Electrical contact phenomena during impact

I -

184

IEEE TRANSACTIONS ON COMPONENTS,HYBRIDS, AND MANUFACTURING TECHNOLOGY,VOL. 15, NO. 2, APRIL 1992

Electrical Contact Phenomena During Impact John W. McBride and Suleiman M. Sharkh

Abstract-The events occurring during the impact of electrical contacts are vital to the long term reliability of switching systems. This papep descnies the theory of impact as it applies to electrical contacts, under the influence of current loading. The theory uses the coefficient of restitution as a means of modeling the events at impact. Design parameters are considered in terms of their influence upon the dynamics of impact and bounce. Experimental studies are presented which include the measurement of impact forces, impact time, and both current and voltage characteristics. The influence of preimpact arcing is evaluated in the medium current range, and is shown to have an effect on the events occurring during the first impact. A semi-empirical mathematical model is proposed for electrical contact bounce.

by current loading. An experimental investigation showed that the magnitude of electrical bounce reduction was a function of both mechanical bounce duration (the bounce duration under identical test conditions but without the influence of current) and current loading. In a second paper [6], further experiments were presented, where a focus was given to the events occurring during the first bounce. It was shown that currents in the range of 10-50A dc were having an effect on the duration and maximum displacement of the first bounce. The main influence was shown to be the velocity of impact, which is independent of current. At low impact velocities the bounce with current was reduced; and this effect was thought to be caused by preimpact NOMENCLATURE arcing, which could be expected to be both a function of impact E Modulus. velocity and circuit conditions. Under most conditions the J Inertia. bounce was increased with the passage of current; this second K Spring stiffness. effect was proposed to result from electromagnetic forces at p Dynamic yield pressure. the end of the first impact period. It was also shown that for an accurate model of the impact mechanics, it is important to I. INTRODUC~ON take the coefficient of restitution “e” (which gives an indication HE impact of electrical contacts when a circuit is closed of the energy absorbed during impact) as a variable, and not can cause severe arc erosion, and therefore, reduce the as a constant as assumed in previous studies. It was shown reliability of a switching system, particularly if the system is that there are substantial changes in e during electrical contact subject to high inrush currents. The traditional approach to bounce, as follows: i) preimpact arcing acts to reduce el (the reduce the arcing has been to increase the closing force, but impact coefficient during the first impact), ii) el reduces as the this simple solution does not account for a second solution, velocity of impact increases, and iii) the current magnitude has that of reducing the impact energy. The work presented in this a large influence on el, e2, and e3, etc. paper identifies the parameters effecting contact bounce and In the present paper the influence of the preimpact arcing also assesses their relative importance. Part of this paper has, is established, assessed, and compared to the influence of therefore, been an investigation of some of the less explained the velocity of impact on e. A model of the impact and aspects of closing contacts for example, preimpact arcing. bounce mechanics is then developed to include the influences This paper is the latest in a series of publications which discussed. have investigated the influence of current on the mechanics of impact and bounce. In the first paper [ 11, mathematical models of the mechanics of bounce developed by Barkan [2], were 11. IMPACT MECHANICSFOR A PIVOTING MECHANISM extended to take into account pivoting arm mechanisms rather The test apparatus used in this paper has been designed to than that of a ball dropped from a height onto a smooth surface reduce the effects from contact supports, elastic vibrations, [3]. The equations desc&ng the bounce mechanics were also developed to relate the bounce height to the bounce time. and electromagnetic forces in switchgear systems [7]. The It was shown that the first bounce duration was effected by schematic diagram of the test system shown in Fig. 1 shows the current; this had not been reported in previous studies, where pivoting contact arm; the test system allows for full control of the first bounce duration was taken as constant irrespective of the test parameters. The parameters measured in the computerthe current loading [4], [ 5 ] . After the first bounce, the so called based test system include the static contact force “F,”, the “subsequent bounces’’ were shown to be significantly reduced dynamic contact force “F,”the bounce displacement “2,” and the current and voltage transients across the contacts. From Manuscript received October 7, 1991; revised December 23, 1991. This work was supported by the SERC and Crabtree Electrical Industries. This these parameters, the software allows for the evaluation of paper was presented at the 37th IEEE Holm Conference on Electrical Contacts, secondary parameters such as velocities, accelerations, event Chicago, IL, October 7-9, 1991. durations, and arc energy. The authors are with the Department of Electrical and Mechanical EngiThe two parameters used to control the bounce characterisneering, University of Southampton, Highfield, Southampton, SO 95 NH, U.K. tics are the static force F, and the drop height, H. IEEE Log Number 9106587.

