Abstract: Blade Tip Timing (BTT) is a promising non-contact method for monitoring blade vibration. Traditional BTT requires a "once-per-revolution" (OPR) sensor for a timing reference, which is often difficult to install in aero-engines due to space and environmental constraints. This research proposes an OPR-less BTT method for synchronous vibration identification. By using three BTT sensors and analyzing displacement differences, the method identifies vibration amplitude and engine order ($N_e$) through an improved two-parameter approach. Numerical simulations and experimental tests validate that this method accurately identifies vibration parameters, offering a viable technical solution for OPR-less BTT in engineering applications.
Introduction
Rotating blades are the heart of an aero-engine, but they are prone to synchronous vibration—vibration that occurs at integer multiples of the rotation speed. If left unmonitored, this can lead to fatigue, cracks, and catastrophic engine failure. Blade Tip Timing (BTT) is the gold standard for non-contact monitoring, using sensors on the engine casing to record when a blade tip passes. However, standard BTT systems require an OPR sensor on the main shaft to provide a timing and speed reference. In actual aero-engines, installing an OPR sensor is often impractical due to space constraints and harsh internal environments. This "reference sensor bottleneck" frequently prevents BTT from being used in the field.
Methodology: The OPR-less Approach
The researchers addressed this challenge by evolving the Two-Parameter Plot Method. Traditionally, this method uses two BTT sensors and one OPR sensor. The proposed innovation uses three BTT sensors and treats one as a "virtual" reference, effectively bypassing the need for a physical OPR sensor.
As illustrated in Figure 1, the traditional workflow (top) relies on OPR sensor data for speed and timing. The proposed workflow (bottom) utilizes an additional BTT sensor to estimate rotation speed and calculate displacement differences between sensors.
When these displacement differences are plotted against each other, they form an elliptical trajectory. By applying a Least Squares fitting algorithm (specifically the Levenberg-Marquardt algorithm) to this ellipse, the researchers can mathematically derive the two most critical vibration parameters: the amplitude ($A_{max}$) and the engine order ($N_e$).
Figure 4 demonstrates this fitting process, where the raw vibration data is mapped to a fitted elliptical curve to extract precise frequency and amplitude information.
Key Results and Validation
The method was validated through both numerical simulation and experimental testing:
- Numerical Simulation: For a simulated blade with a target engine order of 8 and amplitude of 1 mm, the system identified an $N_e$ of 8.07 and an amplitude of 1.01 mm—results nearly identical to traditional OPR-based methods.
- Experimental Success: On a specialized rotating disk rig equipped with three fiber-optic BTT sensors, the method successfully identified vibrations with an engine order near 30 and amplitudes around 163 $\mu m$.
Consistency between this OPR-less approach and the standard "Two-Parameter Method" proves that the OPR sensor can be safely removed without sacrificing data quality.
Conclusion
This research provides a significant practical advancement for aero-engine health monitoring. By eliminating the OPR sensor, BTT systems become much easier to install in cramped engine environments. While sensor layout remains critical to avoid "blind spots" where data might degenerate, the proposed method offers a robust, high-accuracy alternative for identifying synchronous blade vibrations in real-world engineering applications.
Summary updated