Propeller inspection is mandatory for the safe operation of aircraft. Non-contact damage evaluation on rotating structures requires dedicated measurement techniques. We report on a non-contacting stroboscopic technique that allows inspection of rotating aluminum propellers. To excite Lamb waves we used a Q-switched Nd:YAG laser in synchrony with data acquisition by a Laser Doppler Vibrometer. We detected, sized, and imaged a surface breaking notch on a sample rotating at 415 rpm. The technique showed potential for automatic non-contacting damage detection on rotating structures such as helicopter blades and turbines.
The samples, manufactured at the Department of Physics of the University of Helsinki were aluminum blades 130 mm x 76.10 mm x 4 mm in size. One of them was used as a baseline while the other one had a surface breaking rectangular defect that was 5 mm x 10 mm x 2.4 mm in size, situated 35.10 mm from the center of the propeller.
The propeller was rotated by a 12 V DC motor connected to a pulse width modulation circuit based on a 555 timer integrated circuit. To detect the movement of the rotating blade, a custom made optical gate was built by using common electronics such as blue LED, a LM311 comparator and a blue enhanced photodiode. All electronics were built in the Electronics Research Laboratory of the Department of Physics. Using C++ programming language we designed an algorithm for scanning the moving target. The code was implemented into an Arduino Mega 2560 microcontroller that was connected to a computer. The user operated the computer which controlled the microcontroller via a Labview program.
Two types of scans of the samples were performed. In one type the TX scanned the blade from edge to center while the LDV pickup spot was stationary at the center. From this experiment we were able to determine the distance from the notch to the center of the blade and the length of the defect from time-of-flight measurements. The second type of scan was done by situating the LDV pickup point close to the center of the blade and the excitation point at the far side of it, close to the edge. This time we did not scan with the lasers but we divided the arc path that they follow into even steps. From this experiment we were able to determine the width of the notch as well as its depth based on time-of-flight analysis. The results agree with the expected values.
Finally, by doing data selection of the delayed Lamb waves with Matlab R2015 and Blender 2.77a, we were able to do image reconstruction of the notch in a 3D model of the aluminum propeller.
