טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentArie Guy
SubjectSurface Morphology Evolution during Fatigue Crack Initiation
DepartmentDepartment of Mechanical Engineering
Supervisor Professor Emeritus Eli Altus
Full Thesis textFull thesis text - English Version


Abstract

Fatigue failure has been studied for more than 200 years, focusing mostly on the stage of crack propagation and final fracture. Interest on the first stage of damage, which includes local plasticity and micro-crack initiation, is growing in recent years. This important stage may span more than half life of the component's service.

Recently advanced high resolution devices have opened up possibilities for accurate measurements and the establishment of sophisticated models on micro-crack initiation. Many researchers have measured Persistent Slip Bands (PSB) during fatigue process and studied their evolution which ends up in the appearance of cracks.

The early assessment of fatigue damage may be used as a Non-Destructive-Test (NDT), detecting potential crack sites prior to their appearance. In this manner, the component could be disassembled and healed by thermal treatment, extending its service life. In addition, the time between inspections may be considerably increased, allowing longer continuous operation. These two issues may have a great economical influence on the aircraft industry, and crucial role in preventing failures and accidents.

In this study, the morphological properties of the surfaces and their evolution were investigated, conducting fatigue experiments on Aluminum-2024-T4. Using a White Light Interferometer (WLI) with high vertical resolution, Surface Topography (ST) was measured during the fatigue crack initiation stage.

Global bulk deformation and local PSB patterns, driven by different mechanisms, were detected and separated using the Fast Fourier Transform (FFT) and Blurring filter based technique. The deformation evolution was investigated with the aim of 'One-Point statistics', such as height distributions and roughness. Higher order stochastic tools, such as the 'Two-Point Correlation', were also implemented for additional information.

Distinct relation between the ST one point correlation, and the fatigue process was found. The ST variance evolution is similar to cumulative damage parameters, increasing with the number of cycles. The ST two point correlation (TPC) could easily identify the PSB orientation and the characteristic length of deformation heterogeneity. Clear ST changes could be detected right from the first few cycles for fatigue life of about 40k cycles.