7.
1
The repulsion of electrical contacts crossed by short-circuit currents
D. PICCOZ - 2), Ph. TESTE"', R. ANDLAUER"', T. LEBLANC"' and J.P. CHABRERIE
(I) Laboratoire de Wnie Electrique de Paris CNRS (URA 127), Plateau du Moulon, 91 192 Gif sur Yvette
T61 :01.69.85.16.761Fax: 01.46.97.30.08
(2) Schneider Electric, 33 Avenue du M d c h a l Joffre, 92000 Nanterre Cedex France
T61 : 01.41.39.62.001Fax : 01.41.32.62.78
Abstract
The study of forces exerted by an arc on silver based electrical contacts is important in order to understand
the dynamic phenomena concerning the opening or closing phases of contacts in a variety of switching applications.
For that, an experimental device was built. It allows us to measure accurately the force on the static electrode and the
displacement of the mobile electrode (and hence the force on it) when the two contacts, crossed by a current pulse
(from 0 to 4 kA) separate. The measured forces are the sum of Lorentz magnetic forces (constriction force and loop
effect force) and blow open forces due to the arc plasma (evaporation and (or) decomposition of contact materials).
We have studied the influence on the measured forces of the contact material, the duration of the current pulse, the
polarity of the electrodes...
When the contacts separate, an electric arc appears.
1. Introduction F L and Fc still act on the contacts. A supplementary
Contactors are often connected into series force will also act on the contacts. This force
with a disjonctor. They must switch currents which (denoted FA) is due to the arc action on the electrode
are often several times greater than the nominal surfaces and may be caused by :
current value I,. Concerning their "dimensioning -the electrostatic strain
rules", there are some important problems which are -the droplet ejection
to be solved, such as : -the plasma pressure action (<<blow open
- contact resistance, force >>)
- welding during the bounce phenomenon, -etc.
- blowing of the arc during the opening contact
phase, A preliminary bibliographical survey showed that
- contact repulsion under short-circuit currents. only a few studies have been done in that field [ 1, 2, 3
,4, 5 , 6, 71. This is why we propose here a new
We present here a study relative to the last experimental set-up, using the same force transducer
point. Indeed, when a contactor is crossed by a very than the one already described in [6, 71. This device
intense current, such as a short-circuit current able to had been called FMD for Force Measuring Device.
reach ten times the nominal value I, or more, Such a set-up enabled us to study the transient
repulsive forces tend to open the contact and generate phenomena relative to the repulsive opening phase of
an electrical arc. Energy is transferred to the zones of two contacts under the effect of an intense current
electrical contact and can lead to melted regions. pulse. Studies have been carried out in air for various
Insofar as it is difficult to avoid a fast reclosing of the electrodes materials and various current pulse
contacts, one can encounter severe welding durations. Contrary to the device described in [3], this
phenomena able to put the apparatus out of use. one enables the imposition of a given and constant
(versus time) contact force (denoted Fo)
When the contacts are closed two kinds of magnetic
forces can lead to a contact opening phase : 2. Experimentation
- a force FL, again called "loop effect", due to the
geometry of the electrical circuit around the 2.1. Descriution of the exuerimental device
electrodes, The scheme of principle of the experimental
- a force F, due to the constriction of the current set-up used is presented in figure 1. The device is
composed of several parts :
lines in the vicinity of each contact spot.
-a motion less set (7) which comprises the
motion less contact (lower contact). The force
0 1999 IEEE
0-7803-5549-0/99/$10.00 129
exerted on this electrode is measured with the
help of a pre-stressed piezoelectric transducer ! n
(2). This quartz transducer is thermally and
electrically insulated with the help of ceramic
pieces (8) and magnetically screened ; this is
achieved utilizing axisymmetric geometry which
enables a very efficient protection of the quartz.
The frequency response of the force
measurement system is limited by the charge
amplifier to 180 kHz. An annular groove
machined in the base piece (1) and filled with a
liquid metal (6) allows the current to flow from
the electrode to the generator. It ensures a good
mechanical uncoupling between test electrode
(7) and the base piece (1).
-a moving set ( 5 ) able to move vertically is
guided with a very low stiffness by two flexible
diaphragms (3) made of composite materials.
