Transport Methods
Deterministic transport solvers for heterogeneous lattice cells. Each method discretises the angular variable differently and uses a geometry-specific kernel or sweep algorithm:
Collision Probability Method — integral transport via the \(P_{ij}\) matrix (slab, cylindrical, spherical kernels).
Discrete Ordinates Method (SN) — differential transport via angular quadrature and spatial sweeps (Cartesian 1-D / 2-D, curvilinear 1-D).
Method of Characteristics (MOC) — characteristic ray tracing with flat-source approximation (2-D pin cell).
Monte Carlo Neutron Transport — stochastic transport via Woodcock delta-tracking with analog absorption and weight-based population control.
- Collision Probability Method
- Key Facts
- Overview
- Architecture: Base Geometry and Augmented Geometry
- The Integral Transport Equation
- Definition of Collision Probabilities
- The Three CP Kernels
- Optical Path Construction Along a Chord
- Slab Geometry: The \(E_3\) Kernel
- Concentric Cylindrical Geometry: The Ki3/Ki4 Kernel
- Concentric Spherical Geometry: The Exponential Kernel
- Geometry Comparison
- The Eigenvalue Problem
- Cross-Section Data Layout
- Verification
- Implementation Details
- Design Decisions
- References
- Discrete Ordinates Method (SN)
- Method of Characteristics (MOC)
- Key Facts
- Overview
- Architecture
- The Transport Equation Along a Characteristic
- Angular Quadrature
- Ray Tracing
- Reflective Boundary Conditions
- Scalar Flux Update
- Power Iteration
- Cross-Section Data Layout
- Verification
- Investigation History: ERR-019 (Weight Factor)
- Design Decisions and Alternatives
- Gotchas and Subtleties
- Open Improvements
- References
- Monte Carlo Neutron Transport