Objective  In laser cleaning, quality monitoring is the current focus of research, traditional quality inspection (visual inspection, chemical analysis, etc.) has the disadvantages of low inspection accuracy, high damage to the substrate, complicated process, high cost, and prone to human error., acoustic signal detection as a new detection method has non-destructive characteristics, without direct intervention in the substrate. In addition, in the cleaning process can be real-time detection, the cleaning process of the abnormal situation can be timely response, high sensitivity, easy to deploy and maintain. At present, the acoustic signal detection laser cleaning of the relevant research for the cleaning material is mainly rust (metal oxides), paint, oil and other inorganic substances, and the research focuses on the principle of detection and feasibility, has not yet established a relationship between acoustic signals and cleaning effect, less research on laser cleaning of marine microorganisms using signal detection, so this study is important for the management of metal-organic pollution of the ocean engineering.

Methods  This article builds an acoustic signal acquisition system (Fig.2). The system consists of a laser (model YLPN-10-30×240-200-R), a microphone, and a sound card (UR22C) to capture acoustic signals during laser cleaning of marine microbial on the surface of high-strength steel (30Cr3).First, the acoustic signal is pre-processed for noise reduction, then the Fourier transform is performed to plot the waveform in the time-frequency domain, and the four characteristic quantities of short-time average amplitude, Root-Mean-Square (RMS), Kurtosis factor, and Instantaneous frequency are extracted.Then a series of characterization of the metal surface before and after cleaning, such as the removal of thickness, roughness, and the degree of damage to determine the cleaning effect. The relationship between acoustic signals and cleaning quality is established through the correspondence between the characteristic quantities and the cleaning effect.

Results and Discussions  Through extracting the acoustic signal feature volume and corresponding to the cleaning effect characterization results, the results of the analysis show that the acoustic signal short-time average amplitude can reflect the thickness of microbial removal, the larger the short-time average amplitude, the larger the thickness of the microbial layer removed, the acoustic signal time-domain crag factor and the root-mean-square can reflect the degree of roughness of the cleaned surface, with the increase of the energy density, the crag factor firstly decreased and then increased, the root-mean-square value firstly increased and then With the increase of energy density, the Kurtosis factor decreases and then increases, the root mean square value increases and then decreases, and the inflection points of the two eigenvalues are close to each other, and the roughness reaches the minimum value at the inflection point, and the cleaning effect is good. Under the larger energy density, the instantaneous frequency of the acoustic signal can reflect the degree of laser damage to the substrate after cleaning, and the more low-frequency components in the transient frequency, the greater the damage to the substrate and the more the average hardness decreases.

Conclusions  There is a "main peak" signal in the frequency domain graph of the acoustic signal of short-pulse laser cleaning, The short-time average amplitude, Kurtosis factorr and root mean square, and instantaneous frequency of the acoustic signal can reflect the thickness of removal, roughness, degree of substrate damage, respectively. As shown in Fig.6, when the short-time average amplitude value is 0.013, the microbial layer on the surface is removed by 1 μm, which reaches the "start cleaning threshold". As shown in Fig.10, when the acoustic signal Kurtosis factor reaches 2.01, the cleaning effect is good and reaches the "optimal cleaning threshold".