Recent experiments on the 1 MA, 100 ns Zebra driver at the Nevada Terawatt Facility at the University of Nevada, Reno, investigated the compression of a deuterium target by a high-atomic-number (Ar or Kr) gas-puff liner. Pinch stability improved with axial premagnetization of 1–2 kG observed as a decrease in magneto-Rayleigh-Taylor instability growth. Implosion dynamics and stagnation conditions were studied computationally with the radiation-MHD code MACH2 using initial conditions that approximate those in the experiment. Typical average and peak implosion velocities exceeded 300 and 400 km/s, respectively, which raised the target adiabat by shock heating as the front converges on axis, at which time the target is adiabatically compressed to stagnation. Experimental fusion yields of up to 2 × 109 for Ar liner on D target implosions were measured, while with a Kr liner yields up to 1 × 1010 were measured. Higher yields in Kr compared to Ar were also calculated in 2-D MACH2 simulations. These observations will be further tested with other radiation-MHD codes, and experiments on the 1 MA LTD-III machine at UC San Diego.

Gas puff Z-pinches driven by pulsed power machines are often utilized in the production of high energy density plasmas for a wide variety of applications.1–3 The self-generated magnetic field associated with the high current creates a J × B force which compresses the cylindrical gas shell. In the late 1970s, Fisher and collaborators at the University of California, Irvine, created the first gas puff Z-pinch using a 200-kA pulsed power generator.4 Later, several refinements were made to the Z-pinch, including the development of the multispecies gas-puff Z-pinch and the gas-mixture Z-pinch.5 These enhanced the pinch stability and increased its radiation efficiency. Pre-ionization by e-beam was used to improve the uniformity of the initial-breakdown process in a gas puff Z-pinch6 and increase magnetic flux compression, allowing for the amplification of an axial magnetic field Bz and stabilization of the Z-pinch.7,8 This led to the pinch-on-fiber configuration and to the development of the Staged Z-pinch (SZP)9 as source of thermonuclear neutrons.10–12 

The Staged Z-pinch is a concept in which a high-Z cylindrical plasma shell implodes on a low-Z plasma target. Its key features are shock heating, and control and mitigation of the magneto-Rayleigh-Taylor instability (MRTI), resulting in stable, hot, and dense target plasma, even though the liner plasma becomes unstable.13 Shocks in Z-pinches were first observed as a uniform axial structure in an unmagnetized gas-mixture (90% D2 and 10% Ar, by mass) Z-pinch.5 Cylindrical liner experiments on the Magpie generator also provide strong evidence for radially converging shocks, using a fast, high-resolution axial-imaging system.14 

Staged Z-pinch implosions were conducted on the Zebra driver at University of Nevada, Reno, by current pulses as high as 1 MA with time to maximum current of 100 ns. The fast rise time results in significant acceleration which can lead to magneto-Rayleigh-Taylor instability growth.15 Two common approaches to MRTI mitigation are axial premagnetization, whereby the tension in the axial magnetic field lines smooths out density perturbations,3,13 and density profile tailoring,16,17 which reduces the instability growth by reducing radial acceleration. In this work, we show experimental images demonstrating MRTI mitigation by staging and axial premagnetization.

The remaining sections are organized as follows: In Sec. II, we discuss simulations using the radiation-MHD code MACH218 which illustrate the central role of shock heating in the Staged Z-pinch, in Sec. III we provide an overview of our recent experiments and compare some MACH2 calculated parameters with corresponding measured parameters, and Sec. IV has concluding remarks.

Understanding of the SZP dynamics requires careful modeling with 2-D radiation-MHD codes. Plasma shocks create very steep pressure gradients at the liner/target interface and are best resolved with adaptive spatial grids, for example, with the Arbitrary Lagrangian Eulerian (ALE) method, which generates fine spatial resolution (∼10 μm) at the shock fronts. These codes solve resistive MHD equations for continuity, momentum, energy, magnetic field, and radiation diffusion. MACH2 is a single-fluid radiation-MHD code that can use either analytic or tabular equation of state and transport coefficients. It calculates electron, ion, and radiation temperatures. In lieu of the radiation diffusion equation, a simple volumetric emission model can be used, which is appropriate when the liner and target are optically thin, but this does not capture the transition of the Zebra SZP liner plasma from optically thin to optically thick near stagnation. We use initial gas density profiles inferred from interferometry measurements19 as code input.

