You are on page 1of 9

IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 38, NO.

4, APRIL 2010

545

Magnetic Field Diffusion and Enhanced Resistivity in 12-cm-Diameter 200-ns 3.5-MA Z -Pinch Implosions
Niansheng Qi, B. H. Failor, J. S. Levine, J. Goyer, H. Sze, and A. Verma
AbstractInvestigations of magnetic eld diffusion and plasma resistivity in 12-cm-diameter triple-gas-puff Ar Z-pinch implosions were carried out by using planar laser-induced uorescence (PLIF), a laser shearing interferometer (LSI), and a laser wavefront analyzer (LWA) on a 3.5-MA 200-ns generator. The PLIF measurements gave the initial Ar gas distributions. The implosion velocity and electron density proles were measured from LWA and/or LSI. From these, the implosion plasma sheath thickness, ion density, mean ion charge states, temperatures, and implosion velocity are obtained, which allows us to calculate the classical plasma resistivity. A 1-D analytic magnetic eld diffusion model is constructed and used to predict the imploding plasma sheath thickness and its resistivity. Based on comparisons of the experimental measurements and the diffusion model prediction, we found out that plasma resistivity is enhanced by the cross-eld diffusion above the classical value, as high as 60 times the Spitzers value. Details are given in this paper. Index TermsConductivity, plasma pinch, plasma sheaths. uorescence, interferometry,

I. I NTRODUCTION SING Z-pinches is an excellent way to produce and conne high-temperature plasmas. These plasmas generate copious X-ray radiations that have numerous applications such as lithography, microscopy, and inertial connement fusion (ICF) for driving high-temperature radiation cavities (hohlraums) [1]. As a less expensive alternative to laser ICF, 65-MA 100-ns Z-pinches are proposed for ICF applications [2], [3]. However, pulsed-power accelerators have to overcome the large dI/dt and the resulting voltage on the insulator stack. With a typical 30-nH inductance in the insulators, the 65-MA 100-ns design will need an insulator stack that can hold off 20 MV. It is a daunting challenge given the state of the art. Thus, advanced Z-pinch loads are desired to reduce the need for fast-rise-time pulsed power, which, in turn, reduces the voltage hold-off on the insulator stack as well as the complexity and the cost associated with a fast-rise-time pulsed-power machine. A longer pulse driver also permits better electrical energy transfer from the Marx bank to the Z-pinch loads. To produce the necManuscript received November 19, 2009; revised January 5, 2010. First published March 8, 2010; current version published April 9, 2010. This work was supported by the Defense Threat Reduction Agency. N. Qi, B. H. Failor, J. S. Levine, J. Goyer, and H. Sze are with the Pulse Sciences Division, L-3 Communications, San Leandro, CA 94577 USA (e-mail: Niansheng.Qi@L-3com.com; Bruce.Failor@L-3com.com; jerrold. levine@l-3com.com; John.Goyer@L-3com.com; Henry.Sze@L-3com.com). A. Verma is with the Defense Threat Reduction Agency, Albuquerque, NM 87117 USA (e-mail: Ajay.verma@abq.dtra.mil). Color versions of one or more of the gures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identier 10.1109/TPS.2010.2041939

essary energy per atom in 200300-ns implosions, largediameter Z-pinch loads have been studied. Using a novel structured density prole to mitigate implosion instabilities, it has been demonstrated that a 200-ns-implosion-time 12-cmdiameter argon Z-pinch can provide the same argon K-shell yield as the one obtained from a 100-ns-implosion-time 2.5-cmdiameter Z-pinch at about the same peak current [4], [5]. Besides the Rayleigh-Taylor (RT) instabilities, plasma resistivity is another important but difcult parameter to measure in Z-pinches. The thickness of the plasma sheath, which is associated with the resistivity and plays a role in the performance of the implosion [6], is formed primarily by the magnetic eld diffusing into the plasma. When the magnetic eld diffuses into the resistive plasma, acceleration is spread over a distance determined by the magnetic scale length. Current ow results in Joule heating of the plasma, which changes plasma resistivity and, in turn, the magnetic scale length. In a situation where the implosion is stable and the rate of plasma heat conduction is relatively small, the thickness of the plasma sheath is mainly determined by the magnetic eld skin depth and can be calculated from a simple 1-D analytical diffusion model. In this paper, we present the following: 1) the characterization of the 12-cm-diameter structured Ar gas-puff implosions (double shells with a central jet) on a 3.5-MA 200-ns generator and 2) a 1-D magnetic eld diffusion model for predicting the imploding plasma resistivity and sheath thickness. Using planar laser-induced uorescence (PLIF) [7], [8], a laser shearing interferometer (LSI) [9], [10], and a laser wavefront analyzer (LWA) [11], the initial gas prole, the imploding plasma density prole, and the implosion velocity were measured, which enabled us to derive the ion charge states and plasma temperature. From these data, a 1-D magnetic eld diffusion model was used to show that plasma resistivity is about ten times the classical Spitzer value. The enhanced plasma resistivity can be explained by the Hall effect. The remainder of this paper is organized as follows. Sections II and III show the experimental arrangements and results, respectively. Section IV presents the 1-D magnetic eld diffusion model and the comparison of the predicted and calculated plasma sheath thicknesses. In Section V, the conclusion is drawn from the experiments, and recommendations for future experiments are given. II. E XPERIMENTAL A PPARATUS The experiments were conducted on the 3.5-MA Double Eagle generator operating in a long-pulse 200-ns-implosiontime mode. Fig. 1 shows the arrangements of the experiments.

0093-3813/$26.00 2010 IEEE


Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

546

IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 38, NO. 4, APRIL 2010

Fig. 2.

Experiment arrangements of LSI and LWA.

