)VIESI_) RETURNING MATERIALS: Place in book drop to LJBRARJES remove this checkout from am your record. FINES Win be charged if book is returned after the date stamped below. NANOSECOND ELECTRIC PULSE DISCHARGE DAMAGE IN ALUMINIUM SINGLE CRYSTALS AT VARIOUS OVERPRESSURES OF PURE NITROGEN By Narendra B. Dahotre A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Metallurgy, Mechanics and Materials Science 1983 ABSTRACT NANOSECOND ELECTRIC PULSE DISCHARGE DAMAGE IN ALUMINIUM SINGLE CRYSTAL AT VARIOUS OVERPRESSURES OF PURE NITROGEN By Narendra B. Dahotre Pulse discharges of 36 nanoseconds duration and 3KV have been produced on single crystal surfaces of pure aluminium. These discharge experiments have been carried at pressures of l, 1.5 and 2 atmospheres of pure nitrogen gas. The nature of the microcrater and the associated plastic deformation modes have been analysed by using SEM and X-ray diffraction. A periodic ripple pattern around the crater is observed. At the same time single crystal of aluminium shows crystallographic slip around the crater and this slip pattern is predominant for a electrical discharge at pressures greater than one atmosphere. A linear relation between the average crater diameter and the nitrogen pressure is observed and a simple thermodynamic model is proposed to explain this correlation. Immediately after a spark damage, only wavy deformation bands and slip traces are observed. However, upon storage of these damaged material in ordinary atmosphere, an enhanced oxidation, adjacent to the damaged sites, and cracking of these oxide layers were observed. ACKNOWLEDGMENTS I would like to take this opportunity to thank my advisor Dr. Kalinath Mukherjee but for whose constant guidance and inspiration this endeavour would not have ended in success. He is also an excellent teacher and a friend. I would also like to thank my friendsand my colleagues, especially Mr. S. Sircar and Mr. S. Shekhar, for their timely help and support when I needed them most. TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES I. II. III. IV. VI. INTRODUCTION LITURATURE SURVEY 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 The Elecric Arc Distribution of Potential Along the Arc High Pressure Arcs and Low Pressure Arcs Cathode Phenomena Cathode Current Density Spark Breakdown The Sparking Potential Quantitative Theory of Cathode Processes in an Arc Features of the Erosion Spots EXPERIMENTAL PROCEDURE 3.1 3.2 3.3 Experimental Set-up Specimen Preperation (a) Mechanical Polishing (b) Electropolishing Optical Microscopy and Scanning Electron Microscopy RESULTS AND DISCUSSION POSSIBLE APPLICATION OF ELECTRIC PULSE TECHNIQUE CONCLUSIONS iii Page iv vi 10 12 15 17 24 28 3O 3O 34 35 35 36 37 61 63 iv Page APPENDIX : Nanosecond Electric Pulse Utilized for Present Investigation 65 LIST OF REFERENCES 66 Figure 1. Schematic characterastic for a gaseous discharge 2. Voltage vs current characterastic of the electrode discharge with tip or edge 3. Distribution of potential along the arc 4. Three temperatures in positive columns 5. Radial distribution of temperatures in positive column 6. A crater as a result of microexplosion 7. Voltage characteristics for positive point-plane gap in compressed air and nitrogen 8. Breakdown voltage curves in nitrogen for platinum and sodium cathodes 9. Relative breakdown strengths of various gases 10. Experimental set-up used for present investigation 11. The discharge chamber 12. Schematic diagram showing experimental set-up for electric pulse discharge 13. Optical micrograph (at 835K Magnification) of a typical microcrater produced-under 2 atmospheric pressure of N2 14. Crater diameter as a function of nitrogen atmosphere pressure 15. The sparking potential vs pd relationship 16. The discharge damage produced in a material 17. The one dimensional representation of a discharge damage and LIST OF FIGURES a sphere of which it is a part Page 11 11 16 21 22 25 31 32 33 38' 39 42 44 44 vi Figure 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Optical micrograph (at 835K Magnification) of a crater (produced under 2 atmospheric pressure of N2) showing ripple pattern around it Optical micrograph (at 415x Magnification) showing a slip pattern around the crater (produced under 2 atmospheric pressure of N2) Optical micrograph (415x Magnification) showing a slip pattern arround the Crater (produced under 1.5 atmospheric pressure of N2) Optical micrograph (at 415x Magnification) of a crater produced under 1 atmospheric pressure of N2 Optical micrograph (at 415x Magnification) of a microcrater (produced under 2 atmospheric pressure of N2) showing cracks around it Optical micrograph (at 415K Magnification) of a microcrater (produced under 2 atmospheric pressure of N2) showing cracks around it SEMInicrograph (at 3000K Magnification) of a.microcrater showing radial cracks in oxide layer around it SEM micrograph of a crater produced under 2 atmospheric pressure of N2 SEM micrgraph showing foldings in the molten material around it Optical micrograph (at 155x Magnification) of a crater Page 49 50 51 52 54 55 56 58 59 vii Figure Page (a) Immediately after the damage has been produced (b) After the around 30 days 60 28. A nanosecond electric pule discharge utilized for present investigation 65 LIST OF TABLES Table Page 1. Properties of arcs at One atmospheric pressure 8 B pd 2. Constants (A,B) of the equation, VS = 19 A pd 1n -———-——-—— ln(1/Y) 3. Electrons emitted from metals per impacting positive ion (Y) 20 4. Minimum parking constants (V8 and pd) 23 5. Electropolishing data for pure aluminium 36 6. The average crater diameter vs. nitrogen atmosphere pressure 40 7. Sparking potential (VS) vs. pd (mm—Hg X cm) data 41 viii I. INTRODUCTION The intraction of high-energy electric pulse discharge, ele— ctron beam and laser light with metal target materials has become an important area of investigation. Essentially all such interactions involve the absorption of a portion of the electric, electron or light energy and its conversion to heat, with or without a change in phase of the surface material. As the materials surface recoils against the large input of energy in a very short time, a high pressure stress wave (or shock wave) is generated. A compressive pressure pulse raises the boiling point and the melting point of the underlying material, which becomes superheated as more heat is conducted into the interior (1-3). In effect the dynamic thermal expansion which follows the rapid absorption of radiant energy, at the surface of an elastic solid can propagate a shock wave into the interior (4). Most of these applications involve rapid melting and soli- dification of a thin surface layer. The area of damage formed and the amount of melting occuring, under the impact of a single and multiple spark discharges, are dependent on various factors such as energy input in the spark gap, current, melting and boiling points of the electrodes, composition of electrodes, thermal and electrical conductivities of electrodes, width of the spark gap, sur- face condition of the electrodes etc. Under impact of single spark discharge of nanosecond duration, single or multiple microcraters are obtained. The periodic surface ripples present around the craters are spread in narrow region (5-7). Thus the appearance of such characteristic ripples may be treated as the signature of melt threshold. In the present experiments aluminium single crystals were spark damaged. The study of the damaged surface of the electrode was done by using optical microscopy and scanning electron microscopy (Hitachi 415A). II. LITURATURE SURVEY 2.1 The Electric Arc. An arc is a discharge of electricity, between