Design and performance of the Z magnetically-insulated transmission lines

We have designed and tested a 10-nH 1.5-m-radius vacuum section for the Z accelerator. The vacuum section consists of four vacuum flares, four conical 1.3-m-radius magnetically-insulated transmission lines, a 7.6-cm-radius 12-post double-post-hole convolute which connects the four outer MITLs in parallel, and a 5-cm-long inner MITL which connects the output of the convolute to a z-pinch load. IVORY and ELECTRO calculations were performed to minimize the inductance of the vacuum flares with the constraint that there be no significant electron emission from the insulator-stack grading rings. Iterative TLCODE calculations were performed to minimize the inductance of the outer MITLs with the constraint that the MITL electron-flow-current fraction be less than or equal to 7% at peak current. The TLCODE simulations assume a 2.5 cm/mu s MITL-cathode-plasma expansion velocity. The design limits the electron dose to the outer-MITL anodes to 50 J/g to prevent the formation of an anode plasma. The TLCODE results were confirmed by SCREAMER TRIFL, TWOQUICK, IVORY, and LASNEX simulations. For the TLCODE, SCREAMER, and TRIFL calculations, we assume that after magnetic insulation is established, the electron-flow current launched in the outer MITLs is lost at the convolute. This assumption has been validated by 3-D QUICKSILVER simulations for load impedances less than or equal to 0.36 ohms. LASNEX calculations suggest that the ohmic resistance of the pinch and conduction-current-induced energy loss to the MITL electrodes can be neglected in Z-power-flow modeling that is accurate to first order. To date, the Z vacuum section has been tested on 100 shots. We have demonstrated we can deliver a 100-ns rise-time 20-MA current pulse to the baseline z-pinch load. We have produced a 1.9-MJ x-ray yield; the project goal was 1.5 MJ. We can reproduce the peak MITL current to within +/-1.6%. Power-flow measurements indicate the vacuum section performs as expected until peak current. Afterward, measurements and simulation results diverge. TLCODE calculations indicate elimination of this discrepancy may increase by 20% the kinetic energy delivered to the pinch.