Fig. 1. Experimental setup. 1) Outer shell gas. 2) Inner shell gas. 3) Center jet gas. 4) Nozzle anode plane. 5) Current return posts. 6) Probe laser path. The argon gas-puff nozzle has an outer annulus spanning from 5 to 6 cm in radius, an inner annulus spanning from 2 to 3 cm, and a central jet of 0.5 cm in radius. The pinch plasma length is 3.8 cm.

A triple nozzle has been used to produce three concentric Ar gas puffs. The inner/outer diameters of the two annular shells were 10/12 and 4/6 cm at the nozzle exit plane, respectively. The center puff was a pencillike jet with a 1-cm-diameter opening at the nozzle exit plane. The gas puff was positioned in the load region of the high-power pulsed electrical generator. The cathode plane was formed by mounting a high-transparency brass mesh on the nozzle exit, and the anode plane was formed by a web of aluminum wires that is located 3.8-cm downstream from the cathode. The wires were attached to 12 return current posts located at a diameter of 15.6 cm. In the experiments, the supersonic Ar gas was injected from the triple nozzle (cathode) into the Z-pinch load region. An electrical breakdown signal was produced by a spark gap located in the gas-puff nozzle throat, which serves as the timing of the gas injection. The pulse generator was discharged 500 s after the breakdown signal. The implosion current reached a peak of 3.5 MA in 100 ns, stayed at that level for 100 ns, and then decreased. The current was measured by a Rogowski coil mounted on the anode feed just outside the current return post array. The time history of the spatially integrated L- and K-shell X-ray emissions was measured using ltered aluminum X-ray diodes. The implosion time, dened as the time from the onset of the current to the onset of the K-shell emission, depended on the gas-puff mass and typically was about 220 ns. Three laser-based diagnostic instruments were used to characterize the implosions. The initial mass distribution of the gas puff, which determines the imploding plasma trajectory, was measured using the PLIF in a separate test bed. The electron density proles during the implosion were obtained by using the LSI and LWA. In the PLIF diagnostic, a frequency-quadruped Nd:YAG laser (7 ns, 45 mJ, and 266 nm) was used. The line-focused laser beam ( 0.2 40 mm) passed through the diameter of the argon gas puff. To produce uorescence, the argon gas was mixed

with 5% acetone by pressure. Fluorescence from the acetone tracer in the argon gas was captured using a gated intensied charge-coupled device (CCD) camera, which has a 576 384 pixel array with a pixel size of 19 m. With an image demagnication of 1 : 16.3, the spatial resolution of the gas density measurements was 0.31 mm, and the sensitivity or minimum detectable density was 5 1013 acetone molecules cm3 for a pure acetone vapor or 1015 Ar atoms/cm3 for Ar gas puffs. Detailed descriptions of the PLIF measurements are reported in an earlier publication [8]. For the LSI and LWA diagnostics, a 4-ns 25-mJ 532-nm 6.5-cm-diameter Nd:YAG laser beam was used. Although the relatively long (4 ns) laser pulse was not ideal to freeze the plasma motion and caused blurring in the LSI/LWA images, we were able to observe the imploding plasma structures and to derive the plasma densities, as shown later. The LSI and LWA measurement arrangements are shown in Fig. 2. The laser probe consisted of a laser, a spatial lter, and a beam expander. It was contained on a 4 8 ft optical table. The laser beam was expanded to 6.5 cm in diameter and directed toward the imploding plasma through the steering mirrors and periscopes shown in Fig. 1. After passing through the plasma, the wavefront of the laser beam was shifted in phase and refracted perpendicular to the propagation direction. It was collected and then split into two, from which the refractions due to the plasma were analyzed by the LSI and LAW. The full length (3.8 cm long), but only half the diameter (12 cm), of the plasma was imaged by these two laser instruments. Narrowband laser line lters were used to preferentially pass the laser light and discriminate against the intense self-emissions from the imploding plasma. The LSI consists of a corner cube beam splitter and two planar mirrors. This shearing causes interference between these two wavefronts and produces an interferogram. The images were captured using a large-format view camera with Polaroid Type-55 positive/negative lm. The sensitivity of the LSI was set at about 1/4 wave of the phase shift, which is equal to a gradient of the density chordal integral of 5 1016 cm3 . The spatial resolutions of the LSI could be 25 m or better, but due to the relatively wide laser pulse (4 ns) and fast implosion velocity (3 107 cm/s), the radial resolution was reduced to 1 mm by motion blurring. In the LWA, a 65 65 microlens array with a 0.4 0.4 mm lenslet size and a 24-mm focal length was used to produce an

