Identification of kite aerodynamic characteristics using the estimation before modeling technique

R. Borobia-Moreno, D. Ramiro-Rebollo, R. Schmehl, G. Sánchez-Arriaga

Research output: Contribution to journalArticleScientificpeer-review

Abstract

The aerodynamic characteristics of a leading edge inflatable (LEI) kite and a rigid-framed delta (RFD) kite were investigated. Flight data were recorded by using an experimental setup that includes an inertial measurement unit, a GPS, a magnetometer, and a multi-hole Pitot tube onboard the kites, load cells at every tether, and a wind station that measures the velocity and heading angle of the wind. These data were used to feed a flight path reconstruction algorithm that estimated the full state vector of the kite. Since the latter includes the aerodynamic force and moment about the center of mass of the kite, quantitative information about the aerodynamic characteristics of the kites was obtained. Due to limitation of the experimental setup, the LEI kite flew most of the time in post-stall conditions, which resulted in a poor maneuverability and data acquisition. This assumption was corroborated by a particular maneuver where the lift coefficient decreased from 1 to 0.4, while its angle of attack increased from 35° to 50°. On the contrary, abundant flight data were obtained for the RFD kite during more than 10 figure-eight maneuvers. Although the angle of attack was high, between 20° and 40°, the kite did not reach its maximum lift coefficient. High tether tensions and a good maneuverability were achieved. Statistical analysis of the behavior of the lift, drag, and pitch moment coefficients as a function of the angle of attack and the sideslip angle allowed to identify some basic aerodynamic parameters of the kite.

Original languageEnglish
Number of pages13
JournalWind Energy
DOIs
Publication statusPublished - 2021

Keywords

  • aerodynamics
  • airborne wind energy
  • kite

Fingerprint Dive into the research topics of 'Identification of kite aerodynamic characteristics using the estimation before modeling technique'. Together they form a unique fingerprint.

Cite this