Advancing Carbon Fiber Composite Inspection: Deep Learning-Enabled Defect Localization and Sizing via 3-Dimensional U-Net Segmentation of Ultrasonic Data.

In Non-Destructive Evaluation (NDE), accurately characterizing defects within components relies on accurate sizing and localization to evaluate the severity or criticality of defects. This study presents for the first time a deep learning methodology using 3-Dimensional (3D) U-Net to localize and size defects in Carbon Fibre Reinforced Polymer (CFRP) composites through volumetric segmentation of ultrasonic testing data. Using a previously developed approach, synthetic training data closely representative of experimental data was used for the automatic generation of ground truth segmentation masks. The model's performance was compared to the conventional amplitude 6 dB drop analysis method used in industry against ultrasonic defect responses from 40 defects fabricated in CFRP components. The results showed good agreement with the 6 dB drop method for in-plane localization and excellent through-thickness localization, with Mean Absolute Errors (MAE) of 0.57 mm and 0.08 mm, respectively. Initial sizing results consistently oversized defects with a 55% higher mean average error than the 6 dB drop method. However, when a correction factor was applied to account for variation between the experimental and synthetic domains the final sizing accuracy resulted in a 35% reduction in MAE compared to the 6 dB drop technique. By working with volumetric ultrasonic data (as opposed to 2D images) this approach reduces pre-processing (such as signal gating) and allows for the generation of 3D defect masks which can be used for the generation of computer aided design files; greatly reducing the qualification reporting burden of NDE operators.

Single-Bit Reception with Coded Excitation for Lightweight Advanced Ultrasonic Imaging Systems.

The demand for an efficient and reliable ultrasonic phased array imaging system is not unique to a single industry. Today's imaging systems can be enhanced in a number of areas including; improving scanning and processing times, reducing data storage requirements, simplifying hardware and prolonging probe lifespan. In this work, it is shown that by combining the use of Coded Excitation with single-bit data capture, a number of these areas can be improved. Despite using single-bit receive data, resolution can be recovered through the coded excitation pulse compression process, and shown to produce high Signal-to-Noise Ratio (SNR) images of Phase Coherence Imaging (PCI) and Total Focusing Method (TFM) of tip diffraction in a carbon steel sample. Comparison with conventional single-cycle transmission pulses has shown that little imaging performance degradation is seen despite a significant reduction in data resolution and size. This has also been shown to be effective at low excitation voltages with gain compensation due to the obsolescence of signal saturation concerns when considering single-bit receive data. The ability to compute high-resolution ultrasonic images from low-resolution input data at low transmission voltages has important implications for data compression, acquisition & imaging performance, operator safety and hardware simplification for ultrasonic imaging systems across industrial and medical fields.

Unsupervised machine learning for flaw detection in automated ultrasonic testing of carbon fibre reinforced plastic composites.

The use of Carbon Fibre Reinforced Plastic (CFRP) composite materials for critical components has significantly surged within the energy and aerospace industry. With this rapid increase in deployment, reliable post-manufacturing Non-Destructive Evaluation (NDE) is critical for verifying the mechanical integrity of manufactured components. To this end, an automated Ultrasonic Testing (UT) NDE process delivered by an industrial manipulator was developed, greatly increasing the measurement speed, repeatability, and locational precision, while increasing the throughput of data generated by the selected NDE modality. Data interpretation of UT signals presents a current bottleneck, as it is still predominantly performed manually in industrial settings. To reduce the interpretation time and minimise human error, this paper presents a two-stage automated NDE evaluation pipeline consisting of a) an intelligent gating process and b) an autoencoder (AE) defect detector. Both stages are based on an unsupervised method, leveraging density-based spatial clustering of applications with noise clustering method for robust automated gating and undefective UT data for the training of the AE architecture. The AE network trained on ultrasonic B-scan data was tested for performance on a set of reference CFRP samples with embedded and manufactured defects. The developed model is rapid during inference, processing over 2000 ultrasonic B-scans in 1.26 s with the area under the receiver operating characteristic curve of 0.922 in simple and 0.879 in complex geometry samples. The benefits and shortcomings of the presented methods are discussed, and uncertainties associated with the reported results are evaluated.

