Qualifying parabolic mirrors with deflectometry

This paper proposes a new shearing approach to eliminate the systematic errors in phase measuring deflectometry (PMD). Besides a normal measurement using PMD, another three measurements are taken when the test surfaces are translated and rotated with an amount of known magnitude in our proposed method. The four measurements give three sheared wavefronts, and then the difference Zernike polynomial fitting and a least squares algorithm are proposed to reconstruct the true test surface figure. The proposed method is demonstrated and verified through computer simulation and experimental results. The error propagation when the sheared wavefronts include random noise and the introduced errors due to the incorrect shearing values are both analyzed.

Phase Measuring Deflectometry (PMD) and Structured-Light Modulation Analysis Technique (SMAT) perform effectively in shape and defect measurements of specular objects, but the difficulty of giving consideration to accuracy and speed has also restricted the further development and application of them. Inspired by recent successes of deep learning techniques for computational imaging, we demonstrate for the first time that deep learning techniques can be used to recover high-precision modulation distributions of specular surfaces from a single-frame fringe pattern under SMAT, enabling fast and high-quality defect detection of specular surfaces. This method can also be applied to recover higher-precision phase distributions of specular surfaces from a single-frame fringe pattern under PMD, so as to realize the 3D shape measurement. In this paper, we combine depthwise separable convolution, residual structure and U-Net to build an improved U-Net network. The experimental results prove that the method has excellent performance in the phase and modulation retrieval of specular surfaces, which almost reach the accuracy of the results obtained by ten-step phase-shifting method.

Street defects, such as potholes and sunken manholes, in general develop quickly compared to other pavement distresses, such as cracking and rutting. Those street defects can result in vehicle damage. This paper proposes an automated and innovative method to obtain up-to-date information about those street defects with the use of a mobile data collection kit mounted on vehicles. In each mobile data collection kit, a triaxial accelerometer and global positioning system sensor collect data for the detection of street defects. A local algorithm is embedded in the mobile data collection kit to increase the efficiency of a local data logging process and to perform a preliminary detection of street defects. At a back-end server, a more precise street defect detection algorithm enhances the performance of the proposed monitoring system by integrating data collected from multiple sensor-equipped vehicles. The street defect detection algorithm at the back-end server relies on a supervised machine learning technique and a trajectory clustering algorithm. The framework of the data collection and integration is developed for the detection of isolated street defects and rough road conditions. The potential of detecting these conditions based on the dynamic responses of vehicles using machine learning techniques is investigated on real road conditions. The preliminary ratings for pavement distress are calculated by integrating the three classification results. Road networks that have isolated street defects and rough road surfaces are identified and visualized on an online map. The proposed system is of practical importance since it provides continuous information about road conditions, which can be valuable for pavement management systems and public safety.

The implementation of online, nondestructive quality control (QC) solutions is a key aspect to consider especially in medical device manufacturing, where the production process should comply to relevant quality-related standards to guarantee a safe product for the patient. For metallic implants fabricated using electron beam melting (EBM), the presence of defects (e.g., porosities, cracks, delamination, balling effect) could affect the final part quality in terms of structural integrity and mechanical properties. In this context, we propose an expert system algorithm capable of automated detection of porosities that could occur in prosthetic components, such as tibial trays, printed using a cobalt-chromium alloy. The choice of focusing only on the porosities was done as these defects are particularly critical for orthopedic prosthesis, where they can negatively impact the fatigue behavior of the printed component. The algorithm was developed ad hoc to analyze images of the manufacturing process taken from a camera embedded in the printer (Arcam Q10plus). Images can be used to identify porosities, performing a nondestructive evaluation that supports the process of part qualification. The developed algorithm automates and improves the visual inspection task conducted by human experts, including quantitative assessment on the size and location of the porosities, and reporting the presence of high porosity density areas. The defect detection performance was evaluated through the design of two tasks: The layer-wise defect detection and the large pore identification. The defect detection was performed with a sensitivity of 91% and a precision of 76%. Furthermore, a comparison with the gold-standard nondestructive evaluation technique, that is, computed tomography evaluation, allowed to validate the algorithm (percent agreement of 98%). The developed expert system allows to quickly evaluate an entire printing volume with several components representing a reliable and fast tool for defect detection and QC of EBM-printed prosthetic components.

