by v. s. komelkov, u. v. skvortsov and s. s. tserevitinov ......v == 30 kv, (dl/dt)màx = 5.5-6x...
TRANSCRIPT
P/2302 USSR
Development of a Powerful Discharge in Deuterium
By V. S. Komelkov, U. V. Skvortsov and S. S. Tserevitinov *
The investigations carried out at the Institute ofAtomic Energy of the Academy of Sciences of theUSSR resulted in the discovery of both neutron andhard X-ray radiation occurring under certain condi-tions in a powerful pulse discharge in deuterium.1-3
In the present work, the investigations in thisfield were continued with a view to studying theseprocesses at greater currents and higher rates ofincrease of the current, by minimizing the circuitinductance and the size of the discharge chambers.Studies were made of the current distribution in thechamber, neutron radiation, electrode-metal vapormovement, and the effect of pre-ionization on theinitial stages of the process.
EXPERIMENTAL APPARATUS
The arrangement consisted of capacitors connectedin parallel with a total capacity of 130/Л. The totalinductance of the circuit leads and the capacitorswas about 15 cm. The spark gaps used for switchingwere of two types : air gaps, and gaps with a breakablehard dielectric. The construction of the latter switchis described elsewhere. Three types of chambersare used in the experiments:
The chambers of the first type had an inside diameterof 60.0 cm, a height of 17.5 cm, and porcelain walls1.5 cm thick, with a number of holes for probes. Ithad two flat aluminum electrodes. In the lower elec-trode there was a cylindrical crater 4.0 cm deep, witha diameter equal to the inside diameter of the porcelaincylinder. This electrode was provided with a window3.4 cm high, whose lower edge was at the level ofthe flat electrode surface. The total length of thewindow was 50.0 cm and it was divided into separatesections by four metal baffles.
The chambers of the second type had flat copperelectrodes, were 4.7-cm high, and were of 18.5-cminside diameter. The walls were 20.0-cm outer diameterglass cylinders, which were destroyed during eachtest.
The porcelain chamber of the third type had flatcopper electrodes and the same diameter as the second
Original language: Russian.* Institute of Atomic Energy, Academy of Sciences of the
USSR, Moscow.
type, but its height was 25.0 cm. In optical measure-ments the porcelain cylinder was replaced by a glassone of the same diameter and height. The currentand voltage were recorded by double-beam oscillo-scopes with the aid of voltage dividers and Rogovskybelts. The accuracy of the measurements was± io%-
Measurements of the field and current distributionin the discharge were carried out by the magneticprobe method evolved by Andrianov, Bazilevskayaand Prokhorov.1 The 3 x 6 mm measuring coilhad 10 to 20 turns. One end of the coil was connectedto a copper tube 4 mm in diameter and 40.0-cm long,serving both as a shield and a conductor. The secondlead of the coil was inside the tube.
Both the tube and coil were placed in a cappedporcelain tube to completely eliminate any contactwith plasma. The connecting cable, amplifier andoscilloscope were insulated from ground. The coilswere used to measure the magnetic field (accuracyi 20%), and its time derivative, which determinesthe time at which the magnetic field appears at afixed point in the discharge (accuracy ^ 0.5 ^sec).The plasma motion and the variations in the diameterof the pinch were registered by a high-speed camera(SFR-2M)t operated either as a streak camera oras a framing camera.
Kerr cells pulsed by a special circuit were also used.Explosive shutters 4 were used to cut oñ the back-ground light.
A 0.1 to 0.2-mm streak camera slit was placed atthe center of the window (chamber 1) or of the trans-lucent cylinder (chamber 2). To synchronize opticalrecordings with the current, a spark discharge with aduration of 0.2/̂ sec was recorded simultaneouslyby the streak camera and the oscilloscope.
The propagation of metal vapor was investigatedas follows : Light from the discharge chamber windowfell on the entrance slit of a monochromator (UM-2)tpassed through the entrance lens, prism, collimatorlens and outlet slit of the monochromator to thephotocathode of a photomultiplier tube (FEU-39) tand through the oscilloscope amplifier (20 Mc^bandwidth). The vapor velocity measurements were
t Ed. note: Transliterated designation.
374
DISCHARGES IN DEUTERIUM 375
lo3-
0
6
/
/
V
f1
(b)
VД
T.me^sec)
Time {fi sec)
Figure 1. Current and voltage characteristics(a) 20 kv, 0.05 mm Hg; (b) 20 kv, 0.1 mm Hg;(c) 20 kv, 0.5 mm Hg;
(d) 30 kv, 0.05 mm Hg; (e) 30 kv, 0.1 mm Hg
carried out with the À 3961 Á aluminum line. Theresolving power of the monochromator in this spectralregion was 30 A/mm.
