Development of Maximum Residual Stress Prediction Technique for Shot-Peened Specimen Using Rayleigh Wave Dispersion Data Based on Convolutional Neural Network.

Shot peening is a surface treatment process that improves the fatigue life of a material and suppresses cracks by generating residual stress on the surface. The injected small shots create a compressive residual stress layer on the material's surface. Maximum compressive residual stress occurs at a certain depth, and tensile residual stress gradually occurs as the depth increases. This process is primarily used for nickel-based superalloy steel materials in certain environments, such as the aerospace industry and nuclear power fields. To prevent such a severe accident due to the high-temperature and high-pressure environment, evaluating the residual stress of shot-peened materials is essential in evaluating the soundness of the material. Representative methods for evaluating residual stress include perforation strain gauge analysis, X-ray diffraction (XRD), and ultrasonic testing. Among them, ultrasonic testing is a representative, non-destructive evaluation method, and residual stress can be estimated using a Rayleigh wave. Therefore, in this study, the maximum compressive residual stress value of the peened Inconel 718 specimen was predicted using a prediction convolutional neural network (CNN) based on the relationship between Rayleigh wave dispersion and stress distribution on the specimen. By analyzing the residual stress distribution in the depth direction generated in the model from various studies in the literature, 173 residual stress distributions were generated using the Gaussian function and factorial design approach. The distribution generated using the relationship was converted into 173 Rayleigh wave dispersion data to be used as a database for the CNN model. The CNN model was learned through this database, and performance was verified using validation data. The adopted Rayleigh wave dispersion and convolutional neural network procedures demonstrate the ability to predict the maximum compressive residual stress in the peened specimen.

Nondestructive Evaluation of Residual Stress in Shot Peened Inconel Using Ultrasonic Minimum Reflection Measurement.

Shot peening is a process wherein the surface of a material is impacted by small, spherical metal shots at high velocity to create residual stresses. Nickel-based superalloy is a material with high strength and hardness along with excellent corrosion and fatigue resistance, and it is therefore used in nuclear power plants and aerospace applications. The application of shot peening to INCONEL, a nickel-based superalloy, has been actively researched, and the measurement of residual stresses has been studied as well. Previous studies have used methods such as perforation strain gauge analysis and X-ray diffraction (XRD) to measure residual stress, which can be evaluated with high accuracy, but doing so damages the specimen and involves critical risks to operator safety due to radiation. On the other hand, ultrasonic testing (UT), which utilizes ultrasonic wave, has the advantage of relatively low unit cost and short test time. One UT method, minimum reflection measurement, uses Rayleigh waves to evaluate the properties of material surfaces. Therefore, the present study utilized ultrasonic minimum reflectivity measurements to evaluate the residual stresses in INCONEL specimens. Specifically, this study utilized ultrasonic minimum reflection measurements to evaluate the residual stress in INCONEL 718 specimens. Moreover, an estimation equation was assumed using exponential functions to estimate the residual stress with depth using the obtained data, and an optimization problem was solved to determine it. Finally, to evaluate the estimated residual stress graph, the residual stress of the specimen was measured and compared using the XRD method.

Containment Liner Plate Void Defect Detection Technique Using Phased Array Ultrasonic Testing and Acoustic Resonance Method.

The CLP (containment liner plate) of a nuclear power plant protects the internal system from the external environment and sudden changes in internal pressure or temperature, and it is a structure that blocks and protects radioactive materials leaking inside and outside in the event of a nuclear accident and is composed of a liner plate, reinforcing bars, tendons, and concrete. Recently, corrosion on the rear side of the liner plate and concrete voids has emerged as a severe defect in nuclear power plants across South Korea. Therefore, in this study, we proposed a new inspection method that a line-type inspection method applied phased array ultrasonic testing and the area inspection method applied acoustic resonance method using developed moveable tapper. The acoustic signals were signal-processed and reproduced to a mapping image following the inspection area, and with the image, it was possible to determine the type of defect. Furthermore, an automated inspection system for within the CLP was proposed.

Field-deployable measurement technique for absolute acoustic nonlinearity parameter values.

Over the past two decades, several researchers have demonstrated that changes in the acoustic nonlinearity parameter ß can be related to changes in material properties due to mechanical and/or thermal degradation processes. Generally, a piezoelectric sensor-based detection method is used to measure changes in ß values because it is operationally simpler than the complex capacitive detection method. However, this method is limited to measuring only relative changes in ß values; whereas, the absolute ß values of components in service often need to be measured in the field to quantify the degradation level. Accordingly, a novel field-deployable method for measuring absolute ß values was developed in this study. Nonlinear ultrasonic experiments were conducted using capacitive detection, conventional piezoelectric sensor-based detection, and proposed detection methods, and the results were compared. The ß values of a copper single-crystal sample measured using the new and the capacitive detection methods were 2.49 and 2.1, respectively, and those obtained using the conventional piezoelectric sensor-based detection method ranged between 90 and 130. The test results confirm that the proposed field-deployable measurement method produces more consistent absolute ß values without involving the complexity of the capacitive detection method.