electrodes in a gas or vapor, which has a voltage drop at the cathode of the order of the minimum ionizing or minimum exciting potential of the gas or vapor. There is often considerable uncertainty in the practical application of this particular definition. However, the phenomena that cathode causes the emission of electrons, which constitutes the largest part by far of the arc current, and therefore, maintains the entire discharge. This arc current is more than the threshold which lies between 0.1 and l ampere and the upper limit is not specified and it is very large. But the voltage drop is in the range between the few volts and few tens of volts. The arc voltage is definitely lower than the values encountered in low current gas discharges such as dark discharges, glow discharges etc. Figure (1) represents the general nature of the V—I chara— cteristics of gaseous discharges, for two parallel plate electrodes. These electrodes avoid the formation of corona or stremer discharges. In the case of the wire and co-axial cylinder, the dark discharge is succ— eded by a discharge in which the voltage increases with increased current as shown in Figure (2). This figure illustrates the formation of a corona discharge on a thin rod ending in a hemispherical surface. The shape of a discharge depends upon the gas purity, the gas pressure, electrode surface conditions, polarity of the electrode and the impressed voltage. 3OHU n""l‘|'l"""'- NEON coaufimcmpe . coaumofifiawuaoe. zoos :oHoumHHoo“ cowuquDmm . aowuomHHou mHaEHm“ ‘A ~—4>- 104 . Amuao>v > 0.1 I(amperes) 10'6 10" 10"2 -8 10’12 10'10 10 : Schematic characteristic for a gaseous discharge. Figure l //\ 20000 15000 t Corona or brush discharge -—> ‘\ 10000 ‘ , Breakdown \ ------ t’ \ D . \ Gas Air q— ark discharge \p = 760 mm Hg \ \ \ \ Arc \ #4 Voltage 5000 J O J 1 '10 '8 10"6 10'4 10’2 1 10 Current (log scale) Figure 2 Voltage vs current characteristic of the electrode discharge with tip or edge. 2.2 Distribution of Potential Along the Arc. The burning voltage of an arc discharge is divided into three distinct regions, as shown schematically in Figure (3). A considerable amount of the voltage is used up in a relatively short distance in front of the cathode in what is known as the cathode drop region. The cathode drop region is one of very high positive ion space charge. In general, the cathode fall in potential predominates. It is of the order of magni- tude of the ionization potential of gas in question and the vapor of the element of which the cathode is made. The voltage used up in the positive column, which is a region of uniform longitudinal voltage gradient whose magnitude depends on the gas, pressure and arc current, depends on the length of the column. Approximate values of cathode and anode drops, cathode and anode temperatures are given in Table (l). The form of arc varies depending on the pressure and current. In general, the positive column is very luminous. Close to the electrodes it contracts and forms extremely luminous spots or patches. The electrodes reach an incandescent state and rapidly evaporate. The gas between the electrodes is raised to a very high temperature when the pressure is fai- rly high, it then becomes the seat of convection currents. The cross section of the column can vary from one point to another and with time. 2.3 High Pressure Arcs and Low Pressure Arcs. Depending upon the pressure conditions of gas, the class "arc discharges" is divided into two groups as "high pressure" arcs and "low pressure" arcs. ""1 I I I I I I I l I I I I I I I I I I I I Positive column J Figure 3 : Distribution of potential along the arc. Table (1): Properties of arcs at one atmospheric pressure. Electrode Gas I (A) TC(OK) Ta(°K) VC(V) Va(V) C-C air 1-10 3500 4200 9—11 11-12 C-C N2 4-10 3500 4000 - - Cu-Cu air,N2 N 5 <2200 2400 8-9 2-6 Fe-Fe air m 5 2400 2600 8-12 2—6 Ni-Ni air W 5 2400 2400 - - w-w air m 5 3000 4200 - - Al-Al air m 5 3400 3400 - — Al—Al N2 m 5 m 2500 m 2500 - - Zn-Zn air m 5 3000 3000 - - Zn-Zn N m 5 small small - - As the positive ions and neutral molecules of the gas have more or less the same mass, they exchange energy readily. As there are a large number of molecules, moleculedmolecule and iondmolecule collisions occur frequently. A state of equilibrium is readily established, the tem- peratures Tg and T of the gas and ions merging. In fact, T is slightly 1 1 higher than Tg; because between two collisions, the ions are accelerated by the electric field. The electrons absorb most energy from the electric field. They impart some of this to the ions and molecules by elastic and inelastic collisions. The former heat the gas while the latter excite and ionize it. The collisional frequency depends chiefly on the tempera- ture of the electrons (Te). The kinetics of this collision is expressed by the following equation 2 = _3RT 2nv (l) 1 --§- mc R O nv where, K is Boltzman's constant = T is the temperature of the relevent kind particles. m is mass of the particles. c is root mean square velocity irrespective of direction. n is concentration of particles. v is volume of one mole of gas. At low pressures, the density of neutral particles diminishes. To maintain a given discharge current, the number of ionizing collisions must be increased. Te must, therefore, increase. But then the fraction of energy lost by the electrons in elastic collisions decreases consi- 10 derably and thus Tg decreases. Conversly when the pressure increases, Te decreases and Tg= T increases until Te= T = Tg' This can be seen in 1 Figure (4). The temperature distribution as a function of the distance i from the axis is given in Figure (5) for various pressures. It can be seen that at medium and high pressures T depends strongly on the distance from the axis. 2.4 Cathode Phenomena. The current density at the cathode of an arc is very much greater than that of the glow discharge. The cathode current density is practically independent of the arc current. Materials such as carbon and tungsten that have very high boiling points are characterized by low current densities at the cathode which is incandescent. The cathode spot can be moved only slowly over the surface of these refractory materials. The low boiling point matals have very high cathode current densities. The cathode spot on a low melting point metals is usually in continuous random motion over the surface. A considerable amount of material is lost from the cathode of an arc. At the high temperature of carbon and the tungsten cathodes especially under continuous operation, this loss is lar- gely by vaporization. For metals having low melting points, considerable material may be melted off the electrodes. In addition, there is a blast of particles leaving the cathode that is present even when the cathode is cooled or when the duration of the arc is too short for any appreciable heating. T(°K) II 10 L\ 10 11 I 1 4 :y, p(mm Hg) 10' Figure 4 Figure 5 102 103 104 1 atmosphere : Three temperatures in positive columns. p low (luminous discharge) 2T. 1 p medium _Ti p high ngzTi : Radial distribution of temperatures in positive column. 