Local Strain Research Articles

The Role of the Horizon in Modeling Failure due to Strain and Damage Localization with Peridynamics

The Role of the Horizon in Modeling Failure due to Strain and Damage Localization with Peridynamics

Numerical Analysis of Strain Localization in Granular Soils by Modified Cam-Clay Model Based on Micropolar Continuum Theory

Numerical Analysis of Strain Localization in Granular Soils by Modified Cam-Clay Model Based on Micropolar Continuum Theory

First report of Colletotrichum queenslandicum associated with chilli anthracnose in Mauritius.

Globally, chilli (Capsicum annuum L.) is one of the most economically important and widely cultivated crop which elicits ethnomedicinal and nutritional potential as well as enhancing the taste and aroma of foods (Ayob et al., 2022; Kiran et al., 2020). Anthracnose disease is regarded as a prime constraint in chilli production, leading to enormous losses in tropical and subtropical countries. In September 2022, chilli fruit displaying sunken, shriveled and dark bown to black lesion with abundant acervuli on the surface was obtained from Flacq, Mauritius. From the symptomatic tissue, small pieces of the diseased tissue were excised, surface-disinfected using 1% sodium hypochlorite, twice rinsed using sterilized distilled water, air-dried and plated on PDA. After 7 days of incubation at room temperature, white to greyish white colony with dense white cottony aerial mycelium was recovered. Out of two isolates, CHF and CH10, the latter was considered for morphological and molecular characterization. The observed conidia (n=30) were unicellular, straight, cylindrical with rounded ends and slight constriction near the centre and had average length and width of 20.5 µm and 6 µm, respectively. For growth rate measurement of the isolate, two 5×5 mm of fungal agar plugs were taken from growing edge of colony, inoculated at centre of individual PDA plate and incubated at room temperature with a natural light/dark cycle. The diameter of the cultures were measured perpendicularly for a period of 7 days and the growth rate was calculated as 7-day average of daily growth (mm day-1). The growth rate of the fungal isolate (CH10) was 13.5 mm day-1 on PDA. Based on the morphological characters, the isolate was classified within the C. gloeosporioides species complex. For precise identification of the isolate, DNA was extracted from fungal mycelium using traditional DNA isolation methods (Ranghoo and Hyde, 2000), followed by PCR amplification and DNA sequencing using primer pairs ITS4/ITS5 (White et al., 1990), GDF/GDR and T1/Bt2b (Gan et al., 2016), respectively. ITS gene sequence (600 bp) confirmed that the isolate was Colletotrichum, with 99.83% similarity to KR704204 while GADPH (277 bp), TUB2 (733 bp) and ApMat (801 bp) gene sequences showed 99.64 to 100% similarity to C. queenslandicum with GenBank reference sequences, KT372374, KU221378 and MG674932 respectively. The gene sequences of isolate CH10 were deposited in GenBank database under the following accession numbers OR681557 (ITS), OR233734 (GADPH), OR475575 (TUB2) and PP622748 (ApMat). Koch's postulates were confirmed by spraying disease-free chilli plants with 10µL of conidial suspension (1 × 106 spores/ml) prepared from 7 days old colony of isolate CH10. Healthy chilli plants inoculated with sterile distilled water served as a negative control experiment. The plants were grown in pots in a moist chamber at 25˚C. After 5 days post-inoculation, anthracnose symptoms were developed on test plants while the control plant remained asymptomatic. The original isolate was successfully recovered from the test fruits, thus fulfilling Koch's postulates. The experiment was repeated thrice and revealed the same results. To the best of our knowledge, this is the first record of C. queenslandicum in Mauritius and is the first time to report anthracnose of chilli caused by this fungus. Colletotrichum queenslandicum has previously been reported in Europe, Mexico, US, Puerto Rico, Australia, Fiji, Brazil, Indonesia and China. Furthermore, the latter was associated with papaya, avocado, cashew, coffee, Persian lime, Licania tomentosa, white mangrove, lychee, mango, Nephelium lappaceum, olive, passionfruit, Dracaena cambodiana and Syzygium australe (Câmara and Vieira, 2022; Shidiq et al., 2024; Wang et al., 2022). This study will allow local farmers training and extension facilities to increase awareness among farmers about this disease-causing agent and allow them to take necessary measures for building up chilli crops resilience against this new and emerging pathogen in Mauritius.