Three-Dimensional Residual Neural Architecture Search for Ultrasonic Defect Detection.

This study presents a deep-learning (DL) methodology using 3-D convolutional neural networks (CNNs) to detect defects in carbon fiber-reinforced polymer (CFRP) composites through volumetric ultrasonic testing (UT) data. Acquiring large amounts of ultrasonic training data experimentally is expensive and time-consuming. To address this issue, a synthetic data generation method was extended to incorporate volumetric data. By preserving the complete volumetric data, complex preprocessing is reduced, and the model can utilize spatial and temporal information that is lost during imaging. This enables the model to utilize important features that might be overlooked otherwise. The performance of three architectures was compared. The first architecture is prevalent in the literature for the classification of volumetric datasets. The second demonstrated a hand-designed approach to architecture design, with modifications to the first architecture to address the challenges of this specific task. A key modification was the use of cuboidal kernels to account for the large aspect ratios seen in ultrasonic data. The third architecture was discovered through neural architecture search (NAS) from a modified 3-D residual neural network (ResNet) search space. In addition, domain-specific augmentation methods were incorporated during training, resulting in significant improvements in model performance, with a mean accuracy improvement of 22.4% on the discovered architecture. The discovered architecture demonstrated the best performance with a mean accuracy increase of 7.9% over the second-best model. It was able to consistently detect all defects while maintaining a model size smaller than most 2-D ResNets. Each model had an inference time of less than 0.5 s, making them efficient for the interpretation of large amounts of data.

Transforming Industrial Manipulators via Kinesthetic Guidance for Automated Inspection of Complex Geometries.

The increased demand for cost-efficient manufacturing and metrology inspection solutions for complex-shaped components in High-Value Manufacturing (HVM) sectors requires increased production throughput and precision. This drives the integration of automated robotic solutions. However, the current manipulators utilizing traditional programming approaches demand specialized robotic programming knowledge and make it challenging to generate complex paths and adapt easily to unique specifications per component, resulting in an inflexible and cumbersome teaching process. Therefore, this body of work proposes a novel software system to realize kinesthetic guidance for path planning in real-time intervals at 250 Hz, utilizing an external off-the-shelf force-torque (FT) sensor. The proposed work is demonstrated on a 500 mm2 near-net-shaped Wire-Arc Additive Manufacturing (WAAM) complex component with embedded defects by teaching the inspection path for defect detection with a standard industrial robotic manipulator in a collaborative fashion and adaptively generating the kinematics resulting in the uniform coupling of ultrasound inspection. The utilized method proves superior in performance and speed, accelerating the programming time using online and offline approaches by an estimate of 88% to 98%. The proposed work is a unique development, retrofitting current industrial manipulators into collaborative entities, securing human job resources, and achieving flexible production.

Automated Real-Time Eddy Current Array Inspection of Nuclear Assets.

Inspection of components with surface discontinuities is an area that volumetric Non-Destructive Testing (NDT) methods, such as ultrasonic and radiographic, struggle in detection and characterisation. This coupled with the industrial desire to detect surface-breaking defects of components at the point of manufacture and/or maintenance, to increase design lifetime and further embed sustainability in their business models, is driving the increased adoption of Eddy Current Testing (ECT). Moreover, as businesses move toward Industry 4.0, demand for robotic delivery of NDT has grown. In this work, the authors present the novel implementation and use of a flexible robotic cell to deliver an eddy current array to inspect stress corrosion cracking on a nuclear canister made from 1.4404 stainless steel. Three 180-degree scans at different heights on one side of the canister were performed, and the acquired impedance data were vertically stitched together to show the full extent of the cracking. Axial and transversal datasets, corresponding to the transmit/receive coil configurations of the array elements, were simultaneously acquired at transmission frequencies 250, 300, 400, and 450 kHz and allowed for the generation of several impedance C-scan images. The variation in the lift-off of the eddy current array was innovatively minimised through the use of a force-torque sensor, a padded flexible ECT array and a PI control system. Through the use of bespoke software, the impedance data were logged in real-time (≤7 ms), displayed to the user, saved to a binary file, and flexibly post-processed via phase-rotation and mixing of the impedance data of different frequency and coil configuration channels. Phase rotation alone demonstrated an average increase in Signal to Noise Ratio (SNR) of 4.53 decibels across all datasets acquired, while a selective sum and average mixing technique was shown to increase the SNR by an average of 1.19 decibels. The results show how robotic delivery of eddy current arrays, and innovative post-processing, can allow for repeatable and flexible surface inspection, suitable for the challenges faced in many quality-focused industries.

