FRETTING CORROSION ANALYSIS FOR ELECTRICAL TERMINALS OF VEHICLE CONNECTORS

FRETTING CORROSION ANALYSIS FOR ELECTRICAL TERMINALS OF VEHICLE CONNECTORS A. Bouzera 1, 3, R. El Abdi 2, *, E. Carvou 1, N. Benjemâa 1, L. Tristani 3 and E.M. Zindine 3 1 IPR, Université de Rennes 1, 35042 Rennes Cedex - France 2 Larmaur, Université de Rennes1-35042 Rennes Cedex - France. 3 FCI, France, ZI des Longs Réages B.P. 50025, 28231 Epernon - France. *E-mail: relabdi@univ-rennes1.fr Abstract: The increasing needs of electrical energy in modern vehicles require a high number of various components such as electrical connectors. A connector is designed to allow a current to circulate between two electric or electronic systems. These connectors submitted to car vibrations can suffer damage. This damage is due to a phenomenon called fretting corrosion and leads to a rapid degradation of the electrical performance which results in a severe increase of the electrical contact resistance. Fretting which is a relative and small amplitude movement between two surfaces inside the connector leads to a development of small debris subjected the environment effects such humidity. These debris can oxidize if the contacting surfaces consist of non-noble metals. The oxidizing debris act such as an insulating layer and induces a connector malfunction. Even for gold or silver coating surfaces, wear process can reach a non-noble underlayer, usually tin, nickel or the substrate which is usually made of copper alloys. Connectors in the automotive vehicles are fitted into their housing to protect them from the external damages and are connected to the other vehicle components by wire cables. Vehicle movements are transmitted to the connector through the cable vibrations and can lead to the connector fritting corrosion damage. If the fretting corrosion phenomenon is intensively studied in automotive systems, the taking the cable vibrations into account on the connector aging is very seldom studied. The aim of this paper is to study the influence of the cable movements on the connector and to understand the fretting corrosion apparition in electrical terminals of vehicle connectors under micro-vibrations. Influence electromagnetic shaker frequencies which produces oscillatory motion is analyzed. Keywords: Contact voltage fluctuation, fretting corrosion, relative movement, cable vibrations. 1. Introduction The increasing needs of electrical energy in modern vehicles require a high number of various components such as electrical connectors. A connector is designed to allow a current to circulate between two electric or electronic systems. These connectors submitted to car vibrations can suffer damage. This damage is due to a phenomenon called fretting corrosion and leads to a rapid degradation of the electrical performance which results in a severe increase of the electrical contact resistance. Fretting which is a relative and small amplitude movement between two surfaces inside the connector leads to a development of small debris submitted to the environment effects such humidity. These debris can oxidize if the contacting surfaces consist of nonnoble metals. The oxidizing debris act such as an insulating layer and induces a connector malfunction [1, 6]. Even for gold or silver coating surfaces, wear process can reach a non-noble underlayer, usually tin, nickel or the substrate which is usually made of copper alloys. Even for materials which resist to wearing like titanium alloys, the Ti alloys contact surfaces are very susceptible to fretting and the engine components are protected using a thick soft CuNiln coating on the top of which is deposited a solid lubricant [7]. But it is observed a fast degradation of the lubricant and CuNiln can cause a severe damage to Ti alloy because of important adhesion mechanisms. The fretting-corrosion problems were intensively studied in the last years and some models were proposed to analyse the electrical contact change according to the environment effects or to mechanicals parameters [8]. It was shown that for a threshold amplitude of slip at which fretting damage becomes apparent is found to be the region of 0.5 The Romanian Review Precision Mechanics, Optics & Mechatronics, 2012, No. 41 57

