The purpose of this paper is to automate the interpretation of data during ultrasonic diffraction using a non-destructive testing (NDT) technique to accurately size defects for assisting in decision-making. A convolutional neural network (CNN) architecture was developed to automatically measure the length of the defect. Using the architecture, the population of A-scan signals in the scanning path was classified. The defect region was extracted and its size in the scanning direction was obtained by the connected region solution algorithm based on the classification results. The arrival time of diffraction waves was accurately identified by the intelligent denoising framework proposed, combined with Hilbert transform, and then the height of defects was calculated by corresponding geometric relations. The estimation results demonstrate that the measurement method can be considered as a useful technique for crack sizing in industrial structures, even in the case of complex noise.

Traditional beat-frequency quartz-enhanced photoacoustic spectroscopy (BF-QEPAS) is constrained by short energy accumulation time and the necessity of attenuation cycles, resulting in weaker signals and longer measurement cycles. In this work, we propose a novel scanning optomechanical enhanced (SOE-) BF-QEPAS technique to achieve ultrafast gas detection. By periodically scanning the modulation signal, we supplement excitation energy before the attenuation of the tuning fork’s vibrational energy, thereby optimizing energy accumulation efficiency. For the first time, we introduce a comprehensive quality factor K to evaluate the integrated performance of the system in terms of energy efficiency and signal-to-noise ratio (SNR). After parameter optimization, the SNR demonstrated a 1.88-fold improvement compared to conventional BF-QEPAS, while the response speed increased by 200-400 times.

This paper proposes a method for the detection and depth assessment of tiny defects in or near surfaces by combining laser ultrasonics with convolutional neural networks (CNNs). The innovation in this study lies in several key aspects. Firstly, a comprehensive analysis of changes in ultrasonic signal characteristics caused by variations in defect depth is conducted in both the time and frequency domains, based on discrete frequency spectra and original A—scan signals. Continuous wavelet transform (CWT) is employed to obtain wavelet time–frequency maps, demonstrating the consistent characteristics of this image with crack depth variations. A crucial innovation in this research involves the targeted design and optimization of the model based on the characteristics of ultrasonic signals and dataset size. This includes aspects such as data preparation, CNN architecture construction, and hyperparameter selection. The model is tested using a random validation set, which effectively demonstrates the CNN model’s validity and high precision. The proposed method enables the recognition and depth assessment of tiny defects on or near surfaces.

The reliable detection of embedded defects is critical for ensuring the structural integrity of metallic components produced by additive manufacturing (AM). However, strong acoustic attenuation in AM materials severely limit the performance of conventional ultrasonic inspection techniques. In this study, a laser–piezo hybrid ultrasonic focusing method based on diffracted shear (DS) waves is proposed to improve defect detectability in metal AM structures. In this approach, a bottom-mounted piezoelectric transducer excites longitudinal ultrasonic waves, while a laser Doppler vibrometer scans the upper surface to acquire transmitted signals. Finite element simsulations reveal that longitudinal waves interacting with internal defects generate DS waves that carry strong feature information related to defect size and position. By leveraging these wave characteristics, a synthetic aperture focusing algorithm was developed to selectively reconstruct DS wave fields and achieve high-resolution imaging of embedded defects. Experiments were conducted on selective laser melting 316L stainless-steel samples containing pre-fabricated through-hole defects. The proposed method achieved clear imaging of subsurface defects with diameters as small as 0.5 mm at depths up to 25 mm. The results demonstrate that focusing on DS waves enables accurate localization of deep-lying defects and provides a potential framework for future in-situ ultrasonic monitoring in metal AM.

Metal additive manufacturing (AM) technology enables the direct fabrication of complex components. However, the formation of internal defects remains a significant challenge, limiting the consistent production of high-quality key parts. Online monitoring of internal defect formation during the manufacturing process is critical for ensuring quality control and enhancing the practical application of AM products. Laser ultrasound has demonstrated significant potential for rapid defect scanning and detection, owing to its non-contact and multimodal capabilities. Additionally, the layer-by-layer accumulation process characteristic of additive manufacturing (AM) provides a flat detection surface during the printing of each layer, effectively addressing the challenges posed by the limited detection space in complex components. Compared to other non-destructive testing (NDT) methods, ultrasonic testing (UT) further enables the correlation of online monitoring data with offline inspection results. However, this approach faces certain limitations. The rough surfaces and complex geometries of additive samples lead to high attenuation, low signal-to-noise ratio (SNR) in ultrasonic signals, and interference from structural waves. Under these conditions, extracting effective defect signals and visually displaying internal defects remain challenging. We propose an ultrasonic online monitoring method that employs a bottom piezoelectric sensor to generate excitation signals and constructs a scanning array on the surface using a laser Doppler vibrometer. This approach enables the timely acquisition of effective ultrasonic signals across varying printing layer thicknesses during the additive manufacturing process while mitigating interference from structural waves. To address the challenges of high ultrasonic attenuation and low signal-to-noise ratio (SNR) caused by the complex geometries of additive samples, the interaction between ultrasound and defects was simulated using the finite element method. Defect characteristic signals were identified and integrated into a proposed signal classification and focusing imaging algorithm, facilitating defect signal extraction and reconstruction. 

Crack formation remains a major barrier to laser powder bed fusion (LPBF) of the high-γ′ Ni-based superalloy MAR-M247 because of severe cyclic thermal loading and interlayer heat accumulation. In this work, an active thermal relaxation (ATR) strategy was introduced into LPBF of MAR-M247 by incorporating an intentional interlayer dwell to regulate the thermal history during multilayer deposition. Through physics-based estimation and experimental screening, a dwell time (Δt) of 10 s was identified as the optimal condition for defect suppression. Under this condition, ATR mitigated heat accumulation, narrowed the heat-affected zone, and reduced residual stress, thereby decreasing the porosity from ~1.52% to ~0.8% and reducing the crack number to only 2–3 in the evaluated regions. Meanwhile, ATR promoted a stronger texture and more continuous columnar solidification features by shifting the solidification path away from the columnar-to-equiaxed transition (CET). These changes in defect state and microstructure resulted in a significant enhancement of tensile performance, with the ultimate tensile strength increasing from 791 MPa to 1027 MPa at room temperature and from 606 MPa to 708 MPa at 900 °C. This work demonstrates that ATR provides a simple and effective thermal-management route for improving the structural integrity and tensile performance of LPBF-fabricated MAR-M247.