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

QI et al.: MAGNETIC FIELD DIFFUSION AND RESISTIVITY IN Z-PINCH IMPLOSIONS

547

array of focused microbeamlets in the image. The positions of these focused microbeamlets were captured by a CCD camera, which has an array of 1300 1300 pixels with a pixel size of 24 m. The 4000 beamlet focal spot displacements followed a Gaussian distribution with a 2 width less than 0.1 pixel (or 2.4 m). With a 3 : 1 image demagnication, the spatial resolution of the LWA was 1 mm, and the angular resolution was 33 rad. This corresponds to a 2.6 1017 cm3 gradient of the density chordal integral. The LSI has the potential of giving details of the plasma shape, with spatial resolutions down to 25 m, using a subnanosecond laser, but fringe analysis for density determination is complicated. In our experiments, the radial spatial resolution of the LSI is limited to 1 mm in the region of interest due to the plasma motion during the 4-ns laser pulse. The LWA gives the density straightforwardly, but the spatial resolution is 250 m, as limited by the size of the microlenslet. The applications of these diagnostics to the measurements of Z-pinch plasma density and current distribution have been reported in detail in [11] and [12]. From the LSI and LWA images, the electron density and plasma sheath thickness were obtained. Knowing the initial gas distribution from the PLIF measurements, the plasma ionization states and, then, the electron temperature of the imploding plasma were derived from an ionization balance model. Electron densities were obtained, with relatively large uncertainties of about 20% due to the blurring discussed earlier. III. E XPERIMENTS AND R ESULTS We have conducted Z-pinch implosions with the following two types of initial density proles: 1) the outerinner-jet prole, with argon in the center jet and both inner and outer shell plenums to produce a stable implosion, and 2) the outer-jet prole, with argon in the jet and outer shell plenums but with the inner shell plenum evacuated, which results in unstable implosions [4], [13]. By using the PLIF, the gas density proles were measured at the same time as the Double Eagle voltage pulse is applied. The axial ow velocity of the gas is 650 20 m/s, which exceeds the limit for ideal gases due to the energy released by argon condensation [8]. The Z-pinch implosions using the aforementioned two density proles have been reported earlier for RT instability mitigation [12]. In this paper, we investigate magnetic eld diffusion. Data analysis is focused on Z-pinch implosions with the outerinner-jet prole only as the implosions with such density prole are stable. Also, we show most of the data analysis at or near the axial locations of downstream Z = 2 and/or 3.5 cm for simplicity. Similar conclusions can be found for other Z locations. In the outer-inner-jet case, the nozzle plenum pressures of the outer, inner, and center jets were 52, 155, and 1140 torr, respectively. The peak radius-weighted mass density is 8 g/cm2 . Fig. 3 shows the implosion current (black solid line) and the relative Ar L- (blue dotted line) and K-shell (red solid line) X-ray radiation powers (solid line) at Z = 2 cm, i.e., near the midplane of the 3.8-cm-long pinch for a typical shot. The current started at t = 0, peaked in 100 ns, and then stayed relatively at for another 100 ns. The Ar L-shell radiation

Fig. 3. Experimental and calculated data at the midplane of the pinch (2 cm from the cathode). (Black solid line) Current, (blue dotted line) L- and (red solid line) K-shell X-ray powers (1-TW peak), (black dotted line) calculated implosion plasma radius, and (blue solid line) implosion velocity. The four solid circles with error bars are the implosion plasma radii measured from the LSI images also at a location of Z = 2 cm.

started at t 200 ns and peaked at 232 ns, and the K-shell X-ray emission started at 220 ns and peaked at 224 ns. The imploding plasma typically has a shell-like shape, which implies that the current ows in the outer boundary region. A 0-D snowplow model was used to calculate the plasma sheath radius, velocity, and acceleration during the implosion [14]. The model ignores axial motion. It assumes that the full mass in the initial gas distribution is swept by the radial implosion into an innitely thin current-carrying layer. The snowplow model cannot give the plasma density distribution within the sheath, but the calculated implosion plasma radius and mean velocity generally agree with the experiments and simulations [15] as the acceleration of the plasma sheath is dominant by the magnetic force. The snowplow model for a given axial Z location is given by 2R 0 I 2 d2 R + = 2 dt 4Rm m with
R_post

dR dt

(1)

m(R) = 2
R

(r)r dr

(2)

where R is the implosion plasma radius, t is time, and (r) is the gas-puff mass distribution. The radius and radial velocity of the plasma sheath are calculated from (1) using the measured radial density distribution, with the initial conditions of R = 6.75 cm and dR/dt = 0. The calculation stops when the plasma radius reaches the experimentally measured pinch plasma radius, typically 1 mm. Fig. 3 shows the calculated implosion plasma radius (black dotted line) and velocity (blue solid line) as a function of time at Z = 2 cm. We assume that the calculated radius corresponds to the approximate position of the current-carrying plasma sheath at that axial location, with an error of less than half of the sheath thickness. The LWA and LSI measurements of the plasma radius during the implosion (four solid circles in Fig. 3) are in good agreement with the calculation, where the errors in the plasma radius are about 1 mm, limited by the spatial resolution of the LSI and LWA. The current waveform, implosion time, and K-shell X-ray power were reproducible in a series of four consecutive shots [12]. The spread in the nal K-shell radiation yield is less than 10% [16] and is mainly due to the RT instability growth during

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

548

IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 38, NO. 4, APRIL 2010

Fig. 4. (a) Gas mass (R) prole at Z = 2 cm (peak is 8 g/cm2). (b) PLIF R contours. (c) and (d) LSI images obtained at t = 122 and 148 ns, respectively; the plasma sheath is highlighted in blue color.