12 2.5 Cathode Current Density. Determinations of the cathode current density are subject to considerable uncertainty. The usual method is to measure the current and then determine the active area either by photographing the cathode spot or by observing the marks left on a polished metal surface. The rapid motion of the cathode spot on the metals and the fact that there are high temperature gradients over the surface of the very hot cathodes such as carbon make it difficult at best to determine more than the order of magnitude of the current density. Thus both photographs of the cathode spot and measurements of the marks left by the spot would probably indi- cate too large an area and therefore, too low a current density. The most obvious mechanism for the production of the large electron currents at the arc cathode is thermionic emission. The high temperature is produced by the energy released by the impacting positive ions which may come from the positive column but probably are produced in the cathode drop region. The current density of positive ions nece- ssary for the maintainance of a sufficiently high temperature for a large thermionic emission may be estimated by the space charge equation for positive ions. Because of the high velocity of the electrons the space charge in the cathode region will be considered as due entirely to the positive ions. Then, by the space— charge equation; 1 2e 2/2 j = (2) p 9 n mp d2 13 where, vC is the cathode drop. dc is the thickness of cathode drop region. mp is the mass of the positive ions. For an arc in nitrogen at atmospheric pressure.vc must be taken as a first approximation as the ionization potential 15.8 volts for carbon, The value of dc may be taken equal to an m.f.p. of anion in a gas at tem- perature near that of the cathode, 20000 K to 30000 K. This gives positive ion current density jp of 180 amperes/cmz, The possibility of high field strengths, at the surface of cathodes that vaporize at low temperatures, has led many investigators(8-10)- to favor the theory that necessary electrons are produced by "field" emission. It as many experiments (11,12) seem to indicate, the tempera- ture, some mechanism other than thermionic emission must be present to supply the necessary electrons. The presence of low work function impu- rities would greatly increase the current density in local regions of the cathode. Any point of increased emission will result in a local increase in the positive ion space charge and a further increase in the field strength. Such impurities are believed to be necessary for the cold-cathode tungsten arcs observed by Newman (13). The experiments of Ramberg (14) indicate that the arc cath-U - odes of C, Ca, Mg, W are thermionic in nature where as those of Cu, Hg, Ag, Au are of field emission type. Cathodes of Pt, Sn, Pb, Ni, Zn, Al, Fe, Cd seem to involve a modification of the field emission process, probably the effects of oxides are important, and also the metals of this series, having higher boiling points may combine the effects of themionic and field emission. 14 Many efforts were made to measure the cathode current density. Measurements taken from the channel diameter give current densities which would probably be a good estimate for the arc column, but due to multi- plicity of craters and the fact that more than one channel can be in existance at the same time (15), the current density in the cathode spot would be impossible to estimate from this measurements.The other method, in which for a given cathode erosion mark the area of emission site was considered as identical and the current density was found out. But the accuracy of the current density have been restricted to accurate size determination of the erosion marks. The use of SEM was made by P.E.Secker and I.A.George (16) to examine the cathode arc tracks. On the same line C.C.Sanger and Secker(l7) tried to measure the total emission area by noting which sites touch or lie across a suitable reference line drawn normal to the track axis. Co- rrelating this emission area with the corresponding instantaneous current value they calculated the required mean current density. According to Sanger et al. (17) markings left on the electr- ode after the arc had been driven magnetically, across it at a high speed are the accurately delineated areas which had supported electron emission. A given small area on the cathode acts as a electron source for a very short time until the area has so modified its surface properties that it ceases to be a good emitter. The microcraters forming a given track are thus records of the series of emission sites consecutively brought into action as the preceeding one on the track fails. Joule heating immedia- tely underneath an operating site raises the local temperature of the 15 metal above its boiling point (18). The resulting microexplosion (19) gives rise to a crater with a surrounding lip of electrode debris(Figure 6). Sanger et al. (17) assumed that the effective emission occur over a region equal to the projected area of the crater and its surrounding lip. According to Sanger et a1. possibly it would have been more justified to have taken an area equal to that of crater bottom (Figure 6) but in practice this would have proved very difficult to measure with any certainty. Basharov et a1. (20) showed that cathode spots are local, independent of experimental conditions and remain fixed. They also found that the cathode spots exhibit a complex structure that consists of a layer of chemical compounds and an eroded central part containing melted areas with microcraters and microcracks, which are visible only with the electron microsc0pe. They descarded the determination of the current density based on the areas of the microcraters located inside the melted region which give results that were too high 1011 amperes/cmz). There- fore, they assumed that the determination of the current density interms of total area of the melted region inside the light area of the trace (slightly eroded area) i.e. by the data on the microstructure of the spots give the most reliable results. Calculations based on the micro- structure of the spots yield current densities which are too low. 2.6 Spark Breakdown As the elecric field in a spark gap increases, the current increases and at some point there is a sudden transition from the 16 WNW +++++++++ ‘3) aim/77777, 1/‘7‘7‘27/‘VII‘I36/‘7‘23r‘227kI777 77‘" H” ’ ’7” A \ I I ’ I \ \ \ \ \ \I ’ r’\’z’ / l: \ \ ‘ \\\“—<“' // i’ I “’\§\ Regions of .- intence \ joule heating / (b) 7777777 Oxidised micro-explosion debris Figure 6 : A crater as a result of microexplosion. 17 Townsend, or "dark" discharge to one of the several forms of self-sustai- ning discharge. This transition, or spark, consists in a sudden change in the current of the gap. The type of discharge depends upon the shape of the electrodes, the gap, the pressure, and the nature of the external circuit. For plane electrodes the result is a spark that initiates an arc discharge. For sharply curved electrodes there may be corona or a brush discharge. In an engineering sence, breakdown is considered to occur only when the entire gap is bridged. This usually means the establish- ment of an arc discharge. The breakdown may be initiated either by maintaining the voltage constant and varying the gap or by changing the voltage for fixed gap. A selfsustained discharge is established only by when the conditions of field, pressure and gap are such that each electron leaving the cathode establishes secondary processes whereby it is replaced by a new electron leaving the cathode. 