A systems model for comparing costs of induction linac drivers for heavy ion fusion and for studying parametric dependencies and sensitivities has been developed. The basic driver architecture in this model is a multi-beam injector feeding an electrostatic transport section, four-to-one beam merging and transition to a magnetic transport section (at an energy to be optimized), drift compression, and final focus with two-sided illumination. The multi-beam array shares common induction cores for acceleration in both the electrostatic and magnet transport sections. The drift compression section is needed to shorten the pulse from the ~100 ns at the end of the accelerator to ~10 ns or less required by the target design (main pulse). We include an option for the high energy portion of the linac (e.g., 0.4 to 4.0 GeV) to be replace by a recirculator in order to compare costs for these two configurations. The code determines accelerator design features (e.g., number and size of quads, core dimensions, accelerator length, efficiency, etc.) and system capital costs as a function of various design variables (e.g., ion mass, charge state, number of beams, initial pulse duration, magnetic field strength, etc.). This integrated model includes constraints on component performance and subsystem interfaces all the way from the ion source to the target. For example, the target performance depends on the total energy delivered by all beams, the range of the ions, and the size of the focal spot. For a given type of ion, the ion range determines the required final ion energy, and total beam energy determines the required total charge. That total charge must be produced by the multi-beam source which is limited in achievable current density (A/cm²), and so we can trade-off initial pulse duration with source area. Beam focusability depends in part on emittance, which is related to the source characteristics, beam size, and the emittance growth in the accelerator. In addition, the target and chamber designs constrain the final focus geometry thus impacting the achievable spot size. It is this type of integrated analysis that is needed to assure that proposed designs are internally self-consistent before cost comparisons are made. The work to be presented includes sensitivity studies, preliminary design optimization results, and comparisons of various driver architectures.

*This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract W-7405-ENG-48.