Photocycloaddition of Zero-Dimensional Metal Halide Hybrids with Reversible Photochromism.

In this work, two zero-dimensional (0D) metal halide hybrids L2ZnBr4 [1, L = (E)-4-(2-(1H-pyrrol-3-yl)vinyl)-1-methylpyridin-1-ium] and L6Pb3Br12 (2) were prepared, which demonstrated photochromism and photoinduced cracking. Upon irradiation at 450 nm, a single crystal-to-single crystal transformation occurred as a result of the [2 + 2] photocycloaddition of L. Interestingly, compared to the complete photocycloaddition of L in 1, only two-thirds of L monomers could be photodimerized in 2 because of the difference in L orientation. 1 shows reversible photochromic behavior including rapid response time, few cracks, high conversion rate, and good reaction reversibility, while 2 exhibits no significant color change but distinct photoinduced cracking because of the large local lattice strain induced by inhomogeneous and anisotropic deformation. Moreover, the photocycloaddition of L results in the distinct shift of photoluminescence of 1 and 2, attributed to the variation in conjugation of π electrons and distortion of metal halide clusters. As a proof-of-concept, reversible optical writing is demonstrated for 1. These findings provide new insights into the design of stimuli-responsive multifunctional materials.

Predicting treatment response in multicenter non-small cell lung cancer patients based on federated learning

BackgroundMulticenter non-small cell lung cancer (NSCLC) patient data is information-rich. However, its direct integration becomes exceptionally challenging due to constraints involving different healthcare organizations and regulations. Traditional centralized machine learning methods require centralizing these sensitive medical data for training, posing risks of patient privacy leakage and data security issues. In this context, federated learning (FL) has attracted much attention as a distributed machine learning framework. It effectively addresses this contradiction by preserving data locally, conducting local model training, and aggregating model parameters. This approach enables the utilization of multicenter data with maximum benefit while ensuring privacy safeguards. Based on pre-radiotherapy planning target volume images of NSCLC patients, a multicenter treatment response prediction model is designed by FL for predicting the probability of remission of NSCLC patients. This approach ensures medical data privacy, high prediction accuracy and computing efficiency, offering valuable insights for clinical decision-making.MethodsWe retrospectively collected CT images from 245 NSCLC patients undergoing chemotherapy and radiotherapy (CRT) in four Chinese hospitals. In a simulation environment, we compared the performance of the centralized deep learning (DL) model with that of the FL model using data from two sites. Additionally, due to the unavailability of data from one hospital, we established a real-world FL model using data from three sites. Assessments were conducted using measures such as accuracy, receiver operating characteristic curve, and confusion matrices.ResultsThe model’s prediction performance obtained using FL methods outperforms that of traditional centralized learning methods. In the comparative experiment, the DL model achieves an AUC of 0.718/0.695, while the FL model demonstrates an AUC of 0.725/0.689, with real-world FL model achieving an AUC of 0.698/0.672.ConclusionsWe demonstrate that the performance of a FL predictive model, developed by combining convolutional neural networks (CNNs) with data from multiple medical centers, is comparable to that of a traditional DL model obtained through centralized training. It can efficiently predict CRT treatment response in NSCLC patients while preserving privacy.

High Performance Mg Alloy with Designed Microstructure and Phases.