T

0148-6411/92$03.00 0 1992 IEEE

-

MCBRIDE AND SHARKH: ELECTRICAL CONTACT PHENOMENA DURING IMPACT

185

‘r‘ Micrometer

~

I

J

Arm\ Pivoting

Displacement Transducer

spring8

ctrical Contact ynamic Force 0.1

1 0

8

-Isrhrw

’

I

2 3 h t l c Contact Coma (N)

4

Fig. 4. The velocity of impact versus static contact force with

6

H = 1 mm.

Fig. 1. The test apparatus.

iii) there are no electromagnetic delaying forces, and iv) the impact is dominated by surface effects and not by stress wave propagation.

A. The Velocity of Impact If the angle “a’’ is small; then Kh2 &+-a!+-=O

J

h’F, J

(1)

which can be solved to give u2

Fig. 2. The nomenclature used to model the test system. Displacement

{

= ht2 h2KH2 ~

2Hh’F,

Jx2

The velocity of impact, therefore, increases with both the initial settings; H and F,. Fig. 4 shows how the velocity of impact varies with changes in F,, with the drop height fixed at 1 mm. The figure also shows the theoretical values based on the solution of (2).

n-2

B. Bounce Times Equation (1) can also be used to solve for a bounce duration as shown in Fig. 3:

A simplification of this equation can be made if the bounce heights are small and the spring force remains constant; then the first bounce time is given by t2 t3

14

(4)

Fig. 3. The notation used in the investigation of bounce and impact characteristics.

C. The CoefJicient of Restitution “ e ” Fig. 2 shows the geometry of the system used as the basis for the analytical model, and Fig. 3 shows the nomenclature used to define a particular bounce characteristic. The dynamics of the impact events are based on the following assumptions: i) there are no frictional losses, ii) there are no elastic vibrations in the contact support or contact arm,

The coefficient is defined as the ratio of the impact velocity to the separation velocity u1

= eui.

(5)

For the more general case where there are a number of

~

I -

IEEE TRANSACTIONS ON COMPONENTS. HYBRIDS, AND MANUFACTURING TECHNOLOGY, VOL. 15, NO. 2, APRIL 1992

186

C o e f f i c i e n t of R e s t i t u t i o n " e " . I n ,

~

First Bounce Ume

.

0.015

-

0.01

-

0.005

-

0.6

I

0.21

0

0.1

0.3

0.4

0.3

0.5

0.6

0.7

0.8

0.9

I

0'

I

0

2

1

Fig. 5. The variation of the impact coefficient e with impact velocity, for a range of K .

impacts during a bounce process: U,

= enun-l.

(6)

e = f(ui).

(7)

where rc is given for a ball of radius T , and mass m falling onto a flat surface of the same material by

In the case of the pivoting arm used in this analysis, the constant K , is given by

T}

p5r3hf2 rc = 2.784{

5

go to zero. Indeed, as the velocity un-l reduces, e will tend to unity, suggesting that the bounce will go onto infinity. To model the process more thoroughly, energy equations could be used, giving = U;

2E +m

for a moving contact of mass m, where during the impact process.

The function given by

+ ~ ~ ( 0 . 0 5 2 -7 e0.4218e4 ~ + 1.125e2 - l} = 0

4

Fig. 6. The first bounce duration versus static contact force, for H = 1 mm.

In a simplified analysis e is often taken as a constant, however, in previous work [6], it has been shown that

.:e8

3

st.Uc Contact lome

Velocity of I m p a c t ( m / s e c )

In both cases K will increase with an increase in p and the radius T . Both (sa) and (8b) show that the surface profile of the impact area is significant and that the radius r needs to be investigated further since on a worn electrical contact the radius would be expected to be dominated by the local roughness rather than the initial radius of the contacts. Equation (7) can be solved numerically to give e as a function of changes in velocity of impact, with )i as a parameter, as shown in Fig. 5, for a range of velocities similar to that measured in the test system. The figure shows that at low velocities e = 1, but as the velocity of impact increases e reduces. Also e is shown to be a major function of K , and reduces for a given impact velocity with K reducing. A reduction in K could be interpreted as a reduction in p or r. At values of impact velocity greater than 1 m/s, e could be taken as a constant. The use of (6) to describe the bounce process has a discontinuity for the last impact; irrespective of e being taken as a constant or not, the equation will fail, because U, cannot