Both liner and target initial radial density profiles are well approximated as Gaussians. The target profile has peak density nD = 1 × 1018 cm−3 with FWHM = 0.6 cm. The liner profiles have peak density nAr = 3.1 × 1016 cm−3 for Ar, and nKr = 1.5 × 1016 cm−3 for Kr, both at r = 1.28 cm, and with FWHM = 0.47 cm. The Zebra pulse generator is modeled as an RL circuit driven by a voltage waveform. An initial uniform seed magnetic field B0z = 1.5 kG is applied in all MACH2 simulations discussed in this section.

One method for visualizing the target dynamics is by plotting the axially averaged total plasma pressure (Ptot = Pi + Pe) as a function of time and radius. Typical SZP dynamics near peak compression are shown in Fig. 1 where the shocked target ions are in the orange (∼10 kbar) band, surrounded by the compressing liner. Electrons are free to conduct thermally ahead of the shock front,20 hence the rising plasma pressure on axis before shock convergence. The final liner radius is ∼0.7 mm, corresponding to a liner convergence ratio (CR) of ∼30, and the minimum target radius is ∼0.11 mm, corresponding to a maximum target convergence ratio of ∼80.

FIG. 1.

Total plasma pressure vs time near stagnation for Ar and Kr plasma shells compressing D target.

FIG. 1.

Total plasma pressure vs time near stagnation for Ar and Kr plasma shells compressing D target.

Close modal

We used this method in a 2016 computational study21 to visualize the shock formation in Ne, Ar, Kr, and Xe on deuterium gas puff implosions in Zebra. Although the physics model in that study was identical to the current physics model, the initial conditions were significantly different resulting in different implosion dynamics. For example, a very thin liner profile (FWHM = 0.32 mm) was used in the previous paper in order to allow significant Bθ diffusion. The resulting shocks, launched by the internal magnetic pressure, were detached from the liner throughout the implosion. Conversely, in this paper, we use a much wider liner (FWHM = 4.7 mm) to match the experimental initial conditions; therefore, the effect of the internal J × B force is negligible and the entire load is essentially snowplowed.

Another method is by plotting the axially averaged ion temperature close to the stagnation time. This is illustrated in Fig. 2 where ion temperature at five different times in the 93–108 ns range is shown. For the Kr liner, 30 eV temperature rise due to shock heating is clearly seen on the leading edge of the ion temperature profile at 93 ns; for the Ar liner, this happens 5 ns later, at 98 ns. Further temperature increase due to shock heating continues until ∼107 ns, for both Ar and Kr liners, at which time the adiabatic compression becomes dominant. Both the width and height of the Ti spike at the liner/target boundary is larger for the krypton case, suggesting more effective shock heating than for the Ar case.

FIG. 2.

Axially averaged ion temperature at five different times near stagnation showing shock heating at the liner-target boundary.

FIG. 2.

Axially averaged ion temperature at five different times near stagnation showing shock heating at the liner-target boundary.

Close modal

We plan to investigate details of shock heating in Staged Z-pinch with other radiation-MHD codes, such as HYDRA22 and TRAC-II,23 and will report on those results in a future publication. MACH2 is using a single-group radiation diffusion equation and is likely inferior to multigroup radiation diffusion models (e.g., like that used in HYDRA). Its equation of state, opacity, and transport coefficients are only as accurate as the SESAME tables or simple analytical models (e.g., magnetized Spitzer-Braginskii transport coefficients). For these reasons, we think that replicating the MACH2 results with different MHD codes would solidify our interpretation of the neutron yield difference between Ar and Kr. All these codes, being radiation-MHD models, cannot account for kinetic effects like the generation of beam-target neutrons, which could be an important neutron producing mechanism, especially when no seed magnetic field is used. Recent fully kinetic simulations24 have shown that beam-target effects dominate the yield in deuterium shell on deuterium target gas-puff z-pinch experiments below 3 MA. However, we want to point out that MACH2 simulations, and imaging of staged Z-pinch plasmas, indicate that the inner liner boundary is relatively stable when a seed magnetic field is applied, in spite of the outer liner boundary being heavily deformed due to the magneto-Rayleigh-Taylor instability, suggesting that the beam-target neutron production is likely a second order effect.