Propagation and Attenuation Characteristics of an Ultrasonic Beam in Dissimilar-Metal Welds.

Ultrasonic inspection of welds joining dissimilar metals in nuclear power plants has proven to be a challenge, because the ultrasonic waves are subject to diffraction, distortion, scattering, and noise. These perturbations are due to their interactions with coarse-grained microstructures having anisotropic and heterogeneous metallurgical properties that can promote ultrasonic attenuation. In this paper, to improve the reliability of ultrasonic testing for dissimilar-metal welds (DMWs), ultrasonic beam characteristics for DMWs with a buttering layer were investigated in order to analyze the beam distortion phenomenon caused by inhomogeneous anisotropic properties and coarse grains. Ultrasonic testing was performed on DMW specimens using single ultrasonic transducers to investigate the behavior of the ultrasonic beam in the welds. According to the anisotropic and heterogeneous properties, when passing through the weld and the buttering layer of the DMW, ultrasonic waves were distorted and attenuation was high. In particular, in the case of using angular incidence that passed through the weld and the buttering layer in turn, the received ultrasonic data did not contain accurate internal information. From this, it was verified that internal defects may be detected by transmitting ultrasonic waves in different directions. Finally, the existing limitations on the application of non-destructive ultrasonic testing to dissimilar-metal welds were verified, and a solution to the measurement method was proposed.

Convolutional neural network for ultrasonic weldment flaw classification in noisy conditions.

Ultrasonic flaw classification in weldment is an active area of research and many artificial intelligence approaches have been applied to automate this process. However, in the industrial applications, the ultrasonic flaw signals are not noise free and automatic intelligent defect classification algorithms show relatively low classification performance. In addition, most of the algorithms require some statistical or signal processing techniques to extract some features from signals in order to make classification easier. In this article, the convolutional neural network (CNN) is applied to noisy ultrasonic signatures to improve classification performance of weldment defects and applicability. The result shows that CNN is robust, does not require specific feature extraction methods and give considerable high defect classification accuracies even for noisy signals.

Nonlinear ultrasonic characterization of early degradation of fatigued Al6061-T6 with harmonic generation technique.

In this paper, a third harmonic was used to investigate microstructural changes in Al6061-T6 due to different fatigue cycles and a relationship between fatigue cycle and third order nonlinearity has been observed. Piezoelectric measurement harmonic generation technique was applied for the specimens with 0%, 55%, 75% and 85% fatigue cycles, respectively. The results shows that the third order harmonics gradually increased up to 55% and rapidly decreased after wards, it was attributed to the behavior of dislocation, dislocation-precipitation interaction and voids with increasing fatigue cycle. Further, it was verified with scanning electron microscope (SEM). We also observed that third order nonlinearity is more sensitive to small change in area of fraction of voids than second order nonlinearity after 55% fatigue life and could be a good candidate to investigate Al6061-T6 specimen with voids.

Characterization of TiN coating layers using ultrasonic backward radiation.

Since ceramic layers coated on machinery components inevitably experience the changes in their properties it is necessary to evaluate the characteristics of ceramic coating layers nondestructively for the reliable use of coated components and the remaining life prediction. To address such a need, in the present study, the ultrasonic backward radiation technique is applied to examine the very thin TiN ceramic layers coated on AISI 1045 steel or austenitic 304 steel substrate. Specifically, the ultrasonic backward radiation profiles have been measured with variations in specimen preparation conditions such as coating layer thickness and sliding loading. In the experiments performed in the current study, the peak angle and the peak amplitude of ultrasonic backward radiation profile varied sensitively according to two specimen preparation conditions. In fact, this result demonstrates a high possibility of the ultrasonic backward radiation as an effective tool for the nondestructive characterization of the TiN ceramic coating layers even in such a thin regime.

An ultrasonic measurement model using a multi-Gaussian beam model for a rectangular transducer.

To date, ultrasonic measurement models have primarily treated systems where circular transducers are used. Recently, however, a highly efficient ultrasonic beam model for a rectangular transducer has also become available where the transducer is represented as a superposition of a relatively few Gaussian beams. Thus, using the multi-Gaussian beams, we developed ultrasonic measurement models for systems where a rectangular transducer is employed. In this paper, we describe the developed models including the beam model, the efficiency factor for a rectangular transducer and far-field scattering models for some standard scatterers. Furthermore, the accuracy of the proposed model is verified by the comparison of the model-based predictions to the experimental measurements.

Evaluation of CVD diamond coating layer using leaky Rayleigh wave.

In the present study, the possibility of using leaky Rayleigh waves as a nondestructive tool for the evaluation of CVD diamond coating layer is explored experimentally. For this purpose, a set of CVD diamond coated specimens are prepared and the leaky Rayleigh waves are measured in an immersion, pulse-echo setup. For the proper analysis of the acquired signals we propose a novel signal analysis approach, namely the "time trace angular scan (TTAS)" image. Then, the proposed approach together with the backward radiation profiles are applied for the analysis of signals acquired in the initial experiments. The TTAS image shows the entire information on both time-of-arrival and angle of incidence of the signals for the proper "time-angle windowing." Then, the backward radiation profile of the windowed signals provides adequate parameters from which nondestructive evaluation of the coated specimens is carried out.