the nal pinch phase. Z-pinch plasmas are susceptible to the RT instabilities, which are driven by plasma acceleration. During the implosion, the RT instabilities grow until the outer shell strikes the inner one. When the plasma shell imploded onto the more massive inner gas puff, there is almost no acceleration of the plasma shell from t = 110 to t = 150 ns, as seen in the velocity curve of Fig. 3. The amplitudes of the RT instabilities are thus suppressed. The late magnetic eld diffusion analysis will focus in this time period since the implosion is relatively stable and the current and velocity are almost constant. Fig. 4(a) and (b) shows the measured gas mass times the radius r at an axial position of Z = 2 cm and the 2-D contour, respectively. The errors of the PLIF gas measurements are 20%, as discussed in [8]. The nozzle plenum pressures of the jet, inner shell, and outer shell were 22, 3, and 1 psia, respectively. This gas prole is optimized for 200-ns implosions. Fig. 4(c) and (d) shows the LSI images obtained at t = 122 and 148 ns, respectively. These images were captured by ring multiple shots at the same gas pressure setting in the nozzle plenums with reproducible current drive waveforms. In these images, the two 5-mm-wide horizontal dark zones at 2.2- and 5.5-cm radii are artifacts of four current return posts that blocked the laser beam. Likewise, a dark shadow of the anode return current wires is seen at Z 3.8 cm. Also, at the bottom left, there is a dark shadow of a 2-mm-diameter alignment pin located at a radial distance of 8 cm away from the pinch axis. The imploding plasma sheath, which is highlighted in blue in the LSI images of Fig. 4(c)(d), shows an annular structure of about 510 mm thick, slanted outward from the cathode to the anode, which is consistent with the injected gas prole and the snowplow calculations. The relatively large fringe changes indicate the boundaries of the imploding plasma front, whereas the background fringes outside the plasma are unperturbed. Although the 4-ns laser pulse limits the fringe resolution to 1 mm, the boundary and sheath of the imploding plasma were clearly observed. The quantitative densities are obtained from the LWA images. In the presence of the plasma, the shift in the focal position of each microbeamlet in LWA corresponds to the gradients of the density integration along the laser path, i.e., a chordal measurement. Again, the focal spot of the microbeamlets in the LWA images are blurred due to the plasma motion, but

Fig. 5. (a) LWA image at t = 148 ns. (b) Expanded view of the boxed region in a). (Green spots) The data are superimposed on (red spots) a ducial image taken without plasma, captured immediately prior to the Double Eagle shot. The focal spots are nominally 0.4 mm apart, and typical displacements of the focal spots in the sheath region are 0.1 mm, which correspond to a deection angle of 1.3 mrad of the probe laser beam.

the relative position changes of these focal spots are measurable. Fig. 5(a) shows a typical LWA image captured at t = 148 ns. Fig. 5(b) is an expanded view of the box region shown in Fig. 5(a), and displacements of the focal spots in the plasma sheath region are clearly observed. Fig. 6(a) and (b) shows the derived electron density prole with an uncertainty of 20% at t = 122 and 148 ns, respectively. The imploding plasma looks like a shell, which is consistent with the LSI images. The peak electron density is 4 1017 cm3 . Since the shadow of the return current posts at r 2.5 cm blocked part of the plasma images, the densities at r < 2.5 cm could not be accurately obtained through able inversion but are extrapolated. Fig. 7(a) and (b) shows the measured plasma density versus the radius at Z = 2 and 3.5 cm, respectively.

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

QI et al.: MAGNETIC FIELD DIFFUSION AND RESISTIVITY IN Z-PINCH IMPLOSIONS

549

Fig. 8. Calculated mean ion charge state of Ar plasmas as a function of temperature at an electron temperature of 3.5 1017 cm3 . TABLE I M EASURED AND C ALCULATED Z-P INCH P LASMA PARAMETERS AT A XIAL P OSITIONS Z = 2 AND 3.5 cm AND I MPLOSION T IMES OF 122 AND 148 ns. AND A RE THE C ALCULATED S KIN D EPTH U SING C LASSICAL S PITZER R ESISTIVITY AND E NHANCED R ESISTIVITY , R ESPECTIVELY. T HE M EASURED P LASMA S HEATH T HICKNESS rmeas I S E XPECTED TO B E H ALF OF THE S KIN D EPTH

Fig. 6. Density proles obtained from LWA (a) at t = 122 and (b) at t = 148 ns. The peak electron density is 4 1017 cm3 .

Fig. 7. Electron density at the implosion time of (a) 122 and (b) 148 ns. The black line is the density at Z = 2 cm, and the red line is at Z = 3.5 cm.

From the measured initial gas and implosion electron density prole, the plasma mean charge states, temperature, and, then, resistivity can be calculated. To estimate the number of the ions in the imploding Z-pinch, we assume that all of the mass originally injected by the gas puff (measured by PLIF) from the radius of the imploding plasma outward has been swept up into the sheath and that the ion density prole is similar to that of electrons. The ion and electron line densities are obtained as the integral of the initial gas and plasma density prole from the location of the inner sheath radius outward, respectively. Thus, the mean charge states of the plasma sheath are derived from the ratio of the electron and ion line density. An ionization model for Ar ions was constructed to infer the plasma temperature for a given ionization state. The ionization model includes the net effect of collisional ionization and collisional, radiative, and dielectronic recombination processes in the Ar plasma. The coefcients of these atomic processes are from [17]. Fig. 8

shows the calculated mean ion charge state as a function of the plasma temperature at a typical measured electron density of 3.5 1017 cm3 . The plasma temperatures are 916 eV when the mean ion charge states are 4 and 6, respectively. In the temperature range of 520 eV, the mean ion charge is proportional to 3.4 ln T . Table I lists the derived plasma parameters such as radius, velocity, density, temperature, charge state, and sheath thickness at implosion times of 122 and 148 ns and at axial positions of Z = 2.0 and 3.5 cm from the cathode. For plasmas where Coulomb collisions dominate all other dissipation processes, including wave and turbulence effects, the resistivity is determined by the collisional drag on electrons moving against the background of ions. If a strong magnetic eld is applied perpendicular to the electric eld direction, the current is not due to the direct acceleration of electrons by the

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

550

IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 38, NO. 4, APRIL 2010

electric eld but is diamagnetic in origin. The transverse or cross-eld resistivity was calculated by Spitzer as the rate of momentum transfer from electrons to ions through collisions. Knowing the mean plasma ionization state (Z) and its temperature (T in electronvolts), the Spitzer plasma resistivity for the motion perpendicular to the B eld is calculated from [18], [19]
1 = = 104

The diffusion equation of the magnetic eld is derived as B 1 2B B +v = 0. t x o x2 (8)