Plots of the mass-averaged target temperature vs target convergence ratio CR = Rinitial/R(t) illustrate the effectiveness of shock heating in raising the target adiabat. In lossless, cylindrical adiabatic compression of a monatomic ideal gas, temperature increases as CR4∕3. Shock heating or ohmic heating can cause the target to heat super-adiabatically, whereas radiative and conductive losses can lead to subadiabatic heating.

Trajectories of mass averaged target ion temperature vs target convergence, for Ar and Kr liners, and their corresponding time evolution are shown in Fig. 3. The implosion can be divided into two major phases: Prominent shock heating (CR3) and adiabatic heating (CR30), with an intermediary phase where the shock heating is competing with the thermal conduction and radiation losses. The shock heating ends when the shock front reaches the axis, at t ∼ 107 ns.

FIG. 3.

Mass-averaged target ion temperature vs target convergence ratio (left panel), and vs time (right panel), for typical Ar and Kr implosions on Zebra simulated with the 2-D MACH2 code. Prominent super-adiabatic heating occurs until ∼100 ns, when the convergence ratio CR3.

FIG. 3.

Mass-averaged target ion temperature vs target convergence ratio (left panel), and vs time (right panel), for typical Ar and Kr implosions on Zebra simulated with the 2-D MACH2 code. Prominent super-adiabatic heating occurs until ∼100 ns, when the convergence ratio CR3.

Close modal

The SZP shock behavior is further clarified by examining the total plasma pressure as a (R-z) contour plot, shown in Figs. 4(a) and 4(c). The radial profiles for the plasma-implosion velocity vR, the Alfven velocity vA, and the sound velocity cs are shown in Figs. 4(b) and 4(d). All profiles are calculated at t = 106 ns. These profiles clearly show that the liner implodes both supersonically and super-Alfvenically, with magnetosonic Mach number M=vR/(cS2+vA2)3.

FIG. 4.

(R-z) total plasma pressure contour plots, and radial (vR), Alfven (vA), and sound (cS) velocity profiles, calculated at t = 106 ns. In the liner region, vR > vA > cS.

FIG. 4.

(R-z) total plasma pressure contour plots, and radial (vR), Alfven (vA), and sound (cS) velocity profiles, calculated at t = 106 ns. In the liner region, vR > vA > cS.

Close modal

Finally, (R-z) contour plots of the total plasma pressure, overlaid with 1-D plots of the axial and azimuthal magnetic fields, during implosion at t = 106 ns (top panel), and at stagnation (108.3 ns for Ar, 108.2 ns for Kr; bottom panel), are shown in Fig. 5. The axial magnetic field is compressed as in a snow-plow model rather than in a self-similar way. At stagnation, the target is fully compressed reaching ion density of ∼5 × 1020 cm−3 and axial magnetic field above 10 MG (Table I).

FIG. 5.

Total (electron + ion) pressure and axially averaged magnetic fields Bz and Bθ at 106 ns (top plots), and at peak compression: 108.3 ns for Ar, and 108.2 ns for Kr (bottom plots). Note the stability of the shock front at 106 ns (innermost pressure gradient) in both cases, and the axial magnetic flux compression occurrence only in the liner and shocked target, not ahead of the shock front.

FIG. 5.

Total (electron + ion) pressure and axially averaged magnetic fields Bz and Bθ at 106 ns (top plots), and at peak compression: 108.3 ns for Ar, and 108.2 ns for Kr (bottom plots). Note the stability of the shock front at 106 ns (innermost pressure gradient) in both cases, and the axial magnetic flux compression occurrence only in the liner and shocked target, not ahead of the shock front.

Close modal
TABLE I.

Calculated SZP parameters for Ar/D and Kr/D implosions (B0z = 1.5 kG) from MACH2 simulations.

