Acoustic Scattering Models from Rough Surfaces: A Brief Review and Recent Advances

Eddy Current Sensors Optimization for Defect Detection in Parts Fabricated by Laser Powder Bed Fusion

The production of parts by additive manufacturing is an important issue for the reduction in manufacturing costs and the creation of complex geometries. Optical inspection is often implemented in the machines during the manufacturing process in order to monitor the possible generated defects. However, it is also crucial to test the quality of the manufactured parts after their fabrication and monitor their health throughout their industrial lifetime. Therefore structural health monitoring (SHM) methods need to be studied or designed. In this paper, the eddy current method is used to control fabricated parts, as this technique is adapted to detect surface and shallow defects in conductive materials. Using simulations with the CIVA non-destructive testing software package, several sensors and their parameters were tested in order to determine the most optimal ones: a separate transmitter/receiver sensor and an isotropic sensor were finally designed. The comparison of these sensors' efficiency was made on the detection of notches and engraved letters based on simulation and experimental tests on parts fabricated by laser powder bed fusion (L-PBF) in order to determine the optimal sensor. The various tests showed that the isotropic sensor is the optimal one for the detection and characterization of defects.

Effects of atmospheric refraction on detecting elevated sources with beamforming algorithms

Classical beamforming algorithms are based on rectilinear propagation of signals that are corrupted by additive noise. However, sound waves propagating in the atmosphere are refracted by temperature and wind gradients and distorted by turbulence and scattering from the ground. These effects can reduce the performance of acoustic systems. To investigate these effects, a simulation capability is used to propagate signals from elevated sources to an array of microphones near the ground. Detection algorithms based on four standard beamforming algorithms, implemented with either no model mismatch error or model mismatch errors generated by atmospheric refraction, are tested. The results are evaluated using the area under the receiver operating characteristic curve statistic. The analysis shows no significant improvement in detection performance when refraction effects are included in the beamforming algorithms over a wide range of frequencies. Any small performance improvements achieved by incorporating refraction effects into the beamforming algorithms are overwhelmed by the negative effects of atmospheric turbulence and rough surface scattering. For the selected scenario, incorporating refraction effects into the beamformers is over-engineering. These results demonstrate some of the limitations of developing new signal processing algorithms or array designs for atmospheric acoustic systems.

On Average Losses of Low-Frequency Sound in a Two-Dimensional Shallow-Water Random Waveguide

For a low-frequency sound signal propagating in a two-dimensionally inhomogeneous shallow-water waveguide, the influence of random bathymetry (fluctuating bottom boundary) was considered based on the local-mode approach and statistical modeling using first-order evolution equations. The study was carried out in shallow sea conditions corresponding to the coastal waveguides of the Russian Arctic seas. Here, a feature was the presence of an almost homogeneous water layer with various characteristics of seabed sediments. To describe the latter, a random model of the impedance was adopted. For the conditions of a strongly penetrable bottom boundary, on average, the calculations predicted adequate weak effects of bathymetry fluctuations on the average sound intensity compared to the effect of fluctuations in the sediment parameters and volumetric random inhomogeneities of the water column. In addition, it was shown that, in terms of statistics, the roughness of the bottom boundary perturbed the average sound intensity in a shallow-water waveguide differently than volumetric fluctuations in the speed of sound. The dependence of the statistical effects (the first and second moments of the signal intensity) on the parameters of the waveguide and the frequency range was studied. As a result of numerical modeling, comparative quantitative estimates of the influence of both the random roughness of the bottom interface and fluctuations of bottom sediment parameters on the average losses of the propagating signal, not presented in the literature, were obtained.

Towards Explainable Augmented Intelligence (AI) for Crack Characterization

Crack characterization is one of the central tasks of NDT&E (the Non-destructive Testing and Evaluation) of industrial components and structures. These days data necessary for carrying out this task are often collected using ultrasonic phased arrays. Many ultrasonic phased array inspections are automated but interpretation of the data they produce is not. This paper offers an approach to designing an explainable AI (Augmented Intelligence) to meet this challenge. It describes a C code called AutoNDE, which comprises a signal-processing module based on a modified total focusing method that creates a sequence of two-dimensional images of an evaluated specimen; an image-processing module, which filters and enhances these images; and an explainable AI module—a decision tree, which selects images of possible cracks, groups those of them that appear to represent the same crack and produces for each group a possible inspection report for perusal by a human inspector. AutoNDE has been trained on 16 datasets collected in a laboratory by imaging steel specimens with large smooth planar notches, both embedded and surface-breaking. It has been tested on two other similar datasets. The paper presents results of this training and testing and describes in detail an approach to dealing with the main source of error in ultrasonic data—undulations in the specimens’ surfaces.