Oils and soaps have traditionally been employed to enhance the performance of lime- and gypsum-based plasters, but the influence of these additives on the structure and physicochemical behavior of gypsum-based materials remains insufficiently understood. Here the effects of adding Ca-soap/oil (soap being prepared with 1or 5 wt% olive oil and limewater) to low or high T (calcined at either 180 or 330 °C) gypsum plasters are investigated with the goal of obtaining an optimized material with enhanced weathering resistance for outdoor finishing applications. Samples were subjected to various tests to evaluate the effect of the Ca-soap/oil on their hydration behavior, hydric properties, textural and structural features, and mechanical strength. Although the incorporation of Ca-soap/oil led to moderate reductions in mechanical strength (5 wt% oil reduced strength from 6.05 MPa to 4.69 and 4.27 MPa to 3.40 MPA in low and high T gypsum, respectively), the modified plasters exhibited improved plasticity, and lower surface hydrophilicity (5 wt% oil increased contact angles from 13.0 ± 1.6° to 71.1.0 ± 6.4° and 13.0 ± 3.0° to 91.8 ± 10.2° in low and high T gypsum, respectively). Calcination T influenced open porosity, being ∼40 and 49% for low and high T gypsum, respectively, additives primarily affecting the pore size distribution. 5 wt% oil addition reduced material loss upon water spraying by 36 and 15% in low and high T gypsum, respectively. Overall, the laboratory study provided encouraging results, but further long-term testing is required to fully assess the potential of Ca-soap/oil as additives for gypsum plaster, particularly in environments involving salt-laden substrates or freeze-thaw conditions.
The extracellular matrix, particularly collagen, undergoes dynamic remodeling during cutaneous wound healing, yet its spatiotemporal organization in living tissues remains poorly understood. Here we used multispectral photoacoustic imaging to non-invasively visualize collagen dynamics in a murine fibrotic wound healing model. By spectral unmixing of collagen-specific absorption signatures from the hemoglobin background, we resolved collagen content and distribution with a resolution of 7.8 μm and an imaging depth up to 2.4 mm. We showed that untreated wounds exhibit progressive collagen accumulation with disorganized architecture, whereas silicone gel-treated wounds display attenuated deposition and more aligned fiber organization, as validated by quantitative histology. These findings establish photoacoustic imaging as a powerful tool for dissecting extracellular matrix dynamics in vivo, with implications for understanding scar pathogenesis and monitoring therapeutic interventions.
Climatic events, such as the arrival of hurricanes or typhoons, or geomorphological events, like earthquakes, can also necessitate the evacuation of cities or, at the very least, parts of them. Additionally, situations of social alarm caused by terrorist attacks may also require the evacuation of cities or, at the very least, parts of them. In this context, efforts should not only focus on facilitating the relocation of potentially affected individuals. They should focus on carrying out these processes in coordination with authorities in the shortest possible time, while avoiding congestion on evacuation routes as much as possible.To this end, we present SIMERSAD, an agent-based co-model implemented in GAMA that couples a simplified, diffusion-based flood propagation model – omitting hydrostatic pressure, flow velocity, momentum, and energy losses, and therefore best suited to slow-onset riverine inundation rather than high-velocity flood scenarios – with pedestrian evacuation dynamics in the urban core of Santo Domingo, Dominican Republic. The modular architecture of GAMA facilitates adaptation to other cities, though application-specific recalibration of demographic, hydrological, and infrastructure parameters would be required.Simulations reveal that demographic heterogeneity in walking speed increases worst-case clearance time by over 60%. Dynamic congestion with real-time rerouting triples mean evacuation time, while increasing shelter count may reduce median evacuation significantly, but paradoxically worsens worst-case performance due to peripheral placement, as we have observed comparing with the unlimited shelter capacity scenario. Based on these results, three actionable policy recommendations are proposed for emergency managers, including a staged alert protocol and a hierarchical shelter routing strategy. Finally, we also discuss the limitations of our modeling approach.
Vortex-based hydropower systems can recover low-head hydraulic energy while potentially improving water quality through passive aeration. This study numerically evaluates the hydrodynamic and aeration performance of a hyperbolic vortex turbine basin using computational fluid dynamics (CFD), with emphasis on blade number and blade profile. A steady single-phase model was first used to characterize vortex structure, pressure, velocity, and torque, followed by a steady multiphase Volume of Fluid (VOF) model to capture the air–water interface, free-surface deformation, and central air core. Dissolved oxygen (DO) transport in the water phase was modeled using a User-Defined Scalar (UDS) with a UDF-based oxygen-transfer source term and an inlet DO concentration of 2 mg L⁻¹. Straight and curved runners with 4, 5, 6, 7, 8, 10, 12, and 14 blades were evaluated at nominal flow rates of 10 and 20 L s⁻¹. The single-phase and multiphase baseline torque predictions were 0.3592 and 0.3401 N·m, respectively. At 100 rpm, the highest torque was 1.641 N·m for the 10-blade straight runner at 20 L s⁻¹, whereas the highest blade-study DO value was 5.97 mg L⁻¹ for the 6-blade straight runner at 10 L s⁻¹. Under the representative DO condition, the local outlet concentration reached approximately 6.27 mg L⁻¹ and the area-weighted average approached 5.0 mg L⁻¹. The results reveal a clear trade-off between mechanical energy extraction and passive aeration and demonstrate the numerical potential of the hyperbolic vortex basin as a dual-function low-head system. Direct prototype validation remains necessary for quantitative engineering design.