µm and low frequencies are particularly damaging in electric contacts. On the other hand, the third body approach shown that the state of the surface in contact and the interface are essential parameters that influence the mode of fretting degradation. I (A) R V Static holder Connectors in the automotive vehicles are fitted into their housing to protect them from the external damages and are connected to the other vehicle components by wire cables. Vehicle movements are transmitted to the connector through the cable vibrations and can lead to the connector fritting corrosion damage. If the fretting corrosion phenomenon is intensively studied in automotive systems, the taking the cable vibrations into account on the connector aging is very seldom studied. The aim of this paper is to study the influence of the shaker frequencies and the cable movements on the connector and to understand the fretting corrosion apparition in electrical terminals of vehicle connectors under micro-vibrations. 2. Fretting set-up The pin of the connector is made of tinned copper with nickel undercoat (CuZn30 + Ni) and the connector spring is made of tined copper (CuZn) only. Both are coated with a tin under-layer with a thickness of 2µm. The used wires 10 cm length has a cross section of 0.5 mm 2. The nominal resistance of the connector is equal to 20 mω. Figure 1 gives a description of the used connector. The used vibratory set-up is shown in Fig. 2. The connector is set on a static holder whereas the wire is attached to a shaker. The cyclic movement is given using an electrodynamic vibrator (model LDS V455) able to produce frequencies between 5 Hz and 7500 Hz with a maximum acceleration 117 g (g = 9.81 m/s 2 ). The maximum imposed vibratory displacement can reach 19 mm. Female parts range Male part Female part Figure 1: Used connector Wire 50Hz-800µm 10 cm Insertion-extraction displacements Wire Laser beam copper lamella Figure 2: Simplified diagram of used vibratory set-up The shaker excites the cables which transmit vibrations to the connector. Since the cables are not rigid, only a part of the displacements provided by the shaker is transmitted to the connector through the cable. To measure the real displacement induced by the cable crimped to the female part, a laser beam gives the displacements of a copper lamella welded onto the female part (Fig. 2). To observe the onset of fretting, the analysis of the contact voltage in a voltage range of 5 V corresponding to a current of 500 ma is considered. Two analysis types of fluctuations in voltage contacts are used. The first is based on the specific voltage levels (histogram of voltage contacts), and the second is the Fourier frequency analysis [9]. Indeed, the use of the histogram of measured voltages is advised to represent their distribution. In our case, the contact stress distribution versus time (cycle fretting) to characterize fretting is represented. The second study of fluctuations in voltage contacts (intermittent) is made by Fast Fourier Transform (FFT) [9]. This frequency representation gives a lot of information (fundamental frequency, resonance frequencies, defects detection...). During the vibration test, the contact voltage is continuously sampled by 200 channels and a histogram of the voltage distribution is determined in real time via a digital oscilloscope (DSO) equipped with a real-time histogram functionality. The histogram is built ten periods of the shaker oscillation. Three channels of the oscilloscope are successively activated; each one during 20 s. Thus, each channel is activated every 60 s. The oscilloscope is used with the following voltages: 0 to 5 mv with a sampling period of 80 ns and the This channel is activated every 60 s during 20 s and is dedicated to the analysis of the constriction resistance fluctuation; 58 The Romanian Review Precision Mechanics, Optics & Mechatronics, 2012, No. 41

