稀有金属(英文版) 2019,38(09),885-891

Surface layer structure change and tribological property of Cu/FeS composite under electric field

Ming-Yu Hu Jie Yu Xiao-Long Zhou Zhen-Jun Hong Wei Ye

Faculty of Material Science and Engineering,Kunming University of Science and Technology

Inner Mongolia First Machinery Group Corporation

作者简介：Jie Yu e-mail:yujieone@163.com;

收稿日期：21 February 2016

基金：financially supported by the State Key Laboratory of Advanced Technologies for Comprehensive Utilization of Platinum Metals(No.SKL-SPM-201510);

Surface layer structure change and tribological property of Cu/FeS composite under electric field

Ming-Yu Hu Jie Yu Xiao-Long Zhou Zhen-Jun Hong Wei Ye

Faculty of Material Science and Engineering,Kunming University of Science and Technology

Inner Mongolia First Machinery Group Corporation

Abstract：

The microstructural evolvement and friction performance of Cu/FeS self-lubricating composites within electric field were studied by molecular dynamics simulation.The atoms distribution and movement of the Cu/FeS composite under different electric field strengths were considered.The results show that some Fe atoms and S atoms break away from the original structure and move along the electric field,but the movement of Fe atoms and S atoms is not synchronized in stronger electric field strength.Thus,the whole material system becomes chaotic and its temperature rises.The unevenly electric field force appears in the material for the nonuniform FeS distribution,so the internal stress of the composite is generated.For the internal solid lubricating phases move along the direction of the electric field to the surface of the composite,the adhesive wear and the friction coefficient reduce.The movement of the phases leaves over some voids in the composite,which induces arc wear easily.

Keyword：

Molecular dynamics; Cu/FeS composite; Electric field; Friction and wear;

Received： 21 February 2016

1 Introduction

Cu/FeS composite materials are solid self-lubricating metal matrix composites with their excellent friction properties and have very good prospects as railroad friction materials [ 1, 2, 3] .In the dry friction experiments,the friction pair Cu//Cu/FeS exhibits good friction and wear properties [ 4] .He et al. [ 5, 6] studied the friction coefficient and contact resistance of copper versus copper/carbon composite material,lubricated copper versus copper and unlubricated copper versus copper with electrical current.It was found that the contact resistance of un lubricated copper versus copper was lower than those of the other two friction pairs,but the friction coefficient was larger,which was just opposite to those of lubricated copper versus copper and copper versus copper/carbon composite.In other words,these friction pairs could not meet the requirements of low contact resistance and low friction coefficient at the same time.However,the materials with a self-lubricating function do not need a greasy film at the interface for reducing the friction coefficient [ 7, 8] and have better conductivity for efficient power transfer.

In addition to dry friction property,current-carrier friction,the friction wearing abrasion in the electric field,has to be considered.It is generally believed that the wear volume losses increase with current intensity (the influence of the arc erosion) [ 9, 10, 11] and the arc erosion is affected by the contact normal force [ 12, 13] .Dong et al. [ 10] investigated the friction and wear characteristics of aluminumstainless steel composite used in conductor rail and collector shoe.And they proved that there was a threshold value of normal stress exiting in the friction and wear with current.If normal stress is larger than or equal to the threshold value,the contact between the friction pairs is resistance contact,thus the sliding wear with adhesion and abrasion is the major wear mechanisms and wear volume losses slightly with the normal force increasing.However,if the normal stress is less than the threshold value,the contact is capacitance contact,so the arc erosion is dominant and the wear volume losses intensely with the normal force decreasing.Mechanical wear and arc erosion would be much less if working under the threshold stress,and it would make electrical power supply stable and wear volume losses lower.Bucca and Collina [ 14] built a model in order to define the relationship between the wear rate of the contact wire and the main parameters of collection (sliding speed,contact force,electric current).All of these investigations were mainly focused on the influence of electric current on the friction coefficient,contact resistance,electric arc,etc.However,the change of micromorphology cannot be displayed in detail by the normal experiment,so the computer simulation becomes an effective analysis method.

Using molecular dynamics,Vigonski et al. [ 15] simulated the behaviors of a near-surface Fe precipitate in Cu material under a high electric field.They considered that the presence of the precipitation caused the nucleation dislocated in the material,leading to the appearance of atomic steps on the surface.The steps could also form protruding plateaus to enhance the electric field locally and then lead to the breakdown of the device.They succeeded in explaining the failure mechanism of the device under high electric field.