E

is the energy lost

D. The Evaluation of Bounce Times Combining (2), (4), (5), and solving will give the relationship between the first bounce time and the static contact force. The results of this analysis are shown in Fig. 6, with e constant, ( e = 0.4), and H = 1 mm. Both experimental and theoretical values are shown. It shows that as the static force increases the bounce time of the first bounce reduces. A more rigorous approach to the modeling includes the variation of the coefficient of restitution; hence, by combining (2),(4),(5), and (7), we obtain the variation of the first bounce time with the static contact force. The results are also shown in Fig. 6, with K taken as 0.3. Over most of the experimental conditions this gives a very good model of the bounce times; but at low impact velocities when the static force is low, there is some variation. E. Total Bounce Times It can be shown from (4) that if e =constant:

T, =

2uien J hf2Fg

~

(9)

from which the total bounce time can be approximated using the binomial expansion to give

If, however, e = f ( u ; ) , then this equation should be rewritten as 2ui J T - -{el elez elezea ...}. -

h'2Fs

+

+

MCBRIDE AND SHARKH: ELECTRICAL CONTACT PHENOMENA DURING IMPACT 0.036 Total Bouaw Time

0.65

-

0.025

-

0.02

-

0.015

-

0.01

-

0.m

187

further, AgNi(80/20) contacts have been tested over a range of impact velocities to assess the influence of preimpact arcing on the first impact period. b) To asses the changes in e with the velocity of impact, e has been evaluated from both the displacement characteristics and the bounce times evaluated from the voltage waveforms. Using the two relations el =

U1 and

e2

Ui

-

U2

=U1

(14)

or

ma011 0

0

I

2

4

3

sI.Uc Contact Forca

6

(N)

Fig. 7. The total bounce duration versus static contact force for a range of H .

Hence, the total bounce times can be evaluated from (10) and (11). Fig. 7 shows the resultant characteristics for when e is constant. This shows that the total bounce time reduces when the static force increases, but it increases with H.

e2

T2 =Tl .

(15)

c) Impact forces: the contact forces have been measured to give an understanding of how the impact is effected by the magnitude of the energy at impact. the effect of increasing d) Impact wear studies: to the spring force on the deformation of the contact surfaces. A series of experiments were undertaken to test the contacts for wear, during make only conditions with dc currents.

F. Impact Times For two contacts of radius T , and mass m, where one contact A. Measurement of Velocity is stationary, the impact time is given by To enable the accurate evaluation of e for all conditions, it is vital to identify the velocities at the instance of separation and make. This leads to a difficult instrumentation problem, which can be discussed with reference to Fig. 3. It is clear Hence, the impact time is shown to be independent of the from the displacement characteristics that signal processing will be needed to evaluate ui and u 1 , since as reported in velocity of impact. A re-evaluation of the impact time for the pivoting mech- [6], during the impact period the displacement characteristic is deformed by the impact shock, and the release of strain anism gives energy. To establish the velocity, the voltage across the contact is used to detect the instant of separation, or impact; 10 data (13) points are then used to give the slope and thus the impact velocity. During the small displacements at the end of the where bounce period, the time method is used as expressed in (15), since it will have a greater accuracy. It should be noted that only (14)can be used to find e l . Under the influence of dc current, the bounce dynamics will and be changed by the presence of additional forces; and with the hI2F, c = -presence of additional forces, (15) will no longer be valid. 2J * Hence, in the form of (13) the impact time is shown to be Iv. RESULTSAND DISCUSSIONS a function of the impact velocity. These results are based on the experiments, listed as follows as A , B, C, and D. G. The Influence of Current 1) Preimpact Arcing: Fig. 8 shows the influence of impact All of the above equations fail to account for the influence velocity on the preimpact arc time Tp, shows that as the Of current On the bouncing and impact process* To further the velocity of impact increases, the duration of the preimpact understanding of the effect of current, a series of experimental arc This is shown to be effected by the supply, investigations have been undertaken using the test system where a larger steady-state current and voltage gives a longer described. duration for the same impact velocity. The regression fit curves presented here are based on a large number of data points. The 'I1* EXPERIMENTALINVESTIGAT1oNs voltage characteristic in Fig. 9 shows that the arc voltage is Four experiments have been undertaken in this paper as representative of the minimum arc voltages for the materials follows. used. The mechanism for the initiation of the preimpact arc a) It was identified in a previous paper [6], that preimpact arcing affected the first impact period. To investigate this requires further investigation,but possible mechanisms are: the

I -

IEEE TRANSACTIONS ON COMPONENTS, HYBRIDS, AND MANUFACTURING TECHNOLOGY, VOL. 15, NO. 2, APRIL 1992

188

0.5

0.4

0.3

0.1

11.1

Mechanical Bounce

0

0.2

0.1

0.3

0.5

0.4

0.6

- E l e c l r i c a l Bounce

0.7

(m/d

0 0

Fig. 8. The preimpact arc duration versus the velocity of impact. Pre-ImDact Arc voltace

---L

0.1

0.2

0.3

0.4

0.5

0.6

0.7

U1 ( 4 s )

Fig. 10. The coefficient of the first impact versus the velocity of impact for pure mechanical impact and impact with arcing.