Z ln TeV
3/2

Changing to the moving plasma frame, i.e., y = x vt, (8) becomes 1 2B B = 0. t o y 2 The boundary condition at y = 0 is B(0, t) = B0 et/ . The general solution of (9) and (10) is B(y, t) = B(t)B(y) = B0 et/ y/ y0 (11) (10) (9)

( m)

(3)

where the Coulomb logarithm ln is given by ln = 23 0.5 ln(Zne ) + 1.5 ln T (4)

with ne as the electron density in per cubic centimeter. For the imploding plasma conditions considered here, ln is equal to ve. The derived Spitzer plasma resistivities are listed in Table I. The relative uncertainty of the measurements is 20% for ion mean charge states (Z/Z). Recall that the mean Ar ion charge state is proportional to 3.4 ln T ; the relative error in the Spitzer resistivity is about Z/4 or 25%. IV. M AGNETIC F IELD D IFFUSION M ODEL Very often, the plasma sheath is referred to as the current sheath, whose thickness is related to the magnetic eld skin depth. During the implosion, the magnetic eld diffuses into the resistive plasma, spreading the acceleration over a distance characterized by the magnetic scale length. Current ow results in Joule heating of the plasma, which changes its resistivity and, in turn, the magnetic scale length. A realistic evaluation of the plasma sheath formation requires that the magnetic eld diffusion and implosion dynamics be simultaneously taken into account. This general problem includes the complete set of MHD equations with an appropriate equation of state and can be solved only numerically. However, when the implosion current and velocity are quasi-constant, an approximate expression for the plasma sheath thickness can be derived, as shown in the following. We start the analysis by limiting the problem to one Cartesian coordinate dimension. This is valid when the magnetic eld penetration depth (skin depth) is much smaller than the radius of the implosion plasma boundary. This simplication enables us to nd a simple analytical solution and thus obtains useful insight into the magnetic eld diffusion and/or Joule heating processes during the implosion. We assume that there is a semiinnite plasma of conductivity . At time t = 0, the boundary of the plasma is located at x = 0, i.e., the plasma has lled the region x > 0. The boundary of the plasma is moving at a constant velocity v along the positive x-axis direction as compressed by an external magnetic pressure starting from the region of x < 0. The magnetic eld inside the plasma sheath is dominated by the diffusion of the eld at the boundary. We start from Faraday, Ampere, and Ohms laws B t B = 0 J E = J = (E + v B). (5) (6) (7)

where is the magnetic eld skin depth. At the plasma boundary of radius R, the magnetic eld is equal to 0 I/2R, and its time derivative is dB =B dt y=0 1 dI 1 dR I dt R dt =B 1 dI v + I dt R (12)

where I is the implosion current. Thus, we have for a quasiconstant current 1 = 1 dI v + I dt R v. = R (13)

Notice that the characteristic time should be time independent to derive (11). This requires that t d dt 1. (14)

At the time of interest, the current and implosion velocity ( 3 107 cm/s) are constant, and the plasma radius is 3 cm (see Fig. 3). The characteristic time is on the order of R/v (100 ns). The time scale t for the B eld diffusion is on the order of /v (10 ns), and d /dt 1. Thus, the condition of (14) is satised. Substituting (11) into (9), the magnetic skin depth is derived = /0 = R/0 v. (15)

Equation (15) implies that a higher implosion velocity produces a narrower skin depth. This agrees with the simulations and observations in the magnetic ux compression experiments [20], [21], where it is found that an efcient compression of the axial magnetic ux by a thin plasma sheath requires the sheath velocity to be sufciently high. Furthermore, a fast rise time of the current pulse gives a smaller and also makes a thinner skin depth. In our measurements, the plasma has a constant implosion velocity during the time from 110 to 150 ns. The zero acceleration of the plasma makes the implosion stable and implies that the magnetic pressure is approximately balanced by the plasma

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

QI et al.: MAGNETIC FIELD DIFFUSION AND RESISTIVITY IN Z-PINCH IMPLOSIONS

551

pressure. Assuming relatively slow changes in the electron temperature and noticing that the magnetic pressure is proportional to the square of the magnetic eld, the skin depth of the B-eld diffusion is expected to be two times thicker than the plasma sheath [22], [23]. Table I shows the calculated skin depths using the Spitzer resistivity . The relative error of the skin depth is about 12%, which is half of the relative error in resistivity. Instead of double, the classical skin depth values are at least two times smaller than the measured plasma sheath thickness. It indicates that the plasma resistivity has to be enhanced. When a magnetic eld is present, conducting electrons move, between collisions, in circular (or helical) orbits with a Larmor radius. In a magnetized plasma, the ratio of the electron gyrofrequency ce to the electron-ion collisional frequency ei is 1, and the Hall effect dominates the penetration of the magnetic eld. Reference [19] indicates that the classical values of the transverse and parallel electrical conductivity of the plasma are of the same order; they only differ by a factor varying from 2 at Z = 1 to 3 as Z approaches innity. This is because the strong Hall eld is produced by charge separation within quasi-neutrality. Nevertheless, the classical value of electrical resistivity, among other transport coefcients, is universally used as an important reference value and as the lower bound whenever transport or dissipation phenomena are discussed. From the plasma sheath thickness measurements, the crosseld diffusion is substantially enhanced above the classical value. This is often called enhanced dissipation or anomalous resistivity . One of the models (see [24]) gives an expression for the enhanced resistivity as = 1 + with ce = eB = 1.76 1011 B(T) (s1 ) me Zne ln ei = 2 106 (s1 ) 3/2 TeV (17) (18) c ei
2

eld is given by e = 4.7 nkT 2 me ce e (20)