The friction and mechanical properties of Cu/FeS by experiment and molecular dynamics simulation with no electric field have been studied [ 1, 2, 3] .However,there are few reports about the influences of electric field on surface layer structure change of Cu/FeS composites and the friction and wear mechanism of Cu/FeS composites in the electric field is unclear.

In this paper,the Cu/FeS composite was prepared by in situ sintering.Based on the scanning electron microscope (SEM,PHILIPS XL30ESEM) analysis of Cu/FeS-15 wt%composite shown in Fig.1,the geometrical configuration of Cu/FeS composites under electric field was simulated by molecular dynamics simulation.Furthermore,the friction and wear mechanism of Cu/FeS composites under electric field is explored and the ways to improve the wear performance with the electric current were sought.

Fig.1 SEM image of Cu/FeS-15 wt%composite

2 Potential function and atomic model

The interaction of copper atoms was modeled with the Morse potential function [ 16, 17] ,which can be expressed as:

where D_e represents the binding energy,αis the elastic modulus,r is the distance between two atoms,r₀ is the equilibrium distance within two atoms,and m is the appointed parameter.For the interaction of copper atoms,D_e=5.6774×10^-20 J,α=13.39253 nm^-1,r₀=0.2885009 nmand m=2.Morse potential is often used to analyze the mechanical performances of fcc metal materials,and the results of simulation were well agreed with experimental data [ 18, 19] .The interactions between Fe-S,Fe-Cu and S-Cu were defined as Lennard-Jones (12-6) potential function [ 20] as following:

whereε_ij andσ_ij are characteristic binding energy and equilibrium distance between atom i and atomj,respectively.As Table 1 shows,these two parameters come from the result of Philpott et al. [ 21] by the use of Lorentz-Berthelot combining rules:

whereε_ii,S_jj andσ_iiσ_jj are characteristic binding energy and equilibrium distance between the same atoms,respectively.Philpott et al. [ 21] validated the interactions between Fe-S by Lennard-Jones potential.And in this paper,before the further simulation,the interactions between Fe-Cu and S-Cu which described by LennardJones potential were verified by testing the stabilization of the lattice and a good result was obtained.Hong et al. [ 1] have also used these potential to describe the system of Cu/FeS,and the simulation result was acceptable.It should be noted that the cutoff radius of Morse and LJ potential was set to 1 nm.

  下载原图

Table 1 Parameters of Lennard-Jones potential

The Materials Explorer software which is widely used in the molecular dynamics calculations was employed.The initial charge of Cu atoms was set to be 0,Fe was+2,and S was-2.The charged atoms are subject to a force that is proportional to the charge,from the uniform electrostatic field,in addition to the Coulomb interactions from the point charges of the surrounding atoms.Figure 2shows the atomic model after 1×10⁴ step crystal relaxation,in which the dispersed solid lubricating phase FeS distributed in the copper matrix with neat and stable interface.The size of the model was 2.53,2.53 and2.33 nm along[100](X-axis), [ 10] (Y-axis) and [ 1] (Z-axis) crystal orientation,respectively.The number of particles,temperature and pressure (NTP) ensemble was selected,and velocity-scaling theory [ 22, 23] was employed as temperature control method.The numerical integral method was used to solve the differential equation,and the result was corrected through 5-order Gear predictor-corrector method.The other specific parameters are shown in Table 2.

Fig.2 Atomic model of Cu/FeS composite

3 Results and discussion

3.1 Atoms distribution of Cu/FeS composite in different strength electric fields

High electric field can change the internal atom distribution even the structure in composite material.The dipole orientation of polar molecule FeS would always tend to be stretched or even broken within electric field.The pair correlation function,g(r),is often used to describe the probability density of atom distribution at r,given that there is an atom at the coordinate origin [ 24, 25] .The g(r) curves of Cu-Fe and Cu-S in the different strength electric fields are shown in Fig.3.In Fig.3a,the g(r) values of Cu-Fe and Cu-S in the electric field E1 are both lower than that in E2,and so is the area of the first peak.This result shows that the numbers of Fe atoms and S atoms decrease in the same radius,and the distances of CuFe and Cu-S increase in E1.In other words,the solid lubricating phase FeS and matric Cu become far away or even separate completely.In Fig.3b,the pair correlation function of Cu-Fe and Cu-S in E2 is similar to that of E5.So the strength of electric field E2 has no effect on the internal structure of the composite and the solid lubricating phase FeS and matric Cu do not separate.

