High-precision wing twist measurement in transonic wind tunnel using DIC based fluorescent marker detection method
Year: 2025
Type: article
Abstract: • Hybrid monocular–binocular measurement system for high-precision wind torsion measurement. • Wind tunnel coordinates calibrated via quadratic cross-product using machete mechanism motion. • Fluorescent markers enable robust DIC tracking under low temperature and pressure variations. • Combined stereo/mono reconstruction resolves non-overlapping FOV with error correction. The videogrammetric deformation measurement technique is well-suited to aircraft wings that exhibit torsional deformation during wind tunnel testing. In a low-temperature and high-pressure environment, conventional ellipse recognition methods have difficulty accurately identifying blurred fluorescent markers, which reduces measurement accuracy. The digital image correlation (DIC) method can greatly improve this situation by calculating subregion correlations to locate and track a blurred fluorescent marker. In this work, a high-precision videogrammetric measurement system for wind torsion analysis is proposed, by utilizing a hybrid monocular-binocular camera configuration based on DIC. The DIC method is used to locate and track the subpixel coordinates of fluorescent points. By calculating the three-dimensional coordinates of targets, tracking the trajectories of marker points and combining with the error correction model, the torsional deformation of the wing is accurately measured. Under actual wind tunnel conditions, the movement of a machete mechanism is introduced to accurately calibrate the wind tunnel coordinate system. The experimental results show that the root mean square error (RMSE) of fuselage torsion angle measurement is below 0.01°, and the RMSE of wingtip torsion is below 0.08°. This performance surpasses comparable studies at NASA Langley Research Center (0.2° wingtip position RMSE) and German Aerospace Center (0.1° RMSE).
Source: Measurement
Authors: Junzhe Liang, Guangxi Chen, Leigang Li, Feng Liang, Jinhong Wang, Shiquan Fan, Guohe Zhang
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,760
Award: G4219906238
OpenAlex ID: W4415601222
