How DragGap estimates acceleration

What the simulator calculates

DragGap numerically integrates each vehicle at 120 steps per second. At every step it calculates available wheel power, the lower of power-limited and traction-limited drive force, aerodynamic drag, rolling resistance, transmission state and distance travelled. Timing gates are interpolated between steps rather than rounded to a rendered frame.

Inputs and source hierarchy

1. Official: EPA configuration and fuel-economy records plus NHTSA vPIC identity.
2. Supplemental: power, torque, mass, dimensions and published reference performance from the ODbL Open Vehicle Specs project when a generation and engine match is available.
3. Modelled: missing drag coefficient, frontal area, gear ratios, traction and shift behaviour use body- and drivetrain-specific engineering defaults.

Reference calibration

When a trustworthy published 0–100 km/h time is available, DragGap numerically calibrates the launch portion of the model to that timing gate. A verified standing quarter-mile result, when available, independently calibrates the post-launch phase beginning above 100 km/h. The two controls are deliberately separated so a quarter-mile figure cannot manufacture a quicker launch, and extreme correction factors disqualify a configuration from the public Top 500 rather than hiding bad source data.

Stock top-speed ceilings

When a matched stock configuration includes a plausible published top speed, the engine model smoothly tapers available drive force near that speed and applies the value as an explicit ceiling. For pure EV records, an electrical top-speed field is the vehicle’s road ceiling; for hybrids, an electric-only ceiling is not substituted for the whole vehicle’s top speed. The ceiling can slow a simulated car but can never make it faster. Modified builds remove the stock ceiling because their gearing, power and limiter may no longer be stock. Provenance distinguishes manufacturer-published ceilings from supplemental values; “governor limited,” “manufacturer claim,” and “drag-limited estimate” are not treated as equivalent to an independently measured top-speed run.

What the results mean

Results are reproducible engineering estimates for comparing vehicles under identical dry, level conditions. They are not certified measurements and cannot represent production variation, tyre condition, launch control, driver technique, elevation, temperature, wind or surface preparation. The interface exposes which fields are exact and which are estimated.

Accuracy validation

The source-linked deterministic audit spans 14 FWD, RWD and AWD cars with manual, torque-converter automatic, dual-clutch, CVT-like and direct-drive powertrains. Against Car and Driver’s published instrumented quarter-mile results, the current raw model records 4.0% mean absolute elapsed-time error and 5.1% mean absolute trap-speed error. The expanded set intentionally exposes difficult outliers and is not used to tune each row to its answer. These figures describe that audit set, not guaranteed error for every vehicle.