Multiaxial Fatigue Damage Prediction and Life Estimation of a Centrifugal Impeller for a Turboshaft Engine

Abstract

The rotating gas turbine engine component such as centrifugal compressor is subjected to various static and dynamic loads during service. These service loads introduce multiaxial stress-strain states and inflict severe structural damage and hence reduce compressor efficiency and service life. In many cases, the service loads are non-proportional which introduces additional cyclic hardening depending on the loading path and the material used (Wu et al., Int J Fatigue 59:170–175, 2014). Studies have revealed that fatigue life is drastically reduced as a result of this additional cyclic hardening phenomenon. In recent decades, numerous multiaxial fatigue damage theories (which are strain based) have been developed taking into account the factors influencing fatigue like operating temperature, anisotropy, microstructure, nature of loading (proportional or non-proportional), and environmental effects. In general, these theories envisage crack nucleation and propagation to take place at a critical plane, but differ from one another the way critical planes are defined and damage parameters are proposed. To estimate the fatigue life under the multiaxial loading conditions, these theories relate multiaxial stress-strain components to uniaxial fatigue properties obtained from test data. But, it is observed that most of these theories are limited in application to a particular material or loading conditions. No general consensus is available on the best available fatigue damage theory for a particular application. Further, literature shows that most of these theories are applied and compared on tubular and notched test specimens under proportional and non-proportional loading conditions. Literatures are seldom available which compares these theories at the application level and validate through experiments. The objective of this paper is to estimate the low-cycle fatigue life of a centrifugal impeller used in a gas turbine application using different multiaxial fatigue damage models in the literature and propose a methodology which closely fits the experimental results. For applications such as gas turbine engines, the margin of safety (with respect to 0.2% PS) varies in the range of 1.05 to 1.15 (Stephens et al. in Metal fatigue in engineering. Wiley, New York, 2001). For this, it is extremely important to estimate the life and damage mechanism affecting the life of the components accurately to ensure the flight safety, reliability, and total service life of the gas turbine engine. The centrifugal impeller was designed for a pressure ratio of 3.2 and maximum rotational speed of 52,500 rpm. The titanium alloy Ti6Al4V was chosen as the possible material for the impeller as it has high strength and stiffness to weight ratio and is widely used in aerospace application (Wu et al., Int J Fatigue 44(12):14–20, 2012). The entire analysis can be divided into three main modules. Firstly, the stress and strain history on the centrifugal impeller was determined under actual operating condition. Finite element structural analysis packages were used for this purpose. Secondly, the stress-strain histories obtained from FE analysis were post-processed using several multiaxial fatigue damage models like Maximum von Mises strain model (Kalluri and Bonacuse in In-Phase and Out-of-Phase Axial-Torsional Fatigue Behavior of Haynes 188 at 760 °C. NASA Technical Report 1991, 91-C-046, 1991), which is based on the equivalent strain and the Smith-Watson-Topper (SWT) model (Shamsaei and McKelvey, Int J Fatigue 59:170–175, 2014), Kandil-Brown-Miller (KBM) model (Kandil et al., Met Struct 280:203–210, 1988), Fatemi and Socie (FS) model (Fatemi and Socie, Fatigue Fract Eng Mater 11(3):149–166, 1987), which are based on the critical plane approach to obtain the fatigue life. Finally, a prototype of the centrifugal impeller was made and tested at our sophisticated cyclic spin test facility to validate the Low-Cycle Fatigue life.