Production of neutrons was observed with a stilbenescintillation counter using the FEU-19MÎ photo-multiplier. The discharge current was simulta-neously recorded on the oscilloscope.
To protect the multiplier and the stilbene crystalfrom hard X-ray radiation, 20-mm thick lead screenwas used. A B-type $ unit with a ^-counter placedin a paraffin block was used as a control. The timevariation of the diameter of the charge channelcurrent, voltage across the electrodes, current distri-bution, neutron radiation and metal vapor velocityalong the discharge channel were recorded in cham-ber 1.
The current, voltage across the electrodes, variationsin the pinch diameter (by streak photographs andKerr cell photographs) and neutron radiations wererecorded in chamber 2.
The discharge in chamber 3 was investigated witha framing camera and with 12.0 and 16.0-cm diameterRogovsky belts.
EXPERIMENTAL RESULTS
The experiments in chamber 1 were carried out at aworking voltage of 20 and 30 kv at pressures between0.05 and 0.5 mm Hg. The initial dl/dt and current
Ed. note: Transliterated designation.
•), a/cm2
800
1 400
U
800
400'
\ \
800.
400i I
10 20 30 10 20 30
200
10
1600'
1200'
800'
400
20 30 10
1600
1200
800
400
20 30
1200
800
400
10 20 3ÔT 10 20
-400-1 1—J -4ÜÜ -400Í
i sec) = 1.45 2.85 4.25 5.7 7.15 8.55
Figure 2. Radial distribution ; (r) of the current density; V = 30 kv, P = 0.1 mm Hg
30
376 SESSION A-6 P/2302 V. S. KOMELKOV et al.
400- 400-
n„ 0
• J400,
гП
101.
20 30" 1
_
10 20
i
1200
900
600
n 30°
300
1200'
900-
600-
300
-11-П .10 I 201 30
• H
2302.3
J
600-
300
10••H
20
0
—J
Цзо]*" л lüJ 20 301"—I 1
t {¡i sec) - 1.45 2.85 4.25 57
Figure 3. V = 20 kv, P = 0.05 mm Hg
7.15 8.55
amplitude for these voltages varied between the follow-ing limits:
V = 20 kv, {dI/dt)m3iX = 3.5 — 4 X 1011 amp/sec,Imax= 540-600 kamp
V == 30 kv, (dl/dt)màx = 5 . 5 - 6 x 1011 amp/sec,/max = 750-900 kamp.
The oscillograms of the current and voltage aregiven in Fig. 1 and show the typical characteristicsfor discharges which contract due to magnetic forcesproduced by the current distribution.
The results of measurements with the aid of inte-grating magnetic probes are shown in Figs. 2, 3, 4, 5
in the form of the radial distribution of the currentdensity Q(T) in the chamber at different times. Theoscillographic measurements were interpreted byassuming that cylindrical symmetry of the magneticfield and current are maintained during the discharge.About 50% of the discharge current remained in alayer 5-cm thick near the wall after the current startedin the central region of the chamber. Typically,for pressures of 0.1 and 0.05 mm Hg, it is observedthat negative currents are generated near the chamberwalls. They appear before the main current reachesthe center and before shock waves are reflected fromthe axis of the chamber.
j ( r ) , a / cm 2
8001
400
800
400
800
400
1.
800
400
213 Ш*
1200
600
1200
600
1200
600
2302.4
10 20 3ÍTr (cm)
t (^sec) = 1.45
Tu 2a зо"
2.85 3.8 4.25 5.7
Figure 4. V = 30 kv, P = 0.05 mm Hg
-300
10 20
7.15300;
8.55
DISCHARGES IN DEUTERIUM 377
j
240
120
0
r),a/cm2 36°
10 20 30
240
120
l>
гЛ10 20 зс
560
240
120 Г
—
480
360
240
—120
1 [..•10 20 30
-
-
20
360
240
120
LF
240
120
, 1 •10 201 ЗОГ
t (¿i sec) =1.45 2.85 4.25 5.7
-150
7.15 - 3 0° 8 .55
Figure 5. V = 20 kv, P = 0.5 mm Hg
2302
Figure 6. Oscillograms of the discharge current and neutronpulse; (a) 30 kv, 0.1 mm Hg; (b) 30 kv ,0.05 mm Hg
The oscillograms of the discharge current and theneutron pulse are shown in Fig. 6. For P = 0.5 mmHg and V = 30 kv, the current appears in the centerof the chamber at 8 /¿sec, at 9 //sec it reaches itsmaximum and remains at a high value until 12 /¿sec.In the external circuit, the current changes its signat 8 //sec.
