Study → Explore → inspect → promote → Compare
- Select Half Car · 4 DOF in Study and configure the baseline.
- Choose Explore. The validated baseline, explicit axle roads, initial states and solver settings are copied into the sweep workspace.
- Select one physical parameter, minimum, maximum and 3–25 cases. Values are evenly spaced in ascending order, including both endpoints.
- Run the deterministic sweep. Each candidate uses exactly the same validated Half-Car transient dispatcher as Study.
- Inspect the candidate table, metric plots and selected candidate response overlays. Select a candidate and choose Promote to Compare.
- Compare receives the completed Explore Baseline and selected candidate as Variant A. Press Run Half-Car Compare explicitly.
Supported parameters
| Physical parameter | Unit |
|---|---|
| Sprung mass | kg |
| Pitch inertia | kg·m² |
| CG to front axle | m |
| CG to rear axle | m |
| Front suspension stiffness | N/m |
| Front suspension damping | N·s/m |
| Rear suspension stiffness | N/m |
| Rear suspension damping | N·s/m |
| Front unsprung mass | kg |
| Rear unsprung mass | kg |
| Front tire stiffness | N/m |
| Rear tire stiffness | N/m |
All values must satisfy the same engineering bounds as Half-Car Study. Duplicate or non-finite candidates are rejected. A sweep retains at most 260,000 total samples across the baseline and candidates to bound memory. Reduce the count or duration when this limit is reached.
What stays fixed
Every candidate changes only the selected physical parameter. Front and rear road definitions, all eight initial states, duration, timestep, model and RK4 identity remain identical. Study speed is a road-timing convenience: its resulting explicit rear-road delay is retained. Geometry sweeps do not regenerate road timing.
Neutral evidence
The baseline runs separately, even when its parameter value is outside the candidate grid. Tables expose authoritative heave, pitch, suspension-travel and tire-deflection metrics. Deltas are candidate minus baseline; relative change is undefined when baseline is zero. Metric plots show the discrete sweep; response overlays compare the selected candidate against baseline for heave acceleration, nose-up pitch and front/rear suspension travel.
Explore does not optimize automatically. Candidate response is not a design recommendation. VehicleLab provides no score, ranking or preferred setup.
Completed, current and edited
A complete sweep owns an immutable definition and ordered completed runs. Editing inputs keeps this evidence visible with “Edited — rerun required”. Restoring the exact definition makes it Current without recomputation. Cancellation terminates the Worker and preserves any prior completed record; partial sweeps are never published. Journey records only successful completion and remains independent of freshness.
Projects and exports
Project files store definitions and metadata only. Import requires a new run. CSV follows the existing Explore convention: a compact deterministic row per candidate with parameter identity, value, unit, model/solver identity and eight metrics. Plot PNG and Engineering Report use completed evidence without rerunning physics. Reports include baseline configuration, sweep values, candidate deltas, selected candidate overlays and provenance. Stale exports explicitly use the last completed evidence.
Model limitations
Generic Half-Car parameters are not measured-vehicle correlation. Linear passive, small-angle, equilibrium-offset assumptions remain. No nonlinear suspension, contact-loss model, active control, multi-variable sweep, optimization or Half-Car Frequency is provided. Frequency Explore is deferred.