Different Foamed Metal–Reinforced Composites: Tribological Behavior and Temperature Field Simulation

This article was downloaded by: [Henan University]On: 02 July 2013, At: 00:12Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Tribology TransactionsPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/utrb20

Different Foamed Metal–Reinforced Composites:Tribological Behavior and Temperature Field SimulationKeju Ji a b , Yanqiu Xia b & Zhendong Dai aa Institute of Bio-Inspired Structure and Surface Engineering Academy of Frontier Science ,Nanjing University of Aeronautics and Astronautics 29 Yudao Street , Nanjing , 210016 ,Chinab State Key Laboratory of Solid Lubrication Lanzhou Institute of Chemical Physics , ChineseAcademy of Sciences Lanzhou , 730000 , ChinaAccepted author version posted online: 05 Feb 2013.Published online: 16 May 2013.

To cite this article: Keju Ji , Yanqiu Xia & Zhendong Dai (2013): Different Foamed Metal–Reinforced Composites: TribologicalBehavior and Temperature Field Simulation, Tribology Transactions, 56:4, 615-622

To link to this article: http://dx.doi.org/10.1080/10402004.2012.748235

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.

Tribology Transactions, 56: 615-622, 2013Copyright C© Society of Tribologists and Lubrication EngineersISSN: 1040-2004 print / 1547-397X onlineDOI: 10.1080/10402004.2012.748235

Different Foamed Metal–Reinforced Composites:Tribological Behavior and Temperature Field Simulation

KEJU JI,1,2 YANQIU XIA,2 and ZHENDONG DAI1

1Institute of Bio-Inspired Structure and Surface EngineeringAcademy of Frontier Science, Nanjing University of Aeronautics and Astronautics

29 Yudao Street, Nanjing 210016, China2State Key Laboratory of Solid Lubrication

Lanzhou Institute of Chemical Physics, Chinese Academy of SciencesLanzhou 730000, China

Four kinds of foamed metals (foamed Ni, FeNi, CuNi, and

Cu) filled with polytetrafluoroethylene (PTFE) and graphite

were developed as rubbing materials. These open-cell foamed

metals had the same interconnected three-dimensional (3D)

metallic skeletons. The friction and wear properties of the new

composites were investigated on an M-2000 friction and wear

tester. To study the influence of the metallic skeleton on the

contact between the sample and the rotating ring, an electric

field was imposed to monitor the formation of transfer film by

means of contact resistance. A thermocouple was embedded

into the sample to measure the temperature variation caused by

friction. Optical and scanning electron microscopes were used

to study the worn surface morphologies. The temperature field

of the sample and the effect of the metallic skeleton on tempera-

ture were calculated using finite element analysis (FEA). It was

found that the foamed metal–reinforced composites possessed

better thermal conductivity and wear resistance than homolo-

gous polymers, which was attributed to the following two main

reasons: first, the metallic skeletons are beneficial for restrain-

ing the plastic flow of polymeric matrix and second, the heat

can be conducted along the 3D supporting skeletons effectively.

KEY WORDS

Foamed Metal; Self-Lubricating Composites; Solid Lubri-cants; Thermal Conductivity; Electrical Conductivity; WearMechanisms

INTRODUCTION

Open-cell foamed metal is a porous structure with an inter-connected 3D metallic skeleton that possesses many fascinat-ing properties, such as a unique network structure, low density,specific mechanical performance, high specific surface area, andhigh conductivity (Kim, et al. (1)). To harness these properties,foamed metal has been applied as the template for the growth

Manuscript received August 24, 2012Manuscript accepted October 31, 2012

Review led by Cris Schwartz

of graphene foam (Chen, et al. (2)), as a structured supportfor chemical reactions (Sanz, et al. (3)), as electrodes for high-energy-density battery applications (Wang, et al. (4)), as chan-nels for heat transfer (Xu, et al. (5); Hong and Herling (6)), asmuffling devices (Alvarez-Arenas and Gonzalez (7)), and as fil-ters (Banhart and Weaire (8)). Recently, similar porous struc-tures have been studied as tribological composites to increasethe thermal conductivity and wear resistance of a matrix (Bur-ris and Sawyer (9); Qu, et al. (10); Wang and Liu (11)). Solid lu-bricants such as polytetrafluoroethylene (PTFE) and graphite areexcellent lubricants, but their thermal conductivity and strengthare low (Blanchet, et al. (12); Mu, et al. (13); Khedkar, et al.(14); Burris and Sawyer (15); Wang, et al. (16)). Considering thatfoamed metal possesses the properties of high thermal conductiv-ity and mechanic strength, embedding the foamed metal in a lu-bricant can reinforce its strength and thermal conductivity three-dimensionally.