Frontiers in NDE research nearing maturity for exploitation to ensure structural integrity of pressure retaining components

Metal Additive Manufacturing (AM) is a pillar of the Industry 4.0, with many attractive advantages compared to traditional subtractive fabrication technologies. However, there are many quality issues that can be an obstacle for mass production. The in-situ camera-based monitoring and detection of defects, taking advantage of the layer-by-layer nature of the build, can be an effective solution to this problem. In this context, the use of Computer Vision and Machine Learning algorithms have a very important role. Nonetheless, they are up to this date limited by the scarcity of data for the training, as well as by the difficulty of accessing and integrating the AM process data throughout the fabrication. To tackle this problem, this article proposes a system for in-situ monitoring that analyses images from an off-axis camera mounted on top of the machine to detect the arising defects in real-time, with automated generation of synthetic images based on Generative Adversarial Network (GAN) for dataset augmentation purposes. The computing functionalities are embedded into a holistic distributed AM platform allowing the collection, integration and storage of data at all stages of the AM pipeline.

Ultrasonic technique has been increasingly used in recent years for nondestructive evaluation of defects and material properties (e.g., stiffness) in structural members made up of concrete, steel, wooden, and composite materials. Ultrasonic evaluation of structural members in the past has relied primarily on time-of-flight and velocity measurements. While these measurements lead to fair estimate of stiffness and detection of defects in structural members, the sensitivity of the measurements is not adequate to detect defects in their early stages. In order to enhance the sensitivity of the ultrasonic technique, this paper proposes to use ultrasonic signal amplitude measurements in conjunction with traditional velocity measurements. The conditions necessary to make repeatable amplitude measurements are described, and the data analysis techniques suitable for defect detection are discussed. Also, the advantages of signal amplitude measurement technique are presented using examples of defective steel, wooden, and composite members. The measurement and analysis techniques presented here can be easily generalized to structural members made from other materials such as concrete.

This paper presents a new approach to improve multiple choice and defect detection in cross-border shipments using deep learning (DRL). The design process involves the integration of real-time data from multiple sources to create comprehensive transportation models, including route optimization, cost reduction, and poor research methods. The DRL project is intended to use a multi-agent design to manage complex decision-making processes in a dynamic logistics environment. The hybrid anomaly detection system combines statistics with machine learning techniques to identify and respond to network disruptions. The system's performance was validated using a database including 185,432 shipment records collected over 24 months across the Asia-Pacific transportation system. The experimental results show that the proposed method has achieved 94.5% correct value in choosing the right path and 45% reduction in processing time compared to traditional methods. The negative detection antibody maintains a 96.2% true positive rate with a 1.8% false positive rate. The system's analysis shows that the growth of the needs in the calculation of the growth in the network, indicating the use of good resources in the large deployment. This research supports the state-of-the-art in cross-border business optimization by providing solutions that integrate real-time optimization methods with negative detection and response mechanisms.

