Quantitative Detection of Rail Head Oblique Cracks by Laser Ultrasonic Surface Wave

Fatigue oblique cracks are the most common type of surface damage on steel rails, although they are microscopic and difficult to detect. Rail crack can occur as the crack length increases. This paper presents a method for detecting laser ultrasonic surface waves that is fully non-contact and unaffected by complicated surfaces. To solve the difficulties of low photoacoustic conversion efficiency and poor imaging impact of laser ultrasound, a solution combining particle swarm optimization (PSO) based VMD parameter optimization algorithm with Sobel image edge processing algorithm is presented. This enhances the signal-to-noise ratio (SNR) and image accuracy while also allowing for the quantitative identification of oblique fissures of varying lengths. The highest quantitative error is less than 4%, and the smallest quantitative error is 1.4%.