We have discussed the tribological properties of the com-posites of foamed copper of different sizes filled with PTFE,MoS2, and graphite in our previous papers (Ji, et al. (17), (18)).For the continuation and expansion of serial experiments, inthis article we introduce four different kinds of foamed metals(foamed Ni, foamed FeNi, foamed CuNi, and foamed Cu) to theself-lubricating metal–polymer composites filled with lubricants(PTFE and graphite). The temperature field of the sample andthe effect of the metallic skeleton on temperature were investi-gated and calculated using finite element analysis (FEA), present-ing the roles that the metallic skeleton played in the friction pro-cess, especially on the heat conduction. In addition, the effects ofdifferent metals on the friction performance was evaluated. Theexperimental and simulation results are expected to provide somebasic guidance on the functional design of the conducting rubbingmaterial.

EXPERIMENTAL WORK

Materials Preparation

Four kinds of foamed metals (foamed Ni, foamed FeNi,foamed CuNi, and foamed Cu) used in the current tests werein our laboratory made using an electrochemical deposition

615

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

616 K. JI ET AL.

NOMENCLATURE

B = Width of the specimenb = Width of the wear scarc = Specific heatD = Distance between the supportsd = DeflectionE = Bending modulusF = Maximum loadH = Height of the bending specimenh = Convection coefficientL = Sliding distanceN = Loadp = Contact pressure

q∗ = Heat fluxR = Outer radius of the ringS = Surface areaT = Absolute temperature of the body surfaceT∞ = Surrounding temperatureV = Wear volume lossv = Sliding velocityW = Wear rateα = Thermal diffusivityη = Efficiency of the heat transformingλ = Thermal conductivityρ = Densityσ = Bending strengthσ0 = Stefan-Boltzmann’s constant

technique. These foamed metals are selected because they arethe main varieties by the same production process and have sim-ilar structures and properties (average pore diameter: 1.5 mm;porosity: 96%). They have open-celled structures composed ofdodecahedron-like cells, pentagonal or hexagonal faces, and hol-low metallic skeletons. These foamed metals have the same geo-metric size, and their characteristics and mechanical parametersare shown in Table 1. PTFE powder has an average particle size60 μm (Shanghai Electrochemical Plant), and graphite flake pow-der has an average particle size 7 μm (Shanghai Colloid-ChemicalPlant).

Foamed metal samples were prepared in suitable size and pre-treated in petroleum ether and dilute hydrochloric acid solutionto remove oil and oxide on the surface. The composites were pre-pared by filtration filling under negative pressure, compressionmolding, and sintering. Solid lubricants were mixed in ethanolusing a magnetic stirrer. Then, the mixed solution was filteredand dried using a vacuum filtration apparatus. During the pro-cess, a flexible pipe was used for imbibing and splashing the mix-ture of solid lubricants into the foamed metal sample while theorganic solvent was drawn through the filter core under negativepressure. Then the specimens were preformed using compressionmolding at 60 MPa for about 5 min to allow the preparation ofrectangular and cylindrical samples of sizes 6 mm × 7 mm ×30 mm and Ø12.5 mm × 2 mm at room temperature. Finally,these blocks were sintered at 330◦C for 60 min and 370◦C for60 min in a sintering furnace and then cooled freely to room tem-perature (Ji, et al. (13)). During the process, the foamed metal,acting as a porous support, was compacted combined with thesolid lubricant fillers and compressed to 55% of its original size.

The homologous polymers without foamed metal were madeusing the compression molding and sintering technique. The com-position of each specimen was distinguished by the graphite con-

tent (0–100 vol%), in which the foamed metal (9 vol%) as the ma-trix was not included (P/G: PTFE + graphite; P/Gx + Ni: foamedNi filled with component PTFE + x% graphite; and so forth).

Friction and Wear Testing

The friction and wear behaviors of the specimens were inves-tigated on an M-2000 friction and wear tester (Fig. 1). The blockspecimen sliding against the AISI 52100 steel in a block-on-ringconfiguration is shown in Fig. 1a. An electric field was imposedbetween the specimen and ring to monitor the tribochemical re-action by means of contact resistance; see Fig. 1b; the inset showsthe surface morphology of a typical composite of foamed Cu filledwith P/G40 (60% PTFE + 40% graphite) before wear testing.We can see that the interspaces of foamed Cu were effectivelyfilled with solid lubricants and the hollow metallic skeletons werewidely distributed throughout the body, which provided conduc-tive pathways and supporting function.

