Fretting wear of low current (DC) electrical contacts: quantification of electrical endurance

Transcription

1 26/9/217 Fretting wear of low current (DC) electrical contacts: quantification of electrical endurance S. Fouvry 1, J. Laporte 1, O. Perrinet 1, P. Jedrzejczyk 1, O. Graton 1, R. Enquebecq 1,3, O. Alquier 2, J. Sautel 3 (contact : siegfried.fouvry@ec-lyon.fr) 1 Laboratoire de Tribologie et Dynamique des Systèmes, Ecole Centrale de Lyon, France 2 PSA, Vélizy - France 3 Radiall,Voreppe, France 1 Ecole Centrale de Lyon Lyon Fretting LTDS Fretting Wear, Fretting Fatigue Fretting & Electrical Contacts Plateforme 2DF Durabilité, Fretting & Fatigue Plateforme numérique Bureaux des chercheur de la thématique fretting 2 1

2 ELECTRICAL RESISTANCE R [] 26/9/217 Industrial context 1 Electrical Contact Resistance : ECR Fretting solicitations Contact wear Loss of electrical conduction Electrical Failure,1 Threshold resistance,1,1, NUMBER OF CYCLES N : ECR (electrical contact resistance endurance) Fretting Wear Micro-displacements R. S. Timsit, IEICE Transactions on Electronics 88 (8), (2) B. H. Chudnovsky, Proc. 48th IEEE Holm, (22), N. Ben Jemaa, 23th ICEC, 26, M. Antler, IEEE Transactions 7, 1984, S. Noël et al. Wear 31, 213, 1-61 J. W. McBride, proc. 2th IEEE conf., 26, /19 TOPICS 1. Experimental Simulation 2. ECR endurance versus sliding amplitude 3. Fretting surface damages & ECR behaviour 4. ECR & Friction energy density concept. Complex Fretting-Reciprocating slidings 6. ECR versus Coatings properties 7. Conclusions 2

3 Tangential force 26/9/217 Experimental setup Cross Cylinder configuration Fretting cycle Ed (J) friction energy Tangential force Q(N) δ displacement δ* Q * (µm) displacement amplitude (test compliance dependent) δ : Sliding amplitude (=> δ when Q=) Ed(J) : dissipated energy Crossed cylinders (9 ) Q * ( N) 2S Fretting Log Q(N) * : displacement 2.a Hertz amplitude Q * displacement r = 2.3 mm (µm) 2 ( µm) : tangential force amplitude Fretting Loop DC micro fretting test LTDS (Ecole Centrale de Lyon) Fretting Connector 1 Fretting : δ* : 1 3 µm Temperature : 2-2 C Frequency: 3 hz Humidity (salt) : 9% RH Fretting Connector 2 Fretting : δ* : 1 3 µm Temperature : 2-2 C Frequency: 3 hz Humidity (salt) : 9% RH Gazes (doping) : H 2 S & S 2 Fretting Connector 3 Fretting : δ* : 1 3 µm Reciprocating : D :.1 to mm Temperature : 2-2 C Frequency: 3 hz Humidity (salt) : 9% RH 6 3

4 electrical resistance R () 26/9/217 Definition of the electrical failure criterion Normal force P=3N Displacement * Four point method Stabilized current : I =. A Electrical failure when: DR> DRc=.4 DRc=.4 Rmin fretting cycles (electrical endurance) S. Hannel et al, Wear 249, 21, W. Ren, et al. Tribology International 83,(21), 1-11 Flowers G.T et al. Proc. of the 1st IEEE Holm Conf., 2, Malucci, R.D., Proc. of the 49th IEEE Holm Conf., 23, 1-1 7/19 Studied materials and test conditions Electrodeposition process Gold Silver Tin Noble Semi-Noble Non Noble Applied test conditions: Temperature: 2 C Relative humidity: % Frequency: 3Hz Normal force & δ : varying 2 µm e (µm) CuSn4 (substrate) Ni Au coatings CuSn4 (substrate) Ni Ag coatings CuSn4 (substrate) Ni Sn coatings crossed cylinders 9 2.a Hertz r = 2.3 mm e (µm) Au coatings Ag coatings Sn coatings 2 µm Ni Ni Ni CuSn4 (substrate) CuSn4 (substrate) CuSn4 (substrate) If P= 3N Sphere/Plane a Hertz =43µm & p H, max =772MPa 8 4