QuantityAr linerKr liner
Initial peak liner Ni (cm−33.1 × 1016 1.5 × 1016 
τstagn (ns) 108.3 108.2 
Peak Vr (km/s) 580 660 
Avg final target R (μm) 120 150 
Avg final liner R (μm) 660 610 
Peak Bz (MG) 14.3 12.9 
Peak Bθ (MG) 1.8 1.8 
Avg final target Ni (cm−35.1 × 1020 3.2 × 1020 
Peak target Ti (keV) 7.2 7.4 
Avg final target Ti (keV) 1.3 4.1 
Neutron yield YDD 1.3 × 109 1.2 × 1010 
QuantityAr linerKr liner
Initial peak liner Ni (cm−33.1 × 1016 1.5 × 1016 
τstagn (ns) 108.3 108.2 
Peak Vr (km/s) 580 660 
Avg final target R (μm) 120 150 
Avg final liner R (μm) 660 610 
Peak Bz (MG) 14.3 12.9 
Peak Bθ (MG) 1.8 1.8 
Avg final target Ni (cm−35.1 × 1020 3.2 × 1020 
Peak target Ti (keV) 7.2 7.4 
Avg final target Ti (keV) 1.3 4.1 
Neutron yield YDD 1.3 × 109 1.2 × 1010 

The modeling results from Sec. II, suggesting more efficient shock heating with increasing liner atomic number, motivated a series of experiments on the Zebra machine at the Nevada Terawatt Facility (NTF). We also experimentally studied the effect of a seed axial magnetic field Bz0. The liner and target gas initial conditions are inferred from interpolating multiple-point 1-D interferometry data of areal density vs time.19 Adjustments to plenum pressure scale approximately linearly with mass-per-unit-length (M/L), and adjustments to valve opening times affect M/L and axial density gradients. A typical implosion has liner mass ∼10–12 μg/cm and target mass ∼1 μg/cm, resulting in stagnation time of ∼120 ns.

Measured and MACH2 calculated currents and radiation pulses are compared in Fig. 6. Shot 4330 was from one of our earlier campaigns, where a lower target density and higher Ar liner density were used (∼5 × 1016 D2/cm3 peak target density and ∼7 × 1016 cm3 peak Ar density), resulting in later implosion time.

FIG. 6.

Experimentally measured current profile (solid line) with simulated current profiles (dotted line) together with measured X-ray pulse and simulated radiation energy from MACH2.

FIG. 6.

Experimentally measured current profile (solid line) with simulated current profiles (dotted line) together with measured X-ray pulse and simulated radiation energy from MACH2.

Close modal

The measured and calculated current waveforms are normalized to 10 for easier comparison, and their peak values are annotated in appropriate color. The measured current is obtained by integrating a pair of differential B-dot signals. The X-ray signals are detected with 1 mm2 Si photodiodes covered with 12 μm Ti and 2 μm Ni filters, and the MACH2 radiation energy plot is normalized to the Ni-filtered signal. The shapes and peak values of the simulated and experimentally measured current profiles are in good agreement, while the simulated X-ray radiation pulse is narrower than the experimentally measured ones, possibly due to the lack of pinch zippering modeling in the code. The second observed radiation peak is likely due to the acceleration of a charged particle beam when the plasma column breaks down after stagnation; upon hitting the metal electrodes near the plasma, these particles are expected to produce a high amount of X-rays.

The pinch stability was experimentally observed with extreme ultraviolet (XUV) and optical streak imaging. We deployed a gated XUV pinhole camera which recorded four images in 10 ns intervals, with 150 μm spatial resolution and 4 ns exposure time. A comparison of four sets of these images for Ar/D and Kr/D, without Bz0 and with Bz0 = 1.5 kG, is shown in Figs. 7 and 8, respectively. Argon shots 4972 (Bz0 = 0) and 5049 (Bz0= 1.5 kG) had the same nominal initial conditions, i.e., identical gas-puff delay times and 5.0 psia plenum pressure. Krypton shots 4981 (Bz0 = 0) and 5020 (Bz0 = 1.5 kG) had identical delay times and slightly different pressures (5.5 psia and 5.0 psia, respectively), resulting in approximately 10% mass difference. The times of the X-ray peaks were: 107 ± 3 ns for shot 4972, 118 ± 8 ns for shot 4981, 118 ± 3 ns for shot 5049, and 115 ± 3 ns for shot 5020.

FIG. 7.