The friction and wear tests were conducted selectively at twoload levels (200 and 500 N) and velocities (0.424 and 0.856 m/s).The testing duration was 60 min. Before each test, the surfacesof the specimen and counterpart ring were polished to a surfaceroughness of Ra 0.10–0.15 μm with No. 1000 grade SiC abrasivepaper. The ambient temperature was roughly 20◦C and the rel-ative humidity was about 20%. The friction force was measuredwith a torque shaft, and the friction coefficient was calculated andrecorded by the computer linked with a sensor. The wear rate ofthe specimen was determined by measuring the volume loss af-ter the friction test. The wear scar width was measured using avernier caliper and three repeated tests were conducted.

The wear rate of each specimen was obtained by the followingformula:

W = VLN

, [1]

TABLE 1—CHARACTERISTIC PARAMETERS OF THE FOAMED METALS

Foamed Element Content Apparent Density Bending Strength Compressive YieldMaterial (%) (kg/m3) (MPa) Strength (MPa)

Ni 100 350 0.65 0.86FeNi 50:50 330 2.28 1.80CuNi 20:80 350 1.10 1.08Cu 100 350 0.52 0.58

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

Foamed Metal–Reinforced Material for Tribology 617

Fig. 1—Schematic diagram of M-2000 friction and wear tester: (a) actual contact; (b) layout for the calculation of contact resistance and the surfacemorphology of a typical specimen; and (c) worn specimen and layout for the thermocouple probe (color figure available online).

where L equals the velocity multiplied by the time of the frictionprocess (s), referring to the distance of the friction (m), N is load(N), and V is the wear volume loss (mm3) and has the relation-ship:

V = B

[πR2

180arcsin

(b

2R

)− b

2

√R2 − b2

4

], [2]

where B is the width of the block specimen (7 mm), R is the outerradius of the steel ring (20 mm in this trial), b is the width of thewear scar (mm), and the result of the arcsine is in degrees (Wang,et al. (19)).

The worn surfaces of specimens and the transfer films formedon the surfaces of the counterpart steel rings were observed usinga scanning electron microscope (SEM, JSM-5600LV, Jeol, Japan)and a digital microscope (VHX-600E, Keyence, Japan).

Friction Heat Testing and Thermal Conductivity

Several researchers have investigated the distribution of tem-perature in friction pairs (Abdel-Aal (20)). The friction heat gen-erates both on the surfaces of the sliding pair and in the interiorof the rubbing viscoelastic-polymer. Wieleba (21) found that theplace where the maximum temperature occurred inside the steelor polymeric element could be 1.5 mm from the surface. In thisstudy, measurement of the friction temperature was carried outusing a thermocouple embedded in the specimen (Fig. 1c). A 3-mm-deep, 0.32-mm inner diameter hole was drilled at axial lo-cations of 2 mm from the surface to receive the thermocouplefor temperature measurement. A small amount of tin grains withan average particle size of 45 μm was placed in the hole to en-hance thermal conductivity before the thermocouple was embed-ded. Then the hole was blocked using a quick-drying adhesive assoon as the thermocouple was inserted. The test results would beadopted as the reference for finite element simulation.

The efficiency of the heat exchange and dissipation capacityof selective specimens (P/G40 + Cu and P/G40) was character-ized by the thermal diffusivity using a testing apparatus (LFA 447NanoFlash, Netzsch, Germany). Xenon flashlamps fired a pulse atthe lower surface of the sample, while the infrared detector mea-sured the temperature increase at the top surface. The thermaldiffusivity of the sample was then determined using the software.Specific heat was measured by comparing the actual temperatureincrease of the sample to the temperature increase of a referencesample of known specific heat. Thermal diffusivity (α) and spe-cific heat (cp) can be measured simultaneously, and the thermalconductivity (λ) can be calculated using these values and the bulkdensity (ρ) from the equation:

λ = α · cp · ρ [3]

The specimens used for testing were Ø12.5 mm × 2 mm in size.Two specimens were used per immersion time, and each exper-imental point was obtained as an average reading of three lasershots per temperature and per specimen.

RESULTS AND DISCUSSION

Friction and Wear Behavior

Figure 2 shows the friction coefficients, wear rates as a func-tion of graphite content, and variations in the friction coefficientswith sliding duration of the foamed metal–reinforced compositesand the homologous polymers at a velocity of 0.424 m/s at load of200 N.

