The CEPTRE code system consists of a parallel implementation of the second-order form of the Boltzmann transport equation. Work is underway to implement a parallel first-order solver as well. We examine the properties of both solution approaches, discussing their theoretical asymptotic runtime performance as a function of various scaling parameters. We also provide results for both the first- and second-order implementations. Understanding of the behavioral differences of these two forms can help guide the selection of an appropriate solver for a given problem. CEPTRE (Coupled Electron-Photon Transport for Radiation Effects) is a deterministic code for solving the linear steady-state Boltzmann transport equation. The current implementation solves the second-order form of the transport equation by means of the multigroup energy discretization, the discrete ordinates angular discretization, and continuous finite element spatial discretization on unstructured meshes. A unique feature in the solution method of CEPTRE is that it solves the spatial and angular dependence simultaneously without the conventional inner iteration. This implementation leads to fast convergence rates for electron transport and good parallel efficiency. CEPTRE's primary application is to predict the effects of x-rays and secondary electrons on cables and other electronic components. The high resolution needed for the modeling of electron boundary layers (sub-micron) often requires the use of large meshes and massively parallel computations. We are currently implementing a parallel solver for the first-order form of the transport equation. The iterative solution process for the first-order form is fundamentally different than that for the second-order form, resulting in a different set of strengths and weaknesses for each approach. It is these differences that we wish to exploit when choosing between the two approaches for a particular application. This can be particularly useful for problems with coupled particle species or a single species with different behaviors at different energies; one can create a hybrid solver that employs both transport forms for different particles/energies in order to minimize the overall runtime.

We present a transport-based method for electrons that incorporates the correct transport mechanics and is computationally efficient for implementation in single event Monte Carlo codes. The method yields accurate dose profiles across a broad range of energies in heterogeneous media and presents a viable alternative to the established condensed history method. Our approach is mathematically rigorous, building on higher order Fokker-Planck and Boltzmann Fokker-Planck representations of the scattering process, and we accordingly refer to it as a Generalized Boltzmann Fokker-Planck (GBFP) approach. We postulate the existence of single collision scattering distributions (differential cross sections) and impose the constraint that the first few momentum transfer moments be identical to corresponding analog values. Details of specific moment-preserving strategies are described. Results are presented for dose in heterogeneous media due to a pencil beam of monoenergetic electrons. The computational efficiency of our GBFP formulations are contrasted against two different condensed history implementations.

A novel approach is proposed for charged particle transport calculations using a recently developed second-order, self-adjoint angular flux (SAAF) form of the Boltzmann transport equation with continuous slowing-down. A finite element discretization that is linear continuous in space and linear discontinuous (LD) in energy is described and implemented in a one-dimensional, planar geometry, multigroup, discrete ordinates code for charged particle transport. The cross-section generating code CEPXS is used to generate the electron and photon transport cross sections employed in this code. The discrete ordinates SAAF transport equation is solved using source iteration in conjunction with an inner iteration acceleration scheme and an outer iteration acceleration scheme. Outer iterations are required with the LD energy discretization scheme because the two angular flux unknowns within each group are coupled, which gives rise to effective upscattering. The inner iteration convergence is accelerated using diffusion synthetic acceleration, and the outer iteration convergence is accelerated using a diamond difference approximation to the LD energy discretization. Computational results are given that demonstrate the effectiveness of our convergence acceleration schemes and the accuracy of our discretized SAAF equation.

Massively-parallel computers allow detailed 3D radiation transport simulations to be performed to analyze the response of complex systems to radiation. This has been recently been demonstrated with the coupled electron-photon Monte Carlo code, ITS. To enable such calculations, the combinatorial geometry capability of ITS was improved. For greater geometrical flexibility, a version of ITS is under development that can track particles in CAD geometries. Deterministic radiation transport codes that utilize an unstructured spatial mesh are also being devised. For electron transport, the authors are investigating second-order forms of the transport equations which, when discretized, yield symmetric positive definite matrices. A novel parallelization strategy, simultaneously solving for spatial and angular unknowns, has been applied to the even- and odd-parity forms of the transport equation on a 2D unstructured spatial mesh. Another second-order form, the self-adjoint angular flux transport equation, also shows promise for electron transport.