The mean square displacement (M.S.D) [ 26, 27] of Fe atoms and S atoms in different electric field strengths is shown in Fig.4.The velocity and displacement of Fe,S atoms in E1 are quite different from others.Especially,after 4 ps,the distance between Fe atom and S atom increases gradually,indicating that the movement of Fe atoms and S atoms is not synchronized in E1.And eventually,FeS may be ionized by the strong electric field and the composite will be destroyed.Compared to that in E1,the movement of Fe atoms and S atoms in E2,E3,E4,E5,as shown in Fig.4b,is synchronized although the displacement of Fe atom and S atom is also large in E2,and so the composite is intact basically.

3.2 Structure and atom movement of Cu/FeS composite in different strength electric fields

In the high electric field,the movement of internal atom results in the structure change.The atomic tracks are shown in Fig.5a-e,the system begins to rotate more obviously with the gradually increased electric field strength E5,E4,E3,E2 and E1.The reason is that the stronger the electric field is,the larger displacement the FeS moves,and this movement of FeS will force the copper matrix to move together.However,for the uneven distribution of FeS in the model which may occur in practice,the different parts in the model suffer different magnitudes of electric field forces.On the one hand,the force on the model is not balanced and with the increase in the electric field strength and this imbalance will increase as well.As shown in Fig.5,the right part of the model should suffer a larger force.On the other hand,there is no FeS in the center of the model.And the Cu atoms in the center part will remain the original position at the beginning of the simulation,but they are attractive to the outer atoms.For the larger force on the right part and the attraction from the center part,the whole system rotates.In practice,when the composite is in the strong electric field,it is inevitable that the internal stress of the composite will not be uniform and the system will be unstable.

  下载原图

Table 2 Calculation parameters

Fig.3 Pair correlation functions of Cu-Fe and Cu-S under different electric fields:a E1 and E2 and b E2,E3,E4 and E5

Fig.4 M.S.D curves of Fe²⁺and S^2-in composite under different electric fields:a E1 and E2 and b E2,E3,E4 and E5

The tracks of Cu,Fe and S atoms in E1 are shown in Fig.6.For the rotation of the system,the tracks of all atoms are fluctuant along the direction of electric field.In Fig.6a,Cu atoms mainly move along the positive electric field direction although the initial charge was set to be 0.In Fig.6b,the initial charge was set to be positive,Fe atoms only move along the positive direction of electric field and some move quite a long way.In Fig.6c,initial charge was set to be negative,S atoms have the same movement direction as Fe,but the distance that they move is smaller than that of Fe.And this is because S atoms are influenced not only by electric field but also by Fe atoms.Thus,the strong electric field can force FeS to move along the direction of electric field and Fe atoms move further than S atoms.

The morphology of the system undergone different time in E1 is shown in Fig.7.From Fig.7a-c,the solid lubricating phase and matrix do not change obviously,except that their interfaces appear slightly chaotic,and the structure of the copper remains fec structure.The configuration of the whole copper matrix becomes serious disorder and loses its original fce structure as shown in Fig.7d.In Fig.7e,some Fe atoms and S atoms have broken away from the original structure and move along the direction of the electric field.These atoms have two kinds of state:One is single Fe atom,which is significantly influenced by the electric field,and it moves much longer along the positive direction of the electric field;the other is the cluster,which contains two Fe atoms and one S atom,it moves along the positive direction,but as the existence of negative S atoms,the movement is shorter.These two states are also shown in Fig.6b,c.

Fig.5 Atomic configurations of Cu,Fe and S atoms in composite under different electric fields:a E5,b E4,c E3,d E2 and e E1

Fig.6 Atomic configurations of Cu,Fe and S atoms in composite under high electric field:a trajectories of Cu atoms,b trajectories of Fe atoms and c trajectories of S atoms

3.3 Effect of electric field on friction and wear properties of Cu/FeS composite

The change of internal structure of Cu/FeS composite has a serious impact on the friction property.The temperature change of the composite with the different strength of electric field is shown in Fig.8.The system temperature does not change obviously when the strength of electric field is smaller than E1.It indicates that the strength of electric field is not strong enough to make the internal atoms rotate obviously.When the electric field is E1,especially after 4 ps,the temperature of the system increases significantly.This indicates that the system deforms severely and the kinetic energy of internal atoms increases with the action of the electric field E1.It is generally believed that the wear volume losses increase with the increase in temperature [ 10, 28, 29, 30] ,and so do when the electric field is E1.

