Grid: Computational Foundation

The computational grid defines the spatial and temporal discretization for k-Wave simulations. It forms one of the four core components (Grid, Medium, Source, Sensor) that define every acoustic simulation.

Key Concepts

Grid Spacing: Determines simulation accuracy. Use \(\Delta x \le \lambda_\mathrm{min}/3\) where \(\lambda_\mathrm{min}\) is the smallest wavelength. k‑Wave: Nyquist limit = 2 PPW; k‑Wave defaults to 3 PPW as a safer minimum — increase to ~6–15+ PPW for nonlinear/heterogeneous problems.

Grid Size: Total number of points. Larger grids provide finer resolution but increase computational cost significantly.

Time Step: Computed from the CFL condition (e.g., \(\Delta t = \mathrm{CFL}\cdot \min(\Delta x)/c_\mathrm{max}\)). You can override CFL for custom temporal sampling. See options.

Quick Start

from kwave.kgrid import kWaveGrid

# 2 cm × 1 cm domain, 100 µm resolution
grid = kWaveGrid([200, 100], [100e-6, 100e-6])

For detailed tutorials and parameter-selection guidelines, see ../fundamentals/understanding_grids.