XUV images of shots 4972 and 4981 using Ar and Kr plasma shells, respectively, imploding on D target plasma with no axial premagnetization. The MRTI is evident, particularly in the last ∼20 ns of the implosion as the load converges on axis.

FIG. 7.

XUV images of shots 4972 and 4981 using Ar and Kr plasma shells, respectively, imploding on D target plasma with no axial premagnetization. The MRTI is evident, particularly in the last ∼20 ns of the implosion as the load converges on axis.

Close modal
FIG. 8.

XUV images of shots 5049 and 5020 using Ar and Kr plasma shells, respectively, imploding on D target plasma with initial axial magnetic field of 1.5 kG. MRTI growth is significantly mitigated, resulting in a more stable plasma column through peak compression.

FIG. 8.

XUV images of shots 5049 and 5020 using Ar and Kr plasma shells, respectively, imploding on D target plasma with initial axial magnetic field of 1.5 kG. MRTI growth is significantly mitigated, resulting in a more stable plasma column through peak compression.

Close modal

Bright vertical columns in the interior of the pinch indicate presence of localized liner radiation, perhaps a signature of the target/liner interface. We want to point out the stability of this inner structure relative to the MRT-unstable outer surface. Significant MRT instability mitigation is observed when an axial magnetic field of 1.5 kG is applied to either Ar/D or Kr/D implosions, as shown in Fig. 8.

Similar stable inner surface structure is also suggested by laser shadowgraphy images of the imploding pinch, shown in Fig. 9 (left side), recorded at 95 ns and 105 ns during Ar/D compression. Shadowgraphs for Kr/D compression are shown on the right side of Fig. 9.

FIG. 9.

Laser shadowgraphy images of Staged Z-pinch of Ar and Kr shell imploding on the D plasma near the peak implosion.

FIG. 9.

Laser shadowgraphy images of Staged Z-pinch of Ar and Kr shell imploding on the D plasma near the peak implosion.

Close modal

Streak image analysis can produce the radial pinch implosion velocity and the minimum outer pinch radius. The plasma boundary is determined by scanning each line in the image, and looking for the left- and right-most points where the image intensity crosses a threshold equal to 1/3 of the peak intensity (only if this value is bigger than the intensity noise in the image). The left panels in Fig. 10 show streak images for two experiments where a high-Z liner imploded on a D target (vertical axis is time, and horizontal axis is radius). The top frame shows an Ar/D experiment (shot 5246), whereas the bottom frame shows a Kr/D experiment (shot 5306). Shot 5246 had liner mass/length of 11 μg/cm at a 5-mm axial distance from the gas injection exit plane (which coincides with the Zebra machine anode), with Bz0 = 1.5 kG. Shot 5306 had liner mass/length of 10 μg/cm at a 5-mm axial distance from the gas exit plane, with Bz0 = 1.0 kG. The target density for both shots was 1.2 × 1018 cm−3 D2 at a 5-mm axial distance from the gas exit plane.

FIG. 10.

Streak images (left) showing the imploding plasma for shot 5246 Ar liner on D target (top) and shot 5306 Kr liner on D target (bottom), and the corresponding approximate implosion velocity and radius (right) vs time. Both shots used a 240 μm nozzle. Shot 5246 had liner mass/length of 11 μg/cm at a 5-mm axial distance from the exit plane, with Bz0 = 1.5 kG. Shot 5306 had liner mass/length of 10 μg/cm at a 5-mm axial distance from the exit plane, with Bz0 = 1.0 kG. Target density for both shots was 1.2 × 1018 cm−3 D2 at a 5-mm axial distance from the exit plane. v refers to average implosion velocity, as estimated from the streak images. The MACH2 radii were calculated from the axially averaged peak liner density.

FIG. 10.

Streak images (left) showing the imploding plasma for shot 5246 Ar liner on D target (top) and shot 5306 Kr liner on D target (bottom), and the corresponding approximate implosion velocity and radius (right) vs time. Both shots used a 240 μm nozzle. Shot 5246 had liner mass/length of 11 μg/cm at a 5-mm axial distance from the exit plane, with Bz0 = 1.5 kG. Shot 5306 had liner mass/length of 10 μg/cm at a 5-mm axial distance from the exit plane, with Bz0 = 1.0 kG. Target density for both shots was 1.2 × 1018 cm−3 D2 at a 5-mm axial distance from the exit plane. v refers to average implosion velocity, as estimated from the streak images. The MACH2 radii were calculated from the axially averaged peak liner density.