The use of an annular groove filled with liquid
I
metal (10) enables to uncouple the moving set I
(5) of the fixed referential. The displacement of i Z
the upper set is measured with the help of an
inductive sensor (4). The sensor electronics
consist of an oscillator (1 MHz), linearization
network, amplifiers and a demodulator which
provides an analog voltage directly proportional fig I :Scheme of principle of the experimental set-up.
to the displacement (1 V/mm). We have verified
1 : base piece, 2 : piezoelectric force sensor, 3 : 2 thin diaphragms
that the current passing through the contacts did guiding vertically the mobile equipment, 4 : inductive sensor of
not affect the inductive displacement sensor. displacement, 5 : moving contact (or moving set), 6 and 10 : two
With this sensor it is possible to measure rings of mercury, 7 : motion less contact (or motionless set), 8 :
displacement with a precision of about 5 pm. two ceramic pieces, 9 : current leads. On this figure the current
path is indicated with the help of arrows.
The current intensity I(t) is measured via a Rogowsky
transducer. It is supplied by the discharge of a
2.2. Recall of the exmession of the force measured bv
capacitor bank in series with an adjustable inductance
the FMD
and in parallel with a free wheel diode. The voltage
drop (contact voltage or arc voltage) was also
2.2.1. Case of closed contacts :(FA= 0 N)
recorded.
When the contacts are closed, the force exerted on the
moving set ( 5 ) and the force exerted on the motion
less set (7) is the sum of:
-the Lorentz forces,
-the forces exerted by the liquid metal filling the
circular grooves (6) and (lo),
-the weight of the moving set ( 5 ) which is the
contact force Fo and its reaction on (7).
The calculation of the two first forces have been
detailed in [6].We can recall that the force measured
by the FMD (denoted FM)may be written as follows :
FM = (FL Fc -I- FHgh + (FL -I- Fc + FHg)5 (1)
130
*The sum FL+ FCis the Lorentz force exerted on (7) When the contact are closed the force transducer
and (5). Its vertical component is equal to : measures then a force FM'= 0 N
(FL+Fc);+s = %TZk.nk.dS (2) 2.2.2. Case of opened contacts
s7+s5
The contact will open when the force exerted on the
where S7 and S5 are the total surfaces of the piece (7) upper set : (FL + Fc + FHJ5 will be greater than the
and (5). Tik is the Maxwell stress tensor for Lorentz
contact force Fo.
forces.
If we make the assumption that :
It may be shown that if the current density in the -the arc spot is circular (its radius is denoted
annular groove is uniform (space), this force may be
b ) and located at the center of the circular
written under the following form :
electrode surface,
-the current distribution in the spot is uniform
(3)
(space),
K7 and K5 depends on the geometry of the two sets then, FL + Fc is only exerted in the vertical direction
(7) and (5) respectively and on the contact area and may be analytically calculated.
radius. However, the algebraic sum K7 + K5 does not The total force exerted on the set (7) (Fn) is the sum
depend on the electrical contact area between the two of :
sets. -the Lorentz forces,
For instance, schematically, if: -the forces exerted by the liquid metal filling the
-there is only one contact aera at the set circular grooves (6),
interface, -the force due to the arc action FA
-it is centered on the electrode surface (the We have then, under the previous assumptions :
contact aera radius is denoted a ) , FW=FA+ ( F L + F c + F , , ) 7 = ( A + C 7 + ~ l n ( ~ ) ) . 1 2 + F A
we would have then :
(equation ( 6 ) )
K7 = A + B In ( a R )
In this case the force transducer will measure :
and (4)
K5 = A' - B In ( u R ) FM = Fl-7 - Fo (7)
because the set (5) has moved and does not exert any
where A, A' and B depends on the geometry of the contact force on the set (7).
set (7) and R is one dimension of this set. We can also note the expression of FT5which is the
force exerted on the mobile set (5) :
In our case we have chosen the geometry so that the FT5 = FA FO (FL 4- FC 4- FH&5 (8)
force is equal to zero when the contacts are closed
(K7 + K5 =o) 2.3 Calibration of the device :
The device was calibrated, in static as well
* F H is
~ the force exerted by the liquid metal (Hg) as in dynamic regime with the help of a liquid
filling the circular groove. Under certain conditions, standard constriction (Hg) schematized on the figure
this force is only in the vertical direction and its z 2. On this figure dimensions are given in 111111.
component, for instance for set (7), may be written
under the form :
=c,.I~
where S7w is the surfaces of (7) wetted by Hg. P is
the pressure, o* is the viscous rate of strain and 6*
the unity tensor. A similar expression is valid for set
(5) (we have (FHg)5= Cs.I*).
