C.P. No. 987

MINISTRY OF TECHNOLOGY

AERONAUTICAL RESEARCH

CURRENT PAPERS

The Influence of Gas

COUNCIL

Streams and Magnetic Fields on

Electric Discharges

Part 4. Arcs Moving

along Straight Parallel Electrodes

by

V. W. Adams

LONDON: HER MAJESTY’S STATIONERY OFFICE

1968

PRICE 9s 6d NET

U.D.C. No. 537.523 : 573.6 : 621.3.032.2

c .P . NO .907” March 1967

THE INFLUENCE OF GAS STW6 AND MAGNETIC FIELDS ON ELECTRIC DISWARGES

PART 4. ARCS MOVmG ALONG STRAIGHT PAPALLEL ELECTRODES

Qr

V. W. Adams

This report is a contlnuatlon of the experimental work jn air at one

atmosphere on the motion of d.c. arcs alang straight open-ended electrodes

already reported in Part 3 of this series. The esrlzcr work with the self-

field due to the arc current m the electrodes in the s?mme direction as the

applied ma&n&x fxld has been ertended to mclude results with fxlds up

to 1.0 wbg. Results are also given for three eiectrod~? materials with

electrical connections azanged so t'nct the self-field is eliminated, and are

shown to differ from those where the self-f:+id iz present and apFrcximately

calculated.

A method, employing similarity paraneters given bjr l?autov and Zhukov, of

collating data on arc n,otlon In transverse mognetrc fieids 1.3 used to coqare

results for ihe present series of report& with other published dat3, and is

shown to be effective for a wide range of parameters frotti elevtin different

papers, including the present results.

Arc notion with a iow-vclocicy iqos;d au flor; alofig and relatlw to

the electrodes has been studled, and the sunple notion that the algebraic

difference between the arc veiocitws relative to the electrodes in still

and moving au is equal to the iirposed air velocity 1s found to be only

approximately true when the air stream acts ?gomst the arc motion.

* Replaces R.A.E. Technical Report 67077 - A.R.C. 29302.

2

I

2

3

4

5

CONrENTS

INTROlXJCTION

APPARATUS AND MEASURING TECHNIQUES

2.1 Electrodes

2.2 Measuring techniques

2.3 Piiotographic observations EXPERIMENTAL RESULTS

3.1 Extension of results with brass electrodes and single

B

3

4

4

4

6

6

end connections

3.2 Electrodes with double end connections 3.3 Brass electrodes with single end comections end ir.lposcd

6

7

air flows

DISCUSSION

4.1 Effective drog arca

4.2 Relations between U, B, I end d

4.3 Analysis using similmity relations

4.4 Effects of imposed air flow

CONCLUSIONS

Table i

Symbols

References

Illustrations

Detachable abstract cards

7 8

8

IO

IO

12

13

15

17 18

Figures l-25

077 3

1 IETROWCTION

* Experimental results on the behaviour of direct-current arc discharges in

air at a pressure of one atmosphere propelled along various electrode systems I,2

. by magnetic fields have been published in previous reports . The present

report is concerned with arcs moving along a pair cf straight, parallel,

horizontal open-ended electrodes at one atmosphere, and falls into four parts:-

(1) Extension of the previous work on straight brass electrodes with

current connections to one end as shown in Fig.l(a) 60 that the self-magnetic

field (i.e. the field at the arc of the arc current in the electrodes) is in

the same direction as the externclly applied magnetic field. In this work an

allowance is made for the self-field using the method of calculation given in

Ref.2. The previous results using an arc current of 3700 amp have been

extended to include results in a total magnetic field (externally applied

field plus self-field) of 1.0 W%/m2, resulting in a maximum arc velocity of

580 m/s (Mach Eo.l.65).

.

(2) The motion of arcs on straight electrodes where the self-field is,

as far as possible, eliminated. In addition to brass electrodes for comparison

with the previous work, aluminium and copper electrodes are used. For the

earher work where the self-field IS calcuialxd rt is thought that the total

magnetic field i.e. applied field plus self-field, cannot be accurately

determined at the arc position because of an asymmetric current distribution

in the electrodes. Thus, in order to eliminate the self-field and obtain

results more nearly representing the effects of the accurately known values

of the applied field, balanced current connections are made to both ends of the

electrodes, (Fig.l(b)). For practical rs;lsons this limited arc currents,

for the present apprrratus, to 1000 amp or abaat one quarter of the previous

maximum. The applied field has not yet bien taken beyond 0.12 Wb/m2.

(3) Collation of data on the motion of arcs in transverse magnetic

fields using non-dimensional simiixity parameters given by Bautov and

Zhukod . The results in the present series of reports are compared with other

published data for arcs moving in air at one atnos?hzre.

(4) The motion of arcs on straight el,ectrode s with an imposed air flow

relative to the electrodes either aiding or, opposing the effect of the Lorentz

force (I X B) on the arc. These experiments necessitated, for practical

reasons, the use of electrical connections only to the do-istream ends of the

electrodes. Hence, it is not possible to eliminate the self-field. With the

Lorentz force and air flowopposed, the self-field is in the same direction as

4

the externally applied field. On the other hand, with the Lorentz force and air

flow in the same direction, the electrode polarity, and hence the Lorentz force,

iS reversed and the self-field opposes the applied field. Corrections for the

self-field were made accordingly; the method of calculation given in Ref.2,

although approximate, was used. The electrodes were enclosed in a duct which

limited arc currents to 600 amp. However, results were obtained for fields up

to I .o wb/m*.

2 APPARATUS AND MEASURING TECHNIQUES

The arc power supplies and electromagnet providing the vertical external

driving field have been described in Ref.2.

2.1 Electrodes and ducts

The electrodes used for experiments in still air conalsted of a pair of

horizontal, flat brass bars, 1.2 m x 0.038 m X 0.0033 m, the arcs being started

at one end and confined to the edges (0.0033 m wide). Arc starting was by

either fusing a short length of wire, or using a piece of foldEd metal foil

between the electrodes2. The electrodes were insulated on the qpcr and lower

surfaces with refractory insulating material.

For the experiments with imposed air flows relative to the electrodes, a

pair of flat, horizontal brass bars 0.0033 mthick were used in a long duct as

shown In Fig.2. The air flour, provided by a centrifuge1 blower was fed to the

duct by a flexible rubberised fabric hose 0.2 m in diameter. Experiments were

made with the imposed flows (including zero flow) either, in the same direction

es or, against the Lorentz force on the arc. The top of the duct was made of

pfrspex to permit optical observations. To pronde higher flow velocities in

the arc gap the duct cross-section was reduced first by wooden spacers and

finally, by a resin bonded paper tube fitted with perspex windows, Fig.3.