5 tangential force amplitude, Q* 26/9/217 Basics of fretting contact : sliding condition normal force P Q transition amplitude, * t Partial Slip Q*,* Closed cycle 2 * Q 2 Gross Slip Q*=µ.P open cycle * e 1 a Reciprocating Q displacement (µm) 2 stick zone 2δ * 2δ * tangential force Q (N) sliding zone full sliding a reciprocating a displacement amplitude, * 9 sliding condition & Electrical contact resistance evolution R ()..8 * 7 µm Gross slip (finite ECR endurance) µm - generalized wear of the interface. - formation of an complete oxide debris layer. * 4 µm Partial slip (infinite ECR endurance) R (Ω) Q */P δ (µm) t 4 µm.8 Q */P.4 δ (µm) ΔR= ΔR th = 4 mω undamaged stick zone (metal/metal) fretting cycles, N 7 µm wear in external sliding zone.1.1 (Sn/Sn (e= 1.3 µm); P= 3N, f= 3 Hz, RH=%, T=2 C).1 partial slip gross slip N = cycles displacement amplitude, δ * ( µm) Fretting mechanical criterion If δ* < δ t (PS/GS transition) => inner stick metal zone => R low & stable (infinite endurance) If δ* > δ t (PS/GS transition) => full sliding => Wear => R rises (finite endurance) S. Hannel et al. Wear 249(9), 21

6 Electrical contact resistance ( -3 ) 26/9/217 Correlation between fretting sliding condition and ECR behaviour R() tangential force amplitude Q* Infinite endurance Q Partial Slip Q*,* t Gross Slip Q Q*,* R() finite endurance low & stable ECR Stick zone 2δ * a ECR increase time (fretting cycle) sliding zone full sliding time (fretting cycle) displacement amplitude, * 11 Comparison between Noble & non Non noble coating Non noble : very fast decay Noble & semi noble : Delay before EC failure! Ag/Ag Au/Au Sn/Sn Sn/Sn Ag/Ag Au/Au Fretting cycles Applied test conditions: Temperature: 2 C Relative humidity: % Frequency: 3Hz Normal Force : 3 N Displacement : 8 µm Thickness coat. : 1.3 µm. 1 Non noble (Sn alloy) *< t Rc GP t Rc GT *> t Noble (Au & Ag) *< t S. Hannel et al. Wear 249(9), 21 Rc GP t Wear (delay) Rc GT Rc *> t When non noble substract reached: ECR failure! 12 6

7 Applied displacement amplitude (µm) 26/9/ Quantification of (Electrical Contact Endurance) versus displacement & sliding amplitude? 13 Electrical Endurances a function of the applied displacement amplitude Au/Au interface Applied test conditions: Temperature: 2 C Relative humidity: % Frequency: 3Hz Normal Force : 3 N Thickness coat. : 1.3 µm 1 Finite endurance Domain (GS) t_au = µm 1 1E+8 fretting cycles, (ΔR>4mΩ) Infinite endurance Domain (PS) Asymptotic decreasing of N C (DRc=.4 threshold): the larger the displacement the smaller the ECR endurance 14 7

