Back to search

IKTPLUSS-IKT og digital innovasjon

Icing effects, detection and mitigation on unmanned aerial vehicles (UAVs)

Alternative title: Ising på ubemannede fly: Effekter, deteksjon og beskyttelse

Awarded: NOK 12.0 mill.

Unmanned aerial vehicles (UAVs) have used numerous applications recently. However, in order to release the business potential, UAVs must be integrated into shared airspace and be able to operate safely beyond-visual-line-of-sight (BVLOS) and above populated areas. It is therefore necessary to ensure that UAVs are airworthy according to emerging industry standards and regulatory frameworks. Detection and mitigation of icing on UAVs is a critical challenge for safety and airworthiness. Yet, the research on the topic is very limited and no commercially available solutions exist. The project embraces autonomy as a key enabling concept and combines fundamental research with innovative ideas in an interdisciplinary collaboration with industry and end users. The project's first objective is to develop validated aerodynamic and thermodynamic models, operational knowledge, and computational tools that are needed to understand and predict icing on UAVs. Based on the improved understanding of the physical processes of UAV icing, the second objective of this project is to develop more energy-efficient autonomous detection methods and optimized control systems for mitigating the adverse effects of atmospheric icing on UAVs. The research in 2021 has focused on preliminary studies, data acquisition, and establishing the necessary test systems. Master student Bogdan Løw-Hansen has developed an algorithm for detection of ice shedding during de-icing cycles, in order to optimize energy consumption. Master student Ruben Kleiven has developed control algorithms for autopilot that are robust under situations with full or partial icing on wings, including asymmetric icing with ice on only one wing. Several wind tunnel tests are conducted in collaboration with UBIQ Aerospace and VTT in Finland, where the power consumption of electrothermal deicing systems are analyzed, and aerodynamic penalties (increased losses due to drag) caused by icing on wings and propellers. The project has developed simulation tools for flying in icing conditions, and estalished an official collaboration with DLR in Germany on this. Maritime Robotics and UBIQ Aerospace has developed a prototype icing protection system on a UAV that will be made available for the project after internal tests in early 2022. Based on the research of master student Johannes Oswald, we have developed artificial ice profiles that can be attached on wings during test flights without being in icing conditions.

Icing on UAVs is an emerging topic that has not been studied in a holistic and systematic way until now. We intend to focus on the following aspects of UAV icing that are highly relevant to enable UAV flight in icing conditions: 1) modelling and simulation of ice accretion, including aerodynamic performance penalties due to icing and the effect of ice on flight behavior, control and stability, 2) methods to autonomously predict and detect icing conditions using sensors and mathematical models, 3) electro-thermal mitigation strategies against inflight icing in autonomous anti-icing and de-icing modes, including mission planning approaches. In addition to the numerical simulations, experimental testing will play a major role for this project. Experiments will be conducted in icing wind tunnels and real flights tests. Icing wind tunnels are well suited to study all research questions in a controlled environment. The data collected in these tests will be used to validate and calibrate the existing numerical icing methods. In addition to these controlled tests, testing will be conducted in collaboration with the UAV Lab at NTNU. Tests will consist of attaching artificial ice shapes (that have been generated numerically or in an icing wind tunnel) to a UAV to study the effects on flight behavior. Furthermore, prototypes of the icing detection and protection system will be tested in icing clouds at test sites near Trondheim and Andøya. The experimental work suggested here will generate data sets that will build confidence in the numerical methods, which in turn is important for understanding the effects of icing and for developing icing protection systems.


IKTPLUSS-IKT og digital innovasjon