Search

Microstructure reconstruction is an important and emerging field of research and an essential foundation to improving inverse computational materials engineering (ICME). Much of the recent progress in the field is made based on generative adversarial networks (GANs). Although excellent results have been achieved throughout a variety of materials, challenges remain regarding the interpretability of the model's latent space as well as the applicability to extremely small data sets. The present work addresses these issues by introducing DA-VEGAN, a model with two central innovations. First, a β -variational autoencoder is incorporated into a hybrid GAN architecture that allows to penalize strong nonlinearities in the latent space by an additional parameter, β. Secondly, a custom differentiable data augmentation scheme is developed specifically for this architecture. The differentiability allows the model to learn from extremely small data sets without mode collapse or deteriorated sample quality. An extensive validation on a variety of structures demonstrates the potential of the method and future directions of investigation are discussed. [Display omitted] [ABSTRACT FROM AUTHOR]

• NDVI differentiation in endorheic basins was described nonlinearly by the TWSA. • Approximately 98% of NDVI in the study area responded nonlinearly to the TWSA. • The nonlinear causal test was improved by considering the optimal moving window. The terrestrial water storage anomaly (TWSA) is an important parameter for assessing the land water budget, and it interacts well with terrestrial ecosystems via complex hydrological processes. Recently, the decline in central Asian terrestrial water storage (TWS) has threatened the health of local ecosystems. Therefore, it is of great significance to adopt an efficient approach to explore and identify the nonlinear relationship between two important indicators, i.e., the TWSA and normalized difference vegetation index (NDVI) in the arid central Asian endorheic basins. In this study, we analysed the long-term trends of the TWSA and NDVI, and identified the lag month (1 month) as the optimal moving window of the nonlinear Granger causality test embedded in random forest to detect the nonlinear NDVI change response of NDVI changes in vegetation to the TWSA from 2003 to 2015. There are decreasing trends in TWSAs over approximately 81.7% of the study area and the NDVI generally decreased resulting in approximately 36% vegetation browning in the study area. The nonlinear Granger unidirectional causes of the TWSA were responsible for 97.9% of the NDVI variation in the study area considering the optimal response time for moving windows. The causes of vegetation browning in the central Asian Aral Sea basin and vegetation greening in the basins of Northwest China could be mostly explained by the changes in TWS. Our findings contribute to understanding the nonlinear causal linkages between vegetation and the TWSA in endorheic basins, and these findings provide insights for obtaining terrestrial water consumption patterns and water resource management under the joint influence of climate change and human activities. [ABSTRACT FROM AUTHOR]

To enhance high-temperature wear resistance and cutting performances of cermet tools, nano-Al 2 O 3 particles were added into four kinds of Ti(C,N)-based cermets. The microstructure, mechanical properties, wear resistance, and cutting performances were studied. The nano-Al 2 O 3 was supposed to distribute in the binder phase, leading to the change of lattice constants and dispersion strengthening. Due to the distribution of nanoparticles at the grain boundaries, the movement of the grain boundaries was hindered, and the mean grain size decreased from 1.03 μm to 0.89 μm. The hardness and fracture toughness showed the change tendency of rising the peak and then falling with nano-Al 2 O 3 addition. The cermet with 2 wt% nano-Al 2 O 3 displayed the topmost hardness of 1654.4 HV 30 and fracture toughness of 9.1 MPa·m1/2, due to the dispersion strengthening and pinning effect of nano particles. During the sliding wear tests at 700 °C, the wear rates of cermet containing 2 wt% nano-Al 2 O 3 decreased by 63% compared to cermet without nano-Al 2 O 3. The superior high-temperature wear resistance was ascribed to the thermal softening resistance of nano-Al 2 O 3 , the strengthening and toughening effect of nano-Al 2 O 3 on matrix and the dense tribo-chemical layers. The abrasive wear of cermet with no nano-Al 2 O 3 addition was more significant, while the wear behavior changed to the removal of tribo-chemical layers when nano-Al 2 O 3 was 2 wt%. The tool life of cermet with 2 wt% nano-Al 2 O 3 was increased by 50% in comparison with cermet without nano-Al 2 O 3 during machining 45 steel. • Ti(C,N)-based cermets are prepared with different nano-Al 2 O 3 additions. • Dispersion strengthening theory shows the mechanism of the enhancement of hardness. • High temperature friction properties and wear morphologies of cermets are analyzed. • A certain amount of nano-Al 2 O 3 improves high temperature wear resistance and cutting performance of cermet tools. [ABSTRACT FROM AUTHOR]