Streak camera photographs of the discharge showthat the optical picture of the motion is in reasonableagreement with the magnetic probe data (Fig. 7).If one takes into consideration the fact that the magne-tic probe records the time at which the internal boun-dary of the moving layer passes a fixed point and thatthe streak photograph shows its external boundary,the agreement between the two measurements iswithin the error of the probe measurements.
With V = 30 kv, the maximum rate of contractionis 107 cm/sec (P = 0.05 mm Hg) and 9 X 106 cm/sec(P = 0.1 mm Hg). In these experiments the inten-sity of neutron radiation was notably smaller than ithad been when chambers 50 to 90 cm high were usedin which the neutron yield sometimes reached 109
neutrons/pulse.1 When the distance between theelectrodes is reduced to 5.0 cm with V = 40 kv andP == 0.1 mm Hg, a small number of neutrons can stillbe observed within the sensitivity limit of the scintilla-tion counter, which is 5 X 105 neutrons.
When the distance between electrodes is 17.4 cm,a neutron yield of 106 in one discharge was recordedwith only 30 kv initial voltage, in the pressure range0.05 to 0.4 mm Hg. The neutrons are produced at3.8 /¿sec and 4.3 /¿sec after the current flow beginsfor P = 0.05 mm Hg and for P = 0.1 mm Hg, res-pectively. In all cases, the neutron radiation startsbefore the maximum of the current in the center of the
378 SESSION A-6 P/2302 V. S. KOMELKOV et ai.
Table 1(cm)
2302.7 Time {¡i sec)
Figure 7. Variation of the pinch radius with time from opticaland probe measurements
1. 20 kv, 0.1 mm Hg; optical. 2. 30 kv, 0.1 mm Hg; optical.3. 20 kv, 0.1 mm Hg; probe. 4. 30 kv, 1.0 mm Hg; probe.5. 20 kv, 0.05 mm Hg; optical. 6. 20 kv, 0.05 mm Hg; probe
chamber is reached. It can also be seen from thestreak photographs (Fig. 8) that the neutrons areproduced before the contracted pinch flies apart.
Spectroscopic measurements of the radiation fromneutral aluminum atoms showed that the metalvapor in such discharges moves with velocities up to106 cm/sec, and the velocity is independent of elec-trode polarity. The results of the measurements aregiven in Table 1, where T = time at which metalvapor appears at the electrodes, vay = averagevelocity of metal vapor (0-1.5 cm from electrode).
It was noted that the aluminum vapor appearedat a time which corresponded to a definite currentdensity in the discharge, 400 a/cm2. It was alsofound that the average velocities of the neutralaluminum atoms increased with increasing voltage.Typical oscillograms of the current and vapor lumi-nosity in a discharge are shown in Fig. 9.
A discharge in small chambers 18.5 cm in diameterand 4.7 cm in height was investigated at a voltage of40 kv, initial dl/dt of 1.4 X 1012 amp/sec and maximumcurrents of 1.4-1.8 X 106 amp. The initial pres-sure in the chamber varied between 0.1 and 10 mm Hg.
P (mm Hg)
0 03 . . .0 1 . . . .0 60 030 10.6
T, fisec
. . . 3.0
. . . 3.3
. . . 3.0
. . . 3.8
. . . 4.1
. . . 3.3
Vav cm/sec.
6 x 105
7.5 x 105
5.5 x 105
1.2 x 106
8 x 105
6 x 105
V, kev. Polarity
30 —30 —30 —30 +30 +30 +
For initial pressures of 10 and 1 mm Hg, manyindependent, relatively bright channels were observedin the constricting layer. Constriction of the gasoccurs at an increasing speed with decreasing pressure.For a pressure of 10 mm Hg the maximum speed is5.4 X 106 cm/sec. Full constriction is reached nearthe time of maximum current, 1.4 X 106 amp.
By the time constriction is complete, the gasvelocity reaches 8 X 106 cm/sec; with a current of1.6 X 106 amp, pinch diameter of 3.0 cm and initialpressure of 1 mm Hg.
Figure 8. Streak camera photograph of a discharge withV = 30 kv and P = 0.1 mm Hg
Figure 9. Oscillograms of the current and luminosity withV = 20 kv, P = 0.1 mm Hg
(a) near the electrode, (b) 1.5 cm from the electrode
Streak photographs, Kerr cell photographs, andoscillograms of the discharge current and voltages areshown in Fig. 10.
