simulation-software-setup-run

name	simulation-software-setup-run
description	Setup and execution workflow for engineering and physics simulation software, including FDTD, FEM, Monte Carlo, multiphysics, radiation transport, and medical-physics studies. Use when configuring environments, building models, choosing solver settings, and running reproducible simulation campaigns.

Simulation software setup and run

Scope

Use this skill for simulation lifecycle work before interpretation: problem framing, tool selection, environment setup, model discretization, solver configuration, run orchestration, and reproducibility capture.

Common solver families:

• FDTD tools (electromagnetics and wave propagation)
• FEM tools (structural, thermal, fluid, EM, multiphysics)
• Monte Carlo tools (radiation transport, stochastic transport, uncertainty propagation)
• Domain-specific packages for medical physics, dosimetry, and treatment-planning research

Use for

• Selecting solver family from objective and physics regime
• Installing and validating toolchains (local or cluster)
• Defining geometry, materials, mesh/grid, and boundary/initial conditions
• Choosing timestep, solver tolerances, and convergence criteria
• Configuring parameter sweeps and batch runs
• Logging inputs, versions, random seeds, and compute environment details

Setup protocol

1. Frame the simulation question

• Identify target observables and acceptable error.
• Define what quantity must be predicted versus calibrated.

2. Select method class

• Use FDTD for time-domain wave propagation where explicit field evolution matters.
• Use FEM for complex geometry, coupled PDEs, and boundary-value formulations.
• Use Monte Carlo when stochastic transport, rare events, or particle interactions dominate.

3. Define model and discretization

• Build explicit geometry and material assumptions.
• Define mesh/grid resolution strategy and expected resolution-induced error.
• State boundary and initial conditions with units.

4. Configure numerics

• Set timestep and stability constraints (for example CFL-like constraints in explicit schemes).
• Set nonlinear/linear solver tolerances and max iterations.
• Define stopping criteria and convergence diagnostics.

5. Plan execution

• Smoke test with a reduced case.
• Run baseline then parameter sweeps.
• For stochastic solvers, set seed policy and replicate strategy.

6. Record reproducibility metadata

• Solver version and plugin/module versions.
• Full input deck or config snapshot.
• Hardware profile and runtime flags.

Medical-physics and radiation notes

• Track phantoms, material libraries, and detector or tally definitions explicitly.
• Separate physical-model uncertainty from numerical uncertainty.
• Do not present simulation output as clinical recommendation without external validated workflow.

Failure modes to catch early

• Unstable timestepping or divergence hidden by automatic damping.
• Mesh too coarse for gradients or interfaces.
• Incorrect boundary conditions producing non-physical reflection or leakage.
• Silent unit mismatch between geometry, source, and material files.

Deliverables

• Reproducible run plan with setup steps.
• Parameter table (mesh, timestep, tolerances, seeds).
• Execution log summary with pass/fail for smoke and production runs.
• Known setup limitations and next-run improvements.