Close modal

From these images, the outer pinch radius and the radial implosion velocity can be measured, as shown on the right panels in Fig. 10. The implosion velocities calculated for these shots, averaged over the time window where the outer radius is between 2 mm and 7 mm (for consistency among different shots), are 309 km/s and 471 km/s, respectively. The Ar pinch reaches terminal velocities up to ∼400 km/s, while the Kr pinch has higher terminal velocity and tighter final radius.

MACH2 calculated radii from simulations reflecting the experimental conditions for these two shots are superimposed. They are obtained from the axially averaged peak liner density and excellent agreement is seen, except for the last few ns before stagnation, where the MACH2 radius is smaller than the radius measured from streak images. This is most likely due to the very intense visible light emission at stagnation that can cause an overestimate of the radius in the streak image. In addition, the axial average of radial locations of peak liner density underestimates the pinch radius close to stagnation because at that time the outer liner boundary is heavily affected by the Rayleigh-Taylor instability, i.e., no longer well defined.

Direct measurement of the target temperature and density is very difficult in high energy density plasmas, so assessment of the final SZP target conditions is usually based on the neutron yield data. Counts from a silver activation detector, for a set of three Ar and three Kr shots, with applied seed magnetic field, are shown in Fig. 11; the corresponding yields are given in Table II and their uncertainty is ±20%. The highest measured Kr yield (shot 5045) is in good agreement with the corresponding MACH2 calculated yield in Table I; the average of the three Ar shots is also in good agreement with the MACH2 calculated Ar yield.

FIG. 11.

Silver activation detector counts for Ar (dotted lines) and Kr (solid lines) plasma shells imploding on D target plasma. Bz0 = 1.0 kG for the Kr shots, and Bz0 = 1.5 kG for the Ar shots. Background counts in the ∼180 range have been subtracted.

FIG. 11.

Silver activation detector counts for Ar (dotted lines) and Kr (solid lines) plasma shells imploding on D target plasma. Bz0 = 1.0 kG for the Kr shots, and Bz0 = 1.5 kG for the Ar shots. Background counts in the ∼180 range have been subtracted.

Close modal
TABLE II.

Neutron yields for shots shown in Fig. 11.

Shot #496150485049503950435045
Yield (×1091.2 2.1 1.5 4.5 2.6 9.1 
Shot #496150485049503950435045
Yield (×1091.2 2.1 1.5 4.5 2.6 9.1 

Most of the shots presented in this paper are from our pre-2018 experimental campaigns. They had limited number of shots without seed magnetic field because we thought that by stabilizing the pinch the yield will increase. The fact that some Bz0 = 0 shots had low yields (for example, YDD = 0.3 × 109 for the Ar shot 4972, and YDD = 1.2 × 109 for the Kr shot 4981 in Fig. 7) does not necessarily mean that our hypothesis was correct. In July 2018, a dedicated experiment using Kr liner was undertaken, with multiple repeat shots for a given shot condition. The Bz0 = 0 shots had twice as large yield compared to the yield from shots with Bz0 ≥ 0.5 kG. Detailed analysis of those shots, including the neutron emission isotropy deduced from three neutron time-of-flight detectors, will be presented in a separate paper.

2-D MACH2 SZP simulations without seed magnetic field predict yield of 5.2 × 1010 for Kr liner and 2.7 × 1010 for Ar liner. These large yield predictions may be due to mismatch between the level of perturbation assumed in the simulation and the growth of the MRT instability in the experiment; perturbation levels above 1% terminate the simulation before the peak neutron producing phase. Simulations with seed magnetic field suggest better control of the MRT instability which leads to better neutron yield predictions.

Experiments using the NTF Zebra driver at the University of Nevada, Reno, showed evidence of formation of uniform and relatively stable pinches when a deuterium plasma column was compressed by a hollow plasma shell made of a high-Z gas, such as Ar or Kr. The pinch stability was improved with an external seed axial magnetic field Bz0 = 1–2 kG. Pinhole camera XUV images suggest that the inner structures remain relatively stable. Neutron yields up to 2 × 109 for Ar shots, and up to 1010 for Kr shots, were observed.