Figure 2a shows that the variations in friction coefficientsof the foamed metal–reinforced composites were not consis-tent with those of the homologous polymers as the graphitecontent changed. When the graphite content was 0%, thefoamed metal–reinforced composites had lower friction coef-ficients than the pure PTFE. When the graphite content wasgreater than about 10%, the friction coefficients of the foamed

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

618 K. JI ET AL.

Fig. 2—(a) Friction coefficients, (b) wear rates as a function of graphite content, and (c) variations in friction coefficients with sliding duration of thefoamed metal–reinforced composites and the homologous polymers at a velocity of 0.424 m/s and load of 200 N (P/Gx: PTFE + x% graphite; P/G+ Ni: foamed Ni filled with component PTFE + x% graphite; and so forth) (color figure available online).

metal–reinforced composites were higher than those of the ho-mologous polymers and decreased almost monotonically withthe increase in graphite content. When the graphite content wasgreater than 60%, the bond strength was not enough to holdthe graphite particles, so the homologous polymers were nolonger sufficient to finish the 60-min test of the fatigue wear.The wear rates of the foamed metal–reinforced composites de-creased compared to those of the homologous polymers (Fig.2b). The foamed Cu–reinforced composite, particularly that filledwith P/G40, showed the best friction and wear performance, andthe others (foamed Ni, foamed FeNi, and foamed CuNi) werethe same. The influence of foamed metal on the performance ofcomposites was related to the ductility and toughness of the par-ent metal. The copper particles were more likely to succumb tothe shear stress and showed a lower friction coefficient, whereasthe others achieved a balance between ductility and toughness.Graphite can improve the compactness of PTFE and PTFE-basedcomposites (Yan, et al. (22)), act as an effective barrier to preventlarge-scale fragmentation of PTFE, and reduce the ultimate elon-gation to fracture in the abrasive wear process, which promotesthe net adhesion of the transferred film to the countersurface (Bi-jwe, et al. (23)).

Figure 2c shows the continuous variations in friction coeffi-cients of the P/G40 and P/G40 + Cu (composite of foamed Cufilled with P/G40) with sliding time at 200 N, 0.424 m/s. The fric-tion coefficients showed similar trends in the steady wear stage,except for some fluctuation of the foamed Cu–reinforced com-posite at the running-in stage. The temporary increase in fric-tion coefficient was caused by the local direct contact betweenthe copper and the steel ring at the running-in stage. Then thelubricants quickly and ceaselessly intervened into the frictionalinterface, forming a transfer layer due to the deformation andshear stress. In addition, the friction process entered the steadystage.

Figure 3 shows the friction coefficients, wear rates, and resis-tances of the foamed metal–reinforced composite (foamed met-als filled with P/G40) and the homologous polymer (P/G40) at500 N, 0.424 m/s. Figure 3a is a supplement showing the tribolog-ical performance under higher load. The variable trends in fric-tion coefficient and wear rate at 500 N exhibited slightly differentcharacteristics than those at 200 N. Both friction coefficients andthe wear rates decreased as the foamed metals were introducedinto the composites compared to the homologous polymer P/G40under higher load.

Fig. 3—Values of the friction coefficients, wear rates, and resistances of the selective specimens (P/G40 and the foamed metals filled with P/G40) at avelocity of 0.424 m/s and load of 500 N (color figure available online).

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

Foamed Metal–Reinforced Material for Tribology 619

The foamed metals with their natural porosity served as 3Dsupporting skeletons and were beneficial for restraining the plas-tic flow of the polymeric matrix and served as a lubricant reservoirfor the contact surface, especially under higher load. This was alsothe main reason why the foamed metal composites were still firmeven at 100% graphite content without the binder PTFE.

Figure 3b shows the variation in DC contact resistance foreach specimen at 0.424 m/s, 500 N. The contact resistance be-tween the friction surfaces was mainly dependent on the contactarea. The two friction surfaces formed a conductivity networkat some convex points and facets named as contact spots. Thecontact spots consisted of dielectric/conducting film (Myshkinaand Konchits (24)) and wear debris. A further study on the con-tact resistance indicated that the frictional interface layer had ahigher resistance during the wear stage and lower resistance be-fore and after the run. An explanation for this is that the poly-meric lubricants quickly moved into the interface by the thermalexpansion of the PTFE matrix and shear stress (McCook, et al.