Collaborative Robotic Wire + Arc Additive Manufacture and Sensor-Enabled In-Process Ultrasonic Non-Destructive Evaluation.

The demand for cost-efficient manufacturing of complex metal components has driven research for metal Additive Manufacturing (AM) such as Wire + Arc Additive Manufacturing (WAAM). WAAM enables automated, time- and material-efficient manufacturing of metal parts. To strengthen these benefits, the demand for robotically deployed in-process Non-Destructive Evaluation (NDE) has risen, aiming to replace current manually deployed inspection techniques after completion of the part. This work presents a synchronized multi-robot WAAM and NDE cell aiming to achieve (1) defect detection in-process, (2) enable possible in-process repair and (3) prevent costly scrappage or rework of completed defective builds. The deployment of the NDE during a deposition process is achieved through real-time position control of robots based on sensor input. A novel high-temperature capable, dry-coupled phased array ultrasound transducer (PAUT) roller-probe device is used for the NDE inspection. The dry-coupled sensor is tailored for coupling with an as-built high-temperature WAAM surface at an applied force and speed. The demonstration of the novel ultrasound in-process defect detection approach, presented in this paper, was performed on a titanium WAAM straight sample containing an intentionally embedded tungsten tube reflectors with an internal diameter of 1.0 mm. The ultrasound data were acquired after a pre-specified layer, in-process, employing the Full Matrix Capture (FMC) technique for subsequent post-processing using the adaptive Total Focusing Method (TFM) imaging algorithm assisted by a surface reconstruction algorithm based on the Synthetic Aperture Focusing Technique (SAFT). The presented results show a sufficient signal-to-noise ratio. Therefore, a potential for early defect detection is achieved, directly strengthening the benefits of the AM process by enabling a possible in-process repair.

Solving ultrasonic ray tracing in parts with multiple material layers through Root-Finding methods.

Ultrasonic testing has been used for material analysis and inspection since 1930's. Nevertheless, the applicability of ultrasonic waves to new complex cases is still growing, thanks to the availability of powerful electronics and software. However, the complication that slows down the deployment of ultrasonic inspection to geometric complex parts and structures arises from the wave refraction phenomenon. A clear understanding of the ultrasound wave propagation, impacted by refractions, is crucial to interpret the data obtained from the inspection of multi-layered/multi-medium test subjects as it is not always possible to assume that mechanical waves travel in straight lines. This work presents suitable approaches for solving the ray-tracing problem in multi-layered structures. Accurate benchmarking shows that the use of the Newton-Raphson root-finding method allows a threefold reduction of the computation time, when compared to the bisection-based root-finding methods. An effective combination of the Newton-Raphson methods with bisection-type iterations is also proposed and discussed. Although the work repeatedly refers to the field of ultrasonic inspection, the presented findings are relevant and applicable to areas beyond material inspection.

Sensor-Enabled Multi-Robot System for Automated Welding and In-Process Ultrasonic NDE.

The growth of the automated welding sector and emerging technological requirements of Industry 4.0 have driven demand and research into intelligent sensor-enabled robotic systems. The higher production rates of automated welding have increased the need for fast, robotically deployed Non-Destructive Evaluation (NDE), replacing current time-consuming manually deployed inspection. This paper presents the development and deployment of a novel multi-robot system for automated welding and in-process NDE. Full external positional control is achieved in real time allowing for on-the-fly motion correction, based on multi-sensory input. The inspection capabilities of the system are demonstrated at three different stages of the manufacturing process: after all welding passes are complete; between individual welding passes; and during live-arc welding deposition. The specific advantages and challenges of each approach are outlined, and the defect detection capability is demonstrated through inspection of artificially induced defects. The developed system offers an early defect detection opportunity compared to current inspection methods, drastically reducing the delay between defect formation and discovery. This approach would enable in-process weld repair, leading to higher production efficiency, reduced rework rates and lower production costs.