The method of arc starting was dependent on the directron of air flow relative

to the arc motion. For flows against the arc motion successful arc starting

was only achieved by providing a stagnant region at the downstream end of the

eiectrodes and hence it was convenient to use a piece of metal foil. For flows

with the arc motion no stagnant region was necessary and it was more convenifnt

to start the arc by fusing a wire.

2.2 Measuring techniques

Measurements of the mean src velocity, the total arc voltage and arc

current were recorded on a Polaroid camera and Tektronix type 502A oscilloscope

using the optical probes and apparatus described in Ref.2. Minimum values Of

77 5

the arc voltage were measured from t‘ne records since it was considered that

these correspond to arc lengths equal to tha ejectrode spacing. Tvo optical

probes fixed 0.5 3~0.001 m apart and about 5 cm above the centre of the arc gap

were used for velocity measurement, and a third probe, placed in front of the

first of the two measurxng probes and 3.2 m from the starting position of the

arc was used to trigger the oscilloscope. In this way sny effects due to the

initial starting period of the arc wara avoided.

Double end current connections necessitated the use of a different

arrzngament for current and voltage measurements, as shown in Fig.l(b). In

this case the arc voltages were meesured by using two Tektronix type 6017 ten

times attenuation probes connected one to each electrode, the output signals

with respect to the earthed connccticn being algebraically added by making

use of the differential operation of the oscilloscope input amplifier. Thus,

with t'flis method the mcasur~d voltage was proportional to the difference between

the two input voltages from the electrodes.

The probe signals, corresponrling to the transit time of the arc past the

two measuring stations, were reduced by a suitabic attenuator and superimposed

on t'ne arc current signal by mskin g use of the differential input to t'ne second

beam of the osciLloscope. In this way, one two beam oscilloscope was used for

recording arc voltage, current and velocity on a corzrion time scale, SEE Fig.4.

Previously it was reported' that the arc current records contalned high

frequency noise signals which were clminoted by means of a low pass fil.ter,

e.g. Fig.4. However, the H.F. noise was subsequently traced to its source and

the source eliminated. No difference in the res~ults was apparent from this

change.

The applied magnetic fields were taken from calibration curves of magnet

supply current against magnetic field obtaxncd by using either, a search coil

and fluxmeter for fields up to 0.15 Wb/n2 or, a E&l1 probe fluxmeter 4

accurate

to _+I% for the higher range up to 1.0 Wb/n2. The two methods were found to be

compatible over an overlappping range up to 0.15 Wb/m2. The self-field due to

the arc current in the e1:ctrode.s was, where appropriate, caiculatcd using the

method given in Ref.2 and a correction made to the applied field.

The accuracy of the mcasurmg techniqxs has bc-tn discussed in the

previous reports of this series end typical observation errors are shown in

tnc plotted experimental results. A list of errors for all ~aremetzrs used

is given with tne list of syt!bols.

6

2.3 Photographic observations

High speed photographs of the monng arcs IR the ~revmus reports were

obtained using a Fastax rotating prism camera (lOL frames per second), and it

was found that changes in tine arc shape occurred between successive frames

probably due to insufficient framing speed of the camera. It was thus decided

that a higher framing rat* of at least IO5 ?~ctures per second would be

required. The sxnplest and most economical camera providing this facility 5 appeared to be the aperture - scanning type used for circuit breaker research,

and a camera of this type was made at R.A.E. using drawings provided by

A. Reyrolle and Co.

3 EXPERIMFPITAL RESULTS

The results are collected together in Figs 5 to 17. The extensions to

the results for single end connections using brass electrodes are shown in

Figs.5 to 7, while Figs.8 to 15 are for electrodes (brass, copper or aluminium)

with double end connections and Flgs.16 and 17 are for arcs with &posed air

flows along the elactrodes. n

It was found’ that the arc velocities were consistent only after several

arc runs had been made end visible tracks appeared on the electrodes. Thus,

the results are for electrodes used between 10 and 100 times.

The photographic observations with the framing speed now avaxlablc

(IO5 pictures per second) showed that the changes m arc shape were gradual

with many fewer abrupt changes between successive frames.

We now consider the results for each of the three types of experiment in

turn before discussing them in a later section.

3.1 Extension of results with brass tfectrodcs and single end connections

The mean arc velocities, U, and minimum values of arc voltage, V, are

platted against magnetic field, B, in Figs.5 and 6 respectively, for an arc

current, I, of 3650 Yi50 amp. They are swn to follow the trend of the previous

results. Calculated values of the induced voltage due to the arc motion in the

magnetic field are shown to be negligible, Fig.6. The earlier work has also

been extended by determining the variation of arc velocity witn current and

field for a larger arc gap, Fig.7.

.

3.2 Electrodes with double end ccnnections

This configuration eliminated as far as possible the self-field due to

the arc current in the clcctrodes , and the results may be compared with those

of the earlier work2 where the self-field was calculated.

The mean arc velocities are plotted in Figs.8 to 12 egainst arc current

for each electrode material and a range of gap widths, d, and applied magnetic

fields. The variation of velocity with ar c current and magnetic field is

different for each material, and in particular the brass results differ from

those for the axperiments corrected for single end cOnncctions2, (See

section 4 and Table I).

Minimum values of arc voltage, V, against applied magnetic field and arc

current are given in Figs.13 to 15. It may be shown that, with the exception

of aluminium, a plot of V against the electrode spacing, d, gives curves with

slopes similar to those obtained for experiments with single end COnneCtiOns 2

.

3.3 Brass electrodes with single end connections and imposed air flows

The connections to these electrodes had to be made to the downstream ends

and corrections for the self-field at the arc gap were made accordingly. With

the air flow and arc motion opposed the arcs were started at the downstream

ends of the electrodes and the self-field was added to the applied field.

Alternatively, with the air flow and arc motion in the same direction the arcs

were started at the upstream ends of the electrodes with reversed electrode

polarity and the self-field was subtracted from the applied field. The maximum velocity of the air flow was calculated from measurements made after the arc had

travelled along the electrodes with a 0.32 cm diameter Pitot tube held at a

point midway between the electrodes. No correction was made for the proximity

of the electrodes to the Pitot tube. Measurements made with the tube adJacent

to the electrode edges gave a value for the velocity about 3/k of that at the

gap centre.