A high strength and ductile Mg-Gd-Y-Zn-Zr alloy was designed and fabricated. The local strain evolution of the alloys during plastic deformation was analyzed using high-resolution digital image correlation (DIC). The results showed that the β particles, nano-sized γ' phases, and LPSO phases were distributed in the as-extruded alloy and a bimodal microstructure was exhibited, including elongated un-dynamic recrystallized grains and fine dynamic recrystallized grains. With increasing extrusion ratio, the grain size remained, with the volume fraction of dynamic recrystallization of the as-extruded alloy increasing from 30% to 75%, and the as-extruded alloy exhibited a high strength-ductility synergy, which is attributed to the grain refinement, extensive β particles, and elongated block-shaped LPSO phases. The strain evolution analysis showed that a strain-transfer from un-DRXed regions to adjacent DRXed regions and LPSO phases can promote uniform plastic deformation, which tends to improve the ductility of the alloy.

In Situ Analysis of Local Strain Distribution of Amorphous Polyrotaxane Adhesives Constrained by Metal Substrates

In Situ Analysis of Local Strain Distribution of Amorphous Polyrotaxane Adhesives Constrained by Metal Substrates

Insights into the fracture properties of recycled ceramic and rubber composite cement-based materials: Fracture mechanics, acoustic emission, and digital image correlation

Recycling plays a pivotal role in construction waste management, offering a path toward sustainability for the industry. Although waste ceramics can compensate for the mechanical properties of rubberized cement-based materials, the smooth glaze surface of waste ceramics increases the risk of brittle failure. This study investigated the effect of ceramic waste aggregate (CWA) on the fracture properties of rubberized mortar through the three-point bending experiment. The fracture parameters were calculated using the Double-K fracture criterion. Simultaneously, acoustic emission (AE) and digital image correlation (DIC) technologies were used to explore the fracture damage mechanism and the crack evolution process. The results indicate that the integration of 15 % and 45 % CWA significantly enhances both initiation (KICini) and unstable (KICun) fracture toughness, whereas incorporating 90 % CWA leads to a 16.8 % and 8.2 % decrease in (KICini)and (KICun), respectively. The fracture energy (GF) of rubberized mortar exhibits a continuous decline as a function of increasing CWA substitution rates. Furthermore, GMM-based RA-AF analysis elucidates a positive correlation between CWA content and tensile crack proportion in specimens. When the CWA content reaches 90 %, the proportion of tensile cracks is quantified at 29.83 %, surpassing that in standard mortar. Crack tip opening displacement (CTOD) at 0.5post-Pmax for specimens with 15 %, 45 %, and 90 % CWA substitution reveal improvements of 49.1 % and 17.2 %, and a decrease of 23.3 %, respectively, compared to standard mortar, indicating that 15 % and 45 % CWA effectively preserve the ductility of rubberized mortar. Elevated CWA substitution leads to an amplified strain localization phenomenon around the crack tip, heightening the propensity for brittle failure.

In Situ Structural Densification of Hydrogel Network and Its Interface with Electrodes for High-Performance Multimodal Artificial Skin.

The multisensory responsiveness of hydrogels positions them as promising candidates for artificial skin, whereas the mismatch of modulus between soft hydrogels and hard electrodes as well as the poor adhesion and conductance at the interface greatly impairs the stability of electronics devices. Herein, we propose an in situ postprocessing approach utilizing electrochemical reactions between metals (Zn, etc.) and hydrogels to synergistically achieve strong adhesion of the hydrogel-electrode interface, low interfacial impedance, and local strain isolation due to the structural densification of the hydrogel network. The mechanism is that Zn electrochemically oxidizes to Zn2+ and injects into the hydrogel, gradually forming a mechanically interlocked structure, Zn2+-polymer dual-helix structural nodes, and a high-modulus ZnO from the surface to the interior. Compared to untreated samples, the treated sample displays 8.7 times increased interfacial adhesion energy between the hydrogel and electrode (87 J/m2), 95% decreased interfacial impedance (218.8 Ω), and a high-strain isolation efficiency (εtotal/εisolation > 400). Akin to human skin, the prepared sensor demonstrates multimodal sensing capabilities, encompassing highly sensitive strain perception and simultaneous perception of temperature, humidity, and oxygen content unaffected by strain interference. This easy on-chip preparation of hydrogel-based multimodal sensor array shows great potential for health and environment monitoring as artificial skin.