• Overlapping heating and cooling demand in the entire European building stock is assessed. • Evaluation is based on hectare-sized units and GIS method. • 8 scenarios representing future climate, renovations, and cooling area are considered. • Less than 0.1% of the current building stock has enough overlapping demand for 5GDHC. • The future potential of 5GDHC considering overlapping demand and density is limited. With warmer climate conditions, the growing need to cool down building environment in Europe calls for efficient solutions to solve the heating and cooling demand altogether. To meet such requirement, the bi-directional fifth generation district heating and cooling (5GDHC) system has been proposed in recent years, while having the capability of integrating the renewable and waste energy sources. The overlapping heating and cooling demand, or referred to as the simultaneous demand, is proved as the key pre-requisite to guide the application of 5GDHC through several case studies. However, how large is this overlapping part in the practical building stock and where are the feasible areas for 5GDHC remain unclear. To close the gap, this study assesses the quantities and geographical distributions of overlapping heating and cooling demand across Europe in EU 27, UK, and EFTA countries. Using GIS-based methods and data sources, the study region is split into 28 million hectare-sized units where the overlapping demand is specifically calculated. Moreover, the possible future changes to the heating and cooling demand alone brought by warmer climate, building renovations, and increased cooling area were investigated. The results from reference condition at year 2016 reveal that less than 0.1% of the building stock has DOC larger than 0.3, which is the threshold for 5GDHC being energy efficient. These potential areas are primarily found in city centres involving cooling demands from commercial and industrial processes. In the future scenarios of year 2050, while a better energy performance of buildings and warmer climate may decrease the heating and increase the cooling demand, the overlapping part is only slightly increased by around 5%. Accordingly, around 2500 ha-sized units from the entire study region are found to meet the DOC limit required by the 5GDHC while having large enough demand density to support district energy systems. The presented works geographically identified the potential areas for 5GDHC and can be extended with regional details to serve as the connection between top-level planning and bottom-level case studies. [ABSTRACT FROM AUTHOR]

Medical image segmentation is a fundamental and critical step in many clinical approaches. Semi-supervised learning has been widely applied to medical image segmentation tasks since it alleviates the heavy burden of acquiring expert-examined annotations and takes the advantage of unlabeled data which is much easier to acquire. Although consistency learning has been proven to be an effective approach by enforcing an invariance of predictions under different distributions, existing approaches cannot make full use of region-level shape constraint and boundary-level distance information from unlabeled data. In this paper, we propose a novel uncertainty-guided mutual consistency learning framework to effectively exploit unlabeled data by integrating intra-task consistency learning from up-to-date predictions for self-ensembling and cross-task consistency learning from task-level regularization to exploit geometric shape information. The framework is guided by the estimated segmentation uncertainty of models to select out relatively certain predictions for consistency learning, so as to effectively exploit more reliable information from unlabeled data. Experiments on two publicly available benchmark datasets showed that: (1) Our proposed method can achieve significant performance improvement by leveraging unlabeled data, with up to 4.13% and 9.82% in Dice coefficient compared to supervised baseline on left atrium segmentation and brain tumor segmentation, respectively. (2) Compared with other semi-supervised segmentation methods, our proposed method achieve better segmentation performance under the same backbone network and task settings on both datasets, demonstrating the effectiveness and robustness of our method and potential transferability for other medical image segmentation tasks. • A novel uncertainty-guided mutual consistency learning framework is proposed for semi-supervised medical image segmentation to alleviate the heavy burden of acquiring expert-examined annotations. • Incorporating both intra-task and cross-task consistency learning with the guidance of estimated segmentation uncertainty to efficiently utilize unlabeled data. • Extensive experiments on two benchmark datasets show that our proposed method outperforms existing semi-supervised segmentation methods with promising performance. [ABSTRACT FROM AUTHOR]

This paper presents a spatial–spectral low-light enhancement approach based on the neural network. Considering that the Image Signal Processor (ISP) which non-linearly maps RAW data to RGB images may introduce additional noise and artifacts, we propose to conduct the enhancement task directly on RAW data. To this end, we propose a cross-scale framework to map the input low-light RAW data to visually pleasing RGB images. The cross-scale framework consists of three branches for low-level, mid-level, and high-level representations of input images. We embed paired Fast Fourier Convolution (FFC) and Transformer in each level for global and local feature analysis and aggregation. Specifically, the FFC enlarges the receptive field of our neural network, exploring the pixel correlations in the spectral space; the following Transformer uses self-attention for feature selection and aggregation in the spatial space. As a result, spatial–spectral information within an image is jointly utilized for the final fusion. Experimental results demonstrate that the proposed approach significantly outperforms state-of-the-art low-light enhancement methods in both full reference assessment metrics, including PSNR, MPSNR, and SSIM, and no-reference metrics, such as NIMA. Moreover, the proposed method produces more aesthetically pleasing RGB images than other methods. [Display omitted] • We propose a cross-scale framework to map the RAW data to RGB images. • We utilize FFC and Transformer to extract latent features in each scale of our framework. • Results show that the proposed method remarkably outperforms sota works. [ABSTRACT FROM AUTHOR]