The results are shown in Figs.16 and I7 where the arc velocity relative

to the electrodes, U, and the minimum arc voltage, V, are plotted against the

magnetic field for each air flow velocity. It can be seen from Fig.16 that

with the air flow and arc motson in the same direction the condition was

reached where there was no net magnetic field and the arc was driven by the

gas stream. The arc velocity was then about $ the air flow velocity.

a

Some attexpts were nlso made to r,pproach the condition of zero arc velocity

with the gas strean opposing the arc notion by removing the externally cpglied

field, so that the arc was driven only by the self-magnetic field. However,

because of the err&x behcviour of the VC: md possTble domnge to the electrodes

no velocity measurerxnts were obtnined.

High speed photographic observations of the arc showed that the shape of

the luminous column when moving under the nctlon either, of the irposed gas

strean in the same direction as the arc notion ?nd no cpplled Icagnetic field

or,of the gcs stream plus zn externally applied field, differed from the

shape for WCS moving through six11 nu In a transverse magnetic field. For

arcs moving through still air the luminous column was zpproxunatcL~ stmlght2,

whereas for w‘cs SubJected to an nddltio no1 gcs flow along the dircctlon of

motion the luminous colun?n was cusp shaped with the apex of the cusp pointing

in the dirfction of motion, cs shown in Flg.18. The rclotive posltions of the

frees (approximately IO5 per second) In this photograph are due to the camera

design, described In Ref.4.

4 DISCUSSION

The results described In scctlon 3 will now be consz.dwed In more detail

and the discussion falls into four pxts.

4.1 Effective drag ?re,-. of crc

Assuming that the arc cauxs n displncawnt of the extcrnrl gcs which

does not close downstream (i.e. c w&e IS fonr.ed) it 1s possible to nssign a 2

chnrwterlstic frontal arec to the crc . This is the effective drag orea, A,

of the arc, so that the drag or retwding fort 2 nzy be equated to the over311

Iorentz force:-

+J2A = Bid

where p is the density of the surrounding gas. Using th;s eyation It is

possible to wlcul?te values of A for the range of U B ad I used.

I.e.,

A = 2BId/p lJ2

(1)

(2)

9

A more detailed onolysis of A requires estimates of the effective arc

width transverse to the direction of motion. High speed photographs taken

(i) along the arc axis, (ii) along the direction of motion and (iii) transverse

to both the arc axis and direction of motion indicated2 that the luminous

cross-sectional shape oP the arc was more elliptical than circular with the

major axis transverse to the direction of motion. This suggestion is in 6,718 accordance with spectroscopic and other recent work . The present results

do not permit estimates of the arc width to be made, and it is only possible

to calCulatc effective drag areas and extend the range slready given in Ref.2.

The results for arcs moving in still air on brass electrodes, including

those for single end connections given in Ref.2, are shown in Fig.19, where A

is plotted against U for different arc currents. It can be seen that the

electrodes with single end connections show practically no variation with arc

current and there is a steady increase of A from 25 m/s to 580 z/s. On the

other hand, the results for electrodes with double end connections show that A

increases with arc current, and for the highest current of 1000 amp A also

increases with arc velocity.

The only difference between the above two sets of results is that those

with little or no variation of A with I include calculated corrections to

B for the self-magnetic field, while t!lose displaying a variation of A with I

are for conditions where the self-field is eliminated. It may be that the

corrections for the self-field are insufficiently cccurnte. The calculations2

for the self-field are approximate because (i) it is assumed that the current

is uniform&$ distributed in the electrodes, whereas in fact the distribution

where the arc joins the electrodes is probably asymmetric, ‘and (ii) only the

self-field at the gap centre has been calculated nnd no account tcken of its

variation across the gap and along the electrodes. These variations cre

indicated dicgrarunaticnlly in Fig.20.

It should be noted, however, that the calculetid sc:f-field is in most

cases small compared with the applied field (6 x -6 10 I Wb/m* for d = 1.27 cm)

so that a small error in the value at the gcp centre is negligible. Hence, it

would seem thct the variation across the arc gap and any signlficent change in

the field close to the electrodes due to the asymmetric current distribution

must account for the diffcrcnce in the results. A comparison of arc shapes

recorded photographically has shown that on the average they were simllnr, but

for electrodes with single end connections (self-field present) a tendency to

produce curved arcs was noticed. This difference was small but could possibly

be accounted for by a variation of the magnetic field across the arc gap.

IO

4.2 Relations between U, B, I and d

For comparison tnth previous results* we now obtain relations of tne form:- m

and

U = KI "I

B"*

A= 21 (I-2n,) (1-2x2)

P K2 B

(3) *

(4)

where K 1s an empirical constant and equation (4) is obtained by comparing

quatlons (I) and (3). The values of K, n, and xi2 have been obtained for a

fixed electrode =qacing by plotting loglo U ng:ainst loglo B at constant values

of I, and loglO U against loglo I at constant values of B, and the results ore

shown in Table 1, together with those given in Rcf.2 for brass.

The variations of xc velocity with electrode qcc~g or ?rc lcagth has

been discussed in the previous reports in this scrics, and It is seen, (Fig.21)

that they are corp?rzble for dlffercnt electrode confrguratlons and materlols,

with the exccptlon of clumir.ium. We mny regresect the varxtion of arc vGloclty

with electrode spacing, d, by relations of the form given by equction (3) end

Table I, where K is now not constant but depends on d. Fig.22 shows the

voriatlon of K rvlth d, and no simple rel,?tlon between t'nem could be determined.

The differences between results with sxgle or double end current

connections Irld the scme electrode mater181 (brass) could possibly be accounted

for by differences in the magnetic fl-lld cs dlscusscd in the j?rzvious section.

4.3 Annlysls using slmilnrlty re!.Ftions

It was shown in Ref.2 that cn ar.olysis of results on L-ovxng arcs using

suClcrity ppremeters given by Lord' for mxform arc colucxx~, or,d which evolved

the arc column voltage gradlent, depended markedly on the valaes used for the

gradlent. The estimated voltage gradients wi-rc affcctL-d by tbc p?rt of the

curve, through a plot of arc voltage against +zlactrode gap, used to estlnatc

the slope. In eddition, it is possible to cstimct2 3 voltcge gradient by

subtrccting cssumed values for the crc electrode fall voltaces and dividing by

the crc length or electrode gap. This yields quite dlfferent values for 'Gas

gradient end i( different relrtlon between the slClnrity pornmeters. Thus, in

view of the uncertainty involved nnd since for short arcs it is doubtful whether

a uniform column theory is applicable, the analysis used 1n Ref.2 is not dalt

with here.

An effective method of collating c wide range of data on the motion of

arcs (see later) is to use the parameters given by Dautov and Zhukov', derived

by non-dimensional analysis. These par,ameters do not include the arc voltage,

but include the electrode spacing so that it rn,2y be possible to use them for

arcs displaying a variation with electrode gap or with non-uniform columns.

The derivation used by Dautov and Zhukov makes no suppositions about the form

and dimensions of the arc cross-section or of the current density distribution.

The conditions of similarity have been described' for arcs in the srme gases

and between the same electrodes, by the following non-dimensional parameters:-

U d p'/~, I', PO I/d B> B2/po P and L/d

where p, is the permeability of free space, P is the ambient pressure, L is the electrode thickness end the other symbol s have meanings as already difmed. By plotting experimcntcl data from Zclesski 2nd Kukekov 10 for arcs moving along

straight elrtrodes in the form of tine dimensionless arc velocity,

U d p'/p, I', against ~~ I/B d, Dautov and Zhukw shoJ that U d/I 1s a function of I/B d at constant ambient temperature and pressure. The collapse of data also indicated that the variation of U d/I with B2/po P was smell. The present

results for brass electrodes are plotted in the ssme form in Fig.23, together

with those from other published data (Refs.2 and 10 to l8), assuning a value for

p at 300%. It is seen that results for straight or rail electrodes and for

arcs rotating in annular gaps using brass or copper electrodes at atmospheric

pressure are approximately collated over a wide range of parameters. This

trend, although approximate, is remarkable when one considers the wide range of

conditions involved, i.e., 3 amp to 20 K amp, I.3 m/s to 900 m/s and crc gaps

from 0.1 to 10 cm. The plotted results may be opproximctely represented by:

U d $/j~, IS = 1.4(l.ro I/B d)-Om6

with a scatter about a mean line of approximately -L50$.

In order to obtain a more accurate equation for the dimensronless arc

velocity we must take into account its variation with B2/po P and L/d. Now, since the experimental results are at atmaspherlc pressure, P is constent but

e change In B causes a change in B2/po P. ALSO, since the results are for different electrode spacings a change in d causes a change in L/d, and since the results shown in F1g.23 are from different sources using electrodes of

different dimensions a change in L also causes a change in L/d. From Table I,

12

we olreody know that U depends on a single function of I end B fcr etch electrode

spccing, and t&t this function is different for cxh electrode configururatlon.

Hence, for each configuration we vrould expect th?t, at constant P, and L/d,the

dimensionless crc velocity depends only on I, B end d. In terms of the

dimensionless parameters we may write:

U d +/pi I = $(B2/p, P) (11, I/B d)! (6)

for each electrode confiyrntion rnd constant L/d, where @ is some continuous

function.

This is demonstrated in Figs.24(c) ad (b) where the dimensionless crc

velocity is plotted agcinst p, I/B d for different values of Busing the results

in the present series of reports for strright brass electrodes. It shoald be

notlced thnt the results in Flg.2b(b), for double end connections, are for

electrode spacings from 1.27 cm to 3.8 cr. Thu.s, the dependence of U d/I on

L/d when d is changed, is mall for the range of gaps consldcred, and now, by

plotting the velocity parameter ngaicst B2 at fued values of PO I/B d,

(Flgs.25(n) end (b)), the form of equation (6) mcy bc deterriincd. By r?easuring

the siopes 2nd extropolat~ng the lines in Figs.24 and 25 the function Q is

determuied for each set of results rt ctnospheric pressure, i.e.,

for the brass electrodes with single end connections rnd,

U d &, I = P;‘(B) -0.18 &, I,B d)-o.68 +a.02

(3)

forthe brass electrodes with double end coIxxctions. The dlffcrence between

the two sets of results can onVJ be cccounted for by differences in the m,?gnetic

fields resulting from variations ln the self-field for the cnse with single end

connactlons, as discussed in section 4.1.

4.4 Effects of Imposed air flows

The effects of an imposed air flow clang the electrodes ore dlffcrent

depending on the direction of the flow rcktive to the dzectlon of the LOrentz

force.

When the flow acts against the Lorentz force the observed arc velocity

relative to the electrodes for constant current and magnetic field is reduced

by approximagely the same amount as the r:easured flow velocrty at the gap

centre (Fig.l6(a)) for -24 m/s and -43 m/s. Thus, for the Lorentz force and

air flow in opposition the arc velocity relative to the oir is ?pproximateljr

equal to the velocity in still air. This is confirmed by the fact that the

tot31 arc voltage, and hence input power at constcnt current, is unaffected by

opposing air flows for constcnt magnetic field or velocity relctlve to the clr,

Flg.li'(a), indicnt:ng no chcnge in the convection loss.

On the other hand, the additional effect of a flow niding the Lorentz force varies both with the applied magnetic field and the imposed elr velocity,

Figs.l6(c) and (b) for +46 m/s and +g6 m/s. e.g.

(i) ct fields between 0.06 Wb/n' 2nd 0.09 We/m 2

and cn imposed air

flow of +46 m/s the arc vcloclty is increcsed by approximately the same amount as the air flow, but at 0.01 Wb/m2 the increase 1s only 28 m/e

(ii) at fields between 0.01 hTb/r? and 0.1 Wb/m2 and an imposed flow of

+$ m/s the arc velocity IS increased by only about 60 m/s, whsrecs at 1.0 Mb/m2 I

the increase is 80 m/s.

Thus, for the Tnrrntz force and air flow in the same direction, the arc velocity

relative to the cir 1s app~ux5mntel.y cquel to the velocity in still air only for

the low air velocity of 46 m/s end comparatively high arc velocities of 110 m/s

to 120 m/s. This is confirmed by the fact that the trc voltages, znd hence

input power St constant current, are the same for on air flow of +46 m/s ud

magnetic fields 0.06 Wb/m2 to 0.09 Mb/m2 as in still air, Fig.lT(a). This

indicates no change in the convection loss. However, for the range of field up

to (i) about 0.06 Wb/m2 with +46 m/s flow, Fig.l6(a),and(ii)l.O Wb/m2 with +g6 m/s

flow, Fig.Ifi(b), the crc velocity relative to the air IS clw~.ys less thzn the

velocity in still CD for the s?me rnnge of fields. The arc voltages 3t constant

current over these r<anges nre greater than for In arc in stiil air, Fig.lqa) and (b'

indicating an mcranse in the power losses or a change in the arc properties, by

the action of the air flow against the arc wake.

