
Knife-edge interferogram analysis for corrosive wear propagation at sharp edges
Author(s) -
Zhikun Wang,
ChaBum Lee
Publication year - 2021
Publication title -
applied optics
Language(s) - Uncategorized
Resource type - Journals
SCImago Journal Rank - 0.668
H-Index - 197
eISSN - 2155-3165
pISSN - 1559-128X
DOI - 10.1364/ao.417572
Subject(s) - enhanced data rates for gsm evolution , materials science , optics , interferometry , diffraction , edge detection , correlation coefficient , similarity (geometry) , computer science , image processing , artificial intelligence , physics , image (mathematics) , machine learning
This paper presents a novel noncontact measurement and inspection method based on knife-edge diffraction theory for corrosive wear propagation monitoring at a sharp edge. The degree of corrosion on the sharp edge was quantitatively traced in process by knife-edge interferometry (KEI). The measurement system consists of a laser diode, an avalanche photodiode, and a linear stage for scanning. KEI utilizes the interferometric fringes projected on the measurement plane when the light is incident on a sharp edge. The corrosion propagation on sharp edges was characterized by analyzing the difference in the two interferometric fringes obtained from the control and measurement groups. By using the cross-correlation algorithm, the corrosion conditions on sharp edges were quantitatively quantified into two factors: lag and similarity for edge loss and edge roughness, respectively. The KEI sensor noise level was estimated at 0.03% in full scale. The computational approach to knife-edge diffraction was validated by experimental validation, and the computational error was evaluated at less than 1%. Two sets of razor blades for measurement and control groups were used. As a result, the lag will be increased at an edge loss ratio of 1.007/µm due to the corrosive wear, while the similarity will be decreased at a ratio of 5.4×10 -4 /µ m with respect to edge roughness change. Experimental results showed a good agreement with computational results.