Use this URL to cite or link to this record in EThOS:
Title: Radiation transport in multiphase and spatially random media
Author: Park, Samuel
ISNI:       0000 0004 6348 468X
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
An important class of problems within radiation (e.g. neutrons, photons) transport are those in which the radiation migrates through a medium which has a random or stochastic composition. The uncertain medium composition introduces an uncertainty in the radiation angular/scalar flux, current, reaction rates and other quantities of interest. Stochastic media play an important role within radiation transport and have numerous applications such as radiation shielding, nuclear criticality assessment, as well as radiative transfer in clouds, stellar atmospheres and plasma physics. Stochastic radiation transport problems reduce to treating the adsorption, scatter and other macroscopic cross-section data as spatially correlated random fields. Several methods for the treatment of these uncertain-ties have been proposed however they are limited in their scope and computational efficiency. Multiphase and spatially random media are often characterised by non-Gaussian random fields which are much more challenging to model than Gaussian random fields. This thesis aims to investigate, develop and implement mathematically rigorous computational algorithms that are more efficient than the current methods for solving radiation transport with multiphase and spatially random media. In particular this thesis applies iso-probabilistic (e.g. Nataf) transforms to transform Gaussian random fields into non-Gaussian random fields. This approach enables the use of optimal spectral stochastic representations, such as the Karhunen-Loève and generalized polynomial chaos methods, to be used to simulate non-Gaussian random fields. This thesis also describes the verification of these iso-probabilistic spectral stochastic projection methods against standard radiation transport in random media benchmarks, such as the widely used Adams-Levermore-Pomraning benchmark. This thesis is the first time the general Nataf method has been applied to model radiation transport through multiphase and spatially random media.
Supervisor: Eaton, Matthew ; Bluck, Michael Sponsor: Engineering and Physical Sciences Research Council ; Rolls-Royce plc
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral