Advanced Reactor Design and Analysis

Research activities under this area encompass the design and analysis of advanced reactors using computational tools in the realm of multi-physics multi-scale computational modeling and simulation (M&S). Research projects we have worked in the past years include:

  • An accelerator driven liquid lead cooled thorium fueled fast reactor design and analysis using MCNPX/SCALE tools
  • Core design studies for a BWR-type small modular reactor with long-life core using CASMO-3/PARCS/RELAP-5 LWR analysis code suites
  • Reactor design and analysis of stationary liquid fuel fast reactor (SLFFR) using ANL MC2-3/DIF3D/REBUS reactor physics code system
  • Hypothetical accident analysis on SLFFR with self-developed single channel T/H transient code
  • Reactor safety analysis on fuel conversion for the NIST plate-type research reactor (NBSR)
  • Feasibility studies on research reactor replacement at NIST using MCNP-6 and PARET code system
  • Integrated 1-D system level thermal stratification model development for sodium fast reactor (SFR)

Computational Methods Development for Nuclear Applications

Linear Boltzmann transport equation is the essential mathematical model that governs particle transport phenomena in many nuclear applications including reactor analysis, radiation shielding, medical radiography, etc. With the growth of high performance computing capability and the ever-improving of numerical methodologies, computational techniques on these problems are constantly demanded with higher accuracy and efficiency. Research topics that we have dedicated in this area include:

  • Applying the lattice Boltzmann method (LBM) to solve the neutron diffusion and transport equations
  • A modified form of the SAAF transport equation with fully void-compatible feature
  • Hybrid Monte Carlo-deterministic methods for reactor analysis
  • Advances in inverse transport problems and applications to neutron tomography
  • A new 1-D Sn analytic solution for heterogeneous problems with no iteration on interfacial fluxes
  • A modified step characteristic method for Sn transport solution that possesses 2nd order accuracy and diffusion limit

Sensitivity Analysis and Uncertainty Quantification

The meaning of a physical quantity predicted by computational models is limited if the uncertainty information associated with the quantity is not provided. This is particularly true in nuclear applications because nuclear data is normally emerged with a statistical nature. Therefore, sensitivity analysis (SA) and uncertainty quantification (UQ) became important in modern predictive science. The research efforts we have made in this area include:

  • GPT-free Sensitivity Analysis for Deterministic and Monte Carlo Models
  • More accurate k-eigenvalue sensitivity estimation using the complex-step derivative method
  • Enhanced 1-D SFR thermal stratification model via advanced inverse uncertainty quantification methods