5 CONCLUSIONS

This report has shown thot there is a difference between results obtained

for arcs moving along straight, parallel electrodes in a transverse mcgnetic

field when either,

(1) the self-field due to the arrr currcnt in the electrodes is eluninated or.

14

(ii) an estimate of this self-field is added to the externally appllcd field.

However, this difference is probably due to the cpproxigote cclculation of the

self-field; in the absence of time-resolved r;agnetlc field measurements con- current with the WC motion it bps not been possible to resolve the difference.

The effective drrg area for the arc has been calculated and it has been

shown that this area incrcnses rrith i‘rc v?1ocity.

Using the results from (i) the present series of reports for brass

electrodes 2nd (ii) nine other pap-pcrs publlshtd by different authors for

brass or copper electrodes, it has been shown that a n&hod, epsploying

similarity parnnaetcrs given by Dzutov md Zhukov, for collating data on Crcs

moving in transverse ma.gnetic fields 1s cff?ctivc for a wide rcnge of arc

velocity, currcnt, field and electrode spacing. In addrtion, the results for

straight brass electrodes In the present series of papers hcvi. been analysed 111

Inore detail.

The effects of low velocity irr?posed air flows relative to the electrodes

hsve been described but only lmited inferences con be mode from this aspect Of

the wark. The siqle notion that the incrcasc or decrmse in crc velocity,

dcqending on the flow direction, 1s equz.1 to the flow velocity iS Only

approxjlnately true when the air stream acts cgainst the Lorcntz force. The

cddition?1 effect of a g?s stream aiding thz Lorcntz force vcries with the

magnetic field or arc velocity in still air, and this vnrlation also depends on

the inposed cir velocity.

High speed photographic observations at \D5 frmcs per second hcve verified

that most of the sudden chcngcs in the arc shape recorded prevXusly2 wers due to

lnsufficicnt framing speed of the camera.

17

A

B

cD cl

I

K

L

Y' "2 P

u

v

P

PO

effective drag area

magnetic field

drag coefficient

electrode g=p width

menn ire current

constant of proportionality

electrode thickness indicies in simple power relations

pressure

menn arc velocity relative to electrodes

minimum nrc voltage

density of cir magnetic permeability of free space (= h X 10-7)

Estimated experimental errors

A +I39

B n$

d 20.05 cm

I *lo%

U +sYJ U

vTR 5% *lop

Ud/I ?I275

I/M c12$

B2/P 27% (const. P)

bd (NliP)

(4

@t/m2 ) b/s 1 (volt) &3h3) (Henry/m)

18

No. Author

7 V.W, Adams

2 V.W. Adems

3

10

G. Yu Dautov

M.F. Zhukov

W.R. MacDonald

C.R. Kirk

J.R. Bsgshaw

A.A. Wells

E.E. Soehngen

W.T. Lord

W.T. Lord

A.M. Zalesski

G.A. Kukekov

Title, etc.

The influence of gas streams and magnetic fields on

electric discharges.

Part 1 Arcs at atmospheric pressure in annular gaps,

June 1963

Part 2 The shape of en arc rotating round en annular

gap, September 1963

A.R.C. C.P. 743 1963

As above.

Part 3 Arcs in transverse magnetic fields at atmospheric

pressure

A.R.C. C.P. 959 December 1965

Coordinated results of research on electric arcs.

Zh. Prikl. Mech i. Tekh. Fiz. 2,97(1965)

(Translated by W.G. Plumtree, Department of Aeronautics,

Daperial College of Science end Technology, London)

A Hall probe fluxmeter.

R.A.E. Technical Note IR 35, (1964)

High speed cinema photography in circuit breaker

research and development. Reyrolle review, 182, (July 1563) (Reyrolle reprint

No.108)

l&published Mintech Report.

J-al of Engineering Power.

hens A.S.M.E. 279 (1966)

Unpublished Mintech Report.

Effects of a radiative heat sink on erc voltage current

characteristics.

Proc. of Specialists Meeting on arc heaters, Belgium

(1964), AGARD ograph 84, Part II, 6~

Characteristics of a sectionally cooled erc.

Proc. Leningrad Institute No.], 410 (1954)

REFERENCES (Contd)

Title, etc.

Properties of a d.c. arc in a magnetic field.

Trans. A.I.E.E. 7J, 143, (1956)

Das Verhalten des Lichtbogens im transversalen

Magnetfeld.

Arch. fiir Elektrotech, 65, 94, (1957)

Wsnderungsvorg'&ge van kursen Lichtbogen hoher

Stroms&ke im eigenerregten Magnetfeld.

Elektrizitatswirtshaft 5J, 196, (1958)

Lichtbogenwanderung an Rundenstgben.

E.T.Z. A, & 132, (1960)

Uber den Einflussdes Laufschienenfeldes auf die

Ausbildung und Bewegung von Lichtbogen Fusspunkten.

Arch. 3% Elektrotech 5 188, (I$O)

Arc travel in the annular clearances of spark gaps in

magnetic rotating erc arrestors.

Elektrichestvo, No.& 56(1963} Electric Tech. U.S.S.R. z 440 (1963)

The movement of high-current arcs in transverse

external and self-magnetic fields in air at

atmospheric pressure.

A.R.C. C.P. 777, I&Q 1964

The magnetic deflection of short arcs rotating between

annular electrodes above and below atmospheric pressure.

A.R.C. C.P. 843, October 1964

J&

11

12

13

14 I. G&enc

15 D. Hesse

16 A. Bronfman

17

18

Author

L.P. Winsor

T.H. Lee

A. Eidinger

W. Reider

L. MZler

H.C. Spink A.E. Guile

E.D. Blix

A.E. Guile

STABILISING

RESISTANCE

I I I ?. I CR 0 (VELOCITY MEAS ) ^I_ A L.“.U I \

TRIGGER

I I ANODE I 9 0 e

1

f CATUODE I @APPLIED UNIFORM MAGNETIC FIELD

CR 0 OPTICAL PROBES

Q SINGLE END CURRENT CONNECTIONS

SHUNT

STABILISING RESISTANCE

CR0 VELOCITY MEAS

OPTI CAL PROBES

b DOUBLE END CURRENT CONNECTIONS

FIG. la 8 b CONNECTIONS TO ELECTRODES

------- ------

BRASS ELECTRODES 0.33 cm THICK

DUCT INSIOE DIMENSIONS :- l2r6cm

FIG. 2 PAIR OF BRASS ELECTRODES IN DUCT. (ELECTRODES INSULATED ON UPPER AND LOWER

SURFACES WITH THIN ALUMINIUM OXIDE COATING)

ORIGINAL DUCT\

ELECTROOES - SPLIT s.u.ap

BE a SUPPORTS

(DUCT SIZE: 4~BCm OIA )

END VIEW

AN0 REDUCING OLtCT

(CARDBOARD)

(DUCT SIZE: 4 Ucm DIA.)

