# Wiki

Clone wiki# supernu / LANL_README

## Explanation of LANS Copyright Assertion:

The Los Alamos National Security (LANS), LLC, copyright assertion pertains to modifications made to SuperNu, Version 3.x, by Ryan Wollaeger, while at Los Alamos National Laboratory (LANL) and while funded by the Office of Advanced Simulation and Computing (ASC). The copyright assertion can be found in LANL_COPYING. The LANS copyright assertion does not pertain to revisions to SuperNu made by SuperNu author Daniel van Rossum. The LANS copyright assertion applies to all revisions to SuperNu made by author Ryan Wollaeger after January 20, 2015.

Please see the main readme file, README, for a description of the entire SuperNu code, including a summary of functionality and setup instructions. A description of the developments that fall under the auspices of the LANS copyright assertion is included in the abstract below.

## Abstract:

# A. Patches to SuperNu, Version 3.x, added at LANL:

(1) 3D Cylindrical IMC-DDMC and minor patches (default branch):

At LANL, modifications to enable 3D cylindrical IMC-DDMC were made to the following files: -- inputparmod.f -- inputstrmod.f -- particle_advance.f90 -- particle_advance_gamgrey.f90 -- SOURCE/boundary_source.f90 -- SOURCE/initial_particles.f90 -- GRID/leakage_opacity2.f90 -- TRANSPORT2/transport2.f90 -- TRANSPORT2/diffusion2.f90 -- TRANSPORT2/transport2_gamgrey.f90

The generalization of 2D cylindrical to 3D cylindrical geometry does not extend the physics capbilities of SuperNu, which had 3D Cartesian and 3D spherical IMC-DDMC implemented prior to SuperNu Version 3.x.

(2) Lambda-T optimization ("wlT" branch):

The following summary pertains to a discontinued branch for an attempted optimization at LANL. The implementation of the dual grid wlT optimization, which was part of the LANL copyright assertion, proved inaccurate in supernova light curve calculations.

An alternate form of an optimization in the IMC-DDMC radiation transport code SuperNu was developed and tested at LANL. This optimization involves storing a non-dimensional wavelength grid for computing opacities and tracking Monte Carlo photon radiation particles. The physical wavelength grid is the non-dimensional grid divided by the local spatial cell gas temperature.

The patch to the optimization is: -- the removal of temperature corrections to DDMC particle weights, -- the introduction of a dual wavelength grid that is dimensional and uniform in space.

The temperature correction to DDMC particle weights seemed to result naturally from the transport equations but could not be theoretically justified and was deemed a bug. The additional wavelength grid is applied in DDMC propagation of particles and calculation of DDMC leakage opacities to avoid in-line Planck integrals at IMC-DDMC interfaces (an expense that would otherwise be incurred by using the non-dimensional grid without temperature corrections in DDMC). The new wavelength grid also adapts in time to cover all dimensional frequencies implied by the non-dimensional grid and the material temperature field.

Despite the advantage of a global group Planck function integral array, the dual-grid patch to the non-dimensional multigroup scheme requires a mapping of multigroup opacity to the dimensional dual grid.

To implement the wavelength dual-grid optimization, modifications included: -- addition of new input parameters and allocatable arrays to handle mapping from nondimensional to dimensional wavelength grid, -- addition of a subroutine to adapt the dimensional wavelength grid each time step (called "group_update"), -- addition of a function (called "regroup") that maps multigroup data from the non-dimensional to the dimensional wavelength grid, -- modification of IMC-DDMC particle exchange conditions to use values mapped from the regroup function.

Updated