2.7 The Sparking Potential. An analytical expression for the sparking potential of a uniform field if it is assumed that the spark is determined by secondary emission of electrons at the cathode is as following (21), B pd V - (3) s - A pd 1n 1n (l/Y) 18 where, A and B are the constants depending upon the gas(values of these are given in Table 2). Y is the electrons emitted from metals per impacting positive ion (probable values of Y for different cathodes and diffe- rent gases are presented in Table 3). p is pressure of the gas. d is the electrode gap. Thus,the sparking voltage is a function of pd alone. It is important to remember that this does not necessarily mean a linear function, eventhough it is linear over some regions. Actually, the gas density 6, rather than the pressure p should be used in Paschen's law to account for the effect of temperature at co- nstant pressure on the m.f.p. in the gas. The number of collisions made by an electron in crossing the gap is proportional to the quantity 6d. Thus,the number of ions produced and also value of VS must depend upon 6d, a (Townsend constant ) and y. Foord (22) has shown the positive point discharge behavior in pure nitrogen according to Figure (7). The breakdown voltage characte- ristic in nitrogen for platinum and sodium cathodes as measured by Ehrenkranz (23) is given in Figure (8). It is seen that the spark breakdown voltage has a minimum value at a critical value of pd. At low pressures, below the pd for minimum Vs’ the spark discharge will take place over the longer of two possible paths, the longer path requires the lower voltage for break- down. Table (4) presents values of the minimum sparking voltage, the corresponding pd. 19 B pd Table (2): Constants of the equation, Gas A B Air 14.60 365 A 13.60 235 C02 20.00 466 H2 5.00 130 H20 12.90 289 He 2.80 34 N 9.34 375 A pd 1n(l/Y) 20 Table (3): Electrons emitted from metals per impacting positive ion (7). Metal A 2 He Air NZ Ne Al 0.12 .095 0.021 0.035 0.10 0.053 Ba 0.14 - 0.100 — 0.14 - C - .014 - — - - Cu 0.058 .050 - 0.025 0.066 - Fe 0.058 .061 0.015 0.020 0.059 0.022 Hg - .008 0.020 - — - K 0.22 .22 0.17 0.077 0.12 0.22 Mg 0.077 .125 0.031 0.038 0.089 0.11 Ni 0.058 .053 0.019 0.036 0.077 0.023 Pt 0.058 .020 0.010 0.017 0.059 0.023 W - - — - - 0.045 21 140 I I I 1 J r I I l T l I I l I T ’ Spark breakdown J 9120 _ ***** Corona onset _ 5 U*-—--— 0.094 cm. diam. m ““'1cm gap . m f_:£:_. $5100 *' __ - :4 _ O p’ «I > g 80 - - o 'o .x - - w o 35 60 fi- 'o C I. m U 3, 40 b c o k m 8 . h 20 o D k 0 L l L l 1 9 4 i i l 1 4 L L L 1 0 20 40 60 80 100 120 140 160 Gas pressure (lbs/inz) Figure 7 : Voltage characterastics for positive point-plane gap in compressed air and nitrogen. 16000 14000 12000 10000 8000 6000 4000 2000 Figure 8 22 Nitrogen 1 § i0 15 20 I 1 1 a 1 A; 40 80 120 160 200 240 280 320 360 380 pd (mm Hg X cm) : Breakdown voltage curves in nitrogen for platinum and sodium cathodes. 23 Table (4): Minimum sparking constants. Gas VS minimum pd (volts) (mm.Hg X cm.) Air 327 0.567 A 137 0.9 H2 273 1.15 He 156 4.0 C02 420 0.51 N2 251 0.67 N20 418 0.5 02 450 0.7 Na(vapor) 335 0.04 802 457 0.33 H S 414 0.6 24 The minimum sparking potential V8m and its corresponding pdm which may be determined by differntiating above equation (3) and setting the derivative equal to zero are given by, _ B l Vsm — 2.718 T In T (4) _ 2.718 1 and pdm - -—-Zr—-ln —?—- (5) It is important to note that a spark breakdown cannot occur at voltages lower than Vsm' Ofcourse, solid insulation interposed between the electrodes, even though its surfaces coincide at every point with the field lines of the gap, will result in a lower breakdown voltage (24). This lowering of VS may be due to the collection of surface changes that distort the field and also to the effects of layers of gas, moisture and dust. Surface flush over of insulation is markedly affected by humidity. An idea of the relative breakdown strengths of various gases may be obtained from Figure (9). It is seen from this figure that nitrogen has highest breakdown strength and hydrogen the lowest while air is almost as good as nitrogen. 2.8 Quantitative Theory of Cathode Processes in an Arc. Many papers devoted to the theory of cathode processes in arcs that treat thermal field emission have been published in recent years. Analysis of cathode processes based on the equations of thermal field emission presupposes simultaneous consideration of the thermal 25 T l l T V I 18 a, 16 _ - 750 mm—Hg 14 — - 5 12 - - I O.) O". m :1 1o - — O > E 8 8 .. '55 a (U G) H m if 6 - 250_ m mm. A H g 4 ” - 100 mm. 2 - /H2 Hg- 0 I l I I I l 0 0.1 0.2 0.3 0.4 0.5 0.6 0. Figure 9 : Gap in cm. Relative breakdown strengths of various gases. 26 processes, accompanied by the fusion and evaporation of the cathode material. However, several simplifications are used in calculating the thermal processes. Thus Lee and Greenwood (25) have determined the thermal conditions at the surface of the cathode spot without considering phase transition and the energy of resistive heating of the material in cathode spot. While Goloveike (26) takes into account the Joule-Lena heat, but uses an arbitrary depth of the fused zone for calculations. Rakhovskii and Beilis (27) have derived a system of equations that describe the emi- ssion processes in the most general form, however, as in (25), the resi- stive heating energy is not considered. Kulyapin (28) did the analysis of this phenomena by taking into account both the emission and thermal phenomena. The fused metal zone was considered for calculating the resistive heating energy. This zone was determined by the magnitude and duration of the thermal flux resulting from the ion bombardment of the cathode. The equation of the energy bala- nce suggested by Kulyapin (28) is Cli+qv=qo+qe+qr (6) where, qi is the thermal flux due to the ion bombardment of the cathode. qv is the thermal flux of resistive heating of the fused metal zone. qo is the thermal flux spent on evaporation, fusion of the cathode material, and heat transfer. 27 qe is the thermal flux carried away by electron emission. qr is the thermal flux of radiation losses. Another concept about cathode erosion was suggested by A.I. Struchkov (29). According to Struchkov, the cathode erosion is governed primarily by thermal conductivity and the resistivity of the material which was neglected by most of the researchers. According to Kantsel' et a1. (30) the destruction of the cathode in vacuum occurs mainly as a result of vaporization. In this case the erosion rate for W, Cu, Ni is written in the form Y = A . 10KI (7) where, K is the constant which does not depend on the material of the electrode and that it characterizes the mechanism of ero- sion in a vacuum arc discharge. Y is the rate of erosion (in ug/sec.). I is the current (in KAmp.). A is the constant whose value is a function of electrode mat- erial preperties such as the specific heat of vaporization or of fusion, temperature of vaporization and temperature of fusion. By other theory proposed by Lebedev et al. (31) and I.G. Nekrashevish (32), the energy of input from the arc column is very high and that cathode erosion is a consequence of thermal microexplosion which 28 removes a certain mass. The highly compressed low temperature plasma which forms as a result of an explosion has low electrical conductivity, so that further input of energy to the crater formed on the cathode is sto- pped and the erosion at the given surface point ceases. A cathode spot initiation mechanism based on the model of explosive electron emission was proposed recently by Mesyats (33). If the field at the cathode reaches 106 V/cm. then field emission from the ind- ividual microprojections becomes substantial. In this case the cathode layer is unstable to fluctuations of the field emission current. The instability leads to an increase of the field emission current and of the ion currents per microprojection, with the result that the microproject- ion explodes and the cathode spot forms. The necessary condition for the development of such an instability is the presence of a strong field at the cathode. 