0 to 200 mv, the sampling rate is 80 ns and the This channel is activated every 60 s during 20 s and is dedicated to the analysis of voltage fluctuations due to the intermittency; 0 to 600 mv, sampling rate is 80 ns and the This channel change according to the time and then its influence on the fretting phenomenon. 3. Results and discussion Connector Cable Distribution of the contact voltage at 50 Hz At first, an experimental test has been made using variable frequencies. This test has shown that the cable resonance frequency is equal to 50 Hz under an acceleration of 4 g for imposed displacement amplitude of 800 µm. Figure 3 shows the cable vibrations for the resonance frequency. Using the resonance frequency under an acceleration of 4 g, different currents and voltages have been used to analyse the resistance Figure 3: Cable vibrations (50 Hz, 4 g, 800 µm) Figure 4 (a, b, c) gives for different voltage range the distribution of the contact voltage versus time. The number of contact voltage obtaining is represented with different colors. A bright color (red for example) indicates a high obtaining number of a contact voltage. In this case, the high population term is used. A dark color (purple for example) will correspond to a relatively small population. 16000 14000 (a) 0 to 5 mv 12000 Time (s) Temps (s) 10000 8000 6000 4000 2000 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 Tension (V) Voltage Vc (V) Constriction voltage Nombre d'échantillons 1 10 100 1000 10000 Number of cycles 10 5 2.10 5 3.10 5 4.10 5 5.10 5 6.10 5 7.10 5 8.10 5 (b) 0 to 200 mv (c) (c) 0 to 0 600 to 600 mv mv 0.0 0.2 0.4 0.6 0.8 1 1.2 Voltage Vc (V) Melting voltage Figure 4: Distribution of the contact voltage Vc versus cycles number for frequency of 50 Hz for an amplitude of 800 μm for three voltage levels (5 V-500 ma) The Romanian Review Precision Mechanics, Optics & Mechatronics, 2012, No. 41 59

The first voltage range (5 mv) have been identified as a voltage constriction which is equal to 10 mv (corresponding to 20 mω) in the first cycles of fretting (Fig. 4 a). We can also see (Fig. 4 b) (0 to 200 mv) that the contact voltage has many fluctuations between 400mV and 800 mv. Note that the melting voltage copper is equal to 415 mv at 1083 C [10, 11] and the phenomenon of fritting (breakdown of oxide films) is observed [6]. The contact voltage reaches 2V and even exceeds 5 V (Fig. 4 c), the current always passes through the connector but with a sharp drop, and in this case the connector has shrunk. Frequency sweep and amplitude analysis To identify the resonance frequencies of the used system, two frequency scans (from 50 Hz to 500 Hz) for a constant shaker displacement of 80 μm and for another constant shaker displacement of 220 μm, were performed on a connector. Fig. 5 gives the shaker movement and the peak to peak contact voltage from 50 Hz to 500 Hz for constant displacement amplitude of 80 μm (0.4 g to 40 g) on a new connector for the test conditions: (20 V-1.6 A). For the same parameters, Fig. 6 shows the frequency sweep. We can note that the maximum of the contact voltage is obtained when the frequency is around 80 Hz, 100 Hz, 110 Hz, 145 Hz, 210 Hz, 250 Hz, 325 Hz and 380 Hz. These frequencies may represent the system resonant frequencies; which are obtained when the copper lamella displacements (Fig. 2) present maximum values (Fig. 6). Figure 5: Screenshot of contact voltage and shaker displacements. Lamella displacement A pk-pk (µm) 560 480 400 320 240 160 80 displacement Peak to peak of contact voltage V Cpk-pk 0 0.0 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 Frequency (Hz) Figure 6: Amplitude of lamella displacements (A) and contact voltage (Vc) with frequency sweep from 50 Hz to 500 Hz for imposed amplitude shaker displacement of 80 μm (20 V-1.6 A). A Vc 1.4 1.2 1.0 0.8 0.6 0.4 0.2 Contact voltage Vc pk-pk (V) Among the frequencies previously cited (80 Hz, 100 Hz, 110 Hz, 145 Hz, 210 Hz, 250 Hz, 325 Hz and 380 Hz) some frequencies are not resonance frequencies as 380Hz. A voltage histogram is given in Fig. 7. When a frequency equal to 380 Hz is applied with an acceleration of 23 g, a shaker displacement of 80 µm is obtained. The applied current is 1.6 A with a limit voltage of 20 V. The constriction voltage is equal to 20 mv (corresponding to 20 mω) (Fig. 7). This voltage is stable and low throughout the fretting cycle until 1.8x10 7 cycles and no fretting corrosion is obtained. Therefore, this frequency can not be a resonance frequency of the system (Fig. 7) and the connector failure does not occur. Figure 7: Distribution of contact voltage Vc with frequency of 380 Hz (80 μm-23 g-20 V-1.6 A). Fourier transform The voltage of the contact system subjected to vibrations is not constant but varies around an average value higher