Dev Roadmap

[ZoeRichter]

This issue is a parent issue to organize the next development steps. Items are generally listed in the order they need to be completed in. Complex items have associated issues.

Improve current code:

• Improve variable names #39 : Improve variable names
• [Feature] Geometry handling simplification/update #19 : Geometry simplification/update (or a decision to not make major changes, but explain why if you close the issue
• Record LAMMPS installation requirements for Ghastly #11 and an improved README with install instructions
• [Feature] When a jinja2 template is used, make the file choice an input parameter with a default setting #24 , improvements to file generation as you work on this
• [Feature] Verbosity specification for runs #18 and other misc. improvements
• consider sorting templates into sub-directories by code (lammps, openmc, etc etc)

Flow data:

• convert to .csv
• velocity profiler
  • Radial/axial binned average velocities (colormap)
  • mass flow average velocity (radially dependent, for low fidelity models)
  • can also flag pebbles with a high kinetic energy in this phase for possible avalanches (print timestep/uid to csv, but stop there)
• Single pass core transit pool
  • save these to csv as well, separate from the csv coord data used to determine these
  • can also generate streamlines at this phase (include optional flag to generate or not)
• transit numbers by radial zone
• residence time by radial zone

Input file additional variables:

• fresh fuel enrichment
• TRISO geometry/comp
  • TRISO pf in pebbles, fueled radius
  • optional: allow a 'simple' non-depleting pebble option (collapse the TRISO layers)
• non-fuel material properties
• burnup steps
• assay delay time

Pebble management:

• add attributes needed to pebble class
  • depleting/non-depleting flag
    • if non-depleting, will need something to point to the correct composition matrix object
  • history
  • radial zone in pass
  • because you convert to (r, z) so often, add r as an additional coord
  • revisit this item during openmc material autogen if you need to add any additional attributes beyond mat_id to link depleting pebbles to their mat.
  • RGB color (optional, only do this if you're certain storing the colors here is better than assigning colors when plotting) (definitely not needed for paraview - only useful (maybe) for generating OpenMC color dicts) (idea: look into function to auto-blend region colors and then darken according to pass (with RGB calculated with math))
• pebble composition matrices code infrastructure
  • create Ghastly class, which needs:
    • history code attribute
    • current pass attribute
    • composition units attribute
    • attribute containing or pointing to the isotopic composition map
      • bin in (r, z). each bin contains a composition dict (or such) in the form of {isotope : concentration}
    • method to return the correct composition vector when given a pebble coordinate (to pass to OpenMC autogen)
      • method to interpolate using nearest neighbor bins
    • method to update composition matrix, given data from an improved depletion sim

OpenMC autogen:

• fuel material autogen (see BCC model for ref):
  • depleting
    • generate fresh fuel given enrichment, after that its tracking the ind. mats as they burn
  • non-depleting
    • see Ghastly composition map class discussion above for how the comp. is stored
    • add function to randomly assign passes/histories (both to bed and to pebbles that have recirculated)
• pebble geometry autogen (see BCC/test models for ref)
  • if the pebble class doesn't store color data, will need to gen. color dict here
  • generate TRISOs in pebs
    • depleting pebs might use different TRISO arrangements, but non-depleting will definitely just use the same one
    • see input file section, add optional flag for simplifying non-depleting TRISOs to one outer layer
• simple vessel autogen
  • basic reflector
    • maybe allow grey skirt iff it is very simple (<30 minutes to add)
  • evenly spaced coolant channels
  • simple RPV (basically a single shell around everything)
  • void outside
• assay period model
  • CRP packed cylinder w/ relatively lower packing fraction (~35%?), decay for X days in void

Low fidelity models:

• CRP model
  • figure out JT-algorithm in cone (OpenMC dev?) OR alternative
  • automatic pebble geometry and material gen should otherwise use same functions
  • move pebbles function
• lattice model
  • convert velocity profile into lattice units
  • autogen lattice positions and pebble unit cell definitions
  • move pebbles functionw