-- I

0.7

0.6 0.5

0

2

4

8

8

10

I2

14

18

18

time (micro sec) Fig. 9. The preimpact arc voltage characteristic.

breakdown of the gap between the contacts momentarily before impact and the spontaneous vaporization of the asperities on impact. The data support the argument presented previously that the events occurring during the first impact are effected by the velocity of impact via the preimpact arcing mechanism. As the contacts come together, the greater the velocity, the smaller the pre-arc duration, then the smaller the effect on the first impact. This is confirmed in Fig. 10, which shows the evaluation of el based on (14), for a range of impact velocities, with and without current. The curves presented are regression fits taken over a large number of data samples. At the low impact velocities (e.g., 0.1 m/s), it is shown that the preimpact arcing sufficiently softens the surface to reduce el, from 0.5 to 0.32, a reduction by 64%. At the higher impact velocities the reduction becomes marginal. 2) The Influence of Velocity During Bounce: Fig. 11 shows the evaluation of e for individual impacts during one closure for both mechanical and electrical bounce. The top trace shows four points corresponding to the evaluation of e based on (14) and (15), for a single mechanical closure. The initial impact of 0.4 m/s gives el = 0.45, which then increases with the subsequent impacts. This curve closely follows Fig. 5, but in this case the surface will change between impacts, leading to work hardening effects which will tend to alter K . The curve shows that as the velocity reduces, e increases, producing increased elastic behavior. The lower curve shows the changes in e under the same

Fig. 11. The impact coefficient changes during a single make operation, with and without current loading.

initial conditions with the same contacts, but with a 15-A dc current passing between the contacts. The initial impact velocity is similar, but this leads to a smaller el than with the mechanical impact, caused by the effects discussed in A. The subsequent impacts shows that e reduces, in a linear relation, but only exhibits a total of three impacts, compared to the four with the mechanical closure. The linear electrical line can be expressed by the empirical relationship: e = 0.235

+ 0.371 U.

(16)

This relationship gives e as a function of velocity for the current of 15 A, for the particular contact conditions used in the experiment. The reduction in e can be explained with reference to (8b) as a reduction in K , since the second impact will occur on a softened surface as shown in Fig. 12. The degree of softening will be a function of the arc energy dissipation, but the impact mechanics will be primarily effected by the solid surfaces rather than the liquid metal interface. The third impact will occur on a surface heated further, leading to a further reduction in p and thus K . 3) ImpactForces: Fig. 13 shows an example trace of the contact displacements, arc voltages, and impact force. It shows that during the first impact for the conditions given, the impact force is 60 N, compared with a static force of 0.344

MCBRIDE AND SHARKH: ELECTRICAL. CONTACT PHENOMENA DURING IMPACT

189

Molten

Fig. 12. Localized molten metal during impact with arcing, surrounded by a much more significant area of softening. Contact Cap (mm)

Imoact Force (N)

0

: *6*0.6 o

0.2

0.4

0.6

0.6

U1 (m/d

TI

I

50..

Fig. 15. The peak impact force during the first impact versus the velocity of impact. I

IO&

......-.........

2

0

6

4

6

12

10

14

16

16

zoo 20

22

24

26 150

t (me) Fig. 13. The contact displacement, arc voltage, and impact force for contact bounce with current loading.

F1 (NI

100

U (micro aec)

-

-

9

-

W -

4Qb

900

206

Fig. 16. The duration of the first impact versus the peak impact force.

LOO

I

0

a1

0.5

1

1.5

P

2.5

3

9.5

4

4.5

(N) Fig. 14. The peak impact force during the first impact versus the static contact force for a range of H.

