Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON PARALLEL, DISTRIBUTED, GPU AND CLOUD COMPUTING FOR ENGINEERING
Edited by: P. Iványi and B.H.V. Topping
Impact risk assessment of small unmanned aerial vehicles
Y. Choi, J.W. Jung, S.E. Lee and J.W. Hong
Department of Civil and Environmental Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea
Y. Choi, J.W. Jung, S.E. Lee, J.W. Hong, "Impact risk assessment of small unmanned aerial vehicles", in P. Iványi, B.H.V. Topping, (Editors), "Proceedings of the Sixth International Conference on Parallel, Distributed, GPU and Cloud Computing for Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 13, 2019. doi:10.4203/ccp.112.13
Keywords: UAV, impact risk assessment, high speed collision, finite element analysis.
Recently, the development and the use of small unmanned aerial vehicles (UAVs) have increased in various fields such as surveillance and logistics. It is necessary to establish safety regulations and standards to secure public safety for the use of UAVs in public areas. Most current regulations entirely prohibit the operation of UAVs in public areas. However, the rising demand for applications of UAVs may change the current regulations, allowing airborne flights in more areas. This proves to be a difficult task as the enforcement of safety regulations and securement of operable airspace for UAVs may be exclusive to each other. To accommodate the use of UAVs in public areas, it is imperative to estimate and define the impact damage that UAVs may cause in malfunctioning as there is limited data elucidating the impact risk of UAVs against people or structures.
In this research, experimental and numerical investigations are performed to estimate the collision impact damage by UAVs to the exterior glass panels of building structures, which are widely used in South Korea. Monolithic panes of heat-strengthened glass of several thicknesses (e.g., 3 mm, 5 mm, and 8mm) are installed in a steel frame structure, and a commercial drone DJI F450 quadcopter that weighs a kilogram is driven to collide with the target at a velocity ranging from 5 m/s to 18 m/s. The interaction of the drone and glass panel is recorded with a highspeed camera, and the applied forces to the targets are measured using load-cells. The recorded collision responses are compared with numerical simulation results. Using a commercial finite element code LS-DYNA, the DJI drone model is discretized with finite elements, and a constant velocity is imposed to simulate the collision velocity. By comparison, the results of experiments and numerical analyses are in good agreement in the collision responses, and the damages of the glass panels in the experiments are simulated accurately.
In order to evaluate the parallelization of the finite element software, the number of threads used for the analyses is sequentially increased from 1 to 32 threads. The wall-clock times for the collision analyses are obtained, and the ceiling of the speedup with an increasing number of threads is estimated. This effective modeling technique for UAVs that ensures both accuracy and numerical efficiency will be very useful in performing future parametric studies to estimate the collision damages caused by many different kinds of UAVs.
purchase the full-text of this paper (price £22)