UAV icing: system identification of leading-edge rime-ice effects on a small fixed-wing drone

Abstract

As small fixed-wing unmanned (or uncrewed) aerial vehicles (UAVs) are increasingly deployed for long-endurance, beyond visual line-of-sight (BVLOS) missions, atmospheric icing has emerged as a critical operational hazard. One of the barriers to safe flight in these conditions is the lack of validated aerodynamic icing-degradation models, which are essential for developing robust mitigation strategies. This paper presents an experimental system identification study quantifying the effects of leading-edge wing icing on the flight characteristics of a Skywalker X8 UAV. Three-dimensional rime ice shapes were generated in an icing wind tunnel (IWT), digitised using photogrammetry, and reproduced via 3D printing for flight testing. Closed-loop excitation manoeuvres were flown at constant airspeed in clean and iced configurations, enabling reconstruction of aerodynamic coefficients and identification of quasi-steady coefficient models. The results show a clear increase in drag (a 50% increase in zero-lift drag coefficient), which roughly halves the maximum lift-to-drag ratio. Furthermore, the iced aircraft exhibits a 23% decrease in lift slope in the primary regime (angle-of-attack above 0 Superscript ring 0 ∘ $0^\circ$ ), a reduced static stability margin and degraded pitch and lift control effectiveness. Linearised analysis further shows a pronounced slowdown of the short-period and roll dynamics as well as reduced Dutch-roll damping. Overall, these models provide a crucial quantitative baseline to assess performance penalties and develop robust icing detection, envelope protection and guidance strategies for small fixed-wing UAVs operating in icing conditions.