N. The waveform has been processed to remove time delays caused by the stress wave propagation. The impact forces are significantly higher than the static forces. This is further emphasized in Fig. 14 which shows the impact force versus the static force. This shows that as the static force and H increase, the maximum impact force increases. Fig. 15 shows a linear relationship between the impact force and the velocity of impact; and the impact force has been used in Fig. 16, to show how the impact time is affected by the impact velocity. As the velocity increases the maximum impact force increases, but the impact time reduces, as suggested in (13). 4) Wear During Impact It was shown in B that as current passes through the contacts the subsequent values of e are effected. It is apparent from (8b) that there are two mechanisms by which e can be reduced. The reduction of K , will lead to a reduced e, caused by: i) the radius of the impact region

reducing, or ii) the heating of the surface causing a reduction in p. Fig. 17 shows the surface profile of a worn contact and compares it to the original profile to demonstrate the changes in contact radius. The radius of the new contact, superimposed upon the worn contact is 7.7 mm, compared to the radius of the highest contact spot on the worn contact of 0.711 mm. If a flat surface were to impact upon the rough surface, the radius used in (8b) would be reduced by a factor of 10. Hence, K could be expected to reduce with surface wear. Fig. 18 shows the result of a single electrical impact onto a new contact, with a worn contact; this shows that, when compared with a similar impact without current, the area affected is quite similar but the depth of the penetration is double with current passing through the contacts. This demonstrates that the impact event is a localized event, only effecting a small part of the contact surface. Fig. 19 shows the result of a wear test taken over 1000 make only operations with a supply current of 15 A. The upper trace is for a contact with a static contact force of 0.344 N, whereas the lower trace is for a static force of 4.33 N; both contacts are anodes and both were tested from new with a new cathode. Thus although the higher force contact has suffered less arc erosion because of the reduced contact bounce, it

I

'

I -

190

IEEE TRANSACIIONS ON COMPONENTS,HYBRIDS,AND MANUFACTURING TECHNOLOGY,VOL. 15, NO. 2, AF'RIL 1992

+5EE.E um

+59E.E u m

+45E.E um

+4SE.E um

+499.E

+4EE.E um

UI

+359.9 um

+ 3 5 E . E um

i

+398.9 um

+ 3 9 9 . E um

i

+ 2 5 E . 9 um

+259.E

+298.9 um

+299.0 um

I

+15E.E u m

i

I I

UI

+8.9 U

+1SE.E u m

+lEE.E

+199.9 u m

+5E.E

UI

U I

+5E.E um

I

+ E . E um

I

Fig. 17. The original contact profile superimposed upon a the worn contact profile, to demonstrate changes in localized contact radii. +48E.E

+48E.9 u m

U I

*46E.E um

+46E.E

U-

+44E.E u m

+44E.E

U-

+429.8 u m

+4EE.0 um

+38E.E um

+389.8

+36E.E

+36E.E um

UI

U I

+34E.E um

+349.E u m

+32E.E um

+32E.E um

*39E.E u m

+3EE.E um

+28E.E um

+289.B um

Fig. 18. The wear profile after a single electrical make operation at 15 A.

shows the most deformation. The lower force contact was severely blackened by the arcing which can be witnessed by the degree of roughness outside the central wear area.

v.

ANALYSIS OF

ELECTRICAL CONTACT BOUNCE

From the results presented in the previous section we are now able to elaborate on the purely mechanical analysis in

Section 11.

A. The First Bounce Although the first impact is effected by the passage of current, it can be seen with reference to Fig. 11, that the reduction in el is small compared to the effects after the first impact and bounce. It is, therefore, proposed that the first

MCBRIDE AND S H A R M ELECTRICAL CONTACT PHENOMENA DURING IMPACT

191

+ 7 0 9 . 0 um

+700.0

-100.0

U I

-100.9

U I

-zeo. E

U I

-280.9

U I

-390. ci

UI

-300.0

U I

,

U I

Fig. 19. The wear profile of two anodes after loo0 make only operations at 15 A, for two different static contact forces.

impact time can be given with a reasonable approximation by (4), in the following form:

In most situations the first electrical bounce will account for most of the arc erosion.

B. Subsequent Bounces where U ; is given by (2) and e is determined numerically from (7). The evaluation of K can be established for a new contact from the materials data; however, with a worn contact, the value would be expected to change. For a new contact, used in this experimentation, K = 0.363. With worn contacts the radius could be taken as 0.1 of the new contact, giving: K. = 0.86. However, this only assumes that r changes and does not account for the heating effects on the contact materials. In previous work, with contacts tested from new contacts, on the same test system, it was shown that the first bounce actually increases with wear [l], which suggests that the changes in r are not important, and the impact mechanics depends more on the macro-surface characteristics rather than the microsurface. Hence, to a first approximation we can assume the K value used for the new contacts remains constant with contact wear. Then in (7), we can substitute for IE; in doing so, only the first and last terms become significant: ,?e8

+ 3.39 x 10-4e2 + 3.01 x

= 0.

