NCRC

INL’s Nuclear Computational Resource Center

About NCRC

The Nuclear Computational Resource Center (NCRC) was established at INL to provide easy access to computational tools, high performance computing (HPC) resources and training.

Modeling and simulation are essential to nuclear energy innovation, as well as the continued safe, secure and efficient operation of existing nuclear systems.

How you can use the NCRC resources

Modeling the behavior of materials in harsh environments (irradiation and high temperatures)
Analyzing the performance of existing light water reactors and advanced nuclear reactors
Multiscale, multiphysics analysis of nuclear fuel performance

Examples of nuclear energy-related applications include:

MODELING AND SIMULATION CODES

MOOSE

MOOSE, the Multiphysics Object Oriented Simulation Environment, is a finite-element multiphysics framework largely developed by Idaho National Laboratory. It offers a high level interface to sophisticated nonlinear solver technology, such as BISON for nuclear fuel, Rattlesnake for radiation transport, MAMMOTH for general reactor physics, and more. MOOSE has a straightforward API well suited to how scientists and engineers tackle real-world problems. It is a fully coupled and fully implicit multiphysics solver that is automatically parallel, making it possible to run large simulations and tackle complicated models.

BISON

BISON is a finite element-based nuclear fuel performance code. It is applicable to light water reactor fuel rods, TRISO particle fuel, metallic rod and plate fuel, and other fuel forms. BISON solves thermomechanics and species diffusion equations for 1D, 2D and 3D geometries, with fuel models that describe temperature properties, fission product swelling and other material aspects. Because it is based on the MOOSE framework, Bison can solve problems efficiently using standard workstations or very large high-performance computers.

MARMOT

MARMOT is a mesoscale fuel performance code. As such, it can predict the evolution of the microstructure and material properties of fuels and claddings due to stress, temperature, and irradiation damage. MARMOT can, therefore, supply microstructure-based materials models to other code that works on engineering scale, which is larger than mesoscale, with an example being BISON. MARMOT solves equations involving solid mechanics and heat conduction using the finite element method.

HIGH-PERFORMANCE COMPUTING

SAWTOOTH

INL’s newest supercomputer operates with a LINPACK rating of 5.6 petaops, which ranked #37 on the November 2019 TOP500 list. The HPE SGI 8600 system comprises 99,792 cores with 403 TB of memory. The system also includes dedicated GPU capability.

LEMHI

The Lemhi supercomputer is a Dell 6420-based system operating on an OmniPath fat tree network. It came online in the fall of 2018, containing 20,160 cores and 94 total terabytes of memory. Lemhi is rated at 1 petaflop and ranked #427 on the November 2018 TOP500 list.

HOODOO

Hoodoo is a Lambda Hyperplane deep learning distributed memory system with 44 NVIDIA A100 tensor core GPUs and 7.2 TB of total memory. The system provides a maximum performance of 429 TFlops double precision or 858 TFlops single precision.