2.9 Features of the Erosion Spots. The erosion spots are produced regardless of the polarity of the voltage applied to the electrode (34). At lower voltages spots are observed only if the tip is the cathode. If the electrodes are having pointed tips, then the tip radius increases from pulse to pulse. None— theless, the microexplosions are initiated. With larger tips, no melting of cathode or anode takes place, and microcraters are observed on the tip of surfaces. The character of the erosion is different on the cathode and on the anode. The cathode surface damage takes the form of bunched molten spheres which are some times scattered over the entire surface, and at 29 other times densely packed. On the anode, however, the traces of the erosion in the form of microcraters are distributed uniformly over the entire surface. The study of the cathode track was carried out by Basharov et al. (20) and they outlined a few prominent features of it. This study was done with an electron microscope which showed that the eroded central area consists of separate pits. There are also microcracks and microcraters inside the spot, whose outline is defined by the position of the outside craters. The number of melted regions increases towards the center of the pit and there is the central area where the melted regions are in contact. The material ejected from the crater forms a small ridge around the crater. It is some times possible to observe that the outside edge of the ridge has a ragged rim. The area around the central area is a layer of chemical compounds. The characteristic features of this layer if its round shape and the absence of any preferential elongation in the direction of motion of the discharge over the surface of the electrode. Thus the cathode track consists essentially of three areas 1. an area covered with a layer of chemical compounds. 2. a slightly eroded area with separate melted spots. 3. completely eroded central area. III. EXPERIMENTAL PROCEDURE 3.1 Experimental Set-up. The experimental set-up used was as shown in Figure (10). Electrical sparks of 36 nanoseconds duration were obtained in a airtight chamber by the discharge of a charged cable. For this purpose a "Nano- second Pulser" (Microwave Associates Inc.) was used. The anode was a tungsten needle with fine tip of 40p diameter and the single crystal alu- minium specimens, to be studied for surface damage, acting as cathodes. A cylindrical copper box (3.5" diameter, 3.5" height) was developed into a discharge chamber (Figure 11). The front side of this box was converted into a window to introduce the sample into the discharge chamber. During discharge this window was covered with a transparent lucite plate which was gasketed against the chamber wall to make it airtight. The discharge chamber was provided with the gas inlet and gas outlet valves and also the pressure gauge calibrated in psi. This discharge chamber was moun- ted on a metallurgical microscope stage whose optical column was conve- rted into the anode. The anode and cathode had a common ground through a thick aluminium cable. The schematic diagram for experimental set—up for pulse discharge is shown in Figure (12). The voltage applied between the electrodes was 3KV and the spark gap of 0.2 mm.was maintained between the electrodes. A very precise adjustment of the spark gap could be obtained by adjusting the fine focu- ssing knob of the microscope stage. This study was conducted at various overpressures (l atmos., 1.5 atmos., 2 atmos.) of pure nitrogen gas. In 30 31 .GOHumeuwm>:H ucmmmun How vow: moluom Hmucosfluoaxm may " 0H muswam 32 Figure 11 : The discharge chamber. 33 .mwumnumfiv amasm ofiuuomao now aslumm Hmucoswuoaxm mafiaosm Emuumwv ofiumEmnom " NH shaman I— mauomc coumw::H IIHI I, « m>Hm> uMHuso Hoasmcu wow :38 as 1!; owsmw ouommopm.llvn w manmo wmumzumflo Hmmfina ncoommlocmc Cu lfl/l/lkV 34 order to do this, initially the air in the chamber was driven out by cir- culating the nitrogen gas in it for few minutes by keeping both valves Open. This ensured that the partial pressure of oxygen in the chamber was reduced. Then by closing the outlet valve the nitrogen pressure in the chamber was raised to required values. The nitrogen atmosphere in the spark chamber avoids the oxidation of the melted material during the spark damage. In order to have a better electrical contact between the specimen and the specimen platform,a silver paint was applied to the bottom of the specimen. As this experimental set—up involves a metallurgical micro— scope stage, the spark gap setting was adjusted with the help of the coa— rse and fine focusing knobs. Arc energies could be varied by changing either or both the discharge duration and discharge current. For a given cable system, the discharge duration could be changed by varying the length of the charging cable, and the discharge current could be varied by changing the electrode gap setting; the gap determines the breakdown voltage which in turn determines the arc current. The breakdown voltage also depends upon the atmospheric pressure under which discharge takes place. In the present experiment a 509 cable was used and electrode separation was set at 0.2 mm. Hence in this experiment the breakdown vol— tage was varying according to variation in atmospheric pressure in discharge chamber. 3.2 Specimen Preperation. Aluminium single crystal specimens of purity 99.999 Z were 35 subjected to spark damage after they were prepared for microscopic exa- mination. As specimens were single crystals, the metallographic pre- peration was done according to following procedure. 3.2(a) Mechanical Polishing. Small pieces were cut from a large aluminium single crystal. The single crystal was cut on a spark cutting machine to get strainfree sample. Being single crystal, the specimens were subjected to the mecha- nical polishing directly on the stationary fine emery belts. This poli— shing was done with light force thereby avoiding inducement of stresses and surface recrystallization ultimately. The grades of the emery belts were 240, 320, 400, and 600 respectively. The fine polishing of the specimens was obtained by using rotating disks covered with linen cloths of grades 5 microns, 0.5 micron respectively. Alumina powder of particle size 0.06 micron was used succe- ssively to get mirrorlike surface. This mechanical polishing is followed by electropolishing. 3.2(b) Electropolishing. Eventhough specimens are polished on fine emery belts and rotating disks, the surface is not scratch free. In order to have smooth mirrorlike and scratchfree surface, the specimens were subjected to ele- ctropolishing. For electropolishing of aluminium, controlled conditions involving, optimum composition and temperature of electrolyte, current 36 density, voltage and time as recommended by Kehl (35) were used. The data containing composition of electrolyte and electropolishing conditions is given in the following table 4. Table 4. Electropolishing Data for Pure Aluminium. Composition of electrolyte Current density Voltage Time (amp./sq.dm.) (volts) (minutes) Methyl a1cohol(absolute) 2 Parts 31-93 4-7 20 Nitric acid (conc.) 1 part Remarks: Stainless steel cathode, ordinary temperature, distance between cathode and anode is 1/2 inch, stirring of electrolyte is not necessary. Immediately after electropolishing, the specimens were rinsed in distilled water, followed by washing in methyl alcohol. In order to avoid subsequent formation of a thin oxide layer on the surface, the specimens were stored in acitonitrile at room temperature. 