Exploring high-temperature deformation and damage behaviour in high-performance ferritic (HiperFerSCR) steel with Laves phase particles

The mechanical behaviour at high temperatures and the corresponding microstructural characteristics of a newly developed high-performance ferritic, salt-corrosion-resistant (HiperFerSCR) steel containing Laves phase particles were studied to investigate the active deformation mechanisms and the evolution of damage. A series of tensile tests were conducted at both room temperature and elevated temperatures (550, 600, and 650 °C), along with detailed microstructural analyses using scanning electron microscopy and electron backscatter diffraction. The tensile flow characteristics of HiperFerSCR at room temperature exhibit significant strain hardening, which is due to the Laves phase particles that encourage planar slip, demonstrated by the formation of numerous slip bands and low-angle grain boundaries. The intersections of these slip bands with grain boundaries appear to be the initiation sites for cracks, which then propagate intergranularly as well as within grain interiors along the slip bands. At high temperatures, the flow characteristics show strain softening, which is attributed to a dynamic recovery (DRV) mechanism, evident in the formation of subgrain boundaries decorated with Laves phase particles. During high-temperature deformation, damage initiates in the particle-free zone (PFZ), which expands at higher temperatures, leading to strain localisation, void formation along the grains, and eventually, grain boundary decohesion.

Stability of Running Stride Biomechanical Parameters during Half-Marathon Race

This study explores the stability of biomechanical parameters of the running stride of male trained athletes during a half-marathon competition. Using a field-based descriptive design, eight male athletes from a local training group were monitored throughout an official half-marathon race under identical conditions, assessing biomechanical parameters including ground contact time (GCT), leg spring stiffness (LSS), vertical oscillation (VO), and stride length (SL) recorded via the Stryd Summit Power Meter. A repeated measures analysis of variance (RM ANOVA) was conducted to detect significant changes in biomechanical parameters as the race progressed. Results demonstrated minimal changes in all parameters, with no significant differences observed for GCT (F = 0.96, p = 0.38), VO (F = 0.23, p = 0.87), and SL (F = 1.07, p = 0.35), and a small (η2 = 0.004) yet statistically significant difference in LSS (F = 5.52, p = 0.03) between the first and second segments, indicating that athletes were able to maintain stable biomechanical parameters throughout the race. The conclusion highlights the need for personalized training programs tailored to the unique biomechanical adaptations and demands of endurance running.

Multi-scale microstructural construction in ultralight graphene aerogels enables super elasticity and unprecedented durability for impact protection materials

Ultralight graphene aerogels have gained extensive recognition in the impact protection field. However, attaining both elasticity and durability at low material density is challenging due to their intrinsic conflicts. Inspired by the mantis ootheca, we present a simultaneous improvement in the elasticity, durability, and density restrictions of ultralight graphene aerogels via constructing a multiscale honeycomb microstructure (MHM) within the graphene skeleton. This approach enables resulting graphene aerogel to achieve a strength per unit volume of 284.6 cm3 mg−1, the ability to recover its shape within 10 ms after an impact at 3.569 m/s, and maintain 97.2 % of its sample height after 20,000 cycles at 90 % strain. The operand analyses and calculation results reveal that the MHM structure facilitates this aerogel’s dual-stage stress transfer pathway. Initially, the macroscale honeycomb structure (millimeter-scale) of the graphene aerogels bear and transmit stress to the surrounding regions, followed by the microscale honeycomb structure (micron-scale) deformation to convert stress kinetic energy into elastic potential energy. This two-stage stress transition mechanism of the MHM structure can effectively mitigate excessive local stress and suppress strain localization, thus providing remarkable elasticity and durability. Ultimately, the obtained graphene aerogel demonstrates promising applications as a fall height detection device and impact protective material.

