2015 11 13_meetingsummary

Jump to bottom

baseendje edited this page Dec 11, 2020 · 1 revision

Present: Robert Clegg, Bastiaan Cockx, Stefan Lang

Bas is making good progress on agent states

States as "primary" (simple values) and "secondary" (calculated from primary states)
No speed loss in this approach (1 million made in 200 ms)! Next needs to focus on
Agent-boundary interactions
Agent-grid interactions (finding the voxel that contains the agent centre is already possible)

Stefan has nice-looking diagrams of cylindrical voxels

Written in MatLab, will port into Java once it's working well

Next challenges are

Iterating through neighbour voxels of a focal voxel
Calculating the surface area neighbouring voxels share

Circle Voxels

Rob has ODE and PDE solvers up and running

So far only Dirichlet and Neumann boundaries
So far only explicit solver for PDEs, Rosenbrock solver for ODEs
So far no way of setting the domain/diffusivity to differ over a compartment
Compartments are able to take a total size and decide a way of dividing it up into a grid
Test run fine in a "Simulator" object with multiple compartments, although the compartments are not yet joined together

Next challenges are

Cyclic boundaries for solutes
Implementing more complex PDE solvers
Implementing domain/diffusivity methods (e.g. diffusivity reduced inside biofilm)