Radiation-MHD simulations using the MACH2 code revealed details of the SZP compression dynamics, in particular, the importance of shock heating during the implosion stages, contributing to significant preheating at the liner-target boundary. In these simulations, the preheating is higher for Kr than for Ar, resulting in higher peak temperatures at stagnation and, consequently, higher neutron yields. Further computational studies of the Staged Z-pinch with other radiation-MHD codes, and experiments on the 1 MA LTD-III machine at UC San Diego, are planned in the near future.

The LTD-III machine is based on the new linear transformer driver technology.25 Compared to the older Marx generator with pulse-forming lines architecture for a similar current and risetime, LTDs have a much smaller footprint, much lower stored energy requirements, and improved coupling to the load; in addition, the anode-cathode voltage in LTD-III will be several times lower than the ∼1 MV voltage in Zebra. These advantages will enable us to perform SZP experiments with a higher shot rate and more extended machine time than on Zebra, allowing higher statistics, more diverse configurations, and reduced risk of damaging components. We will also use a more complete set of diagnostics to better measure the pinch conditions.

Funding for this work was provided by ARPA-E, Grant No. DE-AR0000569, and by Strong Atomics.

1.
D. D.
Ryutov
,
M. S.
Derzon
, and
M. K.
Matzen
,
Rev. Mod. Phys.
72
,
167
(
2000
).
2.
M. G.
Haines
,
Plasma Phys. Controlled Fusion
53
,
093001
(
2011
).
3.
J. L.
Giuliani
and
R. J.
Commisso
,
IEEE Trans. Plasma Sci.
43
,
2385
(
2015
).
4.
J.
Shiloh
,
A.
Fisher
, and
N.
Rostoker
,
Phys. Rev. Lett.
40
,
515
(
1978
).
5.
J.
Bailey
,
Y.
Ettinger
,
A.
Fisher
, and
N.
Rostoker
,
Appl. Phys. Lett.
40
,
460
(
1982
).
6.
E.
Ruden
,
H. U.
Rahman
,
A.
Fisher
, and
N.
Rostoker
,
J. Appl. Phys.
61
,
1311
(
1987
).
7.
F. J.
Wessel
,
F. S.
Felber
,
N. C.
Wild
, and
H. U.
Rahman
,
Appl. Phys. Lett.
48
,
1119
(
1986
).
8.
F. S.
Felber
,
F. J.
Wessel
,
N. C.
Wild
,
H. U.
Rahman
,
A.
Fisher
,
C. M.
Fowler
,
M. A.
Liberman
, and
A. L.
Velikovich
,
J. Appl. Phys.
64
,
3831
(
1988
).
9.
H. U.
Rahman
,
F. J.
Wessel
, and
N.
Rostoker
,
Phys. Rev. Lett.
74
,
714
(
1995
).
10.
P.
Ney
,
H. U.
Rahman
,
F. J.
Wessel
, and
N.
Rostoker
,
Phys. Plasmas
8
,
616
(
2001
).
11.
H. U.
Rahman
,
F. J.
Wessel
,
N.
Rostoker
, and
P.
Ney
,
J. Plasma Phys.
75
,
749
(
2009
).
12.
H. U.
Rahman
,
P.
Ney
,
N.
Rostoker
, and
F. J.
Wessel
,
Astrophys. Space Sci.
323
,
51
(
2009
).
13.
H. U.
Rahman
,
P.
Ney
,
N.
Rostoker