Recently, Morel and McGhee described an alternate second-order form of the transport equation called the self adjoint angular flux (SAAF) equation that has the angular flux as its unknown. The SAAF formulation has all the advantages of the traditional even- and odd-parity self-adjoint equations, with the added advantages that it yields the full angular flux when it is numerically solved, it is significantly easier to implement reflective and reflective-like boundary conditions, and in the appropriate form it can be solved in void regions. The SAAF equation has the disadvantage that the angular domain is the full unit sphere and, like the even- and odd- parity form, S{sub n} source iteration cannot be implemented using the standard sweeping algorithm. Also, problems arise in pure scattering media. Morel and McGhee demonstrated the efficacy of the SAAF formulation for neutral particle transport. Here we apply the SAAF formulation to coupled electron-photon transport problems using multigroup cross-sections from the CEPXS code and S{sub n} discretization.

This report contains the notes from the second session of the 1997 IEEENuclear and Space Radiation Effects Conference Short Course on ApplyingComputer Simulation Tools to Radiation Effects Problems. Part A discussesthe physical phenomena modeled in radiation transport codes and varioustypes of algorithmic implementations. Part B gives examples of how thesecodes can be used to design experiments whose results can be easily analyzedand describes how to calculate quantities of interest for...

A hybrid multigroup/continuous-energy Monte Carlo algorithm is developed for solving the Boltzmann-Fokker-Planck equation. This algorithm differs significantly from previous charged-particle Monte Carlo algorithms. Most importantly, it can be used to perform both forward and adjoint transport calculations, using the same basic multigroup cross-section data. The new algorithm is fully described, computationally tested, and compared with a standard condensed history algorithm for coupled electron-photon transport calculations.

The MITS multigroup/continuous-energy electron-photon Monte Carlo transport code system has matured to the point that it is capable of addressing more realistic three-dimensional adjoint applications. It is first employed to efficiently predict point doses as a function of source energy for simple three-dimensional experimental geometries exposed to planar sources of monoenergetic electrons up to 4.0 MeV due to simulated uniform isotropic fluences. Results are in very good agreement with experimental data. It is then used to efficiently simulate dose to a detector in a subsystem of a GPS satellite from the natural electron environment, employing a relatively complex model of the satellite. The capability for survivability analysis of space systems is demonstrated, and results are obtained with and without variance reduction.

Hermes III is a pulsed power, Bremsstrahlung simulator used for radiation-hardness testing of electronics components [19-MeV spectrum, 20-ns pulse duration, and typical doses (silicon) ≤100 krad (1 kGy)]. CaF2:Mn thermoluminescent dosimeter chips (TLDs) have been compared to a set of x-ray calorimeters in the Hermes III environment for doses between 10–75 krad. Similar to a design reported by Murray and Attix, this set of detectors included different dosimetric materials (silicon and aluminum) and two independent temperature sensors (thermistors and thermocouples). The electronic recording system was also updated. The average disagreement between TLDs and calorimeters was 1%–3%. Radiation transport calculations, however, suggest a possible bias of 4%–6% (source unknown). With the silicon calorimeter the ac bridge, which measured the resistance of thermistor temperature sensors, was extremely sensitive to EMP.

A general adjoint coupled electron-photon Monte Carlo code for solving the Boltzmann-Fokker-Planck equation has recently been created. It is a modified version of ITS 3.0, a coupled electron-photon Monte Carlo code that has worldwide distribution. The applicability of the new code to radiation-interaction problems of the type found in space environments is demonstrated.

integrated into the ITS code package. Multigroup data produced by the CEPXS cross-section-generating code is needed to operate the BFP codes in adjoint electron-photon mode. In this paper, we present adjoint electron-photon transport results obtained with a new version of CEPXS and a new multigroup version of ITS.