The maximum compression ratio of the plasmawas estimated to be 140 from the square of the ratioof the initial to final optical pinch diameter. Theexpansion is accompanied by stratification of the gas,part of which moves back to the center of the chamber,where the luminous pinch and some current stillremain. The stratification takes place three timeswith a time interval of 0.2 to 1.0 /¿sec. The Kerr cell
DISCHARGES IN DEUTERIUM 379
Figure 10. (a) Streak camera photographs, (b) current and voltage osciliograms, (c, d) Kerr cell pictures. P = 1 mm Hg
photographs show that the stratification is symme-trical. As the pressure fell to 0.1 mm Hg the rateof contraction rose to l . l -1.3xlO7 cm. Threesuccessive constrictions were observed, each originat-ing from the wall. In the first constriction theminimum pinch diameter was 1.5 cm and the totalcurrent was 1.1 x 106 amp. In the second and thirdconstrictions the currents were 1.2 x 106 and 1.3X 106 amp, respectively. Neutrons were not observedwith a lower limit of sensitivity of 106 neutrons/pulse.
Experiments with a discharge chamber 18.5 cmin diameter and 25.0 cm in height were aimed largelyat finding the efíect of pre-ionization on the subsequentdevelopment of the discharge. Pre-ionization wasaccomplished by means of a discharge from an auxi-liary 22 [xî condenser bank rated at 20 kv.
After the pinch formed by the auxiliary currentdetached itself from the wall and traveled somedistance towards the center, the main 20 kv bankwas discharged. The maximum dl/dt and currentamplitude for the first and second discharges were1.4 X 1010 amp/sec, 5.5 X 104 amp, and 1011 amp/sec,4.5 X 105 amp, respectively. The experiments werecarried out at an initial pressure of 1 mm Hg. Typicalosciliograms which show the total current and thecurrent in the Rogovsky belt, with the main and auxi-liary capacitor banks operating separately and jointly,are shown in Fig. 11. The experimental results aresummarized in Table 2, where т is the time lag betweenthe auxiliary and main discharges, t0 is the time atwhich current appears in the Rogovsky belt and Iis the total discharge current at this time.
With pre-ionization, the period during which theconducting sheath remains near the wall is cut atleast by half. The velocity of the current sheath,as obtained from oscillographic measurements of thetime at which current appears in the belts and fromoptical measurements of the motion of the luminouslayer, diners by not more than i 15%.
DISCUSSION
In discharge chambers where the space betweenthe electrodes is small, the influence of such processesas heat conduction from the constricted pinch to theelectrodes and the occupation of the chamber bymetal vapors should be taken into account. As hasbeen demonstrated by S. I. Braginsky and V. D.Shafranov, heat conduction to electrodes during thecurrent flow is insignificant for times
t< {3ll5){e*N¡c*kT)K
where I is the chamber length, T the particle tempera-ture, с the velocity of light, e the electronic charge,
Table 2Belt
diameter Energy Source t(nsec) to{f*sec) I(amp)
160
120
Auxiliary dischargeMain dischargeJoint 6.4 ±Auxiliary dischargeMain dischargeJoint 6.9 ±
0.1
0.4
5.81.40.7
10.02.71.2
±±-j-
++±
0.40.10.10.30.20.3
42190
50200
XX
XX
103
103
103
103
380 SESSION A-6 P/2302 V. S. KOMELKOV et al.
Figure 11. Oscillograms of (1) total discharge current and (2)current in a Rogovsky belt obtained with (a) auxiliary dischargeand 16 cm belt, (b) main discharge and 12 cm belt and (c) joint
discharge and 12 cm belt
k the Boltzmann constant, and N the number of parti-cles in the discharge chamber. For chambers 4.7 cmin height, at a temperature of 300 ev and an initialpressure of 0.1 mm Hg, this time will be about 3 //sec.
High-velocity metal vapors which move into thedischarge are observed in regions where the currentdensity is relatively high immediately before, or atthe time of, maximum constriction. However, thesevelocities are such that in chambers of the first type,at a pressure of 0.03 mm Hg, the influence of themetal vapor is important 8 /¿sec after the time ofmaximum constriction. In chambers of the secondtype at a pressure of 0.1 mm Hg, metal vapors appear3 //sec after the time of maximum constriction.