ELECTRODES PERSPEX WIN DOW 1N TlJEiC ;! 0

PLAN VIEW WITH COVER REMOVED

FIG.3 ARRANGEMENTS FOR REDUCING DUCT SIZE ( @ AND a )

With low-pass filter

Without low-pass filter

Records for arcs on pair of brass electrodes.

upper traces: arc voltage (50 V/em1

Lower traces: arc current (100 ,A/cm) and optical probe signaLs superimposed (probe signals reduced by 4, dB attenuator)

B' 0.023 k'b/m' GAP = 1.27 cm Time base = 2 msee/cm

(eero time at R.H.S.)

%

0 In +t

(s/w) n

FROM REFZ

_____ -se--- --

0---= -.------ ___.____ ---- ----- - _ - =----~------- === .

0.1 0.2 O-3 04 0.5 0.b 0.7 08 09 IO

13 (Wb/m’)

INDUCED VOLTAGE (OPPOSES ARC

VOLTAGE ) dr 2-54CW d: 1.27 Cm

FIG. 6 MINIMUM ARC VOLTAGE AGAINST MAGNETIC FIELD FOR BRASS ELECTRODES WITH SINGLE END CONNECTIONS

140

120

100 x --

E I I I- 81 I xl -4-x IX I +-ti I

40 --

20

0 200 400 600 800 1000 1200 1400 lb00 1800 2000 2200

I (amp)

0 002 0 04 006 OC

B (Wb/m*)

FIG. 7 ARC VELOCITY AGAINST ARC CURRENT AND MAGNETIC FIELD FOR BRASS ELECTRODES WITH SINGLE END CONNECTIONS (d a3.5 cm)

I20

LOO

80

';;

z y 60,

40

0'

20 ?-

0

IZC

* IOC

6C

-7 z ’ 4c

2c

C

d-0

I I I I I 1 . L--l

I I I I + /

s-f Mf3OL 0 0 I X +I* 6 (Wb[,,,Z)lOOl55 (0 031 1 lo

d t 0706 0.1075

I.27 cm 0.05

Is--

)-

,-

,_

)-

)-

,,

x0/ --- -0 _ o------

_--” 0--

--o---- -----0

,-e---- -Q--.0-

0--

0 200 300 400 500 600 700 800 900 IO< -I -\ 1 (amp)

SYMBOL 0 I 0 I X + 1 . B (Wb/mz) 10.015 lo.025 1OG43510 081 10.109

d. 2.54cm

FIG.8.ARC VELOCITY AGAINST ARC CURRENT FOR DOUBLE END CONNECTIONS TO BRASS ELECTRODES

100

80

- 60 *

z

3 40

20

O+ 100 5

‘L#-- -0--i -.

-./ * ,.- -

- _--

,+- -. +

l // /- .

,P--+- +e- l-T-4

I . # -

-- ,xX----::

t 2.

I - ---0 o--

“, - z -_- J-0 __._--

--a----- - -0 0, _ __- (j-m -

a-4 e---P

I I I I I I I zoo 300 I 400 500 600 700 a00

*(amp) 900

SYMBOL 8 0 X + l 1 1 @(Wb/ mZ) 0 015 0.026 0 0505 0 0805 O-1075 ds 3.2cm

60

c & --- 3

40

SYMBOL 0 0 I X + . 8 ( Wb/m=) 0.0155 cm 0

03, ooso ds3.a

0 081 0 109

FIG.9.ARC VELOCITY AGAINST ARC CURRENT FOR DOUBLE END CONNECTIONS TO BRASS ELECTRODES

--• . --

r I I . I 1 ,---I .

I I I I I

120

IOC

SC

P

z - 6C 3

4c

5%

0

, 0 100 200 300 400 500 600 700 800 900 1000

1 C-P) SYMBOL 0 0 Y + .

B (wblm’ ) ~Oaz 1 d= 1.27cm

o.o*e o.os35 a~oal 0.108

SYMBOL 0 0 x + ’ dz 1-91cm

e (wb/ln2 ) 0.0125 0.026 0.0535 0.032 0 lo,

FIGIO ARC VELOCITY AGAINST ARC CURRENT FOR DOUBLE END CONNECTIONS TO COPPER ELECTRODES

4-t 7

,.----- T

/-

/RR

.x R . ---+-----

-+

_c +-

, /. f--

.. --- I T II I--

-- x +’

‘+1 *

:

I 1

I 300 400 500

1(amp) 60

SYMBOL 1 0 0 I x I + I . I I I -- I B (Wb/m*) 10.0125 IO.027 10 058 10 091 )O.lOe d = 254cm

v I 0 __- - vb- --_

8

A A/ -0

20 ---__ A-O- _ - - -O--n- 1i

*

I (amp)

SYMBOL 0 0 x + 0 8 (Wb/m*)lo-0125 10 024 10.0~1 d= 8*2cm 0.05~

o-108

SYMBOL 1 A 1 v c1 A v

8 (Wb/m*) IO.012 b d= 3.ec.m

.024’S 0 0525 0.0805 0 IO8

FlG.ll. ARC VELOCITY AGAINST ARC CURRENT FOR DOUBLE END CONNECTIONS TO COPPER ELECTRODES

100

80

- 60- ul

3

’ 40

c

20

0 100

1 zc

1 .

I I I I I I I I J 01 IO 300 400 500 600 700 800 900 IO

J (amp1

SYMBOL . + B(Wb/mL) 0.1055 0 081 lo z5, ‘oo;5~ d: l*27cm

5VM00L . . 0 v P dt 254cm

8 (Wb/rr?) O-109 0 082 0052 002

SYMBOL 0 o x + l -d: 32cm

8 ( Wb/m2 ) 0.014 0.026 0 OS3 0 080 0 IOS

SYMBOL V 0 * .

,9(Wb/m*) 0.0235 0.06’3 OOSl5 0 1075 .dr37cm

3

FIG.12 ARC VELOCITY AGAINST ARC CURRENT FOR DOUBLE END CONNECTIONS TO ALUMINIUM ELECTRODES

90 90

. . . . . . . . . . , , . 0 . 0 l + l + . .

. .

. . . . . .