• Spatiotemporal variation in vegetation on the Loess Plateau was investigated. • Precipitation was the greatest driving factor. • There was a bivariate or nonlinear enhancement between all factors. • Key factors' adaptation range or type suitable for vegetation cover was identified. Vegetation, being a core component of ecosystems, is known to be influenced by natural and anthropogenic factors. Studying vegetation cover dynamics variation and its drivers is critical to understanding the interactions between vegetation and ecosystems. The Loess Plateau (LP) is located in a semi-arid and semi-humid region with severe soil erosion and fragile ecology. This paper used the annual maximum Normalized Difference Vegetation Index (NDVI ymax) and growing season mean NDVI (NDVI gsmean) as the vegetation cover indicator. The vegetation cover variation of LP from 2000 to 2020 was analyzed using Sen's slope and Mann-Kendall test. Then the influence of natural and anthropogenic factors on the driving mechanisms of spatial vegetation differentiation was explored by the Geodetector model. The results revealed that the growth trends of NDVI ymax and NDVI gsmean were 0.075/10 years and 0.038/10 years, respectively, and the areas with improved vegetation cover accounted for 92.67 % and 88.58 % of the total area. The vegetation cover of the southeastern and northern parts of the LP decreased significantly; however, the central, southwestern, and northeastern parts exhibited a remarkable improvement. Precipitation, vegetation type, soil type, temperature, and land use type were the key driving forces, ranked differently on NDVI ymax and NDVI gsmean. However, precipitation was the most critical factor both on NDVI ymax and NDVI gsmean. The interaction detection showed non-linear and mutual enhancement, with no independent factor. The findings of our study can assist in identifying the vegetation cover status of the LP, as well as the driving forces, which can provide theoretical support for the formulation of environmental conservation policies. [ABSTRACT FROM AUTHOR]

Recently, deep convolutional neural networks have achieved great success for medical image segmentation. However, unlike segmentation of natural images, most medical images such as MRI and CT are volumetric data. In order to make full use of volumetric information, 3D CNNs are widely used. However, 3D CNNs suffer from higher inference time and computation cost, which hinders their further clinical applications. Additionally, with the increased number of parameters, the risk of overfitting is higher, especially for medical images where data and annotations are expensive to acquire. To issue this problem, many 2.5D segmentation methods have been proposed to make use of volumetric spatial information with less computation cost. Despite these works lead to improvements on a variety of segmentation tasks, to the best of our knowledge, there has not previously been a large-scale empirical comparison of these methods. In this paper, we aim to present a review of the latest developments of 2.5D methods for volumetric medical image segmentation. Additionally, to compare the performance and effectiveness of these methods, we provide an empirical study of these methods on three representative segmentation tasks involving different modalities and targets. Our experimental results highlight that 3D CNNs may not always be the best choice. Despite all these 2.5D methods can bring performance gains to 2D baseline, not all the methods hold the benefits on different datasets. We hope the results and conclusions of our study will prove useful for the community on exploring and developing efficient volumetric medical image segmentation methods. • We summarize the latest developments about 2.5D segmentation methods for volumetric medical image segmentation tasks. • We make a large-scale empirical comparison of 2.5D segmentation methods with 2D and 3D methods on three representative public datasets involving different modalities (CT and MRI) and targets (cardiac, prostate and abdomen) to systematically evaluate the performance of these methods on different segmentation tasks. • We make conclusions base on the experimental results which we hope will prove useful for the community on exploring and developing efficient volumetric medical image segmentation methods. [ABSTRACT FROM AUTHOR]