Correction: Ghanei et al. Effect of Nano-CuO on Engineering and Microstructure Properties of Fibre-Reinforced Mortars Incorporating Metakaolin: Experimental and Numerical Studies. Materials 2017, 10, 1215.

The authors regret that Figures 6 and 11b in [...].

A Study of the Automated Eddy Current Detection of Cracks in Steel Plates.

Applying life estimation approaches to determine in-service life of structures and plan the inspection schedules accordingly are becoming acceptable safety design procedures in aerospace. However, these design systems shall be fed with reliable parameters related to material properties, loading conditions and defect characteristics. In this context, the role of non-destructive (NDT) testing reliability is of high importance in detecting and sizing defects. Eddy current test (ECT) is an electromagnetic NDT method frequently used to inspect tiny surface fatigue cracks in sensitive industries. Owing to the new advances in robotic technologies, there is a trend to integrate the ECT into automated systems to perform NDT inspections more efficiently. In fact, ECT can be effectively automated as to increase the coverage, repeatability and scanning speed. The reliability of ECT scanning, however, should be thoroughly investigated and compared to conventional modes of applications to obtain a better understanding of the advantages and shortcomings related to this technique. In this contribution, a series of manual and automated ECT tests are carried out on a set of samples using a split-D reflection differential surface probe. The study investigates the level of noise recorded in each technique and discuss its dependency on different parameters, such as surface roughness and frequency. Afterwards, a description of the effect of crack orientation on ECT signal amplitude is provided through experimental tests and finite element simulations. Finally, the reliability of each ECT technique is investigated by means of probability of detection (POD) curves. POD parameters are then extracted and compared to examine the effect of scanning index, frequency and automation on detection reliability.

Finite Element Modelling of a Reflection Differential Split-D Eddy Current Probe Scanning Surface Notches.

Differential eddy current probes are commonly used to detect shallow surface cracks in conductive materials. In recent years, a growing number of research works on their numerical modelling was conducted since the development of analytical or semi-analytical models for such a sensor may be prone to intractable complications. In this paper finite element modelling (FEM) has been employed to simulate the interaction of a reflection differential split-D probe with surface electrical discharge machined (EDM) notches in 3-dimensional (3-D) half-space. In order to attain a better insight into the correct setup of the FEM parameters, a simple multi-turn cylindrical absolute coil has also been modelled. The outcome generated through the simulated scan of this absolute coil over a surface notch in aluminum is validated with existing experimental impedance data taken from the literature. Parameters contributing to reliable FEM simulation results, such as maximum mesh size, mesh distribution, the extent of the surrounding air domain and conductivity of the air are investigated for the 3-D modelling of both absolute and differential probes. This study shows that the simulation results on a commercial reflection differential split-D surface pencil probe closely estimate the experimental measurements of the probe's impedance variations as it scans three EDM notches having different depths in aluminum. The simulation results, generated by Comsol Multiphysics FEM package (COMSOL I, COMSOL multiphysics reference manual, version 5.3, COMSOL AB, 2018, www.comsol.com), for the cases of absolute and differential probes are checked for their extent of validity.

Adaptive Neuro-fuzzy Inference System Trained for Sizing Semi-elliptical Notches Scanned by Eddy Currents.

The present study explores the capability of COMSOL Multiphysics, as a finite element modelling (FEM) tool, to model the interaction between a split-D differential surface eddy current (ECT) probe and semi-elliptical surface electrical discharge machined (EDM) notches. The effect of the small probe's lift-off and tilt on its signal is investigated through modelling and subsequently, the simulation outcomes are validated using the probe's impedance measurements. In the next stage, an adaptive neuro-fuzzy inference system (ANFIS) is designed to take the signal features as inputs and consequently, provide the length of the scanned notch as the system's output. The system is trained by extracted features of thirty model-generated signals obtained from scanning of the same number of semi-elliptical notches by means of the split-D probe. The trained ANFIS is tested afterwards using the measured signals of 3 calibration EDM notches together with 5 model-based ones. A very low average estimation error is observed with regard to the length estimation of the test notches and the accuracy of the length estimation is found to be quite reasonable.