than the constriction voltage. Fig. 8 shows obtained fluctuations for a few vibrations. However despite these fluctuations, it seems that the contact voltage is periodic (Fig. 8), and an analysis by the Fast Fourier Transform FFT analysis was made. The Fast Fourier Transform is based on the fact that any periodic function of time, x(t) can be split up into an infinite sum of sines and cosines using a base frequency f 0 = 1/T, where T is the period of x(t): x( t) = a + ( ak cos (2π k f0 t ) + bk sin (2πk f0 = 0 t k 1 )) (1) Contact voltage is continuously measured and analyzed by a Fast Fourier Transform (using a direct digital oscilloscope mathematical function). The FFT is used to determine the fundamental frequency f 0 (Fig. 8) and harmonic frequencies of the signal k f 0. 60 The Romanian Review Precision Mechanics, Optics & Mechatronics, 2012, No. 41

Figure 8: Evolution of the contact voltage Vc at 5 s (50 Hz, 800 µm). To understand the fretting phenomenon, the signal superposition was undertaken as shown in Fig. 9 after a time of 5 s and 3600 s. Amplitude of contact voltage A(V) and of lamella displacement A(µm) Figure 9: FFT of contact voltage A(V) and displacement A(µm). In Fig. 9, x-axis gives the harmonic frequencies (50 to 625 Hz) and the y-axis represents the harmonic amplitudes. The main frequency peaks are multiple of 50 Hz which is the shaker vibration frequency. Thus, the contact voltage has a period of 20 ms. A change of the shaker frequency leads to variations of all the harmonic frequencies which are multiple of the fundamental vibration frequency. In addition, the amplitudes of harmonic frequencies are related to fretting states. Indeed, the contact voltage increases during fretting because the oxide particles were building up on tracks of particles and modify the geometry and the electrical conduction of the contact interface. So, the frequency fluctuations are due to the contact interface wear. 4. Conclusion Zoo Times (s) A(µm) at t=3600s A(V) at t=3600s A(µm) at t=5s A(V) at t=5s Frequency (Hz) Using the histogram statistical analysis, two voltage levels are identified. The first particular voltage at the beginning of the test is the constriction voltage. After a higher number of vibration tests where the wear increased, a second voltage level is observed and reaches several hundred mv (500 mv). The second study of contact voltage fluctuations is made using the FFT analysis during the vibration test. This analysis is a powerful tool to better understand the origin of the frequency fluctuations and their relationship with the vibration frequency. We can note that when the contact moves along the track the contact conduction remains constant. Therefore, the voltage presents the same evolutions for consecutive vibration periods. 5. References [1] R. B. Waterhouse, Fretting Wear. Proc. Int. Conf. on Wear of Materials, 1981, American Society of Mechanical Engineers, New-York, 1981; 17. [2] G. A. Tomlinson, The Rusting of Steel Surfaces in Contact. Proc. R. Soc. London, Ser. A, 1927; 115, 472. [3] R. D. Mindlin, Compliance of Elastic Bodies. J. Appli. Mech., 1949; 16, 259. [4] L. Tristani, E. M. Zindine, L. Boyer, G. Klimek, Mechanical Modelling of Fretting Cycles in Electrical Contacts. Wear, 2001; 249, 12. [5] M. Antler, Survey of Contact Fretting in Electrical Connectors, IEEE Trans., CHMT 8 (1), 1985, 87-104. [6] L. Féchant, Le Contact Electrique. Editions Hermès I, 1996; 172. [7] C. Mary, S. Fouvry, J. M. martin, B. Bonnet, Pressure and Temperature Effects on Fretting Wear Damage of Cu-Ni-ln Plasma Coating Versus Ti17 Titanium Alloy Contact. Wear, 2011; 272, 18. [8] S. Hannel, S. Fouvry, P. Kapsa, I. Vincent, The Fretting Sliding Transition as a Criterion for Electrical Contact Performance. Wear, 2001; 249, 761. [9] E. Carvou, N. Benjemaa, Time and Level Analysis of Contact Voltage Intermittence Induced by Fretting in Power Connector. Proc. 53 rd HOLM Int. Conf., Pittsburgh, USA, 2007; 221. [10] L. Féchant, Le Contact Electrique: Phénomène Physiques et Matériaux. Collection SEE, Editions HERMES, Septembre 1996; 60-137. [11] C. Maul, Intermittent Electrical Discontinuities Tin- Plated Automotive Contacts. Southampton, England, PhD Thesis. September 2001. The Romanian Review Precision Mechanics, Optics & Mechatronics, 2012, No. 41 61