- - 30. 30. t-•+ t-•+ + + 1 1 + + xt xt x x

! i ! i 0 0 > >

x+x x x+x x x-x-4 x-x-4 x x II II

7 7 70 70 I I

4:) 4:) 0 0 C C Q Q 00 00 00 00 00 OCO 00 OCO 0 0 0 0 / /

@ @ bO_ bO_ Q Q 0 0 0 0 00 00 0 0

0 0 100 100 200 300 200 300 400 400 500 500 600 600 700 700 800 900 800 900 loo0 loo0

SYMBOL 0 0 x + .

8 (Wb/m” d =

) 1.27c:m

0.0~35 0 031 0.05 0.070s 0.1075

140. I401 . .

-120 -120.

2 2 > > + +

+ + -ITiT 7 7 x x

100 100.

0 0

BOL I I “) .’

I,. I 0 100 200 300 400 500 600 700 8

1 (amp)

SYM80L 0 0 x t I .

8 (Wb/rn*) 0.0435 0.Oa.l d :: 2*34cm

O-015 0.025 o-109

I + +

.: w-0 x

8 0

0

I +

x

---l-- 0

J&z- 900 IOC

1601 I I I I I I I I I

0 100 200 300 400 600 600 700 800 900 IO

1 (amp)

SYMflOL 0 0 Y + 1 l

8 (Wb/m*) d: 32cm

0.016 0.026 0 0505 0 0806 0 1075

SYMBOL A I 0 n I * . B(Wb/m2 ) d=3.8cm 0 0155 0 031 0050 0.081 0.109

FIG.13 MINIMUM ARC VOLTAGES FOR DOUBLE END CONNECTIONS TO BRASS ELECTRODES

140

120

. .

*. 100

5

2 0

> a0

E .

I!

60

40 0 2 100 !O ,O 300 400 500 600 700 a00 900 too0

-I-~~-.

.

B(Wb/m*)lO 0123 IO 026~0OS3+0a2]0 IOU

aYMIc3L I A I w I m I * I .

d 91. 91 cm

d. 264cm

FIG. 14 MIMIMUM ARC VOLTAGES FOR DOUBLE END CONNECTIONS To COPPER ELECTRODES

t 100

0’ 7

7 80

60 100 200 300 400 500 600 700 600 900 1000

sYm8oL 0 x + . 8 (Wb/m*) 0.0255 0 OS3 0.051 O~lOSS

d s t*27cm

da 2-54cm

3-

3-

3-

o-

>- 0

I--

.

7, &

_--- I- -.

.A . v 1, A 7”.

. * +-- A ___.--__ - -

XA u ” ot m . 1 0 +

! x 0

- ---_-.-

v 0 00

Tq---

W

0 0 -_---

0 0

0 )

100 200 300 400 500 600 700 800 ‘5

1 (amp)

SYMBOL 0 0 x + l

B [ Wb/m2)10.014 10 026~0~063~0~08O~O 108 ds 3.2 cm

0 u A (WI,/,,,*) 0 0235 0.053 0 815

c.

FIG.15 MINIMUM ARC VOLTAGES FOR DOUBLE END CONNECTIONS TO ALUMINIUM ELECTRODES

0’

,

I.0 I I I I I I

3 0.04 005 0 06 007 008 009 c

B ( Wbimz ) cl I 3 290 t JOamp

*-- SYlvleaL

d = I.59 cm 0 l x +

AIR FLOW REL TO

DIRECTION OF ASC MOWJN (&) 0 -M -43 +& (ATGAP CENTRE)

0-O

/ / 4

/

i

--t-

-FLOW VELOCITY + 96 m/S

AT UP CENTRE

NO AIR FLOW

00 IO

1 : 600 t 100 amp

d: 1.59cm

FlG.karb ARC VELOCITY AGAINST MAGNETIC FIELD FOR ARCS WITH IMPOSED AIR FLOW

100 +

t

T d t >

60

0.02 O-04 0.06 008 0-I

B(Wb/m*)

a I = 290 t 30 amp

d = I.59 cm

/

200

I I I /I’- I A-

120

80

40

FLOW VELOCITY AT GAP CENTRE

(m/s)

+46

FLOW VELOCITY + 96 m/s

AT GAP CENTRE

0.2 0.4

0 (Vfb,m

‘1” . 0.0 IO

b I L 600 k 100 amp

d: I.59 cm

.NO AIR FLOW

FIG. 17a ab MINIMUM ARC VOLTAGES FOR ARCS WITH IMPOSED AIR FLOW

9 = 0

I = 635 amp Nos. indicate sequence of images. u, = s+s3 m/s

IJ = 60 m/s

.l t-C Avis ir str

1.6 - “E 20 1.2 -I

$200 -2400 ,3000-aloo

A la600 - 3700

oou- CM coNNscTlows

SYM.OL lrnc I (amp)

01 I I I I I I I I I I I 60 100 150 200 250 300 350 400 460 so0 630 600

UP/s) 1 I 0.5 I I.0 I.5

MACH No

FROM RWF 2

FIG. 19 A AGAINST ARC VELOCITY FOR BRASS ELECTRODES id= b27cm)

f TOTAL FIELD

___-_----------

1 __-_---------_

SELF-FIELD DUE SELF- FIELD DUE

TO TO ARC CURRCNT ARC CURRCNT

IN ELECTRODES IN ELECTRODES

L------- -

EXTERNALLY APPLIED /

FIELD

I 4 \, \

t; \ ---__

c V DISTANCE ALONG

ELECTRODES

! SELF - FIELD DUE TO ARC

I- ANODE I 00 BARC ‘4

l- CATHOCJP

i 1-f 0

AT

ARC

AXIS

- OISTANCE

l-4 ACROSS ELECTRODE

FIG.20 DIAGRAM OF VARIATION OF MAGNETIC FIELD IN ELECTRODE GAP DUE TO SELF-FIELD

150

100

h m

2 V

3

50

a

cc

t

+ ----_ -.

--._ -.

_.