0.5 = 8.6 1015 n2 (cm3 )TeV B 2 (T) ln . e

The estimated thermal diffusion length is less than 0.1 mm. Compared with the observed sheath thickness of several millimeters, the thermal diffusion is negligible, and heat conduction plays a limited role, i.e., Joule heating stays within the plasma sheath. As discussed earlier, plasma resistivity is increased by a factor of 2060. The effects of the enhanced resistivity can be examined from the energy balance of the implosion plasma. As shown in Table I, the plasma temperature decreased a few electronvolts at Z = 2 cm and stay more or less the same at Z = 3.5 cm from t = 122 to 148 ns. This implies that the Joule heating in the plasma sheath is comparable to the radiation loss and internal energy changes. In the conditions considered here, the dominant energy loss is the plasma radiation [25]. The power of Joule heating (PJoule ) and radiation loss (Prad ) per unit length are given by
R

PJoule = 2

j 2 rdr

2R

1 B 0 y

dy =

I2 8Rr

(21)

(16)

Prad = 2C
0

ni (R)ne (R)r dr RrCne ni (22)

where ne is the electron density in per cubic centimeter. is a geometrical parameter on the order of one, which is used to take into account the departure from the mean values of the density, temperature, and magnetic eld in the sheath. = 0.75 is a good t to the measured plasma sheath thickness and is used in the calculation. The calculated enhanced resistivities are shown in Table I. They are 2060 times higher than the classic values. Using the enhanced plasma resistivity , the resulting skin depth is about twice the plasma sheath thickness measured from the LWA images. This agrees with the theoretical predictions [22], [23]. The implosion plasma sheath might be thicker due to the plasma thermal diffusion. However, this is not the case here. The thermal diffusion length is given by th = th /ne (19)

where th /v 10 ns, and the electron thermal diffusivity (in per centimeter seconds) in the presence of the magnetic

where C is the radiation loss coefcient. r is the plasma sheath thickness and is equal to half of the skin depth. The expression of the magnetic eld distribution B(y) given by (11) is used in (21), and the electron or ion density is proportional to the square of the B eld. In [26], the coefcient C (in watt per cubic centimeter per electron per ion) has been calculated for Ar plasmas up to an electron density of 1016 cm3 using the FINE code [27]. In the electron temperature range of 820 eV, the coefcient C is a weak function of density. To estimate the radiation power, the radiation loss coefcients at 1016 cm3 are used, and the estimated errors could be 50%. Table I also lists the radiation coefcients and calculated powers of Joule heating and radiation loss at t = 122 to 148 ns and two axial locations (Z = 2 and 3.5 cm). The power of Joule heating, using enhanced resistivity, is comparable to the radiation loss. At Z = 3.5 cm location, the Joule heating (24 GW/cm) is more or less the same as the radiation loss (22 GW/cm) at t = 122 ns, and the temperature remains the same at t = 148 ns, i.e., 16 versus 15 eV. However, at Z = 2 cm and t = 122 ns, the 23-GW/cm Joule heating is 40% less than the radiation loss (32 GW/cm). The plasma temperature decreases from 16 to 9 eV at t = 148 ns. Notice that our magnetic diffusion model predicts a factor of 2060 enhancement in the plasma resistivity. Using the classic resistivity, the Joule heating power would be

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

552

IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 38, NO. 4, APRIL 2010

too low compared with the radiation power to maintain the plasma temperature. This effect has been observed in simulations where, using the classical resistivity, Joule heating is often found to be too low and/or plasma sheath is too narrow [13]. V. C ONCLUSION We have characterized a 3.5-MA 200-ns Z-pinch implosion plasma by using three laser diagnostic instruments, i.e., PLIF, LSI, and LWA. From these measurements, the electron density prole, mean ion charge state, temperature, velocity, and resistivity of the imploding plasma were obtained. An analytical model of the magnetic eld diffusion into the imploding plasma was discussed and has been used to predict the magnetic eld skin depth. Using classical resistivity, the calculated skin depth is two to four times narrower than the measured plasma sheath, instead of twice the plasma sheath thickness. Including the Hall effects, plasma resistivity is enhanced by a factor of 2060, and the thickness of the skin depth is increased by a factor of 48. Using the enhanced plasma resistivity, not only do the predicted (twice of the skin depth) and measured plasma sheath thickness agree with each other, but also the Joule heating power is substantially enhanced (2060 times) and is close to the radiation power losses. In the future, the PLIF measurement can be combined with a laser interferometer to reduce the errors in the gas density from 10% down to < 5% [28]. The resolution and sensitivity of the LSI and LWA measurements can be increased by using a short (100 ps), multiframe (24) laser to freeze the plasma motion and reduce the shot-to-shot variations. The current/magnetic eld distributions in the plasma, thus the skin depth, can be measured by adding laser Faraday rotation diagnostics [10]. The temperatures can be derived from the plasma spectrum in the visible or XUV region [17]. The measurements of the total radiation power directly gives the plasma resistivity for comparisons. These experimental improvements in the initial gas distribution and implosion phase provide much more accurate needed data for guiding Z-pinch experiments and simulations [29]. ACKNOWLEDGMENT The authors would like to thank A. Velikovich for the many suggestions and discussions and the L-3 Communications Pulse Sciences accelerator crew for the fast and careful execution of these experiments. R EFERENCES
[1] N. R. Pereira and J. Davis, X rays from Z-pinches on relativistic electron-beam generators, J. Appl. Phys., vol. 64, no. 3, pp. R1R27, Aug. 1988. [2] T. C. Sangster, R. L. McCrory, V. N. Goncharov, D. R. Harding, S. J. Loucks, P. W. McKenty, D. D. Meyerhofer, S. Skupsky, B. Yaakobi, B. J. MacGowan, L. J. Atherton, B. A. Hammel, J. D. Lindl, E. I. Moses, J. L. Porter, M. E. Cuneo, M. K. Matzen, C. W. Barnes, J. C. Fernandez, D. C. Wilson, J. D. Kilkenny, T. P. Bernat, A. Nikroo, B. G. Logan, S. Yu, R. D. Petrasso, J. D. Sethian, and S. Obensc, Overview of inertial fusion research in the United States, Nucl. Fusion, vol. 47, no. 10, pp. S686S695, Sep. 2007. [3] C. L. Olson, G. Rochau, S. Slutz, C. Morrow, R. Olson, M. Cuneo, D. Hanson, G. Bennett, T. Sanford, J. Bailey, W. Stygar, R. Vesey, [16]