We have chosen the geometry so that C7 + C5=O.
131
< 30
>, lead to a gap lower than 5 % with the measured
26
r( r: signals of displacement or of force respectively.
This demonstrates also that the mobile equipment is
correctly guided vertically by the two thin
diaphragms and that the friction forces during its
motion is negligible for the electrode displacement
values used here.
6 - . , . , . , . , . 08
A: current \slue
B :force signal daiwed bythetransducer
C :displacement of the mobile equipment
fig 2 :Scheme ofprinciple of the device in its calibrcrlin phase
2.3.1. The static calibration
The static calibration has been described in [7]. In
this case, the mobile equipment is fixed (z(t)= Cst and -21 . , . , . . -0.2
id$<< 1) where I$ is the diameter of the two 0,oo 0,Ol 0,CQ 0.b 0,b 0,L
electrodes. time(s)
As it has been recalled in 2.2, the total force F(t), fig 3 :temporal evolution of the current value (curve A), of the
measured by the FMD,is of the form K I*. We found force signal delivered by the sensor (curve B) and of the
dkplacement of the mobile equipment (curve C)
a good agreement (within 3%) between the calculated
and the measured value of K.
2.3.2. The dynamic calibration
At t=O the position of the mobile set is z = 0. The
upper set (5) is free to move along the vertical axis.
1 t
On figure 3, we give an example of the triplet I(t),
x(t) and F(t). On figure 4 we have plotted the force
signal delivered by the force transducer versus the
value of the intensity square. [ap] contact tehwm the contacts
As long as the contact between the two electrodes is [p y s]theelectrodes areseparated
established, the signal force delivered by the force
transducer is equal to zero (cf. 2.2.1.). That can be -2
0 1 2 3 4 5 6 7
verified on figures 3 and 4 (see segment [a-@ on
figure 4). current value square ( kA* )
fig 4 :evolution of the force signal in terms of the current value
When the mobile equipment (5) separates from the square
motion less set (7),the theoretical expression of FMis
given in by equations (6) and (7) with FA= 0 and FO= 3. Examples of experimental results
0.65 N (weight of the mobile set (5)).
These equations can be verified on the segments f3y As on the curves presented hereafter, the
and yS of the figure 4.From this figure we can deduce time origin does not correspond to the current pulse
that the factor K7 + C7 is equal to 1.8 l o 7N/A2. This origin, we will denote t the time taken with the origin
value fits the theoretical one within 2% (Cf [ 6 ] ) . of the figures and 6s the time having.for origin the
beginning of the current pulse ;6s = 0for I = 0'
.
The dynamic calibration of the device has been The results presented below are extracted
realized for values of dI/dt in the range [0-3.lo6A.s-'] from a series of tests carried out with Ag-Sn02 (88-
Moreover, the double integration of the force signal 12) electrodes coming from the same fabrication
or the double derivation of the displacement signal batch. The diameter of the electrodes is equal to 8
mm. The electrodes are crossed by current pulses for
132
which the current peak occurs approximately at 6t = resulting from the action of (< plasma pressure >> on
4.7 ms. The contact force FO between the two the contact surfaces.
electrodes is equal to the weight of the mobile
equipment (1.2 N for figures 5 and 6 and 2.4 N for On figure 4, the first peak presented by the measured
figure 7). force F M at t = 0.05 ms (or 6t = 0.0005 ms) may be
attributed to a collapse of asperity in the apparent
3.1. Example no1 contact area. Such a peak disappears after some
The first example of contact dynamics is current shots.
given on figure 5. When the current value reaches
1530 A (curve A) the anode takes off (curve C), 3.2. ExamDle n"2
generating an arc (curve D). If we examine the arc Figure 6 illustrates a case of an "early"
voltage evolution, we can see that during opening of the contacts. Indeed, the contact opening
approximately 2 ms the voltage value (about 15 V) occurs for a current intensity of 270 A. When the
corresponds to the case of a ((metallic B arc. contact opening occurs we have :
10 that is to say
8
6 However, as soon as the contact is opened, we have
FM= -1 N which means that Fn = 0.2 N and then has
4
collapsed. It increases again while the upper electrode
2 embeds itself of 60 pm in the lower one.
0
-2
-10, . I . I . I . I .