• High-speed impact dynamics and temperature changes are measured experimentally. • Theoretical models of drop and surface temperature changes (Δ T d and Δ T s ) are given. • Impact outcomes significantly affect the temperature variations. • Δ T d scaling laws are found for non-Leidenfrost drops: Δ T d ∼ t 3 / 2 first and then ∼ t. • Negligible temperature changes for Leidenfrost droplets due to a vapor layer. [Display omitted] Drop impact on heated surfaces plays a vital role in numerous applications, such as spraying, cooling, and combustion. In addition to impact dynamics, we investigate both droplet and surface temperature changes using a high-speed infrared (IR) camera and thermocouples when a Milli-Q water droplet impacts on a heated surface under various Weber numbers (1.6 ≤ W e ≤ 129 , which compares the drop kinetic to surface energy) and initial surface temperature (100 ∘ C ≤ T s i ≤ 445 ∘ C). Theoretical models of droplet and surface temperature changes are deduced using the energy conservation principle. Both theoretical and experimental results show that temperature variations are dramatically influenced by impact dynamics, which shows spreading, spreading/splashing atomization, or complete rebound depending on W e and T s i. For non-Leidenfrost droplets, the experimental results show that the mean droplet temperature change (Δ T d) is increased with increasing W e and T s i and scales with Δ T d ∼ t 3 / 2 and Δ T d ∼ t in the initial spreading and later-time sticking stages, respectively. The experimental results of solid surface temperature change, Δ T s c ≡ Δ T s (z = 0) , show that Δ T s c alters sharply with W e for W e < 30 , whereas insignificantly for W e > 30 with an identical T s i. Under the same W e , Δ T s c increases first for 100 ∘ C ≲ T s i ≲ 300 ∘ C and then reduces when 300 ∘ C ≲ T s i ≲ 445 ∘ C because of enhanced atomization at elevated temperatures. Finally, a negligible Δ T d and a nearly constant Δ T s c (≲ 4 ∘ C) are observed for Leidenfrost droplets due to the formation of an insulating vapor layer formed underneath the bouncing droplet. [ABSTRACT FROM AUTHOR]

With rapid urbanization, the building stock and related winter heating demands are developing fast in the urban regions of Northern China. In addition to the national strategy to alleviate the air pollution, China has also pledged to achieve carbon neutrality in the future. Although the planning of heating system has been studied from national level, bottom-level regional plans are still lacking, which is especially crucial for a huge country as China that has significantly different characteristics across regions. Therefore, this study examined the transitions of the heating sectors towards 2035 and 2050 for 1047 third-level urban units in Northern China. Future changes in heating resources under the low-carbon transitions and future heating demands with the urbanization trend were investigated. Through geographic mapping of resources and demands for county-level units, the regional differences in future scenarios owing to the uneven socioeconomic development of Northern China were presented, which fill in the knowledge gap. By combining supply-side investigations and demand-side forecasts, various roadmaps were evaluated in terms of primary energy consumption, pollution emissions, and economic benefits. Although significant pollution reductions can be achieved in all proposed roadmaps, the clean heating roadmap with the optimal utilization of available resources presents a lower annualized cost than the business-as-usual roadmap, which simply relies on biomass and supplementary sources such as natural gas. In 2050, the primary energy consumption can be reduced to approximately 3 kgce/m2, with an annualized cost of 6 CNY/m2. Balancing the carbon emissions and economic performance, the optimal renovation pattern, together with supply side technologies, were identified in this study. • Regional investigations in 1047 urban units in Northern China. • Forecast of heating resources and demands in future conditions. • Heat roadmaps were evaluated for environmental and economic implications. • The CHR has lower annualized costs and carbon emissions than the BAU roadmap. • Optimal renovation pattern is identified based on climate and economic implication. [ABSTRACT FROM AUTHOR]

Minuscule pressure/stress detection with high sensitivity can recognize and deal with potential hazards timely, which is crucial for guaranteeing the security of human and stability of social functioning. Lithium niobate (LiNbO 3) single crystal possessing excellent piezoelectric and ferroelectric properties is suitable for fabricating piezoelectric pressure sensors. Therefore, in this paper, the pressure sensors based on LiNbO 3 single crystals with parallel connected structure were researched. Sensors with parallel connected structure can improve the sensitivity from 4.86 V/N of single sheet structure to 8.71 V/N under the stress between 0.01 N and 0.2 N at 30 °C. Both types of the sensors can detect the minuscule stress low to 0.01 N and minuscule pressure generated by balloon pump, blowing largely or slowly. The influence of temperature on sensors was also researched, and the results exhibited that sensors performed well from 30 °C to 60 °C and still sustained high sensitivity of 5.26 V/N of parallel connected structure and 3.71 V/N of single sheet structure at 60 °C. Besides, the response time of sensors to the pressure of balloon pump is 60 ms under different temperature. These excellent properties indicate the great potential applications of fabricated sensors in practical minuscule pressure/stress detection. [ABSTRACT FROM AUTHOR]