Moreover,according to the previous analysis of microcosmic evolvement,it can be inferred that the solid lubricating phases FeS exhibit three state variations in the strong electric field.Firstly,on the surface layer of the composite,Fe atoms move along the direction of the electric field,and at the same time S atoms cannot move synchronously,thus Fe atoms will touch the other side of the friction pair directly,as indicated by A in Fig.9.According to the parameters of the potential function,the binding force of Cu-Fe is smaller than that of Cu-S.Therefore,the adhesive wear will be reduced in the high electric field.Secondly,as indicated by B in Fig.9,the internal solid lubricating phases move along the direction of the electric field to the surface of the composite,thus reducing the wear and tear.However,the movement of the phases will leave over some voids in the composite and then induce arc wear easily.Finally,Fe clusters and S clusters in copper matrix can damage the combination of the copper matrix,and so the mechanical property of the composite will be reduced,as indicated by C in Fig.9.In the process of friction,the position where these clusters exist will tend to be fatigue easily.

Fig.7 Geometrical configuration of Cu and FeS under electric field of E1:a primordial structure of matrix material;b,c snapshots of atomic configurations of matrix material after 2 and 3 ps;d configurations of Cu after 5 ps and e structure of FeS after 5 ps

Fig.8 Curves of temperature versus time under different electric fields

Fig.9 Sketch map of reinforcing phase (FeS) in composite under electric field

So,in the strong electric field,due to the movement of the lubricating phase FeS,the friction coefficient will decrease,but the wear volume losses will increase as the results of the high temperature,the destruction of the matrix structure and the arc wear.Therefore,a threshold value of electric field intensity should exist in order to lower the friction coefficient and the wear volume losses.

4 Conclusion

Small distortion and rotation in the Cu/FeS composites occur under a lower electric field,and the length of system is stretched along the direction of electric field.With the increase in the electric field intensity,the solid lubricating phase FeS and matric Cu become far away or even separate completely.The structure of system is destroyed badly and becomes very confused when the electric field intensity reaches 1×10¹⁰ V·m^-1.Moreover,the movement of Fe atoms and S atoms is not synchronized and the existing state of the solid lubricating phase FeS changes.And the FeS is forced to move to the interface of friction pairs by the strong electric field,and this has a lubrication effect on the friction pairs,thus the friction coefficient and adhesive wear decrease.On the other hand,for the increasing temperature in the system and the destruction of the matrix structure and arc wear,wear volume losses increase.

Acknowledgements This study was financially supported by the State Key Laboratory of Advanced Technologies for Comprehensive Utilization of Platinum Metals (No.SKL-SPM-201510).

参考文献

[1] Hong ZJ,Chen JC,Feng J,Du YP,Yu J,Zhang LJ,Wu SZ,Xu FT.Computer simulation of mechanical properties of nano-scale Cu/FeS composite.J Mater Sci Technol.2010;26(4):337.

[2] Hong ZJ,Chen JC,Feng J,Hu FL.Molecular dynamic simulation of friction performance of Cu//FeS composite by in situ method.Tribology.2008;28(5):406.

[3] Chen BB,Chen S,Yang J,Li HP,Guo S,Tang H,Li CS.Tribological properties of Cu-based composites with S-doped NbSe_2.Rare Met.2015;34(6):407.

[4] Liang N,Kang B,Chen JC.Friction and wear properties of Cu/FeS composite.Lubr Eng.2007;32(10):90.

[5] He DH,Manory RR,Grady N.Wear of railway contract wires against current collector materials.Wear.1998;215(1-2):146.

[6] He DH,Manory R,Sinkis H.A sliding wear tester for overhead wires and current collectors in light rail systems.Wear.2000;239(1):10.

[7] Al-Qutub AM,Khalil A,Saheb N,Hakeem AS.Wear and friction behavior of A16061 alloy reinforced with carbon nanotubes.Wear.2013;297(1-2):752.

[8] He L,Tan YF,Wang XL,Jing QF,Hong X.Tribological properties of laser cladding TiB2 particles reinforced Ni-base alloy composite coatings on aluminum alloy.Rare Met.2015;34(11):789.