Ultrasonic non-destructive evaluation of composites: A review

Active thermography is a widely used non-destructive testing and evaluation technique (NDT&E) for evaluating the properties of materials without impairing its future usefulness. In this work, a mild steel sample made of artificial flat-bottom holes at varied depths, was examined with the emerging non-stationary thermal wave imaging (NSTWI) technique, i.e. frequency modulated thermal wave imaging (FMTWI). The pulse compression favorable of NSTWI technique is eminent for compressing the applied thermal energy into a narrow-compressed pulseto enhance the depth resolution and sensitivity. In this work, pulse compressed thermographic data generated from FMTWI experimentation is analyzed with the unsupervised learning approach independent component analysis (ICA) to test their mutual return in the detection of the deep defects in a mild steel sample and this proposed technique was referred to as frequency modulated independent component thermography (FMICT). In comparison, the effect of FMICT was contrasted with othermethodsi.e. pulse compression of time domain and ICA of feature space by considering the signal-to-noise ratio (SNR) as a figure of merit. Furthermore, a probability of detection (POD) analysis framework based on the minimum threshold SNR criteria for apparent visibility of the defects has been presented to assess the probability of identifying defects at various depths using such approaches. The influence of the SNR threshold value for the above strategies on the POD curves has also been presented.

Additive manufacturing (AM) processes have revolutionized manufacturing industries by enabling the production of complex geometries with reduced material waste and lead times. However, ensuring the quality of AM parts remains a significant challenge due to the complexity of the process and inherent variability in material properties. This review investigates the use of artificial intelligence (AI) to enhance quality assurance in AM processes, focusing on specific machine learning techniques such as convolutional neural networks for defect detection, support vector machines for classification of material properties, and reinforcement learning for real-time process optimization. The AI-driven methodologies are applied to predict defects, optimize process parameters, and monitor real-time production quality, utilizing large datasets generated from sensors and in-situ monitoring systems. The study demonstrates significant improvements in the accuracy of defect detection, the reliability of material property classification, and the efficiency of process optimization. In addition, it addresses challenges such as data pre-processing, model interpretability, and integration with existing AM systems. The findings highlight the potential of AI to transform quality assurance in AM and outline future research directions for further integration and enhancement of AI techniques in AM.

Eddy current pulsed thermography(ECPT) is an emerging Non-destructive testing and evaluation(NDT & E) technique, which uses hybrid eddy current and thermography NDT & E techniques that enhances the detectability from their compensation. Currently, this technique is limited by the manual selection of proper contrast frames and the issue of improving the efficiency of defect detection of complex structure samples remains a challenge. In order to select a specific frame from transient thermal image sequences to maximize the contrast of thermal variation and defect pattern from complex structure samples, an energy driven approach to compute the coefficient energy of wavelet transform is proposed which has the potential of automatically selecting both optimal transient frame and spatial scale for defect detection using ECPT. According to analysis of the variation of different frequency component and the comparison study of the detection performance of different scale and wavelets, the frame at the end of heating phase is automatically selected as an optimal transient frame for defect detection. In addition, the detection capabilities of the complex structure samples can be enhanced through proper spatial scale and wavelet selection. The proposed method has successfully been applied to low speed impact damage detection of carbon fibre reinforced polymer(CFRP) composite as well as providing the guidance to improve the detectability of ECPT technique.

Real-time monitoring of carbon fibre composites during Automated Fibre Placement (AFP) manufacturing remains a challenge for non-destructive evaluation (NDE) techniques. A directional eddy-current (EC) probe with asymmetric transmit and differential receive (Tx-dRx) coils is designed, constructed and characterized to evaluate the detectability of out-of-plane wrinkles. Initial studies were conducted to determine suitable excitation frequencies and to analyse the impact of relative orientations of driver and pickup coils on wrinkle detectability. The probe configurations are evaluated experimentally and employ a new finite element modelling approach to better understand the relationship between eddy-current density and defect detection. The findings indicate that a probe configuration with an asymmetric driver coil normal to the material surface and aligned with the fibre directions, and with differential pickup coils 90° to the scanning direction, shows the best capability for out-of-plane wrinkle detection, with SNR >20 for wrinkles over 1.3 mm in amplitude.