Investigation of the Role of Nano-Titanium on Corrosion and Thermal Performance of Structural Concrete with Macro-Encapsulated PCM.

The present study aims to investigate the impact of thermal energy storage aggregate (TESA) and nano-titanium (NT) on properties of structural concrete. TESA was made of scoria encapsulated with phase change materials (PCMs). Coarse aggregates were replaced by TESA at 100% by volume of aggregate and NT was added at 5% by weight of cement. Compressive strength, probability of corrosion, thermal performance, and microstructure properties were studied. The results indicated that the presence of TESA reduced the compressive strength of concrete, although the strength was still high enough to be used as structural concrete. The use of TESA significantly improved the thermal performance of concrete, and slightly improved the resistance of corrosion in concrete. The thermal test results showed that TESA concrete reduces the peak temperature by 2 °C compared to the control. The addition of NT changed the microstructure of concrete, which resulted in higher compressive strength. Additionally, the use of NT further enhanced the thermal performance of TESA concrete by reducing the probability of corrosion remarkably. These results confirmed the crucial role of NT in improving the permeability and the thermal conductivity of mixtures containing PCM. In other words, the charging and discharging of TESA was enhanced with the presence of NT in the mixture.

Effect of Nano-CuO on Engineering and Microstructure Properties of Fibre-Reinforced Mortars Incorporating Metakaolin: Experimental and Numerical Studies.

In this study, the effects of nano-CuO (NC) on engineering properties of fibre-reinforced mortars incorporating metakaolin (MK) were investigated. The effects of polypropylene fibre (PP) were also examined. A total of twenty-six mixtures were prepared. The experimental results were compared with numerical results obtained by adaptive neuro-fuzzy inference system (ANFIS) and Primal Estimated sub-GrAdient Solver for SVM (Pegasos) algorithm. Scanning Electron Microscope (SEM) was also employed to investigate the microstructure of the cement matrix. The mechanical test results showed that both compressive and flexural strengths of cement mortars decreased with the increase of MK content, however the strength values increased significantly with increasing NC content in the mixture. The water absorption of samples decreased remarkably with increasing NC particles in the mixture. When PP fibres were added, the strengths of cement mortars were further enhanced accompanied with lower water absorption values. The addition of 2 wt % and 3 wt % nanoparticles in cement mortar led to a positive contribution to strength and resistance to water absorption. Mixture of PP-MK10NC3 indicated the best results for both compressive and flexural strengths at 28 and 90 days. SEM images illustrated that the morphology of cement matrix became more porous with increasing MK content, but the porosity reduced with the inclusion of NC. In addition, it is evident from the SEM images that more cement hydration products adhered onto the surface of fibres, which would improve the fibre-matrix interface. The numerical results obtained by ANFIS and Pegasos were close to the experimental results. The value of R² obtained for each data set (validate, test and train) was higher than 0.90 and the values of mean absolute percentage error (MAPE) and the relative root mean squared error (PRMSE) were near zero. The ANFIS and Pegasos models can be used to predict the mechanical properties and water absorptions of fibre-reinforced mortars with MK and NC.

Chloride Diffusion and Acid Resistance of Concrete Containing Zeolite and Tuff as Partial Replacements of Cement and Sand.

In this paper, the properties of concrete containing zeolite and tuff as partial replacements of cement and sand were studied. The compressive strength, water absorption, chloride ion diffusion and resistance to acid environments of concretes made with zeolite at proportions of 10% and 15% of binder and tuff at ratios of 5%, 10% and 15% of fine aggregate were investigated. The results showed that the compressive strength of samples with zeolite and tuff increased considerably. In general, the concrete strength increased with increasing tuff content, and the strength was further improved when cement was replaced by zeolite. According to the water absorption results, specimens with zeolite showed the lowest water absorption values. With the incorporation of tuff and zeolite, the chloride resistance of specimens was enhanced significantly. In terms of the water absorption and chloride diffusion results, the most favorable replacement of cement and sand was 10% zeolite and 15% tuff, respectively. However, the resistance to acid attack reduced due to the absorbing characteristic and calcareous nature of the tuff.