of brake padamples tested under wet conditions (Fig. 7a and b) show alsotrong evidences of a continuous washing and removal of theebris from the interface (i.e. cleaning) besides no transferlms as observed in the dry braking micrographs, (El-Tayebnd Liew, 2007). Accordingly, there was always fresh brakead surface exposed to rotor disc as shown in Fig. 7a and b.hese of course affect the friction and wear quantitatively andualitatively as indicated in the results, Fig. 5. It should be

mphasized that this process also suppresses the formationf transfer film and leads to more frequent solid contacts, i.e.s in mixed lubrication mode where asperities can highly con-ribute to the sliding behaviour. Blau and McLaughlin (2003)

ig. 8 – Typical variation of wear rate versus nominal contact pret 1.3 and 2.1 m/s.

rfaces with no debris produced during wet braking tests.

reported that the formation of friction films was suppressedby water, and that in turn led to a friction spike just prior to thecessation of relative motion. In another work by Eriksson et al.(2001) concerning the effects of relative humidity on frictionand brake squeal, the tribofilms were more easily formed indry environments than in wet. Thus growing evidences (Blauand McLaughlin, 2003; El-Tayeb and Liew, 2007; Eriksson etal., 2001) including the present work suggest that water caneliminate the formation of friction-induced films which ofcourse affect the stability of tribo braking behaviour. However,it should be also emphasized that the removal of the debrisfrom the interface and suppressing the transfer film formationby spraying water (Figs. 6a–c and 7a and b) are not the onlyaspects of determining the tribo behaviour at wet interface,but changing of lubrication mode from one to another is alsoa key role of determining this tribo process, which remains forfurther studies.

3.2. Wear results

Figs. 8 and 9 show the wear rate for all FBP materials tested

under dry and wet continuous braking against GCI disc inwhich wear rate increases with increasing nominal pressureand this trend is more pronounced under wet condition. Fromthe comparison of Figs. 8 and 9, one can see that the wear

ssure for FBP materials tested under dry braking conditions

142 j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144

Fig. 9 – Typical variation of wear rate versus nominal contact pressure for FBP materials tested under wet braking

conditions at 1.3 and 2.1 m/s.

rate of all FBP materials is almost linear with contact pres-sure under wet condition. This interesting feature is probablyassociated with continuous washing the debris from the inter-face and suppressing the formation of friction film by theeffect of spraying water. The work introduced by Blau andMcLaughlin (2003) concluded that the formation of frictionfilm is dominated in dry braking while under wet sliding

this friction film is suppressed by the water. Also from theresults it appears that the wear rate is determined by thetype or ingredient of the FBP materials. For instant, highestwear rates were observed for NF5 and CMB at 1.3 and 2.1 m/s

Fig. 10 – SEM micrographs showing rubbing surfaces of brake padisintegration of patches (plateaus) and (d) debonded fibres.

respectively. Meanwhile, NF1 consistently showed the bestwear performance at both speeds suggesting this type of FBPsample has longer life under wet condition. But in terms offriction, it showed the poorest performance (Fig. 5) which incontrast to the reported results by Cho et al. (2005). Accord-ing to Cho et al. (2005) friction coefficient and wear resistanceshould increase with increasing the phenolic resin content.

On the other hand, improvement in the wear resistance ofFBP materials was associated with higher contents of pheno-lic as reported in Cho et al. (2005). Apparently, in the currentwork, the observed behaviour of NF1 is related to the higher

d samples tested under wet braking conditions (a–c)

j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144 143

F ncea bris a

cF

rcos

ipatt(tootwpabFagww

ig. 11 – Schematic illustration of agglomeration and adherend debris accumulation and (b) instant removal of wear de

ontents of Aramid pulp and phenolic resin compared to otherBP materials.

Qualitative assessment performed by SEM indicated thatemoval of material was associated with either mechanicalrushing action performed by entrapped wear debris (Fig. 6a–c)r due to disintegration of plateaus which are accelerated bypraying the water, Fig. 10a–c.

The formation of the plateaus depends on pressure andts associated contact temperature, shear strength of thelateaus, and some other sliding contact conditions suchs dry or wet contact. Under wet braking condition (Fig. 9)he increase of wear rate with increasing the nominal con-act pressure was more pronounced compared to dry resultsFig. 8). In addition scattering of the data was larger comparedo wet results. This is most probably attributed to the removalf wear debris from the interface by water. Also, the absencef debris agglomeration and transfer film formation con-ribute partially. Additionally, under certain circumstances,hen the protective fibre or particles worn away, the contactlateaus are disintegrated. The tendency of debris to adherend stick around the fibres to form patches or plateaus maye explained with the aid of a model shown schematically inig. 11. It should be realized that accumulation of wear debris

nd formation of transfer film during dry sliding (Fig. 11a) canive support to the steel fibres at the sliding interface. Buthen the water was sprayed into the interface, the generatedear particles were instantly removed (washed out) and the

Fig. 12 – SEM micrographs showing the process of fi

of debris around the steel fibres (a) transfer film formationnd elimination the transfer films.

transfer film formation was eliminated as illustrated schemat-ically in Fig. 11b, and evidenced by the SEM morphologies inFigs. 7 and 10a–c. Removal of material under wet conditionmay be enhanced by debonding some of the steel fibres andbecoming loose due to absence of accumulated debris as it canbe seen in the micrographs of the worn surface, Fig. 10d. Thusthere was always a fresh clean surface of the brake pad spec-imen that experienced to wear. This, in other words, meansthat the tribofilms are easily formed in dry environment andhardly formed in wet environment.