3.3 Optical Microscopy and Scanning Electron Microscopy. In the metallographic study, an Optical microscope and a scanning electron microscope are most important tools. In the present investigation some photomicrographs were taken using an optical microscope and some are scanning electron micrographs using a Hitachi 415A model. IV. RESULTS AND DISCUSSION In the present experiment, the arc discharge was fixed atIMB nanoseconds and the electrode separation was 0.2 mm. The specimen ser- ved as a cathode and anode was a pointed tungsten needle. The discharge was made in overpressures (1 atmos., 1.5 atmos., 2 atmos.) of pure nitrogen gas. The surface damage produced by discharge with this setting was exa- mined by using a scanning electron microscope and an optical microscope. A series of photomicrographs were taken at different magnifications. Figure (13) shows a typical microcrater produced on an aluminium single crystal by 3KV arc in 2 atmospheres of nitrogen gas pre- ssure. As described earlier by many researchers (20,36,37), the micro- crater is well defined. The microcrater is more or less hemispherical in shape with eroded central area. There are also microcraters inside this central area and whose outline is defined by the position of the outward craters. The average diameter of the crater was measured to be 76 microns. The photomicrograph also shows the material ejected from the crater for- ming a small ridge around. The crater is surrounded by the melted thin surface layer. In the vicinity of the main crater some microcraters of smaller dimensions are also seen. The average crater diameter vs. nitrogen atmosphere pressure data is presented in table 5. The crater diameter as a function of nitro- gen atmosphere is shown in a Figure (14). Because of the design limita— tion of the specimen chamber, pressures more than two atmospheres could not be used. Within this experimental range, there exists a linear 37 38 Figure 13 : Optical micragraph (at 835K Magnification) of a typical microcrater produced under 2 atmospheric pressure of N2. C .muammmua wpmcomosum cmuouuwc mo sewuucsu mm umumEmuv umumuo " ca wuawwm Amuucamoeuwv ouammoum ma o.“ m; 04 9o 0 39 _ a a _ . o (su0131m) Janamsra 183913 Hmuauwuomch -I..II HmucmE«uonxm mm 40 relationship between the average crater diameter and nitrogen atmosphere pressure for a constant spark gap and voltage as shown in Figure (14). Table 5. The Average Crater Diameter vs. Nitrogen Atmosphere Pressure. Average crater diameter (microns) 34 t 4 43 i 5 56 i 5 Nitrogen pressure 1 1.5 2 (atmospheres) The relationship between the average crater diameter and nitro- gen atmosphere can be explained by using the following model. The sparking voltage is a function of gas pressure multiplied by the spark gap distance as discussed earlier. An anlytical expression for the sparking potential is then given by : S (3) ln 1n (1/Y) where, V8 is sparking potential, A and B are constants depending upon the gas (38) (for nitrogen A = 9.34, B = 375.01), Y is electrons emitted from metals per impacting positive ion (39) (for nitrogen Y = 0.10). With the help of the above expression, the sparking potential for 0.2 mm. 41 spark gap at different nitrogen pressures was calculated and these values are presented in table 6. Table 6. Sparking Potential (VS) vs. pd (mm-ng cm) Data.- Sparking potential(VS) 1382.59 1888.19 2367.19 (volts) pd mm-ng cms. 15.2 22.8 30.4 The values of minimum sparking voltage (Vsm) and corres— ponding pdm for nitrogen atmosphere are 251 volts, 0.67 mm-ng cm. resp- ective1y(40). Using this data the sparking potential vs. pd relationship is plotted in Figure (15). Within the present experimental range this relationship shows an approximate linear behavior. The process at the cathode is considered to be based on the equations involving thermal processes, accompanied by the fusion and eva- poration of the cathod material. On this basis, a thermodynamic approach is taken and the total heat produced during this process is given by following simple thermodynamic relation. m.p. Tb.p Q=mS deT +mSdeT + mAHf (8) T TR.T. m.p. where, m is mass of the metal melted in crater. Breakdown Voltage (volts) 42 2250 2000 1750 1500 1250 1000 750 500 250 I I I I I I 0 5 10 15 20 25 30 pd (mm Hg X cm) Figure 15 : The sparking potential vs. pd relationship. 43 Cp is the specific heat at constant pressure. A Hf is the enthalpy of fusion. TR T is the specimen temperature before discharge damage. Tm p is the melting point temperature of metal. Tb p is the boiling point temperature of metal. The mass of the material is expressed interms of density and volume of the metal removed from crater. m = v p (9) where, p is the density of metal. v is the volume of metal removed from the crater. In order to calculate the volume of metal removed from a crater, a simple model is suggested. A approximate hemispherical shape of the damage can be ass- umed as the first approximation. Since the damage zone is controlled by the thermal destruction, a temperature rise greater than or atleast equal to the boiling point can be assumed. This high temperature is achieved in a skin depth L, (Figure 16), according to (41), L = (a to)1/2 (10) where, a = thermal diffusivity. 44 -------.. ----- Sample (cathode) Figure 16 : The discharge damage produced in a material / Figure 17 : The one dimensional representation of a discharge damage and a sphere of which it is a part. 45 to = pulse duration. The thermal diffusivity is given by a = _..____ (11) where, A is the thermal conductivity. For aluminium, A 0.9 i 0.3 watts/cm.oK. p 2.3 gm./cc. cp = 7.00 ca1./mole/OK. For present experiment the pulse duration was 36 nanoseconds. By using these values, the calculated value of the skin depth (L) is 1.18 x 10-4cms. The one dimensional representation of a damage and a sphere of which it is a part is done in Figure (17), where R is the radius of a sphere and r is the radius of damage. The volume of damage which is eq— uivalent to the volume of sphere segment with height L (VL) is given by, (R-L) 2 3 2 VL =3—1TR-S‘Irydx (12) 0 The equation of a circle is given by, x2 + y = R (13) 46 Using equation (13) and eliminating y2 from (12), (R-L) V = Z—- n R3 - S n (Rz- x2) dx (14) L 3 0 In the figure (17), R2 = (R - L)2+ r2 (15) By using equation (15) and eliminating R2 from equation (14),it is simplified to, _ 1 3 1 2 VL---6L1r+-——2Lr1r (16) As the volume of metal removed from the crater (v) is equal to volume of the crater (VL), hence equation (9) transforms to, m = p ( —%-L3n +-—%—~L rzn ) (17) Where, r is the radius of crater. Using equation (17), the equation (8) for total heat involved in discharge is modified as follows, 47 r. T T 3 2 m.p. b.p. Q= (L"+l‘-§-—l—) CdT+ CdT-I-AH (18) p 6 2 4 p P f TR.T. Tm.p. k During discharge it is assumed that total electrical energy is converted to heat energy because of which crater is produced. Under this assumption the equation (18) is written as n -\ m.p. Tb.p. L3H ern VI=O(6 +——2-——) (19) T T \. R.T. m.p. J where, V VS = sparking potential. H II current . The current in the circuit was measured to be 3 amperes and the breaking potentials already calculated. Hence by substituting corresponding values in equation (19), the crater diameters at different atmospheric pressures were calculated. These values are very close to the experimental obser- ved values. Calculated values of the crater diameter are plotted against the nitrogen atmosphere pressure in Figure (14). The relation between the theoretical values of crater diameter and the nitrogen atmosphere pressure also shows a linear behavior. 48 Figure (18) shows the periodic ripple pattern around the microcrater. Several investigators (42-47) have reported the similar kind of patterns induced by high energy pulse laser. Recently a similar peri— odic structure in laser treated aluminium has also been reported (48). Some of these authors have suggested that this pattern is produced by a interference between the incident and the reflected laser pulse (42-44,47,48). But in present experiment, there is no such possibility of an interference and hence surface rippling can only be related to the shock deformation phenomena. Also according to T.H.Kim et al.