3324 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 40, NO. 12, DECEMBER 2012 Experimental Analysis of the Acceleration Region in Tungsten Wire Arrays Simon C. Bott, Member, IEEE, Derek Mariscal, Kanchana Gunasekera, Jonathan Peebles, Farhat N. Beg, David A. Hammer, Fellow, IEEE, B. R. Kusse, J. B. Greenly, T. A. Shelkovenko, S. A. Pikuz, I. C. Blesener, Ryan D. McBride, Member, IEEE, J. D. Douglass, K. S. Blesener, and P. F. Knapp Abstract—We present the first analysis of the ablated plasma flow acceleration region in tungsten cylindrical wire arrays within 1 mm of the wire core. We apply a recently developed modification to the Lebedev rocket model to infer the 2-D distribution of effective velocities which redistribute the array mass as a function of time. From these data, it is possible to directly observe the accel- eration region in a wire array. Analysis of radiography data from the 1-MA Cornell Beam Research Accelerator machine suggests a region of rapid acceleration extending up to 300μm from the wire core in 16 wire tungsten arrays. Index Terms—Precursor plasma, wire array Z-pinch. I. I NTRODUCTION W HILST the dynamical evolution of wire arrays is well understood [1]–[4] and multidimensional magnetohy- drodynamic (MHD) modeling has demonstrated significant progress [5]–[8], a comprehensive predictive capability has not been realized to date. Experimental investigations have contin- ued to highlight the need to more closely examine the ablation structure and its dependence on the initial parameters of the array. In particular, the range over which the ablated plasma is accelerated, and hence the extent to which magnetic flux is convected into the array, is often a disputed point in the comparison simulation and analytical work [9]–[11]. Recent work at the University of California at San Diego [12] examined interferometer data taken for nonimploding arrays on the 250-kA GenASIS device [13] for four-wire Al and Manuscript received May 1, 2012; revised July 3, 2012 and July 26, 2012; accepted August 5, 2012. Date of publication September 4, 2012; date of current version December 7, 2012. This work was supported in part by the NSF/DoE Partnership in Basic Plasma Science under Contracts NSF-PHY- 0903876 and DE-SC-0001992 and in part by a grant from the Center of Excellence for Pulsed-Power-Driven High-Energy-Density Physics, Cornell University, through DOE Cooperative Agreement DE-FC03-02NA00057. S. C. Bott, D. Mariscal, K. Gunasekera, J. Peebles, and F. N. Beg are with the Center for Energy Research, University of California at San Diego, La Jolla, CA 92093 USA (e-mail: sbott@ucsd.edu; dmarisca@ucsd.edu; kgunasek@ucsd.edu; jpeebles@ucsd.edu; fbeg@ucsd.edu). D. A. Hammer, B. R. Kusse, J. B. Greenly, T. A. Shelkovenko, S. A. Pikuz, I. C. Blesener, and K. S. Blesener are with the Cen- ter for Pulsed-Power-Driven High-Energy-Density Plasmas, Cornell Univer- sity, Ithaca, NY 14850 USA (e-mail: dah5@cornell.edu; brk2@cornell.edu; jbg2@cornell.edu; tchel55@mail.ru; pikuz@yahoo.com; icb3@cornell.edu; ksblesener@gmail.com). R. D. McBride and P. F. Knapp are with Sandia National Laboratories, Albu- querque, NM 87123 USA (e-mail: rdm27@cornell.edu; pfknapp@gmail.com). J. D. Douglass is with the Tri-Alpha Energy Inc., Rancho Santa Margarita, CA 92688 USA (e-mail: jdd32@cornell.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPS.2012.2213307 W arrays. Two-dimensional areal electron density images al- lowed analysis of flare structures at various radial positions. The average density contrast, i.e., the average ratio of the peak flare density to the minimum density between flares, was ∼2 for W and ∼1.3 for Al, in line with previous studies [14], [15]. This work led to a modification of the standard rocket model [16] which uses an additional sinusoidal axial variation of the ablation velocity to describe both the axial and radial density variations. The form of the velocity is given in (1). Here, the y-offset is the density obtained from the standard rocket model calculated as the mean of V abl,1 and V abl,2 , λ abl is the average flare wavelength, and the amplitude is the range of effective ablation velocities needed to describe the data. This is then substituted into the standard rocket model for the radial density profile V abl (z)=  V abl,1 −V abl,2 2  sin  2πz λ abl  +  V abl,1 +V abl,2 2  (1) ρ mod (r, z)=  μ 0 8π 2 R 0 rV 2 abl (z)  I  t −  R 0 −r V abl (z)  2 (2) where R 0 is the initial array radius. An advantage of this method is that an automated fit routine can be constructed using (1) and (2), which can then be used to assess the variations of the fit parameters required to match data from calibrated radiographs. The remainder of this paper presents the first results from such a process. II. ANALYSIS OF THE FLOW ACCELERATION REGION A previous study at the MAGPIE facility examined inverse wire-array systems using interferometry [14]. Comparison of the experimental data to analytical theory and numerical sim- ulations suggested that, in an aluminum array, the acceleration region extended ∼1.8 mm from the wire. Here, we examine 16 × 13 μm tungsten arrays on the 1-MA 100-ns Cornell Beam Research Accelerator (COBRA) generator [17] at Cornell Uni- versity using X-pinch-based radiography. We make use of the COBRA-STAR [18] system which allows up to five calibrated radiograph frames per experiment with < 5-μm spatial reso- lution and ∼1-ns temporal resolution. An example image is shown in Fig. 1. 0093-3813/$31.00 © 2012 IEEE