[Display omitted] • The PM 2.5 associated premature deaths were estimated to be 965 thousand in 2019 in China. • A range of 816–1304 thousand PM 2.5 associated deaths was projected in 2035 under the indoor guideline of 35 µg/m3. • More stringent policy on indoor PM 2.5 is recommended to bring more benefits. In 2020, China for the first time developed guidelines for indoor fine particulate matter (PM 2.5) in the draft document of indoor air standards, while the associated health implication remains unclear. Here, we first estimated the PM 2.5 associated premature deaths was 965 thousand in 2019, with the indoor PM 2.5 of outdoor origin accounting for 72.9%. Then, we examined the dynamic mortalities under a scenario matrix of 36 conditions, by incorporating various shared socioeconomic pathways in 2035, the draft guidelines and the contributions of ambient PM 2.5 to indoor exposure. Although it may be improbable, the averages of premature deaths associated with ambient PM 2.5 will be 1018–1361 thousand in 2035 when the worst-case scenario of guidelines mandating a yearly (rather than daily) indoor PM 2.5 concentration of 75 µg/m3, compared to the averages of estimation were 816–1304 thousand for better-case scenario of 35 µg/m3. Under these scenarios, the increase in the number of premature deaths was mainly driven by population aging. In 2035, an ambitious target of yearly indoor PM 2.5 concentrations of 15 µg/m3 is anticipated to reduce the number of deaths associated with ambient PM 2.5 by approximately 25% of the 2019 baseline. Stricter guidelines to restrict the indoor PM 2.5 concentrations are recommended to mitigate the mortality risk in the future. [ABSTRACT FROM AUTHOR]

[Display omitted] • WZ/TCM catalyst was suitable for co-catalytic removal of NO and dioxins. • NH 4 + at the Brønsted sites participated in the reaction at high temperature. • Chlorine atom was removed to reduce toxicity at low temperature. Co-catalytic removal of multiple pollutants was able to reduce operating costs of enterprises, which was the main development direction for the waste incineration industry. In this work, series W-Zr-O x /Ti-Ce-Mn-O x (WZ/TCM) catalysts were designed to investigate independent NH 3 -SCR, dioxins oxidation and co-catalytic removal of NO and chlorobenzene (CB, the substitute of dioxins). The W-Zr-O x mainly provided solid super acids; CeO 2 and MnO 2 enhanced the redox performance; and TiO 2 provided a large specific surface area. All these factors guaranteed the excellent performance of WZ/TCM catalyst. The NO conversion of WZ/TCM reached 100 %, and the oxidation activity of CB was also higher than 85 % at 350–500 ℃. Meanwhile, the WZ/TCM had high N 2 selectivity and excellent resistances to sulfur and water. In situ infrared diffuse reflectance spectra confirmed that ammonia gas mainly adsorbed on the Lewis acidic sites of the WZ/TCM catalyst and participated in the SCR reaction at low temperature. The ammonia species (NH 4 +) at the Brønsted sites participated in the reaction at high temperature. Furthermore, the chlorine atom on the chlorobenzene molecule was mainly removed to reduce toxicity when the oxidation of CB was at low temperature, which was deeply oxidized and had a variety of intermediate products at high temperature. [ABSTRACT FROM AUTHOR]

• Cumulative and time-lag effects combined impact vegetation activities. • Solar radiation and temperature are the dominant climate factors in taiga and tundra. • The duration of temporal effects corresponds with ecosystem complexity. • The tundra converted to forests caused by warming enhanced the climate temporal effects. The Northern Hemisphere is experiencing amplified warming and greening; however, the responses of greening to warming remain unclear in terms of the cumulative and time-lag effects of climate factors. We chose Siberia, a hotspot in the Northern Hemisphere, as a study area and identified the relationships between the enhanced vegetation index (EVI) and climate factors (e.g., surface temperature, precipitation, and solar radiation) during 2000–2016. In the growing season, the average cumulative duration of the vegetation response to climate factors was 3.6 ± 2.0 months, and the time-lag period was 1.5 ± 1.2 months, with the longest duration in forests and the shortest duration in tundra. Solar radiation was the dominant driver of vegetation greening in 56% of the study area, followed by temperature (41%) and precipitation (3%). In addition, the temporal effects of the climate factors were separated through partial correlation analysis; vegetation responded most rapidly to solar radiation variations with a cumulative duration of 2.2 ± 1.2 months and a time-lag period of 1.8 ± 0.8 months, followed by temperature (2.3 ± 1.3 and 1.9 ± 1.0 months, respectively). Precipitation showed the longest cumulative effect (3.7 ± 1.5 months) and the shortest time-lag effect (1.2 ± 1.1 months). Our results suggest that vegetation activity is influenced by both cumulative and time-lag effects. The temporal effects vary among the land cover categories, and complex ecosystems generally correspond to long-term temporal effects, and vice versa. Thus, the conversion of tundra to boreal forest due to warming in Siberia enhances the temporal effects of climate. These temporal effects could be incorporated into Earth system models to improve our understanding and monitoring of terrestrial ecosystems. [ABSTRACT FROM AUTHOR]