(49) the spacing between the ripple markings decays with the radial distance from the laser induced crater and this spacing varies from metal to metal and also it is not re- lated to the wave length of the laser used by them. Two effects could account for the amplitude decay. One is the dissipation of elastic energy in plastic work as the wave front moves out and the other is that because of the higher temperature near the edge of the crater the yield stress of the metal is low and thus a given amount of distortional elastic energy can produce a greater plastic yielding. Slip deformation produced by the shock wave has been studied by various investigators (44,47,50-52) for laser induced damage. On the same line, the crystallographic slip pattern around the crater, in alumi- nium single crystal, damaged by nanosecond electrical pulse discharge is observed. Figures (19-21 ) show symmetric six fold slip pattern on the (111) surface of an aluminium single crystal. X-ray diffraction analysis shows that this slip pattern is on {111} planes which is the normal slip plane in a fcc metal. These slip lines are spread all around the crater 49 Figure 18 : Optical micrograph (at 835x Magnification) of a crater (produced under 2 atmospheric pressure of N2) showing ripple pattern around it. 50 Figure 19 : Optical micrograph (at 415K Magnification) Showing a slip pattern around the Crater (produced under 2 atmospheric pressure of N2). 51 Figure 20 : Optical micrograph 415x Magnification) showing a slip pattern around the crater (produced under 1.5 atmospheric pressure of N2). 52 Figure 21 : Optical micrograph (at 415x Magnification) of a crater produced under 1 atmospheric pressure of N2. 53 over the region of 60 microns from the edge of the crater. As seen in Figures (19—21 ), the slip pattern is dominant at higher nitrogen pressure. The possible explanation for the above phenomena is based on the absorption of electrical energy. The sparking potential (VS) is a function of the pressure and the spark gap. In the present experiment, spark gap is kept constant at 0.2 mm. And according to Figure (15), with- in the present experimental pressure range sparking potential increases as the pressure increases. Hence the electrical discharge energy (VSI) which is converted to heat energy (to produce a crater in the material) increases with the pressure. This explains that an increased amount of electrical pulse discharge energy is coupled on the aluminium single crystal surfaces thereby giving a predominant or distinct slip pattern at higher atmospheric pressure. This increased coupling is also evident from the increased crater size with increasing nitrogen pressure. Figures (22,23) show the optical micrographs of a crater in an aluminium single crystal at 2 atmospheric pressure of nitrogen gas. The history of these samples are as follows. They were produced in a pure nitrogen atmosphere and the samples were exposed to air for various leng- ths of time. These mirographs show radial as well as some circumferential cracks produced from and around the crater after some length of time up to few days. SEM micrograph of another such crater with these types of cracks is shown in Figure (24). Figure (24) is the magnified SEM micro- graph of the crater in Figure (23). From this figure it is vivid that the radial cracks are originating from the edge of the crater and they are deep in the surrounding area. These cracks are produced in thin layer of oxide on the surface of aluminium single crystal, and they are not in the 54 Figure 22 : Optical micrograph (at 415x Magnification) of a microcrater (produced under 2 atmospheric pressure of N2) showing cracks around it. 55 Figure 23 : Optical micrograph (at 415K Magnification) of a microcrater (produced under 2 atmospheric pressure of N2) showing cracks around it. 56 .Figure 24 : SEM micrograph (at 3000X Magnification) of a microcrater showing radial cracks in Oxide layer around it. 57 base material. Particularly Figure (24) shows the oxide layer and crack in it. The different facets Of the crater are clearly Observed in the SEM micrograph in Figures (25,26). The surface foldings are created in the molten zone around the edge Of crater. These foldings are the result Of surface tension. When the aluminium single crystal sample is electropolished and immediately subjected tO electrical pulse discharge, it showed the deformation bands and slip traces (Figure27a). But when these damaged samples are stored in air, after around 30 days both radial and circum- ferential cracks are Observed and also slip line density around the crater is decreased as shown in Figure (27b). In due course Of time the thin oxide layer is built up on the surface and as the aluminium oxide is brittle it accomodates stresses by producing cracks in it. It is interesting to note that a preferential oxidation occ- urs in and around the crater. Such an enhanced oxidation may imply the presence of excess lattice defects around the damage spots. These defects then act as catalytic centers for an enhanced chemical reaction. A simi- lar effect Of discoloration around laser damaged spots in pure Cu has also been reported (53). 58 Figure 25 : SEM micrograph of a crater produced under 2 atmospheric pressure of N2. 59 Figure 26 : SEM micrograph showing foldings in the molten material around a crater. Figure 27 60 (a) (b) : Optical micrograph (at 155x Magnification) of a crater (3) Immediately after the damage has been produced (b) After around 30 days. V. POSSIBLE APPLICATION OF ELECTRIC PULSE TECHNIQUE. Abrasion and wear resistant surfaces are Of enormous impor- tance in various fields Of engineering design. Various types of surface hardening techniques are in existance such as case hardening, induction hardening, plasma jet spray, ion implantation, laser treatment etc. Some of these techniques are economically not attractive while some produce a very shallow surface hardening. The present experimental set-up can be used for practical application of surface hardening. The metal which is to be surface treated is initially prepared by usual metallographic technique. The prepared surface is coated with a thin lacquer solution and then metallic or non— metallic particles to be bonded with the base metal are spread on the surface. The surface at ordinary temperature is then subjected tO extre- mely short duration electrical pulse discharge through a very fine tungsten tip(tip diameter approximately 0.45 mm.) with a 3KV discharge potential and a discharge of approximately 36 nanoseconds. Because of the sublimation Of the binder during the discharge and simultaneous melting Of the under- lying metal, the metallic/non-metallic particles spread on the surface are fused on the surface Of base metal. The sample is mounted on X-Y scanner and the pulsing is conducted in repeating mode at a pulse repeat- ition rate rate Of 10 pulses/second which allows an extended surface to be bonded with particles. The uniqueness Of this technique is that, because Of very rapid heating and cooling there is sufficient time for the metallic or 61 62 non-metallic particles to dissolve into the base metal. By this method all types of refractory particles such as silicon carbide, titanium car- bide, tungsten carbide, ceramic particles, boron nitrides, alumina parti- cles and similar known hard substances metallic and non—metallic are bonded