(25)) and then the area and depth of contact on the frictional in-terface decreased correspondingly due to the strain hysteresis ofthe polymeric composite. The strain hysteresis occurred to dis-sipate the mechanical energy when cyclic loading was applied(Wieleba (21)). The different compositions and properties of thewear debris were responsible for the specific variations among thefoamed metal–reinforced composites.

Wear Mechanism Analysis

Figure 4 shows the micrographs of the worn surfaces ofP/G40 and composites of foamed metals filled with P/G40 un-der a load of 500 N at sliding speeds of 0.424 and 0.856 m/s,respectively. Figures 4a, 4b, 4c, 4e, and 4f show that the fric-tion surfaces of foamed metal–reinforced composites were spreadwith prominent veins of various metals. Some fragmentary lu-bricants attached to and/or penetrated the surfaces and cavitiesof the visible metallic skeletons. These cavities belonging to thehollow skeletons served to collect and store lubricants on the

Fig. 4—Micrographs of the worn surfaces of specimens and steel rotating rings under a load of 500 N at sliding speeds of 0.424 and 0.856 m/s: compositesof (a) foamed Ni filled with P/G40 at 0.424 m/s; (b) foamed FeNi at 0.424 m/s; (c) foamed CuNi at 0.424 m/s; (d) P/G40 at 0.424 m/s; (e) foamed Cu at0.424 m/s; (f) foamed Cu at 0.856 m/s; and (g) and (h) SEM micrographs of two characteristic regions on the worn surfaces; the sliding directionis shown by white arrows (color figure available online).

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

620 K. JI ET AL.

friction surfaces (Fig. 4h). Metal and polymeric wear debris wasadhered on the surfaces of the rotating rings, which can bedistinguished by their colors. As a representative of foamedmetal–reinforced composites at 500 N, 0.856 m/s, the wear mor-phologies of foamed Cu–reinforced composite and its rotatingring are displayed in Fig. 4f. The ductile deformation and adhe-sion of copper were more severe due to the higher frictional heatcompared to those at 0.424 m/s. The wear morphology of the ho-mologous polymer P/G40 is shown in Fig. 4d as a reference.

A copper–polymer interfacial transition zone on the worn sur-face of composite of foamed Cu filled with P/G40 is shown inFig. 4g. The worn surface can be divided into three areas: metal-lic skeleton area, polymeric area, and transitional area. It can beseen that the wear scar on the metal was smoother and showedless apparently plucked and ploughed marks than the polymericarea, except for some small cavities. Some marks of hard parti-cle erosion and plastic furrow deformation can be seen on thepolymeric areas, and the plastic flow can also be seen along thesliding direction through the contacting surface layer originatedand extended by the friction heat and compressive and tensilestress. The damage feature in the polymeric area was consistentwith the wear track of PTFE, which also presented ploughing andplastic flow (Lu, et al. (26)). The metallic skeleton, acting as theporous support, and the border resisted plastic flow of the poly-

meric matrix, so the foamed metal–reinforced composites showedgood wear resistance.

Thermal Effect Analysis

In order to better understand the influence of metallic skele-tons on the heat transfer performance of the composites, the tem-perature field of the composites and the effect of metallic skele-ton on temperature were simulated using FEA. According tothe morphology of foamed metal (here foamed Cu) with hollowmetallic skeletons (Fig. 5a), a hollow geometry was modeled asthe cell of foamed metal. This cell was shaped in accordance withthe connection between the exact center and the vertices of theregular tetrahedron structure. The circle section was used herebecause the polyurethane template of foamed metal possessed amildly disordered structure, and self-adaption and correction oc-curred during the foaming procedure of polyurethane. The com-bination of foamed cells with circle section had great flexibility insimulating skeleton topology on computer.

To model the specimen cell, the composite of foamed Cu filledwith P/G40 was selected as the simulated target. A packagedspecimen cell and the combination of friction pairs are shown inFig. 5b. The proportion of copper exposed on the friction surface�2 was about 3% (S�1 /S�2

, where S�1 is the surface area of �1, and

Fig. 5—Steady-state thermal analysis of the friction pair: (a) morphology of foamed Cu and the cell of foamed metal, (b) cell combination and package offriction pairs, (c) simulative temperature distribution in the frictional system, and (d) variations in the measuring temperature with sliding durationat the location of 2 mm at 500 N, 0.424 m/s (color figure available online).