Influence of high-strain-rate compression and subsequent heat treatment on (TiNbZr)89(AlTa)11 refractory high-entropy alloys: Dynamic-mechanical behavior and microstructural changes

This study explored the dynamic-mechanical behavior of a novel low-density (TiNbZr)89(AlTa)11 refractory high-entropy alloy (RHEA) across strain rates ranging from 1.0 × 103 to 3.5 × 103 s−1. A significant increase in the yield and ultimate compressive strengths with rising strain rates up to 3.0 × 103 s−1 was observed and attributed to enhanced dislocation activities and stress-induced microstructural transformations. The formation of the B2 phase and Zr5Al3 precipitates was found to be crucial in bolstering the alloy strength at high strain rates. Beyond strain rates of 3.0 × 103 s−1, a decrease in strength occurred due to thermal softening and strain localization. Microstructural analyses at 3.5 × 103 s−1 revealed grain refinement, the development of micro shear bands, and dislocation tangles, which were indicative of dynamic recrystallization. Besides, the findings also revealed that the post-dynamic compression heat treatment further enhanced the hardness and microstructural stability of the alloy. These results highlight the potential of the (TiNbZr)89(AlTa)11 RHEA for applications requiring materials with high strength-to-weight ratios, particularly in dynamically loaded environments. It is expected that the results of this study will further advance our fundamental understanding of the behavior of RHEAs under extreme conditions, thereby opening new avenues for material innovation.

Effect of thermomechanical processing on mechanical properties and microstructural evolution of Al-Mg-Zn alloy

This study investigates the impact of thermomechanical processing route on the microstructural evolution, and mechanical properties (tensile strength, hardness and elastic plastic (J1c) fracture toughness) of the 7075-aluminum alloy. As received 7075 Al Alloy underwent for the solution heat treatment (SHT) followed by artificial aging and cold rolling (CR) process, respectively. Generally, it was observed that cold rolling of 7075 Al alloy is very challenging, but in this work 90% cold rolling successfully achieved by optimizing the thermo-mechanical process. The novel heat treatment process for achieving the 90% cold rolling reduction as follow: firstly, SHT was performed at 470°C for 8hours(h), there after aging at 140°C for 21h was performed. Characterization techniques like X-ray diffraction (XRD), optical microscopy (OM), scanning electron microscopy (SEM), transmission electron microscopy (TEM) was employed to assess the microstructure and phase constituents. Elastic plastic (J1c) fracture toughness was studied well on SHT, peak aged and rolled sample. Additionally, Vickers hardness and tensile test were performed. Aging and Cold rolling treatment effectively enhanced tensile strength and hardness ascribed to formation of fine rod shape precipitates of η"(Mg2Zn3) and formation of sub grains with localized strain accumulation, respectively. Split diffraction spots with satellite pattern in long range ordering has also observed in selected area electron diffraction (SAED) pattern of η" attributable to stacking faults and periodic arrangement of precipitates, respectively, as a consequence of this 90% cold rolling of 7075 Al alloy was accomplished. The maximum Vickers hardness, Tensile strength and Elastic plastic (J1c) fracture toughness values were achieved after SHT (470°C for 2h), peak aged (PA) (140°C for 21h) and 90% cold rolling are 226 HV, 526MPa and 344.54kJ/m2, respectively.

A review of minimal access surgery provision and training within the United Kingdom

When combined with healthcare pressures, the exponential growth of robotic-assisted surgery (RAS) has impacted UK-based training outcomes, including the learning curve to competency. Aim: To ascertain the current provision of RAS and investigate differences in access to minimal access surgical (MAS) facilities and training across the UK. A two-armed electronic survey was conducted. The first arm questioned clinical leads regarding robotic practice and future training provisions. The second investigated trainee and trainers’ perceptions of MAS training and facilities. 64% (52/81) of responding trusts utilise a robotic system. The majority (68% [55/81]) have plans to expand or acquire a system within 3 years. 171 responses from 112 UK and Republic of Ireland hospitals were collected for Arm 2. Laparoscopic categories queried whether trainees had access to a formal curriculum, training days and sim-boxes. Most consultants (51.9%) and trainees (51.6%) reported that there was no formal local training curriculum for robotic surgery. Combined responses demonstrated 42.1% (n = 195/463) said “yes”, 39.5% (n = 183) “no” and 18.4% (n = 85) “don’t know”. For combined robotic categories (simulation, training days and operative lists) 28.3% (n = 134/473) responded “yes”, 51.6% (n = 244) said “no” and 20.1% (n = 95) said “don’t know”. This study provides insight into the current provision of robotic-assisted surgery at UK trusts and highlights the need to facilitate regular clinical training and equitable access to MAS simulation within a formal curriculum. This may aid regulation of training in parallel with the expansion of robotic practice and avoid a significant skill acquisition gap and risks to patient safety.