8 ln() applied displacement amplitude: * (µm) 26/9/217 Quantification of the electrical endurance Curve : Fatigue like approach : B application to Au/Au interface -n y = x R²= ln( * ) t perfect correlation between experiments and the exponential formulation (only 3 variables : δ t, n, δ * ln( ) n ln( t ) B With n = 3.18, B = 16.8 and µm Power law exp(b) ( * ) * ( when 1µm) t t t ( * ) n t n S. Fouvry et al. Wear 271 (9 ), (211) 1 1E+8 1E+ fretting cycles, (ΔR>4mΩ) 1 Comparison between coatings Applied test conditions: Temperature: 2 C Relative humidity: % Frequency: 3Hz Normal Force : 3 N Thickness coat. : 1.3 µm - Small difference between δ t transitions - Large difference between non noble Sn and noble Au&Ag ECR endurances - All the ECR endurance can be formalised using a simple power law function S. Fouvry et al. Wear 271 (9 ), (211) 16 8

9 26/9/217 Sliding amplitude formulation The measured displacement depends on the test compliance => Results affected by the test signature! C A C QC δ A : tangential test apparatus accommodation C A : test apparatus compliance A test apparatus accommodation δ A = Q x C A real contact displ. δ C (± µm) measured δ (± µm) Tangential force Q(N) δ displacement δ* Q * (µm) (±µm) Ag/Ag ( ) n power law function δ : Sliding amplitude (=> δ when Q=) * t S. Fouvry et al. 8th IEEE Holm, 212, What about herogeneous interfaces? ( µm) x / fretting cycles, Ag/Ag Ag/Sn Sn/Sn O. Perrinet et al., ICEC 214, p Ag/Sn still controlled by noble/noble fretting wear response (ECR failure is delayed by the coating wear) : But the formation of abrasive Sn oxides accelarate the Ag surface wear => mitigate benefit of Ag (only x compared to Sn/Sn!) 18 9

10 Résistance (Ω) 26/9/ Correlation between fretting damage (surface wear & oxide debris) & Electrical behavior 19 Study of the endurance degradation for homogeneous Ag/Ag interface δ g = 9µm ΔR<ΔRc ΔR>ΔRc 7 Ag/Ag e= 2µm P=3N f=3hz RH=% T=2 C δ = 9µm) Fretting cycles J. Laporte et al., Wear (21), Interrupted tests at different fretting cycles to follow the electrical degradation Characterization of fretting scars (SEM, EDX 3D profil) J. Song et al. Wear , (21 2 Y.W. Park et al. Tribology International 41(7),(28) 2

11 concentration (At.%) concentration at% 26/9/ R(mΩ) 2 4 ΔR<ΔRc ΔR>ΔRc fretting cycles O Ag ECR failure Investigation of fretting wear damages Supérieur upper specimen δ * g Inférieur lower specimen δ * g N=2 cycles R=,13mΩ N= cycles R=,8mΩ EDX analysis in the central zone (2% of fretting scar) ΔR<ΔRc=4mΩ N=4 cycles R=,48mΩ Ø c Ø a ΔR>ΔRc=4mΩ =87 cycles R=,4mΩ Ø a =,2.Ø c [O] th = 4at% Electrical failure : ΔR>ΔRc=4mΩ related to a threshold chemical composition of debris layer : when [Ag].2 <[Ag] th %at & []>[] th > 4% then ECR failure. [Ag] th = at% fretting cycles S. Fouvry et al. Wear, 271 (9-), 211, J. Laporte et al., Wear (21), Stability of the proposal? Ag/Ag, e=2µm, P=3N, f=3hz, T=2 C, RH=%, δ =±4µm to ±16.7µm averaged EDX analysis Ø f d =.2 Ø f [O] th =4At.% [O]% [Ag]% [Ag] th =At.% At.% 3.At.% [Ag] [Ni] [Cu] EDX mapping ΔR< ΔRc ΔR>ΔRc upper lower upper lower upper lower ECR endurance, (cycles) [] upper lower Stable if similar area is analyzed (inner 2% of the fretting scar radius observed 22 At the ECR failure) 11