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

Foamed Metal–Reinforced Material for Tribology 621

TABLE 2—THERMAL PROPERTIES OF FRICTION PAIRS AND AIR

ThermalDensity Specific Heat Conductivity

Materials (kg/m3) [J/(kg·K)] [W/(m·K)]

P/G40 + Cu 2,800 850 2.4P/G40 2,200 912 0.96Copper 8,900 385 401Rotating ring cell 7,810 490 45Air 1.29 1,005 0.027

S�2 is the area of �2), and the thermal properties of specimen cell,rotating ring cell, and air are shown in Table 2.

Steady-state thermal analysis rather than transient analysiswas performed using the finite element software ANSYS for dis-cussion of heat distribution. During sliding, the mechanical en-ergy is mainly dissipated as heat (Rubtsov and Kolubaev (27)),which may lead to the deterioration of mechanical properties,which in turn often decreases the friction and wear performance.This is a prominent problem in applications involving high con-tact pressures (p) and sliding velocities (v), because this results ina high heat flux (q∗); that is, the rate of dissipated energy per unitapparent contact area (Larsen, et al. (28)):

q∗ = μ pv [4]

Here μ = 0.14 is the average friction coefficient of the copperskeleton and homologous polymer based on the test, p = 8.9 MPa(500 N loaded) is the contact pressure, and v = 0.424 m/s is thesliding velocity.

According to Sorina-Muller, et al. (29), the fraction of friction-ally dissipated energy converted into heat, η, is about 90%. Tak-ing the viscoelasticity of the polymer into consideration, a smallproportion of frictional energy is converted into internal energyof plastic strain. Therefore, the actual efficiency of the heat trans-forming is η< 90% and the 90% used here is present as the upperbound.

The heat distribution ratio between friction pairs can be cal-culated using the following formula (Gao, et al. (30)):

q1

q2= q∗

1

q∗2

=√

c1ρ1λ1

c2ρ2λ2. [5]

Here, q1 and q2 are the quantity of heat of each friction pair; cis the specific heat; ρ is the density; and λ is the thermal con-ductivity. The indices 1 and 2 denote the friction pair specimencell and ring cell, respectively. The total heat flux of dissipation isη· q∗ = q∗

1 + q∗2. Because the steel rotating ring was sliding against

the specimen at 0.424 m/s, the heat flux loaded onto the frictionalsurface �3 of the ring (Fig. 5b) was determined by an equivalentheat flux E−q∗

2 = q∗2 ·S�2

Scircle, where Scircle is the area of external sur-

faces of the ring.The effective heat dissipation mainly relied on the natural

convection heat transfer for the specimen cell and the forced con-vection for the ring cell. When the sliding velocity v = 0.403 m/s,the heat convection coefficient of the rotating ring hring was de-termined by

hring = Nu ∗ Kair/2R. [6]

Here, Nu is the Nusselt number, Kair is the thermal conductivityof air, and R is the radius of the specimen (here to the outer radiusof the ring; Incropera and DeWitt (31)).

The radiation boundary condition is defined by

q∗r = σ0ε

(T4 − T4

∞), [7]

where q∗r is the heat flux (heat flow velocity per unit area) for radi-

ation (W/m2), σ0 is the Stefan-Boltzmann constant (5.669 × 10−8

W/(m2·K4)), T is the absolute temperature of the body surface,and T∞ is the surrounding temperature (K).

Figure 5c shows the temperature distribution of the frictionpair on the basis of FEA. The temperature on the upper surfaceof the specimen cell (140 ± 5◦C) was higher than the actual mea-sured value of 120◦C (shown in Fig. 5d). This is acceptable be-cause the heat efficiency η = 90% is present as the upper bound.This simulated image reflects the influence of the metallic skele-ton on the temperature gradient of the specimen cell. Consider-ing that the wear debris with a high temperature will remove heatfrom the friction interface, and the wear rate of P/G40 generatedmore wear debris than P/G40 + Cu, the heat was conducted alongor through the 3D supporting skeletons effectively, regardless ofthe heat source effect of the metallic skeletons on the friction sur-face. The variations in temperature with sliding duration shownin Fig. 5d show the temperature reaches to a stable state a halfhour later, and a balance between heat generation and dissipa-tion can be seen in the friction process. During the early stage,the heat source effect of the skeletons was dominant; then withthe formation of a transfer layer, the heat dissipation emerged asthe heat source effect weakened; eventually, a dynamic equilib-rium was reached. This is also why the composite of foamed Cufilled with P/G40 had a slightly lower friction temperature thanthe homologous polymer P/G40.