In a further approximation, we can obtain a solution for e, as follows: K.

which can then be substituted into (17) to give

To account for the subsequent bounces, we can make two further assumptions: firstly, that there are never more than three electrical bounces, and secondly, that the values of e can be expressed in the form of (16), e.g.:

e=a+b.u. Then by substitution into (11) and including (18), we obtain the following: 2r;J T, M -. up18{1 a bKu:"}. (19)

hI2F,

+ +

C. Design Parameters It has been demonstrated in this paper that the simple solution to the contact bounce problem of increasing the static force will have consequences in the degree of contact deformation. The argument is that it will react as follows by: 1) increasing the static force; 2) reducing the bounce time; 3) increasing the impact velocity; 4) leading to increased plastic deformation; 5 ) leading to increased surface deformation. A second more satisfactory solution to the problem would be to reduce the closing velocity. This is easily achieved in the test system, but is not necessarily a reasonable parameter, for design. Hence, the most direct variable which can effect the bounce is the inertia of the moving contact J.

192

IEEE TRANSACTIONS ON COMPONENTS, HYBRIDS, AND MANUFACTURING TECHNOLOGY , VOL. 15, NO. 2, APRIL 1992

VI. CONCLUSIONS

There are a number of methods of reducing contact bounce in simple spring-based systems without the addition of further components: reducing the velocity of contacts at make, reducing the moving inertia, and increasing the static contact force; however, some may hold hidden penalties in the effect on contact wear, for example, increasing the contact force reduces the bounce time, but at the expense of additional plastic deformation of the contact surface. In addition it has been shown that by taking the coefficient of restitution “e,” as a variable, it is possible to evaluate the impact mechanics of contacts with and without a dc current. The evaluation with current is based on an understanding of the effects occurring during the first bounce, and making use of an empirical relationship for the subsequent bounces. The impact forces have been shown to be significantly higher than the static forces, and the dynamic impact force can be expected to have a major influence on the deformation of the contact surface. The influence of preimpact arcing has been shown on the first impact of contacts. It has been shown that the pre-arc is a function of the closing velocity, and supply conditions. REFERENCES [l] J. W. McBride, “Electrical contact bounce in medium duty contacts,” IEEE Trans. Comp., Hybrids, Manuf:Technol., vol. 12, pp. 82-90, Mar. 1989. [2] P. Barkan, “A study of the contact bounce phenomena,” IEEE Trans. Power Appar. Syst., vol. PAS-86, pp. 231-240, Feb. 1967. [3] E. Walczuk, “Einfluss der veranderlichkeit der stosszahl auf die berechnungsergebnisse der kontaktaabhebungen,” Kontakte in der Elektrotechnik, pp. 161-166, 1967.

[4] V. A. Erk and H. Finke, “The mechanism of bouncing contacts,” Trans. E.T.2 Electrotech. A , vol. 86, pp. 129-133, 1965. [5] -, “The behavior of different contact materials for bouncing contacts,” Trans. E.T.Z Electrotech. A, vol. 86, pp. 297-306, 1965. [6] J. W. McBride, “An experimental investigation of contact bounce in medium duty contacts,” IEEE Trans. Comp., Hybrids, Manuf: Technol., vol. 14, pp. 319-326, June 1991. [7] -, “Computer controlled testing of electrical contact mechanisms,” in Proc. 13th Int. Conf: on Electrical Contacts, Lausanne, Switzerland, , 1986, pp. 432-436. [8] F. P. Bowden and D. Tabour, The Friction and Lubrication of Solids Oxford, U.K.: Oxford University, 1950, pp. 260-264.

John W.McBride received the degree in aeronautical engineering from the University of Southampton and the Ph.D. degree from Plymouth Polytechnic. He has lectured in the Mechanical Engineering Department at Plymouth Polytechnic and in instrumentation and measurement with the Department of Mechanical and Electrical Engineering, University of Southampton. His research interests include electrical contacts, automotive electronics, and medical instrumentation. Dr. McBride is a member of the IEE and a Chartered Engineer.

Suleiman Sharkh received the B.Eng. degree in electrical engineering from the University of Southampton in 1990. He is currently pursuing a Ph.D. degree. Mr. Sharkh is a associate member of the IEE.