[4]

[5]

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14] [15]

[17] [18] [19]

T. Mehlhorn, K. Struve, M. Mazarakis, M. Savage, T. Pointon, M. Kiefer, S. Rosenthal, K. Cochrane, L. Schneider, S. Glover, K. Reed, D. Schroen, C. Farnum, M. Modesto, D. Oscar, L. Chhabildas, J. Boyes, V. Vigil, R. Keith, M. Turgeon, B. Cipiti, E. Lindgren, V. Dandini, H. Tran, D. Smith, D. McDaniel, J. Quintenz, M. K. Matzen, J. P. VanDevender, W. Gauster, L. Shephard, M. Walck, T. Renk, T. Tanaka, M. Ulrickson, W. Meier, J. Latkowski, R. Moir, R. Schmitt, S. Reyes, R. Abbott, R. Peterson, G. Pollock, P. Ottinger, J. Schumer, P. Peterson, D. Kammer, G. Kulcinski, L. El-Guebaly, G. Moses, I. Sviatoslavsky, M. Sawan, M. Anderson, R. Bonazza, J. Oakley, P. Meekunasombat, J. De Groot, N. Jensen, M. Abdou, A. Ying, P. Calderoni, N. Morley, S. Abdel-Khalik, C. Dillon, C. Lascar, D. Sadowski, R. Curry, K. McDonald, M. Barkey, W. Szaroletta, R. Gallix, N. Alexander, W. Rickman, C. Charman, H. Shatoff, D. Welch, D. Rose, P. Panchuk, D. Louie, S. Dean, A. Kim, S. Nedoseev, E. Grabovsky, A. Kingsep, and V. Smirnov, Development path for Z-pinch IFE, Fusion Sci. Technol., vol. 47, no. 3, pp. 633640, Apr. 2005. H. Sze, J. Banister, B. H. Failor, J. S. Levine, N. Qi, A. L. Velikovich, J. Davis, D. Lojewski, and P. Sincerny, Efcient radiation production in long implosions of structured gas-puff Z pinch loads from large initial radius, Phys. Rev. Lett., vol. 95, no. 10, p. 105 001, Sep. 2005. J. S. Levine, J. W. Banister, B. H. Failor, N. Qi, H. M. Sze, A. L. Velikovich, R. J. Commisso, J. Davis, and D. Lojewski, Implosion dynamics and radiative characteristics of a high yield structured gas puff load, Phys. Plasmas, vol. 13, no. 8, p. 082702, Aug. 2006. M. R. Douglas, C. Deeney, and N. F. Roderick, The effect of load thickness on the performance of high velocity, annular Z-pinch implosions, Phys. Plasmas, vol. 8, no. 1, pp. 238248, Jan. 2001. B. H. Failor, S. Chantrenne, P. L. Coleman, J. S. Levine, Y. Song, and H. M. Sze, Proof-of-principle laser-induced uorescence measurements of gas distributions from supersonic nozzles, Rev. Sci. Instrum., vol. 74, no. 2, p. 1070, Feb. 2003. N. Qi, B. H. Failor, J. Banister, J. S. Levine, H. M. Sze, and D. Lojewski, Two-dimensional gas density and velocity distributions of a 12-cm-diameter, triple-nozzle argon Z-pinch load, IEEE Trans. Plasma Sci., vol. 33, no. 2, pp. 752762, Apr. 2005. N. Qi, M. Krishnan, R. Prasad, and S. Fulghum, Space and time resolved electron density and current measurements in a dense plasma focus Z-pinch, IEEE Trans. Plasma Sci., vol. 26, no. 4, pp. 11271137, Aug. 1998. N. Qi, J. Schein, J. Thompson, P. Coleman, M. McFarland, R. R. Prasad, M. Krishnan, B. V. Weber, B. G. Moosman, J. W. Schumer, D. Mosher, R. J. Commisso, and D. Bell, Z pinch imploding plasma density prole measurements using a two-frame laser shearing interferometer, IEEE Trans. Plasma Sci., vol. 30, no. 1, pp. 227238, Feb. 2002. N. Qi, R. R. Prasad, K. Campbell, P. Coleman, M. Krishnan, B. V. Weber, S. J. Stephanakis, and D. Mosher, Laser wavefront analyzer for imploding plasma density and current prole measurements, Rev. Sci. Instrum., vol. 75, no. 10, pp. 34423445, Oct. 2004. N. Qi, H. Sze, B. H. Failor, J. Banister, J. S. Levine, J. C. Riordan, P. Steen, P. Sincerny, and D. Lojewski, Magnetic RayleighTaylor instability mitigation in large-diameter gas puff Z-pinch implosions, Phys. Plasmas, vol. 15, no. 2, p. 022703-1, Feb. 2008. H. Sze, J. S. Levine, J. Banister, B. H. Failor, N. Qi, P. Steen, A. L. Velikovich, J. Davis, and A. Wilson, Magnetic RayleighTaylor instability mitigation and efcient radiation production in gas puff Z-pinch implosions, Phys. Plasmas, vol. 14, no. 5, p. 056307, Apr. 2007. D. Mosher, B. V. Weber, B. Moosman, R. J. Commisso, P. Coleman, E. Waisman, H. Sze, Y. Song, D. Parks, P. Steen, J. Levine, B. Failor, and A. Fisher, Laser Part. Beams, vol. 19, p. 579, 2001. J. Schumer, D. Mosher, B. Moosman, B. Weber, R. J. Commisso, N. Qi, J. Schein, and M. Krishnan, Two-dimensional MHD simulations of a neon Z pinch on Hawk, IEEE Trans. Plasma Sci., vol. 30, no. 2, pp. 227497, Apr. 2002. B. H. Failor, H. M. Sze, J. W. Banister, J. S. Levine, N. Qi, J. P. Apruzese, and D. Y. Lojewski, K-shell and extreme ultraviolet spectroscopic signatures of structured Ar puff Z-pinch loads with high K-shell X-ray yield, Phys. Plasmas, vol. 14, no. 2, p. 022703, Feb. 2007. N. Qi and M. Krishnan, Theoretical and experimental study of UV lasers in Be-like ions pumped by resonant photoexcitation, Phys. Rev. A, Gen. Phys., vol. 39, no. 9, pp. 46514667, May 1989. L. Spitzer, Physics of Fully Ionized Gases. New York: Interscience, 1956. S. I. Braginskii, Transport processes in a plasma, in Reviews of Plasma Physics, M. A. Leontovich, Ed. New York: Consultants Bureau, 1965, p. 205.