Movement disorder Parkinson's disease (PD) is the second most common neurodegenerative disease in the world after Alzheimer's disease, which severely affects the quality of patients' lives and imposes an increasingly heavy socioeconomic burden. Aureusidin is a kind of natural flavonoid compound with anti-inflammatory and anti-oxidant activities, while its pharmacological action and mechanism are rarely reported in PD. This study aimed to explore the neuroprotective effects and potential mechanisms of Aureusidin in PD. The present study demonstrated that Aureusidin protected SH-SY5Y cells from cell damage induced by 6-hydroxydopamine (6-OHDA) via inhibiting the mitochondria-dependent apoptosis and activating the Nrf2/HO-1 antioxidant signaling pathway. Additionally, Aureusidin diminished dopaminergic (DA) neuron degeneration induced by 6-OHDA and reduced the aggregation toxicity of α-synuclein (α-Syn) in Caenorhabditis elegans (C. elegans.) In conclusion, Aureusidin showed a neuroprotective effect in the 6-OHDA-induced PD model via activating Nrf2/HO-1 signaling pathway and prevented mitochondria-dependent apoptosis pathway, and these findings suggested that Aureusidin may be an effective drug for the treatment of PD. • Aureusidin showed therapeutic potential to prevent 6-OHDA-induced PD model. • Aureusidin inhibited mitochondria-dependent apoptosis in vitro. • Aureusidin activated the Nrf2/HO-1 antioxidant signaling pathway in vitro. • Aureusidin restored the pathological features of the 6-OHDA-induced PD C. elegans. [ABSTRACT FROM AUTHOR]

Understanding the geological information of fractures from borehole images is essential for conducting a thorough geological analysis in engineering. Recently, the convolutional neural network (CNN) has been employed to extract high-level features of fractures from images. We propose an improved Faster R-CNN (region-based CNN) model for fracture detection, which is suitably equipped with a tailored region proposal network f -RPN for borehole fractures. The anchors generated by f -RPN are specifically designed with a fixed width and varying heights, and only the anchors with the maximum intersection-over-union values are labeled as positive. By taking advantage of the cyclic feature of borehole images, the images are expanded in four orientations to increase the amount and diversity of the model input. The detection results of the model can further be utilized to simplify the Hough transform for characterizing dip angles and dip directions of fractures. The proposed method is validated with borehole images from multiple projects in hydraulic engineering. Experiments reveal that the average detection accuracy is approximately 91.5%, which exhibits an improvement of 11.06% compared to the original Faster R-CNN. Further analysis indicates that both f -RPN and the multi-orientation expansion contribute to improving the prediction accuracy. It is also shown that this method demonstrates robustness against blurred image background and outperforms traditional image processing methods in the presence of noise. The fracture characterization by the simplified Hough transform turns out to be computationally efficient and accurate, with the absolute errors less than 5° compared to the manual identification results. • Faster R-CNN is improved with a tailored f -RPN for borehole fracture detection. • Borehole images are expanded in multiple orientations for data augmentation. • Simplified Hough transform based on Faster R-CNN results to characterize fractures. • The method is robust with blurred and noisy image data from real hydraulic engineering. [ABSTRACT FROM AUTHOR]