12 Concentration (At.%) 26/9/217 Investigation of fretting wear damages EDX Analysis at the failure Fretting scar at (ECR failure) Ø f. At. % [O] J. Laporte et al., Wear (21), d diameter of EDX spot d/ø f The most stable criterion [O] th = 4 At.% The ECR failure is reached when the fretting scar is fully covered by an oxide debris layer [Ag] 23/19 Investigation of fretting wear damages EDX Analysis at the failure Chemical concentration profiles at the ECR failure () EDX scan Fretting track Ni O Ag Cu Position in the fretting track (µm) Test conditions: P=3N δ = 9µm RH=% f=3hz T=2 C e=2µm Analysis zone e=1µm Cu Ni Ag J. Laporte et al., Wear (21), The failure (ΔR>ΔRc=4mΩ) is reached when Ag is remove from the center of the fretting scar (still present in the lateral sides) A. Kassmann Rudolphi et al. Wear 21, (1996) S. Noël et al. Proc. 2nd IEEE Holm, (26) 24/19 12

13 sliding amplitude, δ ( µm) 26/9/ Quantification using a friction energy wear approach 2 Prediction of ECR endurance: energy approach Normal force influence on ECR endurance (6 N ) ( ) (6 N ) n(6 N ) increasing P (1 N ) ( ) (1 N ) n(1 N ) P=6N P=N P=4N P=3N P=2N P=1N Ag/Ag HR=% e=2 µm f=3hz T=2 C An increase of P decrease for a given δ 1.E+3 1.E+4 1.E+ 1.E+6 1.E+7 1.E+8 1.E+9 1.E+ ECR endurance, (cycles) Need a global wear parameter to combine δ and P! S. Fouvry et al. 8th IEEE Holm, 212, J. Laporte et al., Wear (21), /19 13

14 26/9/217 Friction Energy Wear Approach (Global Wear Volume) S Friction Work SEd Q (N) fretting cycles SEd * (m) S Surface transformation & degradations Wear Volume wear volume S. Fouvry et al., Wear, 2 (1996), p electrical contact resistance Friction Energy Local Approach Wear Volume Substrate interactions need to predict wear depth! accumulated friction energy density profile max modification of the energy density distribution FEM RF1 wear box coupled problem modification of the contact pressure RF2 Wear profile maximum wear depth (h) modification of the contact geometry S. Fouvry et al. Wear 2, 23, C. Mary et al. Wear263 (1-6), 27,

15 26/9/217 Prediction of ECR endurance: energy approach Application of friction energy approach prediction requires a local wear approach max (x) energy density Wear Profile z (µm) h max with wear ϕ(x) converge to a flat profile C. Mary et al. Wear 263 fretting loop Ed f A contact area A f =A final f (cycles) Ag/Ag HR=% e=2 µm f=3hz T=2 C β N c =. φ f ( f =6. 12 ϕ = 6 12 cycles and β=2.8 β=-2.8 P=6N P=N P=4N P=3N P=2N P=1N 2 3 φ f = E d /A f (J/m²/cycle) ) Very nice prediction (low dispersion) J. Laporte et al., Wear (21), Prediction of ECR endurance: energy approach Simplified approximation Previous analysis requires the measurement of A f (long & fastidious) A 16 (µm²) Possibility to approximate A f using a single power law function of normal force: A f = A. P m with A =72,6 µm² and m= P(N) The analysis requires the computation of E d! (integrale of the frettting loop) 2 Considering a quadratic shape: Ed 4 P µ.p Ed 3 1

16 exp (cycles) 26/9/217 Prediction of ECR endurance: energy approach Simplified approximation Af A ( ) 1 f Ed (1m ) 4 µ P A f = A. P m E d. µ.. P 4 Very good correlation between experimental lifetime and theoretical prediction e=2µm, 2µm<δg< 16µm P=6N P=N P=4N P=3N P=2N P=1N J. Laporte et al., Wear (21), th (cycles) 31 Prediction of ECR endurance: energy approach Influence of the coating thickness 14 (cycles Ag/Ag P=3N δ = 9µm HR=% f=3hz T=2 C Parabolic evolution: = e ref e ref p p=2.8 e ref =2µm N cref =N c (2µm) 2 J. Laporte et al., Wear (21), Coating thickness (e) e A thicker coating induce a significant lateral contact extension! V Ag α e γ with γ>