0,oO 0.01 0.02 0,03 0,04 0.05
time ( s )
fig 5 :temporal evolution of the current value (curve A), of the
force signal delivered by the sensor (curve B), of the ? 20 4
displacement of the mobile equipment (curve C ) and the voltage rn
a,
(curve 0) -g 10 2
2
After, the electrode gap increases and reaches a value
high enough to let the air enter the arc column. The 0 0
voltage value becomes comparable to value observed
, c , E ! ,
for arc in air [8], that is to say about 22 V for a -10
0,oO 0.01 0.02 0.03 0.04 0.05
electrode gap of 2 mm. time ( s )
The contact is restored a long time (20 ms) after the
fig 6 :temporal evolution of the current value {curve A), of the
end of the current pulse. There is no welding effects : force signal delivered by the sensor (curve B), of the
the upper contact bounces three times before coming displacement of the mobile equipment (curve C ) and the voltage
to a standstill. The electrode erosion is weak. (curve 0 )
We observe that the maximum value of FMis about
5.1 N and then the maximum value of FV is about 6.3 We are in presence of a very short metallic arc. Then
N. This maximum happens slightly after the FV becomes sufficient to make the mobile equipment
maximum value of the current. In [9] and [lo] coming up again until x = 100 pm and then falls back
measurement of the force exerted on the electrodes by on the fixed electrode. Here again, no welding effect
an electric have been realized in the case of constant appears. For I = 2400 A, FTl is maximum and equal
electrode gap (in the range 2 to 8 nun) and for similar to 2.3 N. If we make the assumption that the current
electrode material and arc current pulses. It has been density in the cathode spot is equal to 5.10' A/m2
observed that the maximal value of the force exerted [ 101, we obtain that (FL + Fc + FHJ7= 1.7 N (this is
in such a case is quite smaller (the maximum is about obtained with the help of equation (3) and (5)) and
2.8 N) than the one observed in the case of opening then we obtained that the force FA due to the arc
contact. Such a difference may indicate the existence action (plasma pressure for instance) is equal to 0.6 N
and the order of magnitude of blow open forces
133
3.3. Examde n"3 certain influence of the current time derivative on the
Figure 7 presents the evolution (versus time) arc appearance time.
of I, FM and x, but, now, for Fo = 2.4 N. For I = 500 A When the duration of the current pulse
the two contacts separate. The measured force signal grows, for a constant current peak value the total
FMdecreases abruptly before beginning to increase force Fn grows too. For example, when the duration
again. The upper electrode (anode) goes down in a of the pulse doubles, Fn is 25 % bigger. Moreover,
first time for an hundred of microns and then goes up Fn is often maximum when the current reaches its
again for fifty microns. Then FMdecreases again with peak value.
I(t) in order to give FM= -2.4 N (that is to say Fn =
0) for I equal to zero. The upper electrode embeds 5. Influence of the Dolaritv of the electrodes
itself in the cathode. The total displacement of the The vertical component of the force exerted
moving equipment is 0.3 mm which corresponds to a on the anode is obtained from the displacement signal
very severe erosion (strong surface deformation) for of the mobile equipment, smoothed and derivated
only one current pulse. The two electrodes are totally twice and multiplied by the value of the mass of the
and strongly welded. The erosion of the fixed mobile equipment (M = 0.12 kg in the case of figure
electrode (cathode) is greater than the one of the 8. This figure represents the time evolution of the two
anode. The "pushing in" of the anode is all the more forces exerted on an Ag-SnO, (88-12) anode and
important than Fo takes a great value. cathode crossed by a current pulse presenting a peak
value equal to 2900 A. It shows us an anode force
signal slightly higher than the cathode force signal.
A cuirentwlue '
ce signal deliwed bythe transducer
E
E 4,1
Y
A: force signal calculated at the anode
B : force signal measured at the cathode
n
0,w 0,Ol o,b2 o,b3 0,b 0,bS -5
.
I
time(s) 0,oo 0.01 0,02 0,03 0.04 0.05
fig 7 :temporal evolution of the current value (curve A), of the time ( s )
force signal delivered by the sensor (curve B), of the
displacement of the mobile equipment (curve C) and the voltage fig 8 :temporal evolution of the force signal delivered by the
(curve 0) piezo-electric transducer and the signal force calculated from
the displacement signal of the mobile equipment.