CONCLUSIONS

Four kinds of foamed metals were filled with solid lubricantsto fabricate the foamed metal–reinforced composites. Becausethe interconnected metallic skeletons acting as the porous sup-port and the border could resist the plastic flow of polymer,the foamed metal–reinforced composites possessed better ther-mal conductivity and wear resistance compared to the homolo-gous polymers. The composite of foamed Cu filled with P/G40achieved the best tribological properties among the compositesdue to good ductility and toughness of copper. The steady-statethermal analysis showed that the foamed metal–reinforced com-posites were effective in transmitting heat along the intercon-nected metallic skeletons, despite the heat source effect of themetallic skeletons on the friction surface. A balance between heatgeneration and dissipation for the skeletons can be seen in thefriction process.

ACKNOWLEDGEMENTS

This work was supported by the National Natural ScienceFoundation of China (Grant No. 90916021 and 50823008), fund-ing from the Jiangsu Innovation Program for Graduate Edu-cation (CXLX12 0141), the Fundamental Research Funds forthe Central Universities, and the “Hundred Talents Program”

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013

622 K. JI ET AL.

of CAS. The authors gratefully acknowledge helpful discussionswith Zhouyi Wang and Weigen Shan.

REFERENCES(1) Kim, J. H., Kim, R. H., and Kwon, H. S. (2008), “Preparation of

Copper Foam with 3-Dimensionally Interconnected Spherical Pore Net-work by Electrodeposition,” Electrochemistry Communications, 10(8), pp1148–1151.

(2) Chen, Z. P., Ren, W. C., Gao, L. B., Liu, B. L., Pei, S. F., and Cheng, H.M. (2011), “Three-Dimensional Flexible and Conductive InterconnectedGraphene Networks Grown by Chemical Vapour Deposition,” Nature Ma-terials, 10, pp 424–428.

(3) Sanz, O., Echave, F. J., Sanchez, M., Monzon, A., and Montes, M. (2008),“Aluminium Foams as Structured Supports for Volatile Organic Com-pounds (VOCs) Oxidation,” Applied Catalysis A: General, 340(1), pp125–132.

(4) Wang, J.S., Liu, P., Sherman, E., Verbrugge, M., and Tataria, H. (2011),“Formulation and Characterization of Ultra-Thick Electrodes for HighEnergy Lithium-Ion Batteries Employing Tailored Metal Foams,” Journalof Power Sources, 196, pp 8714–8718.

(5) Xu, J. L., Ji, X. B., Zhang, W., and Liu, G. H. (2008), “Pool Boiling HeatTransfer of Ultra-Light Copper Foam with Open Cells,” InternationalJournal of Multiphase Flow, 34(11), pp 1008–1022.

(6) Hong, S. T. and Herling, D. R. (2006), “Open-Cell Aluminum FoamsFilled with Phase Change Materials as Compact Heat Sinks,” Scripta Ma-terialia, 55(10), pp 887–890.

(7) Alvarez-Arenas, T. E. G. and Gonzalez, I. G. (2007), “Spatial Normaliza-tion of the High-Frequency Ultrasound Energy Loss in Open-Cell Foams,”Applied Physics Letters, 90, p 201903.

(8) Banhart, J. and Weaire, D. (2002), “On the Road Again: Metal Foams FindFavor,” Physics Today, 55(7), pp 37–42.

(9) Burris, D. L. and Sawyer, W. G. (2008), “Hierarchically ConstructedMetal Foam/Polymer Composite for High Thermal Conductivity,” Wear,264(3–4), pp 374–380.

(10) Qu, J., Blau, P. J., Klett, J., and Jolly, B. (2004), “Sliding Friction andWear Characteristics of Novel Graphitic Foam Materials,” Tribology Let-ters, 17(4), pp 879–886.

(11) Wang, Y. J. and Liu, Z. M. (2008), “Tribological Properties of High Tem-perature Self-Lubrication Metal Ceramics with an Interpenetrating Net-work,” Wear, 265(11–12), pp 1720–1726.

(12) Blanchet, T. A., Kandanur, S. S., and Schadler, L. S. (2010), “Coupled Ef-fect of Filler Content and Countersurface Roughness on PTFE Nanocom-posite Wear Resistance,” Tribology Letters, 40, pp 11–21.

(13) Mu, L. W., Feng, X., Zhu, J. H., Wang, H. Y., Sun, Q. J., Shi, Y. J., and Lu,X. H. (2010), “Comparative Study of Tribological Properties of DifferentFibers Reinforced PTFE/PEEK Composites at Elevated Temperatures,”Tribology Transactions, 53(2), pp 189–194.

