Search Results (7)

The automated production of large-scale labeled datasets from concrete X-ray computed tomography (CT) images is a fundamental prerequisite for training and validating deep learning-based segmentation models. However, existing methods either require extensive manual annotation or rely on domain-specific deep learning models that themselves demand labeled data—a circular dependency. This paper presents a parameter-driven three-class segmentation framework that automatically classifies each pixel in a concrete CT slice into one of three material phases: void (air pores and cracks), coarse aggregate, and mortar matrix, generating annotation masks suitable for large-scale dataset production without manual labeling. The proposed method combines: (1) fixed-threshold void detection calibrated to concrete CT grayscale characteristics; (2) adaptive percentile-based initial segmentation responsive to image-specific statistics; (3) multi-criteria connected component scoring based on area, shape descriptors (circularity, solidity, compactness, extent, aspect ratio), intensity distribution, and boundary gradient; (4) material science-informed size constraints aligned with concrete phase volume fractions; and (5) a material continuity enforcement module that applies topological hole-filling and conditional convex-hull consolidation to eliminate internal contamination within accepted aggregate regions, reducing boundary roughness by 7.6% and recovering misclassified boundary pixels. All parameters are centralized in a configuration file, enabling reproducible batch processing of 224 × 224 pixel CT slices at 0.07–1.12 s per image. Evaluated on 1007 224 × 224 concrete CT patches cropped from 200 representative scan frames, the framework produces three-class segmentation masks with physically consistent void fractions (mean 3.2%), aggregate fractions (mean 32.4%), and mortar fractions (mean 64.4%), all within ranges reported in the concrete CT literature (used as a dataset-scale QC screen, not a validation metric). Primary outputs and the archived image–mask pairs for this work are provided as an 8-bit patch archive. For pixel-wise validation, we report IoU, Dice, and pixel accuracy on an independently labeled subset that can be unambiguously paired with the released predictions: averaged over 57 matched patches, mean pixel accuracy is 88.6%, macro-mean IoU is 74.7%, and macro-mean Dice is 84.9%. The framework provides a fully automated annotation pipeline for dataset production, eliminating manual labeling costs for concrete CT image collections. The generated datasets are suitable for training semantic segmentation networks such as U-Net and its variants. Full article

Electric submersible progressive-cavity pumps (ESPCPs) deliver high lifting efficiency but are prone to failure in the high-temperature, high-pressure, and multiphase down-hole environment, leading to production losses and elevated maintenance costs. To achieve reliable condition recognition under these noisy and highly imbalanced data constraints, we fuse deep residual feature learning, ensemble decision-making, and generative augmentation into a unified diagnosis pipeline. A class-aware TimeGAN first synthesizes realistic minority-fault sequences, enlarging the training pool derived from 360 field records. The augmented data are then fed to a CNN backbone equipped with ResNet blocks, and its deep features are classified by a Random-Forest head (CNN-ResNet-RF). Across five benchmark architectures—including plain CNN, CNN-ResNet, GRU-based, and hybrid baselines—the proposed model attains the highest overall validation accuracy (≈97%) and the best Macro-F1, while the confusion-matrix diagonal confirms marked reductions in the previously dominant misclassification between tubing-leakage and low-parameter states. These results demonstrate that residual encoding, ensemble voting, and realistic data augmentation are complementary in coping with sparse, noisy, and class-imbalanced ESPCP signals. The approach therefore offers a practical and robust solution for the real-time down-hole monitoring and preventive maintenance of ESPCP systems. Full article

Electric discharge machining (EDM) is widely employed for machining hard, conductive materials. Tool rotation has emerged as an effective strategy to enhance debris flushing and improve stability during deep-hole EDM drilling. This study proposes a machine learning-based approach to evaluate the influence of tool rotation and predict the unstable machining conditions in EDM of ultrafine grained tungsten carbide. A structured analytical workflow, combining Taguchi–Grey optimization, regression analysis, and classification models, was designed to capture discharge dynamics across macro- and micro-timescales. Classification models trained on raw and processed electrical signal features achieved over 88% accuracy and 90% recall. SHAP analysis revealed that the relationship between key discharge events such as sparks and short circuits varied significantly across stable and unstable machining phases, underscoring the importance of phase-specific modeling. While correlation analysis showed weak global associations, phase-dependent SHAP values revealed opposing feature effects, allowing the context-informed interpretation of model behavior. Phase segmentation revealed that, compared to 1000 RPM, short circuits were reduced by about 40% during stable machining at 8000–9000 RPM. Conversely, during unstable phases, spark effectiveness dropped by nearly 45%, and secondary discharges increased throughout this range. These insights support the design of adaptive control strategies that adjust the rotation rate in response to detected phase changes, aiming to sustain machining stability. The findings support the development of dynamic control frameworks to improve EDM performance, particularly for mold fabrication using tungsten carbide. Full article