[9] Ding T,Chen GX,Li YM,He QD,Xuan WJ.Friction and wear behavior of pantograph strips sliding against copper contact wire with electric current.AASRI Proc.2012;2:288.

[10] Dong L,Chen GX,Zhu MH,Zhou ZR.Wear mechanism of aluminum-stainless steel composite conductor rail sliding against collector shoe with electric current.Wear.2007;263(1-6):598.

[11] Yang ZH,Zhang YZ,Zhao F,Shangguan B.Dynamic variation of arc discharge during current-carrying sliding and its effect on directional erosion.Tribol Int.2016;94:71.

[12] Chen GX,Yang HJ,Zhang WH,Wang X,Zhang SD,Zhou ZR.Experimental study on arc ablation occurring in a contact strip rubbing against a contact wire with electrical current.Tribol Int.2013;61:88.

[13] Ding T,Chen GX,Zhu MH,Zhang WH,Zhou ZR.Influence of the spring stiffness on friction and wear behaviours of stainless steel/copper-impregnated metallized carbon couple with electrical current.Wear.2009;267(5-8):1080.

[14] Bucca G,Collina A.Electromechanical interaction between carbon-based pantograph strip and copper contact wire:a heuristic wear model.Tribol Int.2015;92:47.

[15] Vigonski S,Djurabekova F,Veske M,Aabloo A,Zadin V.Molecular dynamics simulations of near-surface Fe precipitatesin Cu under high electric fields.Modell Simul Mater Sci Eng.2015;23(2):025009.

[16] Morse PM.Diatomic molecules according to the wave mechanics.II.Vibrational levels.Phys Rev.1929;34(1):57.

[17] Milstein F.Theoretical strength of a perfect crystal with exponentially attractive and repulsive interatomic interactions.J Appl Phys.1973;44(9):3833.

[18] Yuan L,Shan DB,Guo B.Molecular dynamics simulation of tensile deformation of nano-single crystal aluminum.J Mater Process Technol.2007;184(1):1.

[19] Komanduri R,Chandrasekaran N,Raff LM.Molecular dynamics(MD)simulation of uniaxial tension of some single-crystal cubic metals at nanolevel.Int J Mech.Sci.2001;43(10):2237.

[20] Lennard-Jones JE.Cohesion.Proc Phys Soc.1931;43(5):461.

[21] Philpott MR,Goliney IY,Lin TT.Molecular dynamics simulation of water in a contact with an iron pyrite FeS_2 surface.J Chem Phys.2004;120(4):1943.

[22] Zhang P,Guo B,Shan DB,Ji Z.FE simulation of laser curve bending of sheet metals.J Mater Process Technol.2007;184(1-3):157.

[23] Rafiitabar H.Computational modelling of thermo-mechanical and transport properties of carbon nanotubes.Phys Rep.2004;390(4-5):235.

[24] Grujicic M,Cao G,Roy WN.Atomistic simulations of the solubilization of single-walled carbon nanotubes in toluene.J Mater Sci.2004;39(7):2315.

[25] Rottereau M,Gimela JC,Nicolai T,Durand D.Monte Carlo simulation of particle aggregation and gelation:II.Pair correlation function and structure factor.Eur Phys J E.2004;15(2):141.

[26] Housley RM,Dash JG,Nussbaum RH.Mean-square displacement of dilute iron impurity atoms in high-purity beryllium and copper.Phys Rev.1964;136(2A):A464.

[27] Cichocki B,Hinsen K.Dynamic computer simulation of concentrated hard sphere suspensions:I.Simulation technique and mean square displacement data.Phys A.1990;166(3):473.

[28] Chen XY,Huang SF,Ma SL,Huang R,Zhang J,Hu HX.Elevated-temperature tribological behavior of Ti-B-C-N coatings deposited by reactive magnetron sputtering.Rare Met.2015;34(12):838.

[29] Shangguan B,Zhang YZ,Xing JD,Sun LM,Qiu M,Niu YP,Hou M.Effect of current density on surface temperature and tribology behavior of chromium bronze/brass couple.Chin J Nonferrous Met.2008;18(7):1237.

[30] Hong ZJ,Chen JC,Feng J.Molecular dynamics simulation of the microstructure evolution of C.u/FeS composite in friction under different tempertures.Tribology.2009;29(1):19.