,
A.
Van Drie
, and
F. J.
Wessel
,
Phys. Plasmas
11
,
5595
(
2004
).
14.
G.
Burdiak
,
S.
Lebedev
,
R.
Drake
,
A.
Harvey-Thompson
,
G.
Swadling
,
F.
Suzuki-Vidal
,
J.
Skidmore
,
L.
Suttle
,
E.
Khoory
,
L.
Pickworth
,
P.
de Grouchy
,
G.
Hall
,
S.
Bland
,
M.
Weinwurm
, and
J.
Chittenden
,
High Energy Density Phys.
9
,
52
(
2013
).
15.
E. G.
Harris
,
Phys. Fluids
5
,
1057
(
1962
).
16.
A. L.
Velikovich
,
F. L.
Cochran
, and
J.
Davis
,
Phys. Rev. Lett.
77
,
853
(
1996
).
17.
N.
Qi
,
E. W.
Rosenberg
,
P. A.
Gourdain
,
P. W. L.
de Grouchy
,
B. R.
Kusse
,
D. A.
Hammer
,
K. S.
Bell
,
T. A.
Shelkovenko
,
W. M.
Potter
,
L.
Atoyan
,
A. D.
Cahill
,
M.
Evans
,
J. B.
Greenly
,
C. L.
Hoyt
,
S. A.
Pikuz
,
P. C.
Schrafel
,
E.
Kroupp
,
A.
Fisher
, and
Y.
Maron
,
Phys. Plasmas
21
,
112702
(
2014
).
18.
M. H.
Frese
, “
MACH2: A two-dimensional magnetohydrodynamic simulation code for complex experimental configurations,” Technical Report No. AMRC-R-874
(Mission Research Corporation,
1987
).
19.
F.
Conti
,
J. C.
Valenzuela
,
N.
Aybar
,
F. J.
Wessel
,
M. P.
Ross
,
J.
Narkis
,
H. U.
Rahman
,
E.
Ruskov
, and
F. N.
Beg
,
IEEE Trans. Plasma Sci.
46
,
3855
(
2018
).
20.
Y. B.
Zel'dovich
and
Y. P.
Raizer
, in
Physics of Shock Waves and High-Temperature Hydrodynamic Phenomena
, edited by
W. D.
Hayes
and
R. F.
Probstein
(
Academic Press
,
New York
,
1967
), pp.
515
522
.
21.
J.
Narkis
,
H. U.
Rahman
,
P.
Ney
,
M. P.
Desjarlais
,
F. J.
Wessel
,
F.
Conti
,
J. C.
Valenzuela
, and
F. N.
Beg
,
Phys. Plasmas
23
,
122706
(
2016
).
22.
M. M.
Marinak
,
G. D.
Kerbel
,
N. A.
Gentile
,
O.
Jones
,
D.
Munro
,
S.
Pollaine
,
T. R.
Dittrich
, and
S. W.
Haan
,
Phys. Plasmas
8
,
2275
(
2001
).
23.
D.
Reisman
,
J.
Hammer
,
A.
Toor
, and
J. D.
Groot
,
Laser Part. Beams
19
,
403
(
2001
).
24.
D. T.
Offermann
,
D. R.
Welch
,
D. V.
Rose
,
C.
Thoma
,
R. E.
Clark
,
C. B.
Mostrom
,
A. E. W.
Schmidt
, and
A. J.
Link
,
Phys. Rev. Lett.
116
,
195001
(
2016
).
25.
R. D.
McBride
,
W. A.
Stygar
,
M. E.
Cuneo
,
D. B.
Sinars
,
M. G.
Mazarakis
,
J. J.
Leckbee
,
M. E.
Savage
,
B. T.
Hutsel
,
J. D.
Douglass
,
M. L.
Kiefer
,
B. V.
Oliver
,
G. R.
Laity
,
M. R.
Gomez
,
D. A.
Yager-Elorriaga
,
S. G.
Patel
,
B. M.
Kovalchuk
,
A. A.
Kim
,
P.-A.
Gourdain
,
S. N.
Bland
,
S.
Portillo
,
S. C.
Bott-Suzuki
,
F. N.
Beg
,
Y.
Maron
,
R. B.
Spielman
,
D. V.
Rose
,
D. R.
Welch
,
J. C.
Zier
,
J. W.
Schumer
,
J. B.
Greenly
,
A. M.
Covington
,
A. M.
Steiner
,
P. C.
Campbell
,
S. M.
Miller
,
J. M.
Woolstrum
,
N. B.
Ramey
,
A. P.
Shah
,
B. J.
Sporer
,
N. M.
Jordan
,
Y. Y.
Lau
, and
R. M.
Gilgenbach
,
IEEE Trans. Plasma Sci.
46
,
3928
(
2018
).