For more than a decade now, geophysical prospecting has been considered an integral part of archaeological research; the ability to quickly investigate large areas and locate objects buried in the ground without directly interacting with it is a key feature that makes such surveys essential for identifying and locating, with good accuracy, buried archaeological remains. In this study, two extensive magnetic and electromagnetic surveys were carried out in two different areas of the Akrai archaeological site, where given the distribution of archaeological remains already found at the site, it was likely that additional buried remains were present. The surveys were performed using a proton precession magnetometer and a multifrequency electromagnetic device with a frequency range of 2.5–250 kHz. By processing the data, the vertical magnetic gradient and electrical conductivity maps were obtained. Furthermore, 3D models of electrical conductivity distribution were reconstructed. On comparing the results, it was notable that many anomalies characterized by low vertical gradient values were identified within areas characterized by low electrical conductivity values. These anomalies detected by both surveys can be associated with good probability with buried archaeological remains made up of limestones. In fact, they exhibit shapes and sizes comparable to those of the wall remains already found at the site.

1. Integration of Data for Resource Evaluation.- Application of recent developments of regression analysis in regional mineral resource evaluation.- Poisson regression analysis and its application.- Measurement, sampling and interpretation.- Levels of information and probability mapping.- Review of geomathematical applications for mineral resources evaluation in China.- 2. Statistical Analysis of Geochemical Data.- New approaches to the recognition of anomalies in exploration geochemistry.- Univariate patterns in the design of multivariate analysis techniques for geochemical data evaluation.- Application of robust statistics in the analysis of geochemical data.- 3. Statistical Mehtods for Oil and Gas Resource Evaluation.- U.S. Geological Survey assessment methodology for estimation of undiscovered petroleum resources in play analysis of the Artic National Wildlife Refuge.- Statistical evaluation of petroleum deposits before discovery.- On assessing dry probabilities in offshore oil and gas exploration: an application of Bayes's theorem.- Statistical analysis of oil and gas discovery data.- 4. Image Processing in Earth Sciences.- Automated integration of mineral resource data by image processing and artificial intelligence.- Principles of mathematical morphology.- The use of geological image analysis in remote sensing and seismic attribute integration for resource assessment.- Statistical image processing.- 5. Statistical Analysis of Spatial Data.- Spatial analysis of patterns of land-based and ocean-floor ore deposits.- Point processes for the earth sciences.- 6. Geostatistics.- An unconventional approach to geostatistical estimation.- Case studies on modelling complex sulphide orebodies for ore reserve estimation using geostatistical methods.- A Bayesian approach to surface estimation.- Conditional simulation of indicator data. Case study of a multi-seam coal deposit.- Factorial kriging as a method to include spatial structure into classification. A case study on a sulphide orebody.- Geostatistical techniques for interpreting multivariate spatial information.- 7. Other Statisitcal Methodologies for Geoscience Data.- The shape of Lloydminster oil and gas deposit attribute data.- Confidence bands for the distribution and quantile functions for truncated and randomly censored data.- Geostatistical crustal abundance resource models.- 8. Artificial Intelligence in Earth Sciences.- The usage of Artificial intelligence in remote sensing: a review of applications and current research.- GEOVALUATOR, an expert system for resource appraisal: a demonstration prototype for kaolin in Georgia, U.S.A.- GEMS: a microcomputer-based expert system for digital image data.- On the understanding of geological cross-sections by expert systems.- 9. Other Contributed Papers.- Examples of spatial analysis and management in the geographical and conformal projection domains.- A classification procedure for exploitable reserves.- Drainage and divide networks derived from high-fidelity digital terrain models.- An algebraic model for data structure and manipulation as a design basis for integrated geoscience processing systems.- Case study on application of qualitative data analysis techniques to an uranium mineralization.- A shell for microcomputer exploration software used in developing countries.- Optimal exploration strategies: the synthesis of resource assessments and operations research.- Advantages and limitations of discovery process modeling: the case of the Northern West Siberian gas plays.- Some aspects of multivariate analysis.- Workshop Reports.- Current problems and future developments in multivariate analysis.- Current problems and future developments in oil and gas resource modeling and forecasting.- Current problems and future developments in spatial statistics and image processing.- Current problems and future developments in mineral and energy resource expert system development.- Current problems and future developments in the use of microcomputers.