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

QI et al.: MAGNETIC FIELD DIFFUSION AND RESISTIVITY IN Z-PINCH IMPLOSIONS

553

[20] F. S. Felber, M. M. Malley, F. J. Wessel, M. K. Matzen, M. A. Palmer, R. B. Spielman, M. A. Liberman, and A. L. Velikovich, Compression of ultrahigh magnetic elds in a gas-puff Z pinch, Phys. Fluids, vol. 31, no. 7, pp. 20532056, Jul. 1988. [21] F. S. Felber, M. A. Liberman, and A. L. Velikovich, Magnetic ux compression by dynamic plasmas. I. Subsonic self-similar compression of a magnetized plasma-lled liner, Phys. Fluids, vol. 31, no. 12, p. 3675, Dec. 1988. [22] R. B. Spielman, R. J. Dukart, D. L. Hanson, B. A. Hammel, W. W. Hsing, M. K. Matzen, and J. L. Porter, Dense Z-pinches, in Proc. 3rd Int. Conf. Dense Z-Pinches, vol. 299, AIP Conf. Proc., M. Haines and A. Knight, Eds., London, U.K., 1993, p. 404. [23] L. I. Rudakov, A. Chuvatin, A. L. Velikovich, and J. Davis, Connement and compression of magnetic ux by plasma shells, Plasma Phys., vol. 10, no. 11, pp. 44354447, Nov. 2003. [24] V. L. Kantsyrev, L. I. Rudakov, A. S. Safronova, D. A. Fedin, V. V. Ivanov, A. L. Velikovich, A. A. Esaulov, A. S. Chuvatin, K. Williamson, N. D. Ouart, V. Nalajala, G. Osborne, I. Shrestha, M. F. Yilmaz, S. Pokala, P. Laca, and T. E. Cowan, Planar wire array as powerful radiation source, IEEE Trans. Plasma Sci., vol. 35, no. 5, pp. 22952302, Oct. 2006. [25] D. Mosher, N. Qi, and M. Krishnan, A two-level model for K-line radiation scaling of the imploding Z-pinch plasma radiation source, IEEE Trans. Plasma Sci., vol. 26, no. 3, pp. 10521061, Jun. 1998. [26] J. Abdallah and R. E. H. Clark, Calculated radiated power loss for neon, silicon, argon, titanium, and iron, in Atomic and Plasma-material Interaction Data for Fusion 11. Vienna, Austria: Int. Atomic Energy Agency, 2003. [27] Los Alamos Manual No. LA-11926, Los Alamos Nat. Lab., Los Alamos, NM, 1990. [28] S. L. Jackson, B. V. Weber, D. Mosher, D. G. Phipps, S. J. Stephanakis, R. J. Commisso, N. Qi, B. H. Failor, and P. L. Coleman, A comparison

of planar, laser-induced uorescence, and high-sensitivity interferometry techniques for gas-puff nozzle density measurements, Rev. Sci. Instrum., vol. 79, no. 10, p. 10E717, Oct. 2008. [29] J. W. Thornhill, Y. K. Chong, J. P. Apruzese, J. Davis, R. W Clark, J. L. Giuliani, R. E. Terry, A. L. Velikovich, R. J. Commisso, K. G. Whitney, M. H. Frese, S. D. Frese, J. S. Levine, N. Qi, H. Sze, B. H. Failor, J. Banister, P. L. Coleman, C. A. Coverdale, and B. Jones, One- and two-dimensional modeling of argon K-shell emission from gas-puff Z-pinch plasmas, Phys. Plasmas, vol. 14, no. 6, p. 063301, Jun. 2007.

Niansheng Qi, photograph and biography not available at the time of publication.

B. H. Failor, photograph and biography not available at the time of publication.

J. S. Levine, photograph and biography not available at the time of publication.

J. Goyer, photograph and biography not available at the time of publication.

H. Sze, photograph and biography not available at the time of publication.

A. Verma, photograph and biography not available at the time of publication.

Authorized licensed use limited to: Univ of Calif Los Angeles. Downloaded on May 07,2010 at 18:54:49 UTC from IEEE Xplore. Restrictions apply.

You might also like