2

‘-- -A

-. -.:-+ &=-

+ ---- ---

-------

ELECTRODES CURRENT SYMeoL

MATERIAL CON,=lG”RATtON CONNECTIONS

. CARBON ROTATIN ARC ’ -

MEAN ARC LEN6-W (Cm)

--._ / --

r CALCULATED FOff INVOCUTC

ARC (RSF 1)

FIG.21 ARC VELOCITY AGAINST MEAN ARC LENGTH. 6x0.047 Wb/m2 1=360amp

38

36

34

32

3c

2e

K

2t

2L

22

2(

I I

I I 0 I 2

d (-1

Y SRASS ( DOUBLE END CONNECTIONS )

0 COPPER (DOUBLE END CONNECTIONS) U==KI “I

gnz

. BRASS (SINGLE END CONNECTIONS )

FIG 22 VARIATION OF K WITH d

i

I 0 ,:

I /o

-

h

0

IO

I

la -3 -I

IO - IO-’ I IO

FlG.24a Ud/I AGAINST VBd FOR BRASS ELECTRODES WITH SINGLE END CONNECTIONS

(d= I 27cm)

IO’

10-Z

-

-

0 = 0.1075 Wb,rn -

\

IdhI . El=‘0 025 v;b/m’

IO -I

I

FIG.24b &l/I AGAINST I/Bd FOR BRASS ELECTRODES WITH DOUBLE END CONNECTIONS

(d = I.27 - 3 8cm)

Prtnted tn England fw Aer Majesty’s Statronely Offtce by the Royal Awcraft Estoblrshment, Famboroufh. Dd.12.9528. X 3.

A.FLC. C-P. No.987 537.5211 :

- 1967 538.6 f 621.3.032.2

A.R.C. C.P. No.987 nwra 1967

537.523 : 538.6 : 621.3.032.2

Adams. VA. Adams. V.W.

TN5 INFLXNCE OF GAS sTREAW.3 AND HADSSTIC FIEIBS OS ELKTRIC DISCRARCSS. lW INFLWNCF, OF SAS S'NtS*pIY AND BXNERC FISKIS 9N FUCTRIC DISCHARGES. PART 4. ARCS NOVING ALONG STRAIORT PARALUL EIXTRODES PART4. ARCS NOVING ALLNG ST%AIGRT PARA"2.L ELEC'IW303S

This report Is a continuation of the experimental work 1" air at One atmosphere on the motion of d-0. arcs along strdght open-ended electmdes already reported I" Part 3 of thls series. 'Ihe earlier m)rk with the slef-field due to the arc current I" the electxvdes I" the same direeclon es the applied magnetic Iield ftxs been extended to include 7suits with fields up to 1.0 blb/& Results are also giV-2" rOr three electrode materials with electrical co"nectlons arranged so that Che self-field is eliminated, and are shorn to differ loom those vher-e the self-field Is present and appmxirately calculated.

This report Is P co"ti""atlo" of the experlmsntal wx-k 1" air at one atmosphere on the motion of d.c. arxs along straight own-ended electrodes already rrported I" Part 3 of this series. The earlier wxk w‘th the self-field due to the am o"rrent 1" tl!a electrodes in t!,e Same dlrectio" as the applied wgnetic field has Me" extended to include results with fields up to 1.0 wbl& Res"1t.S are also SiYe" ior three electrode naterlals rith electrical connections arranged so that the self-field 1s eliminated. and are shown to differ fran those where the self-field is present end appmxlmately calculated.

oJvw) (Over-)

A.R.C. C.P. No.987 537.523 : narch 1967 535.6 :

621.3.032.2 Adams. V.W.

TEE JN%UENCE OF GAS SW ANB "AGNETIC FlWS PN ELECTRIC DISCHAROES. PART 4. ARCS PIOVING AlON STRAIGHT PARALUL EIECTRQXS

This mport 1s a contlmatlcm or the experimental work in air at one atmosphere on the motlo" of d.c. nro.s along straight open-ended eiectmdes already repxted in PWt 3 01 this ssrles, Ths earlier wrk with t,x? seli-fleld due to tl.3 arc c"rre"t I" the electrodes in Lhe SBme directlo" as the applied roeS"etiC field IBS bee" extended to lncluds rewlts with fields up to 1.0 Wb/m2. Res"lts We also RiW3" for three electrode wterlals wiCh electrice connections an-awed so that the self-field is ellmin:ad. and are shove to differ from chose where the self-rfold is present and approxi~tely calculated.

(Over-)

A method, employlw slnllarity ~rarceters given by Cautov and Ztnkov of collatlllg data on arc motion in transverse metic fields Is used to compare results for the present series of reports with other wbllshed data.. and Is shovn to be effectim for a wide We of pamraters from eleven different pprs, including the present msults.

AR: mtlon with a loiv-felocity Imposed alr flow along and relative tr, the electmdes has been Studied, and Che simple notIon that the algebraic difference betmen the arc velocltles relative to tlm electrodes In still and moving air Is equal to the Imposed sir velocity 1s found u) be only approximptely tNB when the air Stmam acts against the am motion.

A method, employing simlla~lty parameters given by Dautov and mv of collating data on PI‘C mtlon In tmnswrs~ metIc fields Is "sad t., Co- r8.wlt.s Ior the present series of repofls with other plbllsbed data. and is shovm to be errsetive rol- a wide Fangs or pol-a75ztePs fmm eleven different pawn, Including the present rsarlts.

Are "h,tlon with a lo,welocity imposed air flow along and relative to the electmdes has bee" studied, and ths slmpla notion that th, algebraic dlflerence between the arc -,elOcltles relative to the electrodes in still and moving ai? Is equI1 to tha imposed air velocity 1s found to be only appmxirW,ely true vhen the air st,-ea,,,a~t~ against the ar.2 moLlcm.

A method, employing similarity prameters given by Cautov and Z~kov of collating data on am motio" In tra"sv~r~e magnetic fields 1s used to

com~re resu1t.s for the present series 01 reports with otiur published daCa, and Is shorn to be effective for. wide mn@ 01 parpmsters from eleven dfrrerent Wpers, including the prese"t results.

Am motion with P low-vwlOCity imposed air rlow aloW and mlatlW to

the electrodes has bee,, studled,a"d the slmgle Dotlon that the algeuraic difierenca between the am relocltles relative to Cha electrodes in still and mO"iW al,- Is equal to the imposed air velocity is foluld to be only appmXlmetelY tme tin the air stream acts agalnst the *IQ rmt1aL

C.P. No. 987

o Crown Copyright 1968

Published by HER MAJESTY’S STATIONERY OFFICE

To be purchased from 49 Hiih Holborn. London w c 1 423 Oxford Street. London w.1 13~ Castle Street, FxJmburgh 2

109 St. Mary Street, Carm Brazennose Street. Manchester 2

50 Fairfax Street, Bristol 1 258-259 Broad Street, Bmmngbam I

7-11 Lmenhall Street, Belfast 2 or through any bookseller

C.P. No. 987

S.O. CODE No. 23-9017-87