This process is roughly similar to that previouslydescribed. г> 7 However, the experiments describedhere have provided the following additional informa-tion:
1. With the exception of the initial time when thedischarge detaches itself from chamber walls, theluminosity and current boundaries coincide withinthe experimental accuracy. The initial discrepancyis explained by the fact that the differentiatingmagnetic probe responds to a build-up of current
inside the stationary conductor. Optical measure-ments with chambers of the second type show thatthe width of the luminous layer between the shockfront and external boundary, for pressures of 10, 1.0,and 0.1 mm Hg are 2.7, 2.2, and 1.0 cm, respect-ively. Should the plasma conductivity be determinedfrom the width of the luminous layer, which can be2 to 3 times the skin-depth, it will be 0.5-1.4 X Ю14
cgse, which corresponds to an electron temperatureof 3.5-6 ev.
The main discharge characteristics are in reasonableagreement with the hydrodynamic theory developedby Leontovitch and Osovetz5 and with cal-culations made by S. I. Braginsky, I. M. Gelfandtand R. P. Fedorenko.
2. The maximum compression ratio obtained fromthe optical measurements with chamber 2 was 140,which greatly exceeds the maximum compressionratio predicted by the hydrodynamic theory. Theactual ratio of the initial and final (maximum) den-sities may be somewhat smaller because all the parti-cles are not swept in by the contracting current sheath.
3. The current in chamber 1, with a pressure of0.05 mm Hg and a voltage of 30 kv persists in thecentral zone (14 cm in diameter) for 4 //sec whichmarkedly exceeds the constriction time obtained fromcalculations. Optical measurements carried out withchamber 2 give similar results.
4. About 50% of the total current is concentratedin the center of the chamber. Characteristically thisportion of the current grows after the maximumconstriction is achieved. In measurements obtainedwith chamber 3, about 20% of the current remainsnear the wall in a layer 0.5-cm thick. This is probablydue to the evaporation of materials with high atomicnumber from the chamber wall. Pre-ionization ofthe gas by a discharge shortens the time during whichlarge currents flow near the wall and this reduces theenergy losses to the wall.
The production of reverse currents at certain timeswill result in the formation of a closed circuit insidethe chamber. This causes the current to grow in thecenter of the chamber at the end of the first and begin-ning of the second half cycles.
As follows from theory,5 the circuit parametersdetermine the rate of contraction and the gas tempe-rature. In Fig. 12 the dependence of the maximumrate of contraction on voltage and the initial induct-
vm o x (106cm/sec)
8—
-¡
- e r - • —
* """-
2302 12
A -
dl/dt (10- 1 1 a/sec)
Figure 12. Maximum speed of contraction as a function of{d\ldt)initiai for P = 1 mm Hg
DISCHARGES IN DEUTERIUM 381
anee is shown. This data was obtained with differentset-ups in chambers with the same diameter at a pres-sure of 1 mm Hg.
According to existing theory, the thermal energy ofthe gas at the time of maximum constriction is appro-ximately twice as great as the maximum kineticenergy during the constriction process. From theseassumptions the maximum temperature in chamber 2at pressures of 10, 1.0 and 0.1 mm Hg is estimatedto be 40, 120 and 340 ev, respectively. In chambersof the first type, the gas temperature, estimated on thesame basis, is 200 ev (at P == 0.05 mm Hg, V = 30 kv).
The neutron radiation from thermonuclear processesthat would be expected at such temperatures couldnot be recorded by the available apparatus. There-fore, the radiation which was detected probably hasthe same origin as described earlier.1' 6 Its occurrencein pinches 5 cm long testifies to the fact that the acce-lerating processes which presumably take place occurin regions of this length or less. A decrease in theradiation intensity with decreasing pinch lengthprobably testifies to the fact that the radiation sourcesare distributed along the entire pinch column, whichis in accordance with data obtained previously.6
2.
REFERENCES
L. A. Artsimovich, A. M. Andrianov, O. A. Bazilevskaya,Y. G. Prokhorov and N. V. Filippov, Atomnaya Energ.,No. 3, 76 (1956).
L. A. Artsimovich, A. M. Andrianov, E. I. Dobrokhotov,S. Y. Lukyanov, I. M. Podgorny, V. I. Sinitsyn andN. V. Filippov, Atomnaya Energ., No. 3, 84 (1956).
3. V. S. Komelkov and G. N. Aretov, Doklady Akad. Nauk.S.S.R., 110, No. 4, 559 (1956).
4. V. S. Komelkov and B. P. Surnin, Pribory i tekhnikaeksperimenta, No. 1, 78 (1956).
5. M. A. Leontovich and S. M. Osovets, Atomnaya Energ.,No. 3, 81 (1956).
6. O. Anderson, W. Baker, S. Colegate, I. Ise, I. Pyle andP. Pyle, Proceedings of the Third International Con-ference on Ionization Phenomena.
7. S. Y. Lukyanov and V. I. Sinitsyn, Atomnaya Energ.,No. 3, 88 (1956).