Nuclear power tube plates are made from the high-strength, low-carbon alloy steel SA-5083, which has high values of toughness and plasticity, though it is forged with poor consistency and entails serious work hardening. It requires a large number of deep holes with a high machining accuracy and high surface quality to be processed. However, the quality of the processed holes is often not up to the standard of the Boring and Trepanning Association (BTA) for the deep-hole drilling of tube plates; this has led to deep-hole processing becoming a bottleneck in the manufacture of steam generators for the main equipment of nuclear power islands. The variation laws of the diameter, roundness, perpendicularity, roughness, microhardness, and residual stress in relation to the feed, speed, and drilling depth are explored in the macro- and micro-dimensions; also explored is the wear morphology of BTA drills. The internal influence mechanisms between them are revealed in order to provide a scientific basis for the control of surface quality and machining accuracy as well as the optimization of process parameters. Our research results indicate that the guide block wear is mainly concentrated at the top 1–2 mm and that the drilling depth and feed have a great influence on the machining diameter. The hole wall roughness is between 0.3 and 0.6 μm, the maximum microhardness is about 2.15 times the hardness of the matrix material, and the residual stress is compressive stress. With increases in the feed and drilling depth, the hole diameter and the roughness increase. With an increase in the speed, the roughness decreases and the compressive stress of the BTA deep-hole drilling wall increases. Full article

Using mechanical drilling to obtain energy resources stored in deep and hard rock layers is becoming increasingly challenging. Therefore, laser irradiation has emerged as a new and promising drilling method. In this study, the effects of immersion conditions on rock-breaking by laser irradiation on the temperature, hole size, rock-breaking efficiency, and macro-fracture after laser irradiation were investigated. Furthermore, the mineral changes and thermogravimetric analysis of rocks were studied. As indicated by the results, the temperature area over 100 °C increases with the increase of irradiation time, and the temperature range of between 2.27 cm² and 13.20 cm² varies with the change of laser power at between 90 W and 135 W. The hole-diameter value of the soaked sample was smaller than that of the dried sample. In addition, the hole depth of the soaked sample reduced by 15% at a power of 90 W and 45% at a power of 135 W, compared with that of the dried sample. The value of the modified specific energy of the soaked sample increased, which was particularly noticeable at low power. The soaked sample had a larger effect on the rate of perforation at high power and a smaller effect at low power. The cracks on the surface of the rock samples became larger after being placed for one month. Fracture length increased from 0.61 to 5.09 mm for dried samples and from 2.24 to 8.7 mm for soaked samples at a laser power of 90 W. Fracture length increased from 6.30 to 9.85 mm for dried samples and from 9.04 to 11.38 mm for soaked samples at a laser power of 135 W. The soaked sample began to show differences when heated at 100 °C, which was caused by the evaporation of some free water molecules in the rock. The main weight loss temperatures of the samples occurred in the range of 640 °C to 900 °C. Full article

Experimental analyses of depth distributions of phase-specific residual stresses after deep rolling were carried out by means of laboratory X-ray diffraction and neutron diffraction for the two duplex steels X2CrNiMoN22-5-3 and X3CrNiMoN27-5-2, which differ significantly in their ferrite to austenite ratios. The aim of the investigation was to elucidate to which extent comparable results can be achieved with the destructive and the non-destructive approach and how the process induced phase-specific micro residual stresses influence the determination of the phase- and {hkl}-specific reference value d₀, required for evaluation of neutron strain scanning experiments. A further focus of the work was the applicability of correction approaches that were developed originally for single-phase materials for accounting for spurious strains during through surface neutron scanning experiments on coarse two-phase materials. The depth distributions of macro residual stresses were separated from the phase-specific micro residual stresses. In this regard, complementary residual stress analysis was carried out by means of incremental hole drilling. The results indicate that meaningful macro residual stress depth distributions can be determined non-destructively by means of neutron diffraction for depths starting at about 150–200 µm. Furthermore, it was shown that the correction of the instrumental surface effects, which are intrinsic for surface neutron strain scanning, through neutron ray-tracing simulation is applicable to multiphase materials and yields reliable results. However, phase-specific micro residual stresses determined by means of neutron diffraction show significant deviations to data determined by means of lab X-ray stress analysis according to the well-known sin²ψ-method. Full article

This research investigates the effect of machining parameters on material removal rate, electrode wear ratio, and gap clearance of macro deep holes with a depth-to-diameter ratio over four. The experiments were carried out using electrical discharge machining with side flushing and multi-aperture flushing to improve the machining performance and surface integrity. The machining parameters were pulse on-time, pulse off-time, current, and electrode rotation. Response surface methodology and the desirability function were used to optimize the electrical discharge machining parameters. The results showed that pulse on-time, current, and electrode rotation were positively correlated with the material removal rate. The electrode wear ratio was inversely correlated with pulse on-time and electrode rotation but positively correlated with current. Gap clearance was positively correlated with pulse on-time but inversely correlated with pulse off-time, current, and electrode rotation. The optimal machining condition of electrical discharge machining with side flushing was 100 µs pulse on-time, 20 µs pulse off-time, 15 A current, and 70 rpm electrode rotation; and that of electrical discharge machining with multi-aperture flushing was 130 µs, 2 µs, 15 A, and 70 rpm. The novelty of this research lies in the use of multi-aperture flushing to improve the machining performance, enable a more uniform GC profile, and minimize the incidence of recast layer. Full article