17 (experimental) (cycles) 26/9/217 th =f(p, δ, µ) = e ref e ref p Prediction of ECR endurance: energy approach Global formulation Normal force Sliding amplitude th =g(p, δ, µ, e) Friction coefficient Thickness A 1 (1 m) 4 µ P e e ref p J. Laporte et al., Wear (21), Eq. (1), e=2µm, 2µm <δ < 16µm P=6N P=N P=4N P=3N P=2N P=1N Eq. (21) P=3N, 2µm <δ < 16µm e=2µm e=3µm e=4µm e=4.8µm Very good correlation confirming the proposal! (predicted) (cycles) 33. Complex Fretting-Reciprocating slidings 34 17

18 R [Ω] 26/9/217 Influence of repetitive clipping & uncliping slidings clip flexible pin clip flexible pin before insertion after insertion ? electrical failure repetitive insertions.1 2 Nombre de cycles N Surface degradations 3 electromagnetic shaker (fretting) Experimental strategy upper sample flexible holder strips weight laser sensor D= 2µm to 1 µm v GC,ref = 8,3µm.s -1 to 124,µm.s -1 D Test stopped when ΔR >ΔRc δ = 9 µm (3 Hz) 9 µm samples Ag/Ag HR=% e=2 µm f=3hz T=2 C flexible strips J. Laporte et al., Wear (217) tangentiel force sensor electromagnetic linear motor (reciprocating) fretting track reciprocating track fretting N f à 6 cycles Real clip assembly 18

19 electrical contact resistance, R (mω) fretting endurance, (cycles) x 26/9/217 Effect of reciprocating sliding regarding ECR fretting response R = 2.14mΩ D=1mm Large sliding B Fretting zone [Ag] 87at% Ag/Ag RH=% e=2 µm f=3hz T=2 C [Ag] 31at% Reciprocating zone δ * g = 9µm N=6 cycles A R = 2.14mΩ δ * g [Ag] 17at% refilling process : Ag is transferred from the external reciprocating track scar to the fretting scar! => The application of reciprocating increase the ECR endurance! 37 Influence of fretting block (periodicity of large sliding)? fretting & reciprocating reciprocating stroke (D= 1 mm) D max () () 1 Nf / N f, tr µm plain fretting (D=µm) =, cycles ΔR=4mΩ Fretting N cycles N f = cycles fretting sequence = 21,7 cycles reciprocating sliding All Ag present in reciprocating track is transferred x 4 max N f,tr plain fretting (N R =), () (Eq.23) : total number of fretting cycles before ERC failure N f too long => no transfer fretting cycles (N) [Ag]=.17at% [O]=48.13at% δ δ * g [Ag]=2.2at% [O]=12.8at% [Ag]=3.4at% [O]=43.6at% 3 µm plain fretting fretting-reciprocating J. Laporte et al., Wear , (217) N R = N R = N R =1 N R =2 N R =2 N R =4 N R =4 x fretting cycles between each reciprocating sliding, N f (cycles) N f 38 19

