7. Getting Started

7.1. Run the cross-plane example

From the repository root:

./build/PhonoMC example/input_cross_100nm.toml

PhonoMC creates an indexed directory such as example/results/Cross_100nm_1 so that an existing result is not overwritten.

7.2. Select the OpenMP thread count

OMP_NUM_THREADS=16 ./build/PhonoMC example/input_cross_100nm.toml

7.3. Start from a copied input

cp example/input_cross_100nm.toml input.toml
./build/PhonoMC input.toml

If no argument is supplied, PhonoMC searches for input.toml near the current working directory and executable.

7.4. Inspect the first results

The result directory contains at least:

  • summary.txt: normalized input, geometry, boundary, and runtime details

  • grid_centers.csv: grid-center coordinates and cell volumes

  • convergence.txt: temperature and transport observables versus time

7.5. Plot a one-dimensional run

python3 tools/plot_convergence.py example/results/Cross_100nm_1

The command writes temperature, heat-flux, and thermal-conductivity figures under plots_1d in the selected result directory.

7.6. Validate before a long run

Before increasing particle count or iteration count:

  1. run with a small particle count

  2. confirm every boundary region matches the intended facet

  3. inspect summary.txt and grid_centers.csv

  4. check particle balance in convergence.txt

  5. then scale threads, particles, grid size, and iterations