A curated collection of AeroVance's peer-reviewed research, technical reports, and conference presentations advancing aerospace engineering, space systems, and defense technologies.
This paper presents comprehensive test data from the AV-9X family of full-flow staged combustion engines. We analyze combustion stability margins, injectors pattern optimization, and regenerative cooling channel geometries that enable 15+ rapid re-flight cycles with minimal erosion. CFD and RANS simulations are validated against hot-fire test data up to 2,800 kN thrust.
Traditional Kalman filtering approaches struggle with cumulative drift in deep-space transfers. We propose a hybrid DRL-PID controller that continuously learns optimal impulse vectors while respecting fuel constraints. Simulated across 10,000 Monte Carlo missions to the lunar south pole, the system reduces ΔV consumption by 18% compared to baseline GNC architectures.
Hypersonic vehicles operating at Mach 6+ face extreme aerodynamic heating. This study evaluates SiC-nanoweave ablative composites under plasma wind tunnel conditions. Microstructural analysis reveals 40% improvement in recession resistance compared to traditional PICA variants, enabling longer loiter times in high-enthalpy environments.
As LEO congestion rises, ground-based TLE updates become insufficient. We implement a distributed swarm algorithm enabling autonomous 6U-12U cubesats to negotiate avoidance maneuvers via optical inter-satellite links. The protocol reduces conjunction assessment latency by 92% and eliminates single-point ground station dependencies.
Directed energy deposition (DED) enables complex internal cooling geometries impossible via investment casting. This paper details laser power, scan speed, and powder feed rate optimization for Inconel 718. High-cycle fatigue testing at 700°C demonstrates 85% of wrought baseline strength with 30% weight reduction.
GPS-denied environments pose critical challenges for autonomous flight. We demonstrate a miniaturized Rb-87 cold atom accelerometer integrated into a 15kg UAV airframe. Field tests show drift rates below 0.1 mGal/s over 4-hour endurance flights, enabling precision navigation without external referencing.
We construct a physics-informed graph neural network that maps component degradation pathways across 2,400+ flight hours. By ingesting real-time QAR data, the model predicts bearing and actuator failures 180 hours before threshold breaches, reducing AOG events by 34% across partner airline fleets.
Single-event upsets (SEUs) remain a critical failure mode for commercial LEO constellations. This comparative study evaluates TMR logic gating against continuous background scrubbing on 28nm FPGA payloads. Results indicate scrubbing reduces power overhead by 22% while maintaining equivalent bit-error-rate performance in trapped radiation belts.