Ultrasonic arrays allow a given scatterer to be illuminated from a wide range of angles and hence are capable of extracting significant information about the scatterer. In this paper a general imaging methodology, termed multi-mode total focusing method, is proposed in which any combination of modes and reflections can be used to produce an image of the test structure. Like the total focusing method, this approach is implemented by post-processing the full matrix of array data to achieve a synthetic focus at every pixel in the image. A hybrid model is used to predict the array data and demonstrate the performance of the multi-mode imaging concept. This hybrid model combines far field scattering coefficient matrices with a ray-based wave propagation model. This allows the inclusion of longitudinal waves, shear waves and wave mode conversions. It is shown that, with prior knowledge of likely scatterer location and orientation, the mode combination and array location can be optimised to maximise the performance of array inspections. A practically relevant weld inspection application is then described and its optimisation is discussed.

Defect detection using ultrasonic arrays: The multi-mode total focusing method

The detection of product defects is essential in quality control in manufacturing. This study surveys stateoftheart deep-learning methods in defect detection. First, we classify the defects of products, such as electronic components, pipes, welded parts, and textile materials, into categories. Second, recent mainstream techniques and deep-learning methods for defects are reviewed with their characteristics, strengths, and shortcomings described. Third, we summarize and analyze the application of ultrasonic testing, filtering, deep learning, machine vision, and other technologies used for defect detection, by focusing on three aspects, namely method and experimental results. To further understand the difficulties in the field of defect detection, we investigate the functions and characteristics of existing equipment used for defect detection. The core ideas and codes of studies related to high precision, high positioning, rapid detection, small object, complex background, occluded object detection and object association, are summarized. Lastly, we outline the current achievements and limitations of the existing methods, along with the current research challenges, to assist the research community on defect detection in setting a further agenda for future studies.

Using Deep Learning to Detect Defects in Manufacturing: A Comprehensive Survey and Current Challenges.

Ultrasonic nondestructive evaluation is used for detection, characterization, and sizing of defects. The accurate sizing of defects that are of similar or less size than the ultrasonic wavelength is of particular importance in assessing structural integrity. In this paper, we demonstrate how measurement of the scattering coefficient matrix of a cracklike defect can be used to obtain its size, shape, and orientation. The scattering coefficient matrix describes the far field amplitude of scattered signals from a scatterer as a function of incident and scattering angles. A finite element (FE) modeling procedure is described that predicts the scattering coefficient matrix of various cracklike defects. Experimental results are presented using a commercial 64-element, 5 MHz array on 2 aluminum test samples that contain several machined slots and through thickness circular holes. To minimize the interference from the reflections of neighboring defects, a subarray approach is used to focus ultrasound on each target defect in turn and extract its scattering coefficient matrices. A circular hole and a fine slot can be clearly distinguished by their different scattering coefficient matrices over a specific range of incident angles and scattering angles. The orientation angles of slots directly below the array are deduced from the measured scattering coefficient matrix to an accuracy of a few degrees, and their lengths are determined with an error of 10%.

Defect characterization using an ultrasonic array to measure the scattering coefficient matrix

https://hal.archives-ouvertes.fr/hal-01322017/document

Plane Wave Imaging for ultrasonic non-destructive testing: Generalization to multimodal imaging

Sizing of flaws using ultrasonic bulk wave testing: A review

Crack detection and sizing technique by ultrasonic and electromagnetic methods

A partial-contact stress corrosion crack (SCC) is electrically modeled as a crack region with non-zero conductivity in eddy current testing (ECT). This partial-contact effect is excluded by an optimally designed crack-conductivity-insensitive depth characterization signal function (DCSF), and consequently the master curves obtained from electric-discharge machining (EDM) notches can be utilized directly in the depth sizing of SCCs. Furthermore, a crack conductivity independent artificial neural network (ANN) is constructed so that the entire depth profile can be reconstructed regardless of the crack conductivity. These two approaches are numerically validated and applied to the characterization of SCCs in SUS304 from measurement ECT signals. The average depth of each SCC is fast estimated from the DCSF, and the detailed depth profile is reconstructed from ANN. The ECT depth-sizing results show reasonable agreement with UT-TOFD measurement.

Depth sizing of partial-contact stress corrosion cracks from ECT signals

A procedure of applying the d-c potential drop technique using the closely coupled probes to NDE of a 3-D surface crack is newly developed. The calibration equation for three sensors which differ in the distance between the probes is derived. Experiments validated the use of the calibration equation for the NDE of cracks. The method to use the three sensors properly based on the measuring sensitivity is shown.

NDE of a 3-D Surface Crack Using Closely Coupled Probes for DCPD Technique

Improved ultrasonic testing by phased array technique and its application

A method is developed to evaluate nondestructively multiple cracks on the surface of materials. The method is based on the d.c. potential drop technique. Multiple two-dimensional cracks of unknown depth are inspected, where the distance between the cracks is known. First the distribution of the potential drop between both sides of each crack is measured on the cracked surface. Next the distribution of the potential drop is calculated numerically by assuming the depth of every crack. Then by comparing the measured and calculated potential drop, a correction factor is obtained for the assumed depth of the respective cracks. Modifying the crack depth by the correction factor is repeated until the difference between the measured and calculated distributions of the potential drop is minimized. It is shown that multiple cracks are sized accurately by the present method.

Ichiro Komura

Papers

Crack detection and sizing technique by ultrasonic and electromagnetic methods

Depth sizing of partial-contact stress corrosion cracks from ECT signals

NDE of a 3-D Surface Crack Using Closely Coupled Probes for DCPD Technique

Improved ultrasonic testing by phased array technique and its application

NDE of multiple cracks on the surface of materials by means of the potential drop technique