with any type Of base metal. This technique covers the surface treatment of ferrous and non-ferrous alloys indicating binary and multi- phase alloys. It also covers pulse discharges at voltages higher or low- er than 3KV and covers shorter or longer pulse duration than 36 nano- seconds used in present demonstration. For surface treatment by this method, apart from X-Y movement, rotational scanning method can be used. This technique not only covers the use Of ordinary atmosphere but also covers the use of an atmosphere such as nitrogen or inert gas during pulsing. Above all, this technique Of surface treatment is inexpensive compared with conventional methods and laser treatment. VI. CONCLUSIONS Based on the information Obtained, the following conclusions can be reached : 1. There exists an approximate linear relationship between the average crater diameter and the nitrogen pressure at which the discharge is carried out. This relationship can be explained by simple themo— dynamic model based on the conversion of electrical energy into heat energy. A periodic ripple pattern around the discharge craters is Observed. A similar type Of periodic structure, Observed in pulse laser treated materials, are considered to be due to an interference between the incident and the reflected laser pulse. But in present experiment there is no such possibility Of an interference and hence the Observed surface rippling can only be related to the shock deformation. Single crystal metallic surfaces show a crystallographic slip pattern around a crater. This slip pattern is more pronounced for electrical discharges at pressures greater than 1 atmosphere. A symmetric slip pattern corresponding to the crystallographic orientation Of the sample is Observed. Immediately after a spark damage, wavy deformation bands and slip traces are present adjacent to the damaged material. Upon storage, at ambient temperature and pressure, a preferential oxidation occurs in and around these damaged sites (craters). SEM and Optical microscopy show that 63 64 numerous cracks form in the surface oxide layer. It is believed that these cracks form as a result Of relaxation Of a very high degree Of intrnal stresses which are introduced during the shock deformation associated with the fast discharge of electrical energy. APPENDIX Figure 28 : Nanosecond electric pulse utilized for present investigation. The pulse duration of the "NanO-second Pulser" is specified (by the manufacturer) to be 36 nanoseconds. However, we do not know the magnitude of the discharge current and thus we can not estimate the energy per discharge. Therefore, it was necessary tO experimentally measure the actual current pulse under the experimental conditions. In order to do this a known resistance (0.186 ohm)was connected through a voltage divider between the cathode and anode in place of the metallic sample. The voltage signal across the known resistance was then recorded by using a Tektronix 545 type oscilloscope. Figure 28 shows the amplitude of this voltage signal. The discharge current was then calculated from these recorded pulses. Please note, that no attempt was made to resolve the shape Of the pulse since Such a resolution requires an oscilloscope with a rise time of abOut l nanosecond. 65 10. ll. 12. 13. 14. 15. 16. 17. 18. REFERENCES Askar'yan, G.A. and Morez, E.M., JETP Lett. 16, 1638 (1961). Ready, J.F., J. Appl. Physics, 36, 462 (1965). Yang, L.C., J. Appl. Phys., 45, 2601 (1974). Bushnell, J.C. and McCloskey, D.J., J. Appl. Phys., 39, 5541 (1968). Williams, J.S.,Brown, W.L., Leamy, H.J., Poate, J.M., Rodgers, J.W., Rousseau, D., Rozgonyi, G.A., Shelnutt, J.A. and Sheng, T.T., Appl. Phys. Lett., 33, 542 (1978). Laser-solid Interactions and Laser Processing - 1978, edited by Ferris, S.D., Leamy, H.J. and Poate, J.M., AIP Conference Proceeding NO. 50 (American Institute Of Physics, New York, 1979), p.457. Oron, M. and Sorensen, G., Appl. Phys. Lett. 35, 782 (1979). Compton, R.T., Phy. Rev. 21, 266, 1923. Langmuir, I.,Gen. Elect. Rev., 26, 732, 1923. Engel, A.V. and Steenbeck, M., "Elektrische Gasentladungen,ihre Physik u Techik," vol. 2, p. 130. Mackeown, S.S., Phys. Rev., 34, 611, 1929. Eyring, C.F., Mackeown, 8.8. and Millikan, R.A., Phys. Rev., 31, 900, 1928. Newman, F.H., Phil. Mag. 14, 788, 1932. Ramberg, W., Ann-d. Physik, 12, 319, 1932. Augis, J.A., Gibson, F.J. and Gray, E.W., Proceedings Of the 16th Annual Holm Seminar on Electrical Contact Phenomena, Chicago 1970, Intern. J. Electronics 30, 315 (1971). Secker, P.E. and George, I.A., Brit. J. Appl. Phys. (J.Phys. D) Ser. 2, vol. 2, 918 (1969). Sanger, C.C. and Secker, P.E., J. Phys. D, Appl. Phys., vol. 4, 1940 (1971). ll'in nad, V.E., Lebedev, S.V., Sov. Phys. Tech. Phys. 7, 717 (1963). 66 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 67 Fursey, G.N. et al., Proc. 9th Int. Conf. on Ionization Phenomena in Gases, Bucharest (Amsterdam, North Holland), 88 (1969). Basharov, R., Gavrilorskaya, E.N., Malkin, 0.A. and Trekhov, E.S., Sov. Phys. Tech. Phys. 10, 1428 (1966). Paschen, F., Wied. Ann., 37, 69, 1889. Foord, F.R., Nature, 166 (1950), 688. Ehrenkranz, F., Phys. Rev. 55 (1939), 219. Maxstadt, F.W., Elect. Eng. 53, 1063, 1934. Lee, T.H. and Greenwood, A., J. Appl. Phys., 32. NO. 5, 916 (1961). Goloveiko, A.G., Inzh.- Fiz. Zhurn., 14, No.3, 178 (1968). Beilis, 1.1. and Rakhovskii, V.l., Teplofiz. Vys. Temp., 7, No.4, 620 (1969) (High Temp., 7, 568, 1969). Kulyapin, V.M., Sov.Phys. Tech. Phys. Vol.16, No.2, 287 (1971). Struchkov, A.I., Sov. Phys. Tech. Phys. Vol.23, No.2, 183 (1978). Kantsel', V.V., Kurakina, T.S., Potokin, V.S., Rakhovskii, V.I. and Tkachev, L.C., Sov. Phys. Tech. Phys. Vol.13, No.6, 814 (1968). Lebedev, S.V., Izv. AN ArmSSR, 3, 33 (1950). Il'in, V.E. and Lebedev, S.V., Zh. Tekh. Fiz., 32 986 (1962) (Sov. Phys. — Tech Phys., 7, 717, 1963). Nekraschevich, I.G. and Bakuto, I.A., IF Zh., 2, 59 (1959) Nekrashevich, I.G., DAN BSSR, 4, 7 (1960). Mesyats, G.A., Pis'ma Zh. Tekh. Fiz. l, 885 (1975) (Sov. Tech. Phys. Lett. 1, 385 (1975). Korolev, Yu.D., Kuz'min, V.A. and Mesyats, G.A., Sov. Phys. Tech. Phys., Vol.24, No.4, 418 (1980). Kehl, G.L., "The Principles Of Metallographic Laboratory Practice," New York (McGraw Hill), 1949. Mukherjee, K., Kim, T.H. and Walter, W.T., "Lasers in Metallurgy", Eds: Mukherjee, K. and Majumder, J., Metallurgical Soc. of AIME, 1982, p. 137. Walter, W.T., Solimene, N., Park, K., Kim, T.H. and Mukherjee, K., ibid, p. 137. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 68 Engel, A.V. and Steenbeck, M., "Elektrische Gasentiadungen,ihre Physik u. Technik," Vol.1, p.98. Knoll, M., Ollendorff, F. and Rompe, R., "Gasentladungstabellen," p.79. Thomson, J.J. and Thomson, G.P., "Conduction Of Electricity through Gases," Vol.2, p.487. Gorbunov, A.V., Leiko, E.B., Nadgorny, E.M. and Valkovskii, S.N., Acta Metallurgica, Vol.14, p.418, 1980. Walters, C.T., Appl. Phys. Lett., 25, 12 (1974). Isenor, N.R., Appl. Phys. Lett., 31, 3 (1977). Emonony, D.C. et al., Appl. Phys. Lett., 23, 11 (1973). Maracas, G.N. et al., Appl. Phys. Lett., 33) 5 (1978). Birnbaum, M., J. Appl. Phys., 36, 3688 (1965). Leamy, H.J. et al., Appl. Phys. Lett., 32, 9 (1978). Jain, A.K. et al., J. Appl. Phys., 52, 4882 (1981). Kim, T.H., Mukherjee, K. and Walter, W.T., Bull. AIME Annual Meeting, Las Vegas, Nevada, 1979. Murphy, R.J. and Ritter, G.J., Nature, April 9, 210, 191 (1966). Vitali, G., Bestolotti, M., Foti, G. and Rimini, E., Phys. Lett., 63A, 351 (1977). Koo, J.C. and Slusher, R.E., Appl. Phys. Lett., 28) No.10, 614 (1976). Mukherjee, K., Kim, T.H. and Walter, W.T., "Lasers in Metallurgy", Eds: Mukherjee, K. and Majumder, Metallurgical Soc. of AIME, 1982.