OPENROCKET VALIDATION

The Methods

To analyze Estes Industries, LLC Athena and XTREME A/B motor class model rockets' flight performance, I measured velocity from high-speed video footage and compared the observed values against OpenRocket theorietical values at corresponding timestamps in the flight. Two pair of frames were extracted from 240 fps slow-motion footage, with frame-to-frame pixel displacement of the rocket measured against a calibrated reference scale (8 foot pole with with 6 inch intervals marks) visible in the same shot. Pixel displacement was converted to real-world distance using the measured pixels-per-foot ratio, then to velocity using the known frame interval. Each measurement's timestamp was then matched to simulated velocity from Open Rocket at the same elapsed time from launch, and percent error was calculated to quantify agreement between the observed and theoretical flight. 

The Data and Analysis

XTREME Results: For the first measurement pair, a travel of 55 px against a calibration of 57 px per 6 in gives a displacement of 0.48 ft, corresponding to an observed velocity of 115.79 ft/s (78.94 mph) at t = 0.293 s. OpenRocket's prediction at this point, 104.14 ft/s, falls 11.18% below the observed value. For the second pair, 50 px of travel at 55 px per 6 in yields a displacement of 0.45 ft and an observed velocity of 109.09 ft/s (74.38 mph) at t = 0.288 s; here OpenRocket predicts 102.86 ft/s, 6.06% below observed. ATHENA Results: For the first measurement pair, 40 px of travel at a calibration of 57 px per 6 in gives a displacement of 0.35 ft, corresponding to an observed velocity of 84.21 ft/s (57.42 mph) at t = 0.342 s. OpenRocket predicts 73.04 ft/s at this point, 15.29% below observed. For the second pair, 37 px of travel at 56 px per 6 in yields a displacement of 0.33 ft and an observed velocity of 79.29 ft/s (54.06 mph) at t = 0.334 s; OpenRocket's prediction of 71.97 ft/s here is 10.17% below observed.

There is a consistent ~10% error of experimental velocity outpacing OpenRocket's theoretical performance at the same timestamp. While the error is indicative of an issue in the methods, the consistent value and OpenRocket-behind-observed pattern offers some potential conclusions. It is possible that OpenRocket's thrust curve for the A8-3 motors did not match the actual thrust curve of the motors used. Moreover, while the dry mass of the rocket was set in OpenRocket to exactly match values measured with a scale, OpenRocket's motor mass may not have matched that of the motors actually flown. In the future, I would measure the motors prior to launch to ensure the wet mass in OpenRocket is correct. I would also like to build a thrust stand so I can test-fire a motor from each order and verify its curve matches the one modeled in OpenRocket. Beyond motor performance, a few other sources could have contributed to a consistent, one-directional offset like this, such as launch rail effects or improper timestamp alignment.

Modeled using available Estes model rocket part piece data sheets where possible, and a pair of calipers to fill in the gaps. This ensured not only maximum accuracy in dimensions, but in material characteristics. Modeling with this level of accuracy was a non-negotiable, otherwise faulty Cg and Cp statistics would lead to systematic error in the accuracy of the predicted flight characteristics.