A new technique for the identification of ultrasonic flaw signals using deconvolution.

The identification of ultrasonic flaw signals is a difficult task in the angle beam ultrasonic testing of welded joints due to the presence of non-relevant signals from the geometric reflectors such as weld-roots and counter-bores. This paper describes a new approach to identify ultrasonic flaw signals in such a problematic situation. A similarity function is defined as the deconvolution of a target signal by a reference signal. The similarity functions for the same type of flaws/references are symmetric bandlimited impulse-like patterns with larger amplitudes while those for different types of flaws/references are asymmetrical broad patterns with relatively smaller amplitudes. Therefore, ultrasonic signals could be identified by the pattern of the similarity function. In the initial experiments, the proposed technique showed great potential for identifying ultrasonic flaw signals in the inspection of weld joints.

Comparison of modeling approaches to ultrasonic testing at near critical angles.

Modeling of ultrasonic testing has been paid a great attention in nondestructive evaluation community recently since it can provide thorough understanding of underlying physics of ultrasonic testing. As a result, there have been developed various modeling approaches up to now. Especially, many practical models have been developed based on either the multi-Gaussian beam or the Rayleigh-Sommerfeld integral. This paper discusses the modeling of ultrasonic testing with oblique incidence at the near critical angles using these two approaches. The theoretical models that can predict the reflection signals from side drilled cylindrical holes in solid specimen immersed in water are developed. Then, the theoretical predictions for the oblique incidence at the near critical angles are compared to the experiments for the investigation of model behavior.

Assessment of material degradation due to corrosion-fatigue using a backscattered Rayleigh surface wave.

Material degradation due to corrosion-fatigue was evaluated nondestructively using backscattered Rayleigh surface wave. A corrosion-fatigue test was carried out for the specimens made of thermo-mechanically controlled process steel in 3.5 wt.% NaCl solution at 25 degrees C. The Backscattering profile, which is the amplitude variation of backscattered ultrasound according to the incident angle, of the specimens were measured in water at room temperature after the corrosion-fatigue test. The velocity of Rayleigh surface wave, determined from the incident angle at which the profile of the backscattered ultrasound became maximum, decreased for the specimen that had the large number of cycles to failure in the corrosion-fatigue test. This fact implies that the corrosion degradation occurred at specimen surface in this specific test is dominantly dependant on the time exposed to corrosion environment. The result observed in the present work demonstrates the high potential of backscattered Rayleigh surface wave as a tool for nondestructive evaluation of corrosion degradation of aged materials.

Determination of ultrasonic wave velocities and phase velocity dispersion curves of an Inconel 600 plate using resonant ultrasound spectroscopy and leaky Lamb waves.

A plate of Inconel 600 was interrogated using the resonant ultrasound spectroscopy (RUS) and the reflected leaky Lamb waves (LLW). It was found that the plate used in the present work has anisotropy in its material properties by the RUS. The longitudinal and the transverse wave velocities of the Inconel 600 plate were determined by the RUS, ultrasonic pulse-echo method and cut-off frequencies of the LLWs. The wave velocities in the direction of thickness determined by the RUS under the assumption of the orthotropic symmetry were quite similar to those obtained by other methods, the pulse-echo method and from cut-off frequencies. The reflected LLW from the plate was measured with varying the incident angle. The dispersion curves obtained from the reflected LLWs show good agreement with the theoretical calculation in general. The mismatches may be caused by anisotropy of the plate.

Evaluation of beryllium stiffening on zirconium plate using backscattered leaky Lamb waves.

The ultrasonic backscattering profiles of the Zr plates with/without Be-Zr alloy layer were measured at various incidence positions to evaluate the embedded Be diffusion layer on the Zr plate. Four principal subprofiles of backward ultrasound radiated from leaky Lamb waves were observed. The angles and the intensities of the subprofile peaks decreased due to the stiffening effect of Be layer. The generation and change of subprofiles can be explained by the influence of acoustical properties, collective group velocities and leaky factors of plates. It is suggested that the backward radiation subprofile is an effective tool for the evaluation of thin diffusion layers on plates.

Simulation of 3-D radiation beam patterns propagated through a planar interface from ultrasonic phased array transducers.

Phased array transducers are quite often mounted on solid wedges with specific angles in many practical ultrasonic inspections of thin plates <10 mm in their thickness or welded joints with convex crowns. For the reliable application of phased array techniques with testing set-up, it is essential to have thorough understanding on the characteristics of radiation beam pattern produced in the interrogated medium. To address such a need, this paper proposes a systematic way to calculate full 3-D radiation beam patterns produced in the interrogated solid medium by phased array transducers mounted on a solid wedge. In order to investigate the characteristics of radiation beam patterns in steel, simulation is carried out for 7.5 MHz array transducers mounted on an acrylic wedge with the angle of 15.45 degrees with various of steering angles and/or focal planes.