(14) Khedkar, J., Negulescu, I., and Meletis, E. I. (2002), “Sliding Wear Behav-ior of PTFE Composites,” Wear, 252, pp 361–369.

(15) Burris, D. L. and Sawyer, W. G. (2006), “A Low Friction and Ultra LowWear Rate PEEK/PTFE Composite,” Wear, 261, pp 410–418.

(16) Wang, L. Q., Jia, X. M., Cui, L., and Chen, G. C. (2008), “Effect of AramidFiber and ZnO Nanoparticles on Friction and Wear of PTFE Compositesin Dry and LN2 Conditions,” Tribology Transactions, 52(1), pp 59–65.

(17) Ji, K. J., Shan, W. G., Xia, Y. Q., and Dai, Z. D. (2012), “The Tribologi-cal Behaviors of Self-Lubricating Composites as Filler in Copper Foam,”Tribology Transactions, 55(1), pp 20–31.

(18) Ji, K. J., Xia, Y. Q., Wang, H. L., and Dai, Z. D. (2012), “Foamed-MetalReinforced Material: Tribological Behaviours of Foamed-Copper Filledwith Polytetrafluoroethylene and Graphite.” Journal of Engineering Tri-bology, 226(2), pp 123–137.

(19) Wang, H. L., Li, H. H., and Yan, F. Y. (2005), “Synthesis and Tribologi-cal Behavior of Metakaolinite-Based Geopolymer Composites,” MaterialsLetters, 59(29–30), pp 3976–3981.

(20) Abdel-Aal, H. A. (1999), “On the Bulk Temperatures of Dry RubbingMetallic Solid Pairs,” International Communications in Heat and MassTransfer, 26(4), pp 587–596.

(21) Wieleba, W. (2005), “The Role of Internal Friction in the Process of En-ergy Dissipation during PTFE Composite Sliding against Steel,” Wear,258(5–6), pp 870–876.

(22) Yan, F. Y., Xue, Q. J., and Yang, S. R. (1996), “Debris Formation Processof PTFE and Its Composites,” Journal of Applied Polymer Science, 61(7),pp 1223–1229.

(23) Bijwe, J., Logani, C. M., and Tewari, U. S. (1990), “Influence of Fillers andFibre Reinforcement on Abrasive Wear Resistance of Some PolymericComposites,” Wear, 138(1–2), pp 77–92.

(24) Myshkina, N. K. and Konchits, V. V. (1994), “Evaluation of the Interfaceat Boundary Lubrication Using the Measurement of Electric Conductiv-ity,” Wear, 172(1), pp 29–40.

(25) McCook, N. L., Burris, D. L., Bourne, G. R., Steffens, J., Hanrahan, J. R.,and Sawyer, W. G. (2005), “Wear Resistant Solid Lubricant Coating Madefrom PTFE and Epoxy,” Tribology Letters, 18(1), pp 119–124.

(26) Lu, X. C., Wen, S. Z., Tong, J., Chen, Y. T., and Ren, L. Q. (1996), “Wet-tability, Soil Adhesion, Abrasion and Friction Wear of PTFE (+ PPS) +Al2O3 Composites,” Wear, 193, 48–55.

(27) Rubtsov, V. E. and Kolubaev, A. V. (2007), “Study of Plastic Shear De-formation in Surface Layer at Friction. Simulation Results. Part I. ModelDescription,” Journal of Friction and Wear, 28(1), pp 65–78.

(28) Larsen, T. Ø., Andersen, T. L., Thorning, B., and Vigild, M. E. (2008),“The Effect of Particle Addition and Fibrous Reinforcement on Epoxy-Matrix Composites for Severe Sliding Conditions,” Wear, 264, pp 857–868.

(29) Sorina-Muller, J., Rettenmayr, M., Schneefeld, D., Roder, O., and Fried,W. (2010), “FEM Simulation of the Linear Friction Welding of TitaniumAlloys,” Computational Materials Science, 48, pp 749–758.

(30) Gao, J. Q., Lee, S. C., Ai, X. L., and Nixon, H. (2000), “An FFT-BasedTransient Flash Temperature Model for General Three-DimensionalRough Surface Contacts,” Journal of Tribology, 122, pp 519–523.

(31) Incropera, F. P. and DeWitt, D. P. (1996), Introduction to Heat Transfer,3rd ed., John Wiley & Sons: New York.

Dow

nloa

ded

by [

Hen

an U

nive

rsity

] at

00:

12 0

2 Ju

ly 2

013