Another feature on the worn surfaces was observed dur-ing the wet tests. The presence of the cavities or craters dueto disintegration (Fig. 10a–c) that scattered over the brakepad surface is likely contributed differently during dry andwet sliding. During dry braking tests they act as debris col-lectors (see for example Fig. 12a and b). On the other hand,during wet tests the situation is different; the process of fill-ing in the cavities with debris is disappeared. Instead, thecavities act as pockets of water to minimize the effect ofheat, remove the loose debris from the interface and fromthe cavities and pumping the water in and out the cavitiesto enforce some debris to become loose. For instant, Fig. 10band c show that some patches were about to disintegrate

from the surface (marked with d) whereas Fig. 10d shows alarge plateau was already disintegrated (marked with p). Allthese aspects contributed to the material removal during wetbraking.

lling in cavities with debris during dry braking.

n g t

r

MMC sliding against phenolic brake pad. Wear 256,1176–1181.

144 j o u r n a l o f m a t e r i a l s p r o c e s s i

4. Conclusions

The friction coefficient and wear rate characteristics underwet braking condition for four different non-commercial andtwo commercial frictional brake pad (FBP) materials weredetermined and the conclusions are summarized as follows:

(1) A dramatic reduction in the friction coefficient to unac-ceptable levels for brake pads applications was evident butthe values were still within the range of dry sliding fric-tion and not in the range of hydrodynamic friction. NF1and NF4 brake pad showed low values of friction coef-ficient (0.02–0.05) at 2.1 m/s which was in the range offriction in mixed lubrication. Introducing water to the discdid not only cool the interface and suppresses the backtransfer film but it also showed a nearly linear relation-ship between wear rate and contact pressure for all FBPmaterials.

(2) The SEM micrographs showed that spraying water resultedin always a fresh surface of the brake pad specimen expe-rienced to wear due to a continuous washing, and removalof the debris from the interface. Besides, washing outdebris and suppressing the transfer film formation at theinterface, by spraying water, were not the only aspects ofdetermining the tribo behaviour of FBP at the wet interface,but the changing of lubrication from one mode to anotheris also a key role of determining this tribo process. Thepresence of the cavities or craters due to plateaus disinte-gration contributed differently during dry and wet braking.During dry braking these cavities collected in debris butduring wet braking, they acted as pockets of water to min-imize the effect of heat, pumping the water in and out toaccelerate removal of debris from the cavities.

(3) In summary, the wet friction and wear results revealedthat regardless the wear rate, the highest friction coef-ficient was consistently exhibited by NF2 at all pressureand speeds whereas the lowest wet friction coefficient wasobtained by NF1. Meanwhile, NF1 showed the best wearperformance among all the brake pads tested. Finally FBPsmaterials performed well during dry is not necessarily toperform well during wet and vice versa.

Acknowledgment

This work was supported by grant No. PR/2005/0450 from CRPP,MMU, Internal Funding Research Program (IFRP).

e c h n o l o g y 2 0 8 ( 2 0 0 8 ) 135–144

e f e r e n c e s

Anderson, A.E., 1992. Anderson, friction and wear of automotivebrakes. ASME Handb. 18, 569–577.

Bhushan, B., 2002. Introduction to Tribology. John Wiley & Sons,New York.

Bijwe, J., 1997. Composites as friction materials: recentdevelopments in non-asbestos fiber reinforced frictionmaterials—a review. Polym. Compos. 18 (3),378–396.

Blau, P.J., McLaughlin, J.C., 2003. Effects of water films and slidingspeed on the frictional behavior of truck disc brake materials.Tribol. Int. 36 (10), 709–715.

Cho, M.H., Kim, S.J., Kim, D., Jang, H., 2005. Effects of ingredientson tribological characteristics of a brake lining: anexperimental case study. Wear 258, 1682–1687.

N.S.M. El-Tayeb, K.W. Liew, May 2007. On dry and wet continuousbraking performance of frictional brake pad materials. J. Wear.Manuscript No. IH-4401, submitted for publication.

El-Tayeb, N.S.M., Liew, K.W., Venkatesh, V.C., 2006. Evaluation ofnew frictional brake pad materials. In: Proceedings of theInternational Conference on Manufacturing Science andTechnology, vols 28–30, Malaysia, pp.380–383.

Eriksson, M., Lundqvist, A., Jacobson, S., 2001. A study of theinfluence of humidity on the friction and squeal generation ofautomotive brake pads. Proc. Inst. Mech. Eng. Part D 215,329–342.

Eriksson, M., Bergman, F., Jacobson, S., 2002. On the nature oftribological contact in automotive brakes. Wear 252 (1/2),26–36.

Gurunath, P.V., Bijwe, J., 2007. Friction and wear studies onbrake-pad materials based on newly developed resin. Wear263, 1212–1219.

Kim, S.J., Jang, H., 2000. Friction and wear of friction materialscontaining two different phenolic resins reinforced witharamid pulp. Tribol. Int. 33, 477–484.

Kim, Y.C., Cho, M.H., Kim, S.J., Jang, H., 2008. The effect ofphenolic resin, potassium titanate, and CNSL on thetribological properties of brake friction materials. Wear 264,204–210.

Mutlu, I., Eldogan, O., Findik, F., 2006. Tribological properties ofsome phenolic composites suggested for automotive brakes.Tribol. Int. 39 (4), 317–325.

Shorowordi, K.M., Haseeb, A.S.M.A., Celis, J.P., 2004. Velocityeffects on the wear, friction and tribochemistry of aluminum

Yafei, L., 2006. A combinatorial approach for automotive frictionmaterials: effects of ingredients on friction performance.Compos. Sci. and Technol. 66, 591–598.