Photovoltaic (PV) power generation can directly convert solar radiation photons into electrical energy, but PV panels produce a large amount of waste heat during absorption of solar radiation, significantly increasing the working temperature and reducing the photoelectric conversion efficiency of the panels. In this study, a phase-change material (PCM) is used to cool the PV panels, and fins are added to enhance PCM heat transfer. Using numerical simulation, the effects of fin spacing, fin height, solar radiation intensity, and ambient temperature on the heat-dissipation performance of the PV/PCM system were then studied. The fin–PV/PCM structure showed a good cooling effect. With decreasing fin spacing, the increase in the panel temperature gradually decreased. Fin spacing of 6 mm exhibited the best cooling effect, with the battery plate temperature reduced by 31.9 °C compared with that of the natural convection condition without fins. The increase in the panel temperature gradually decreased with increasing fin height, and the cooling effect of the panel improved significantly when the fin height was increased from 30 to 70 mm. The heat-dissipation effect of the fin–PV/PCM system was better with higher solar radiation intensity and higher ambient temperature. The results of this study will have important reference value for performance improvement of PV panels. [ABSTRACT FROM AUTHOR]

In view of the problems that the surface of radiant air conditioning is prone to condensation and the cooling capacity is limited by the dew-point temperature, a model for radiant air conditioning with an infrared-transparent membrane has been developed which installs an infrared-transparent membrane vertically below the radiant cooling plate of a radiant air conditioner to form a closed gas interlayer between them. On the basis of this, the effects of the gas type (vacuum, nitrogen, air, argon, and carbon dioxide) and relative air humidity in the interlayer on the radiative heat transfer, convective heat transfer, and radiative heat-transfer ratio of the radiant cooling plate, and the surface temperature of the infrared-transparent membrane were investigated by CFD (Computational Fluid Dynamics) numerical simulation. When the thickness and gas type of the interlayer were 20 mm and vacuum, respectively, the radiant cooling plate showed the most radiative heat transfer, the least convective heat transfer, the largest radiative heat-transfer ratio, and the highest surface temperature of the membrane, followed by when the gas type was argon. There was no risk of condensation in the air conditioning when the relative air humidity in the interlayer did not exceed 80%. The results of the study provide a theoretical basis for the optimal design of radiant air conditioners. [ABSTRACT FROM AUTHOR]

[Display omitted] • The photoresponsive microcapsules were prepared via a one-step method. • The coordination polymer assembled to form nanofibers at the interface. • The nanofibers, as emulsifiers, exhibited a Janus structure. • The microcapsules can be used as photocontrolled microreactors. In recent decades, stimuli-responsive microcapsules have attracted increasing attention due to their potential applications as microreactors for drug delivery, catalytic reactions, etc.; however, the complicated preparation process and the difficulty of size fixation and recycling need to be addressed. In this work, we developed a one-step preparation method to fabricate photoresponsive microcapsules based on the interfacial self-assembly of an azopyridine coordination polymer. The coordination of silver ions and ligands was performed in situ at an oil–water interface, and the coordination polymer self-assembled into nanofibers, which stabilized the Pickering emulsion and produced water-in-oil microcapsules. Additionally, microcapsules with a uniform size were prepared based on a microfluidic process. Both the crystal structure and the Janus surface of nanofibers were determined to elucidate the mechanism of self-assembly and interface stabilization. When the microcapsules were irradiated with 365 nm UV light, the cis isomerization of the azo group in the coordination polymer led to the disassembly and dissolution of the nanofibers, resulting in fusion of the microcapsules and demulsification of the Pickering emulsion. Conversely, visible light and heating triggered the trans isomerization of the azopyridine coordination polymer and thus caused regeneration of the nanofibers and recovery of the microcapsules. Finally, microcapsules encapsulating several reactants were employed in the pH-indicator discoloration reaction and the enzyme-catalyzed decomposition of H 2 O 2 to demonstrate their feasibility as photocontrolled microreactors. This kind of photoresponsive microcapsule with one-step preparation and good recyclability might have potential applications in the field of microreactors. [ABSTRACT FROM AUTHOR]

Adaptive building envelopes with advanced optical materials have received much attention as they can actively regulate solar heat gain. In present work, an adaptive building roof (ABR) combining variable transparency shape-stabilized PCM (VTSS-PCM) was proposed. A numerical model was developed firstly for dealing with time-space effect of radiation properties of VTSS-PCM and validated by experiments. The difference in performance between ABR and the common insulation roof was simulated. Then, a global sensitivity analysis based on the Monte-Carlo method was conducted to determine key parameters among 22 system parameters. Finally, the influences of key parameters on the performance of ABR were studied in detail. The results show that the ABR can reduce the heat gain by 35.25 % compared with the common insulation roof. The thickness of VTSS-PCM, the extinction coefficient of VTSS-PCM in the transparent state, and the albedo of reflective film are three key parameters. The thickness of the VTSS-PCM layer and the extinction coefficient of VTSS-PCM in the transparent state have positive correlations with the heat gain, whereas the albedo of the reflective film has a negative correlation with the heat gain. The time-space effect of radiation properties of VTSS-PCM during the phase change causes the solar absorptance of ABR to be unstable. This research guides the modeling of VTSS-PCM and facilitates the application of ABR. [ABSTRACT FROM AUTHOR]

