Publications

Results 26–50 of 128

Search results

Jump to search filters

Developing and evaluating Malliavin estimators for intrusive sensitivity analysis of Monte Carlo radiation transport

Bond, Stephen D.; Franke, Brian C.; Lehoucq, Richard B.; Smith, J.D.

We will develop Malliavin estimators for Monte Carlo radiation transport by formulating the governing jump stochastic differential equation and deriving the applicable estimators that produce sensitivities for our equations. Efficient and effective sensitivity can be used for design optimization and uncertainty quantification with broad utilization for radiation environments. The technology demonstration will lower development risk for other particle-based simulation methods.

More Details
TYPE Other Report YEAR 2019
DOIOSTI

Electron transport algorithms in the integrated tiger series (ITS) codes

20th Topical Meeting of the Radiation Protection and Shielding Division, RPSD 2018

Franke, Brian C.; Kensek, Ronald P.

We describe the three electron-transport algorithms that have been implemented in the ITS Monte Carlo codes. While the underlying cross-section data is similar, each uses a fundamentally unique method, which at a high level are best characterized as condensed history, multigroup, and single scatter. Through a set of comparisons with experimental data and some comparisons of purely numerical results, we discuss various attributes of each of the algorithms and show some of the defects that can affect results.

More Details
TYPE Conference Poster YEAR 2018
OSTIScopus

Adjoint-enabled multidimensional optimization of satellite electron/proton shields

20th Topical Meeting of the Radiation Protection and Shielding Division, RPSD 2018

Pautz, Shawn D.; Bruss, Donald E.; Adams, Brian M.; Franke, Brian C.; Blansett, Ethan

The design of satellites usually includes the objective of minimizing mass due to high launch costs, which is complicated by the need to protect sensitive electronics from the space radiation environment. There is growing interest in automated design optimization techniques to help achieve that objective. Traditional optimization approaches that rely exclusively on response functions (e.g. dose calculations) can be quite expensive when applied to transport problems. Previously we showed how adjoint-based transport sensitivities used in conjunction with gradient-based optimization algorithms can be quite effective in designing mass-efficient electron/proton shields in one-dimensional slab geometries. In this paper we extend that work to two-dimensional Cartesian geometries. This consists primarily of deriving the sensitivities to geometric changes, given a particular prescription for parametrizing the shield geometry. We incorporate these sensitivities into our optimization process and demonstrate their effectiveness in such design calculations.

More Details
TYPE Conference Poster YEAR 2018
OSTIScopus
Results 26–50 of 128
Results 26–50 of 128
Top