A novel OPR-free method for blade tip timing based on adaptive variable reference blades

Monitoring rotor blade vibrations is imperative in aero-engine testing and service. Blade tip timing (BTT), a promising technique for such monitoring, assesses the operational status of rotor blades through non-intrusive measurement of blade tip displacement. Traditional BTT technology relies on the Once-Per-Revolution (OPR) probe, which poses installation challenges and introduces measurement uncertainties in practical applications. This paper introduces a novel OPR-free approach, employing Adaptive Variable Reference Blades (AVRB) for non-intrusive measurement of rotor blade vibrations. The method assumes a stochastic mistuning of blades within the same stage, resulting in non-simultaneous resonance of blades at a given speed. In each revolution, the vibration level of individual blades is determined by assessing the deviation between the average speed of a blade, as measured by two adjacent BTT probes, and the average speed across the entire circumference. Consequently, a non-vibrating blade is adaptively identified and selected as the reference blade. The relative vibration levels among blades vary dynamically during different revolutions, influenced by mistuning features, leading to variability in the reference blades. These adaptive variable reference blades can then replace the OPR probe, aiding in the computation of vibration displacement and further parameter identification of blades. To validate the effectiveness of the proposed method, a five-degree-of-freedom numerical model is utilized for simulation. Additionally, experiments involving both a five-bladed disk and an eight-bladed disk are conducted to validate the proposed approach. Numerical simulations and two sets of test results demonstrate a fundamental agreement between the blade vibration displacement calculated using the proposed method and the outcomes obtained through the traditional BTT method with OPR. The relative error in the identified blade vibration frequency derived from the blade vibration displacement, as determined by both methods, does not exceed 1 %. Additionally, the relative error in the amplitude of the blade vibration displacement is within the range of 10 %.