• The novel phosphate group intercalated g-C 3 N 4 nanosheets is successfully fabricated. • It exhibits the improvement of photocatalytic activity. • The introduction of phosphate group establishes interlayer channels for faster electron transfer. • The ultrathin nanosheets could shorten charge carrier diffusion path and offer more active reactive sites. • A possible photocatalytic mechanism is proposed. [Display omitted] The development of effective methods that simultaneously regulate energy bands and electron transfer to realize high photocatalytic performance of graphitic carbon nitride (g-C 3 N 4) remains a major challenge. Herein, we developed a novel strategy to construct a phosphate-group-intercalated g-C 3 N 4 nanosheet photocatalyst that is driven by visible light. A series of characterizations confirmed that phosphate groups were inserted between the g-C 3 N 4 layers and formed a vertical channel. A porous ultrathin nanosheet structure was formed. The synergistic effect of the phosphate-group intercalation and porous ultrathin nanosheet structure endowed this material with controllable electronic structures and upshifted the conduction bands, while facilitating photogenerated charge transfer and providing numerous active reactive sites. As a result, the g-C 3 N 4 intercalated with phosphate groups photocatalyst exhibited high photocatalytic activity for H 2 evolution and degradation of a coloured dye rhodamine B (RhB) under visible light. The maximum H 2 production rate and degradation rate constant of the phosphate-group-intercalated g-C 3 N 4 nanosheets (CNP-5) were 17-fold and 6-fold higher than those of primal g-C 3 N 4 , respectively. This study provides new insights for the incorporating atomic groups with modification effects and represents a giant leap in the construction of two-dimensional (2D) materials. [ABSTRACT FROM AUTHOR]

The world is now vigorously developing commerce and industry, which is due to the needs of modern humans. But this demand causes many environmental problems. Multi-element metal composites have good electrochemical performance and controllable microstructure, which proves to be a material with excellent performance. Among them, the modified multi-element metal composites used as electrodes of energy storage devices (LIBs, SCs, SIBs) will exhibit good cycle stability and higher capacity. The modified multi-element metal composite material also has good mechanical stability, and when used as a catalyst and adsorbent, the material structure will not be destroyed. When multi-element metal composites are used as catalysts, the synergistic effect of multiple metal elements increases the catalytic efficiency of the catalyst. The nano-level multi-element metal composite materials with large specific surface area show more excellent adsorption capacity in adsorption applications. High-efficiency catalysts and adsorbents are of great significance to the restoration and protection of the environment. [Display omitted] [ABSTRACT FROM AUTHOR]

[Display omitted] • CM-CNT/graphene electrode was directly constructed by core-sheath 3D printing. • Electrode features hierarchical pores and highly interconnected conductive network. • Electrode enables adequate channel and pathway for ultrafast ion/electron transport. • Pseudocapacitor exhibits prominent stability and electrochemical performance. Pseudocapacitance holds great promise for energy density improvement of supercapacitors. However, how to enable a pseudocapacitor with combined feature of high mass loading of pseudocapacitive material, efficient ion/electron transport, as well as facile and scalable manufacturing, still remains a significant challenge. Herein, we demonstrated an efficient and scalable extrusion-based core-sheath 3D printing strategy for directly and controllably constructing pseudocapacitive carbon nanotube (CNT)- MoS 2 (CM)-CNT/graphene electrode, with high mass loading up to 55% and uniform distribution of MoS 2 , hierarchically porous configuration and highly interconnected conductive network framework built by coupling graphene scaffold located in shell layer and CNTs situated in core layer. The unique architecture featuring adequate channels and pathways can act as a "superhighway" for ultrafast ion diffusion and electron transport throughout the entire device, thereby enabling fast kinetics of Faradaic reactions and abundant availability of electrochemical active sites. A symmetric pseudocapacitor assembled with 3D-printed electrode and in situ electrospun poly(vinylidene fluoride) (PVDF) nanofiber separator delivered high specific capacitance (558 mF cm−2) and energy density (49 µWh cm−2), remarkable cycling stability (87% after 8000 cycles), and superior capacity even at large electrode thickness. This work has shed light on new strategies for designing and fabricating high-performance pseudocapacitors toward future uses. [ABSTRACT FROM AUTHOR]