Fatigue damage assessment method based on infrared polarization thermal image characteristics of metal surface

Abstract

The invention discloses a fatigue damage assessment method based on infrared polarization thermal image characteristics of a metal surface. The method comprises the following steps of: directly acquiring infrared polarization information images of a metal test piece at different azimuth angles in a fatigue process through an infrared polarization camera; performing polarization analysis on the acquired infrared polarization information images at different azimuth angles; segmenting a test piece area of the infrared polarization information images, representing the overall infrared polarizationcharacteristic quantity of the test piece by using a test piece area mean value, representing the difference between the local infrared polarization characteristics and the overall infrared polarization characteristics of the test piece by using a test piece area variance, and acquiring infrared polarization characteristic description in a metal fatigue damage process; performing principal component analysis on the infrared polarization characteristic quantity, and extracting characteristics closer to the fatigue damage degree of the test piece; and constructing a BP neural network by takingthe selected main characteristics as input, and performing fatigue damage prediction. According to the method, real-time, non-destructive and non-contact detection is realized, and the detection reliability can be improved by integrating the infrared thermal characteristics and the infrared polarization characteristics.

Description

technical field [0001] The present invention relates to the field of metal fatigue damage detection, in particular to a fatigue damage assessment method based on infrared polarized thermal imaging features of metal surfaces. Background technique [0002] Fatigue damage is the main form of failure of metal specimens. How to prevent the possible damage caused by the failure of metal specimens and how to qualitatively and quantitatively detect the degree of fatigue damage of metal specimens is an important topic in industrial non-destructive testing. [0003] The fatigue damage process of metal specimens is accompanied by changes in various characteristic quantities such as sound, heat, magnetism, and surface texture. Many scholars have proposed their own metal fatigue damage evaluation models for the above-mentioned characteristics. such as Liu Tao [1] Using thermodynamic entropy and temperature two-dimensional information entropy to extract thermal image features of metal su...

AI Technical Summary

Problems solved by technology

[0008] The above-mentioned detection methods have a good representation of metal fatigue damage in the corresponding feature dimensions, but there are still some deficiencies. First, when extracting feature quantities, such as surface texture and magnetic signals, specific contact detection equipment is required, which is difficult to detect. Achieve industrial non-destructive and real-time monitoring

Second, it is difficult to guarantee the reliability of fatigue damage using only acoustic, thermal, magnetic and other characteristics.

Therefore, infrared detection mainly distinguishes objects by brightness differences. When the radiation difference between objects is small, it will become difficult to use infrared characteristics to detect

Images