We have noted an important dispersion of the results
However, the difference (about 5 %), found in the
particularly with regard to the displacement of the
calibration phase described in paragraph 2.2. between
upper electrode. That may be explained by the fact
the curve resulting from the two derivations of the
that the surface geometry (roughness, for instance)
displacement signal and the recorded force signal,
and surface composition may change after several
allows us to say that the forces exerted on both
tests. For instance we have observed a decrease of the
electrodes are quite similar.
silver quantity on the electrode surface after several
arcs. Moreover, it seems that the displacement
6. Influence of the contact material on the opening
amplitude decreases with the number of operations.
phase. '
The Ag-SnO, (88-12), Ag-CdO (88-12) and
4. Influence of the current time derivative
For the totality of the tests realized with the Ag-CW (60-40) electrodes are crossed by a current
current peak intensity at 6t = 4.7 ms, we noticed that pulse (current peak value at 6t = 4.7 ms). The contact
the separation of the Ag-SnO, electrical contacts force between the 2 electrodes is equal to the weight
of the mobile equipment (Fo = 1.2 N). On figure 9 we
happened solely when the current peak intensity was
have plotted for these three electrode material the
higher than 1200 A. These results seem to show a
maximal value of Fn reached versus the peak value
of the current pulse.
134
On figure 10 we have plotted the maximal value of to be able to use the blow open forces to try to avoid
the displacement of the mobile anode versus the peak electrode severe welding.
value of the current pulse. We remember that the tests
showed us that the measured forces are often 8. Bibliomaphy
maximum when the current value is maximum.
maximal force (N) [ 11 J. J. Shea, B. de Vault and Yun-Ko Chien. << Blow
71
open forces on double break contacts >) IEEE Trans
0 0
on CPMT, part A, Vol 17, no 1, March 94, p. 32
0 AgSn02 (8812)
0 AgCW (60-40)
[2] F. A. Holmes and P. G. Slade. asuppression of
pip and crater formation during interruption of
4-
moo 0 alternating current circuits >> IEEE Trans on CHMT
3- 0 0 Vol-1 n"1 March 78 p. 59
0 0QJA
2- 0
gw 0 [3] R. Robertson, <<Theforce on the cathode of a
1- 0 copper arc >>, Phys. Rev., vol. 50, p. 578, 1938
0 1
0 500 loo0 1500 2000 2500 3000 3500 4000 [4] L. Tonks, <<Theforce at an anchored cathode
current peak value ( A ) spot P, Phys. Rev., vol. 50, p. 226, 1936
fig 9 :maximal value of F n (cathode force) versus the current
peak value [5] E. Kobol, << Pressure and high velocity vapor jets
at cathodes of mercury vacuum arc >>, Phys. Rev., vol.
36, p. 1636, 1930
[6] 3. P. Chabrerie, Ph. Test6, R. Andlauer, Th.
c
Leblanc. << Experimental study of contact opening >>.
1500
0 Proc. of the 41" Holm Conference on electrical
o
0 contacts, 1995, p. 194
A
0
.-
U OO [7] J. P. Chabrerie, J. Devautour, A. M. Gouega, Ph.
A o
Test6 << A sensitive device for the measurement of the
'n force exerted by the arc on the electrodes >> IEEE
Trans on CPMT, Part A, Vol 18, no 2, June 1995.
0
0 500 looo15002ooo25003ooo35004oOo
current peak value ( A ) [8] A.M. Gouega, Ph. Teste, R. Andlauer, J.-P.
j i g 10 :maximal displacement of the mobile anodic equipment Chabrerie and J.Devautour << Study of the electrode
in terms of current peak value gap influence on electrode erosion by an electric arc >)
The results show that the forces exerted on to be published
the Ag-SnO, electrodes are higher than the forces on
the Ag-CW contacts, the last forces being higher than 191 A. M. Gouega, PhD Thesis, University Paris VI,
the forces on the &-CdO electrodes. 1995.
7. Conclusion [ 101 J. Devautour, PhD Thesis, University Paris VI,
The experimental device presented in this 1992.
paper enables the study of the dynamics of the
opening phase of contacts under the effect of an
intense current. The force exerted on the motion less
contact and the displacement of the other contacts are
measured. Given Lorentz loop effect so as given
contact force may be imposed.
It seems that the current time derivative has an
influence on the repulsive forces, so as the electrode
material. Nevertheless, the contact opening dynamics
is too dependent on the contact surface state to hope
135