A federated learning model for integrating sustainable routing with the Internet of Vehicular Things using genetic algorithm

A distributed machine learning technique called federated learning allows numerous Internet of Things (IoT) edge devices to work together to train a model without sharing their raw data. Internet of Vehicular Things (IoVT) are an important tool in smart cities for moving objects, such as knowing the traffic patterns, road conditions, vehicle behavior, etc. To enhance traffic management and optimize routes, federated learning, and IoT must work jointly, which may achieve sustainable development goals (SDG) in many ways. This research outlines a system for federated learning in vehicular networks in smart cities. The suggested architecture considers the difficulties presented by such situations’ restricted network connectivity, privacy issues, and security concerns. The framework employs a hybrid methodology integrating federated learning on a centralized server with local training on individual cars. The proposed framework is assessed based on a real-world dataset from a smart city through IoT devices. The findings demonstrate that the suggested method successfully increases model accuracy while preserving the confidentiality and security of the data. In this investigation, we incorporated the Federated Learning model, which can fetch all the information between arbitrary nodes and derive the Traffic, Fuel Cost, Safety, Parking Cost, and Transportation cost for a better routing approach. The suggested framework can be utilized to increase the effectiveness of the transportation system, decrease congestion in smart cities, and improve traffic management. We employ an improved genetic algorithm (iGA) with generation-dependent even mutation to tackle the emission in the smart environment.

Assessment of antimicrobial activity and GC-MS using culture filtrate of local marine Bacillus strains

Secondary metabolites produced by Bacillus species from marine sources encompass a variety of compounds such as lipopeptides, isocoumarins, polyketides, macrolactones, polypeptides and fatty acids. These bioactive substances exhibit various biological activities, including antibiotic, antifungal, antiviral, and antitumor properties. This study aimed to isolate and identify a particular species of Bacillus from marine water and organisms that can produce bioactive secondary metabolites. Among the 73 Bacillus isolates collected, only 5 exhibited antagonistic activity against various viral and bacterial pathogens. The active isolates were subjected to 16S rRNA sequencing to determine their taxonomical affiliation. Among them, Bacillus tequilensis CCASU-2024-66 strain no. 42, with the accession number ON 054302 in GenBank, exhibited the highest inhibitory potential. It displayed an inhibition zone of 21 mm against Bacillus cereus while showing a minimum zone of inhibition of 9 mm against Escherichia coli and gave different inhibition against pathogenic fungi, the highest inhibition zone 15 mm against Candida albicans but the lowest inhibition zone 10 mm was against Botrytis cinerea, Fusarium oxysporum. Furthermore, it demonstrated the highest percentage of virucidal effect against the Newcastle virus and influenza virus, with rates of 98.6% and 98.1%, respectively. Furthermore, GC-MS analysis was employed to examine the bioactive substance components, specifically focusing on volatile and polysaccharide compounds. Based on these results, Bacillus tequilensis strain 42 may have the potential to be employed as an antiviral agent in poultry cultures to combat Newcastle and influenza, two extremely destructive viruses, thus reducing economic losses in the poultry production sector. Bacteria can be harnessed for the purpose of preserving food and controlling pathogenic fungi in both human and plant environments. Molecular docking for the three highly active derivatives 2,3-Butanediol, 2TMS, D-Xylopyranose, 4TMS, and Glucofuranoside, methyl 2,3,5,6-tetrakis-O-(trimethylsilyl) was carried out against the active sites of Bacillus cereus, Listeria monocytogenes, Candida albicans, Newcastle virus and influenza virus. The data obtained from molecular docking is highly correlated with that obtained from biology. Moreover, these highly active compounds exhibited excellent proposed ADMET profile.

