Accurate Blade Tip Timing Placement on a Centrifugal Impeller Using As-Manufactured Modeling

Abstract

Non-intrusive stress measurement systems (NSMS) are commonly used during rig and engine tests to ensure safe operation of the test asset. Blade tip timing (BTT) is one form of NSMS that estimates operational stresses of a bladed rotor using time-of-arrival (TOA) data coupled with finite element analysis (FEA) predictions. Traditional FEA techniques assume nominal airfoils to generate stress-to-deflections ratios used with TOA data to predict blade stresses. Recent research has been conducted showing the significant variability in stress-to-deflection ratios when accounting for geometric variations in the blade geometry. As-manufactured finite element modeling has been shown to be a prudent way to account for these geometric variations when developing instrumentation placements and safety limits for rotors. Literature on this topic tends to focus on integrally bladed rotors where every blade is notionally identical. Centrifugal impellers alternatively have a main and splitter blade with significantly different geometries making blade tip timing placement more challenging on these components. Traditional BTT placement techniques will be tailored in this work to achieve optimal probe placement to detect specific vibratory modes of interest for both the main and splitter blades. BTT safety limits will be generated for each mode of interest, and optical scanning and mesh morphing approaches will be used to generate an as-manufactured model of the centrifugal impeller. The BTT safety limit variability due to geometric variations will be assessed for both the main and splitter blades. It will be shown that due to geometric variations within the impeller that the BTT limits can vary between blades. This research further validates the importance of using as-manufactured modeling during component design, test, and throughout the life-cycle.