Precise measurement and atomic scale investigation of the growth of short cracks under Corrosion-Fatigue conditions

The continual need for gas turbines to operate with higher efficiency has resulted in the demand for increase in operating temperature of components throughout the gas stream. A consequence of this is that components that were once considered ‘low-risk’ are becoming susceptible to high-temperature corrosion, stress-corrosion and corrosion-fatigue damage and, in some instances, failure.

Introduction

The continual need for gas turbines to operate with higher efficiency has resulted in the demand for increase in operating temperature of components throughout the gas stream. A consequence of this is that components that were once considered ‘low-risk’ are becoming susceptible to high-temperature corrosion, stress-corrosion and corrosion-fatigue damage and, in some instances, failure. Understanding how cracks form and what drives or arrests them in these environments is critical to lifing and can have dire consequences if not understood. Current lifing predictions are based on empirical statistical crack growth rates and statistical assumptions of corrosion incubation. Although these models represent some improvement over recent years there is a desire to improve these predictions by improving mechanistic understanding and having actual crack growth measurements in these harsh environments.

Approach

The thesis will focus on two key areas of interest. Firstly, to understand the chemistry/mechanism at the atomistic scale as to what is initiating and driving short cracks that are generated. Ideally, this will be done on both engine run specimens and lab generated specimens and should enable the test to be optimised to replicate the ‘in-service’ damage as reliably as possible. It is hoped that the main output from this strand will be models that can more reliably predict short crack initiation and propagation. In addition, a possible output that may become apparent, is that if the crack growth mechanism is understood sufficiently, then it may be possible to influence it provided it is practicable to do so in a gas turbine engine.

Secondly, a potential difference measurement technique will be developed that can precisely measure crack growth rates on the test specimens. This will present many challenges due to the harsh environment, multiple cracks and corrosion of the surface. The work here, if successful, will have a significant benefit to OEM lifing algorithms that are currently based on empirical and extensive, relatively expensive, testing and post-test analysis.