FedKG: A Knowledge Distillation-Based Federated Graph Method for Social Bot Detection.

Malicious social bots pose a serious threat to social network security by spreading false information and guiding bad opinions in social networks. The singularity and scarcity of single organization data and the high cost of labeling social bots have given rise to the construction of federated models that combine federated learning with social bot detection. In this paper, we first combine the federated learning framework with the Relational Graph Convolutional Neural Network (RGCN) model to achieve federated social bot detection. A class-level cross entropy loss function is applied in the local model training to mitigate the effects of the class imbalance problem in local data. To address the data heterogeneity issue from multiple participants, we optimize the classical federated learning algorithm by applying knowledge distillation methods. Specifically, we adjust the client-side and server-side models separately: training a global generator to generate pseudo-samples based on the local data distribution knowledge to correct the optimization direction of client-side classification models, and integrating client-side classification models' knowledge on the server side to guide the training of the global classification model. We conduct extensive experiments on widely used datasets, and the results demonstrate the effectiveness of our approach in social bot detection in heterogeneous data scenarios. Compared to baseline methods, our approach achieves a nearly 3-10% improvement in detection accuracy when the data heterogeneity is larger. Additionally, our method achieves the specified accuracy with minimal communication rounds.

Global prototype distillation for heterogeneous federated learning.

Federated learning is a distributed machine learning paradigm where the goal is to collaboratively train a high quality global model while private training data remains local over distributed clients. However, heterogenous data distribution over clients is severely challenging for federated learning system, which severely damage the quality of model. In order to address this challenge, we propose global prototype distillation (FedGPD) for heterogenous federated learning to improve performance of global model. The intuition is to use global class prototypes as knowledge to instruct local training on client side. Eventually, local objectives will be consistent with the global optima so that FedGPD learns an improved global model. Experiments show that FedGPD outperforms previous state-of-art methods by 0.22% ~1.28% in terms of average accuracy on representative benchmark datasets.

Structural concrete measurements: New distributed approach for standard specimens

Concrete structures are widely used in construction because of their many advantages, including high corrosion resistance, durability, low operating costs, and the ability to design any geometry. However, they also have some disadvantages, such as their high dead weight and susceptibility to cracking. Although this material is widely used globally, new types of concrete, such as high-strength or lightweight aggregate-based, are still being developed. This requires constant control and testing in both laboratory and field conditions. Concrete is a heterogeneous material due to the influence of aggregate grains and local discontinuities, voids or pores. Given these characteristics, it is clear that most conventional diagnostic solutions, such as spot foil strain gauges or extensometers, are unable to accurately analyse these localised effects as they average local strains over the measurement base. In contrast, distributed fibre optic sensing (DFOS) enables strain readings to be taken with millimetre geometric resolution, significantly extending the structural assessment capability. The article describes research on standard concrete cylinders equipped with surface-mounted optical fibres for simultaneous measurements of distributed strains in both the longitudinal direction (compression zone) and circumference direction (tension zone). The novelty of the proposed approach, apart from the way of arranging the sensing fibres on the specimens, lies mainly in the new diagnostic approach that allow (1) the evaluation of the homogeneity of the concrete using statistical parameters, (2) the identification of microcracking and (3) the identification of near-to-failure state based on Poisson’s ratio analysis. Twenty cylindrical specimens were laboratory tested using reference techniques to validate the performance of the DFOS-based system and to provide data for statistical evaluation. The results indicate the possibility of detecting the approaching damage even without knowledge of the material properties of the specimen or the initial stress level at the moment of starting the measurements. The discussion also covered future research directions and potential improvements.