20 fretting endurance, (cycles) x c wear rate, f (µm 3 /cycle) c exp (cycles) 26/9/217 Influence of reciprocating stroke? 4 x plain fretting, PF D c D th 2 7 1, 1,2 1, 1,7 reciprocating stroke, D (µm) Non monotonic evolution 39 wear rate Vc V (µm 3 /cy.) V V N prediction (global formulation) V f, PF 1 h V V f, PF V f, ref hv h v f, PF with h=1.66 D 1 determined for D ref =1mm D tr Steady-state N f =N f,ref =,cycles 2µm D 1,µm plain fretting D th 1, 1, 2, reciprocating stroke, D (µm) Vc : total Ag volume involved in fretting wear process f Vc k V[ Ag] k e f D 2 Vc V K f k e f f D * 4 g P 2 N f 1 N f, PF total Ag volume involved in fretting wear Vc V x 4 D=D ref =1mm,cycles N f 6,cycles N f =N f,ref =,cycles 2µm D 1,µm plain fretting V : wear rate per fretting cycle, tr 1 h h w v D 1 Dtr k =.94 (proportion of volumetric Ag volume involved) 2 3 x 4 4 pred (cycles) J. Laporte et al., Wear , (217)

21 26/9/ Simplified strategy to compare coatings : ECR versus Coatings properties? 41 ref : Reference ECR endurance defined for a reference sliding amplitude δ =± 9µm 3 2 (±µm) Definition of two driving parameters to describe GS ECR endurance & Cold welding cold welding index _ref =1/µ max (δ =9µm) µ=q*/p µ max δ =9µm 4 ref δ * =±9µm ECR endurance, power law function fretting cycle the larger ref and _ref, the better the electrical performance! 21

22 26/9/217 Comparison of coatings New coating via Ag PVD Possibility to explore new hardnesses and new conductivities! AgSn Ni e=2µm 2 µm AgSnIn Ni e=2µm 2 µm CuZn37 (substrat) CuZn37 (substrat) AgCrN Ni e=2µm 2 µm AgaC Ni e=2µm 2 µm CuZn37 (substrat) CuZn37 (substrate Comparison of coatings cold welding index ref =1/µ max ref (cycles) 1 2 x GS fretting endurance index AgSn AgCrN AgaC AgSnIn AuNi Ag 2 1 e=2µm et e AuNi =1,3µm Conditions de test: P=3N δ = 9µm RH=% f=3hz T=2 C considering ref & ref the best compromise is obtained with AgSnIn (conductive ITO oxydes) Laporte et al., IEEE 61st Holm Conference, 21,

23 26/9/217 Correlation of _ref versus Hardness & conductivity 2 (x ) 2 (x ) 2 1 Ag AuNi AgSnIn conductive oxydes (ITO) AgaC AgSn Hardness, H (hv) AgCrN High scattering : Hardness is not a relevent parameter 2 1 AgCrN AgSnIn conductive oxydes AgSn C V AuNi AgaC Ag V electrical conductivity, σ (µs.cm-1) Removing (AgSnIn) => Linear increase ) C V ( V C V c Cst (base Argent) ref is proportional to the coating conductivity Tribological interpretation : V ( x 4 µm 3 ) Ag AgSnIn 1 AgaC AuNi AgCrN (cycles) AgSn wear volume measured at is proportional to the K =1,82µm 3 /cycle V (µm 3 ) x V ( x 4 µm 3 ) AgCrN AgSnIn AuNi AgaC AgSn V, σ (µs.cm -1 ) 46 C V the wear volume at the ECR failure is proportional to the coating conductivity Ag The higher the coating electrical conductivity, the larger the wear volume required to reach ECR failure is proportional to the coating conductivity 23

24 26/9/217 Conclusions - The partial slip/ gross slip displacement transition controls the transition from infinite to finite ECR endurance ( : N_fretting when Δ R>4 mω) t - is controlled by surface wear processes: ECR failure is reached when [O] > 4 at% (noble metal eliminated and replace by a oxide layer) [O] - can be expressed as a power law function of friction energy density ϕ (Wear depth is controlled by ϕ) ( f ) - can be expressed as a power low function of sliding amplitude (deduced from the general friction energy density formulation) - Application of large sliding induced a noble metal refilling process of fretting scar (Increase of ) - The global response of a coating can described by two variables - ref (GS endurance index) & χ ref (cold welding index) ( ) n - _ref is controlled by the electrical conductivity of the coating 47 Thank you for your attention! 24