Project · Flight dynamics
6-DOF Flight Simulator
Six-degree-of-freedom trajectory simulation for the same vehicle, with Monte Carlo dispersion analysis. It answers the questions that decide whether a launch happens: where does it land, how much margin is left, and how confident can we be in either number.
Fully interactive. Set the airframe, engine and environment, run a nominal trajectory, then run a Monte Carlo campaign over the uncertainties. Best viewed full screen on a desktop display.
The model
The vehicle is integrated as a rigid body with six degrees of freedom, so attitude is a solved quantity rather than an assumption. That matters for a sounding rocket: tilt off vertical is what turns a vertical flight into a downrange one, and it comes out of the aerodynamics and the rail exit conditions rather than being prescribed.
| Inputs | Airframe geometry, mass properties, engine and propulsion parameters, fin definition, environment, and solver settings. |
|---|---|
| Trajectory outputs | Altitude, altitude versus downrange, speed and Mach, dynamic pressure, acceleration magnitude, thrust and mass, chamber and tank pressure, angle of attack, body rates, tilt from vertical, static margin, and ground track. |
| Dispersions | Per-parameter standard deviations feeding a Monte Carlo campaign, with a trajectory envelope, percentile bands, correlation and sensitivity views, and histograms of apogee, Mach, dynamic pressure and impact range. |
| Safety products | Impact footprint with 1σ, 2σ and 3σ contours, and a static margin history checked against the stability limit across all cases. |
| Export | CSV and MATLAB output for downstream analysis. |
Why dispersions, not a single trajectory
A nominal trajectory is the least interesting output of a flight simulator. Every parameter it consumes is uncertain: propellant load, thrust, drag coefficient, centre of mass, wind, rail angle. A single run tells you what happens if all of those are exactly right, which is the one case that never occurs.
The Monte Carlo layer assigns a standard deviation to each of those inputs and runs the campaign, so the deliverable is a distribution rather than a point. The impact footprint contours come out of that distribution directly, which is what a range safety conversation actually needs. The sensitivity view then closes the loop by ranking which uncertainties are driving the spread, so effort goes into tightening the parameters that matter instead of the ones that are easy to measure.
Relationship to the engine model
Both tools descend from a single integrated MATLAB simulation covering engine burn and flight. Splitting them was deliberate. Engine trades and trajectory trades happen on different cadences and are run by different people, and a combined tool forces everyone through the same interface at the same fidelity. Kept separate, the propulsion output feeds the flight model as a thrust and mass schedule, and each side can be revised on its own.