We examined the structural properties of the interview questions in large language model (LLM)-based systems for job interview training, by a job interview through a role-play simulation using a ChatGPT-based chatbot. The 53 trainees were randomly assigned to one of two groups: 24 trainees to the Directive Questions group, and 29 to the Non-Directive Questions group. Trainees underwent a performance review after the interview and then repeated the job interview simulation. The results demonstrate that trainees in the Directive Questions group invested more time answering each question and had longer training time compared to the Non-Directive Questions group. Responses were longer, and the avatar asked fewer questions, in the second session compared to the first session. Our findings support the hypothesis that LLM-based interviews with directive questions make interviewees think more about each question. In addition, the duration of the response and its length may reflect two different cognitive processes.
Explainable artificial intelligence (XAI) is increasingly used in high-stakes applications involving tabular machine-learning systems, yet the reliability and practical utility of explanations depend not only on explanation methods but also on the quality, structure, and preprocessing of the underlying data. This systematic literature review examines the relationships among data conditions, explanation reliability, diagnostic use, corrective intervention, and methodological advances in XAI for tabular data. The search covered January 2015 to July 2026 and identified 27 eligible primary empirical studies published between 2020 and 2026, which were synthesized through a structured narrative approach addressing four research questions. Evidence indicates that explanation reliability, particularly stability and fidelity, can be affected by noise, class imbalance, feature correlation, distribution shift, sampling and neighborhood-generation strategies, and preprocessing choices, whereas practically important conditions such as missing values remain under-investigated. XAI can also support the identification and localization of prediction errors, biases, distributional changes, and other model failures. However, a substantial gap remains between diagnosis and corrective intervention, with relatively few studies translating explanatory insights into corrective actions followed by systematic re-evaluation. Recent methodological advances emphasize structured evaluation frameworks, reliability metrics, aggregation strategies, intrinsically interpretable alternatives, and integration of XAI into monitoring and intervention-oriented workflows. Overall, the findings support a data-centric, lifecycle-oriented perspective in which data conditions, explanation reliability, diagnosis, intervention, and re-evaluation are interconnected. Future research should prioritize standardized evaluation protocols, under-studied data-quality conditions, XAI-aware preprocessing pipelines, and empirically validated closed-loop frameworks for defensible corrective intervention.
Serial tendon routing couples shape, tendon loading, and motor stroke in multisection tendon-driven continuum robots. When distal tendons pass through a proximal section before becoming active distally, one physical route governs tendon length, generalized tendon loading, and motor-side stroke. A geometry-first inverse solution can therefore reach a target while remaining statically inconsistent or actuator-side infeasible. A route-aware actuator executability analysis is formulated for a two-section, six-tendon elastic rod. The contribution is that serial pass-through routing makes one route field enter the length, load, and stroke maps together, so that freezing the backbone shape before force recovery imposes a nonzero static-residual floor. The inverse problem solves reduced Cosserat backbone coordinates, nonnegative tendon-tension increments, and a common-within-target retraction variable simultaneously under tip-position, static-balance, and stroke-feasibility constraints. A fixed-shape consistency analysis explains why sequential inverse pipelines cannot remove force-closure errors after the backbone shape is frozen. In circular radius-sweep benchmarks, the joint formulation reduces the average static residual from 0.356 to 0.100 and the average minimum bias requirement, represented by the lower endpoint BT−, from 22.9 mm to 1.14 mm compared with traditional sequential inversion, while fixed-bias feasibility increases from 1/4 to 4/4 trajectories and the root-mean-square axial tip-position error remains below 0.03 mm. A routing-layout ablation shows that the nominal inner pass-through reduces the average fixed-bias demand from 1.44 mm for an outer-radius pass-through to 1.14 mm, while the route scan retains a localized, non-monotone accepted-solution corridor. Cross-task trajectories, non-ideal residual tests, and external multibody replay provide complementary robustness and model-form checks.
Accurate reconstruction of the complete air-blast pressure history generated by a moving explosive charge is challenging because the resulting response varies jointly with propagation distance, observation direction, and the translational velocity of the charge immediately before detonation. Across these operating conditions, the pressure waveform exhibits condition-dependent first-arrival timing, a localized positive-pressure peak, and a longer post-peak attenuation stage. The prediction problem considered here is therefore to reconstruct the continuous conditional pressure field p(t,r,θ,v) from operating coordinates alone, where v denotes the pre-detonation charge velocity, without using pressure observations from the target condition. To address this problem, a structure-aware multiscale Fourier physics-informed neural network (SPF-PINN) is developed. Within a single shared pressure field, multiscale Gaussian Fourier features provide the representation basis, while a localized first-arrival propagation prior and region-specific front, peak, and attenuation constraints regulate complementary spatial and temporal characteristics. It is evaluated on 420 numerically generated pressure histories and six independent experimental signals. On the numerical test set, SPF-PINN achieves a root mean square error (RMSE) of 22.48 kPa, relative RMSE of 0.60%, peak pressure error of 23.78 kPa, relative peak pressure error of 0.63%, and rise time error of 4.45 μs. For the six experimental cases, the mean RMSE, relative RMSE, and rise time error are 1.19 kPa, 6.07%, and 4.50 μs, respectively. These results show that combining multiscale coordinate representation with localized physical regularization improves full-waveform reconstruction and supports transfer beyond the numerical propagation-distance range, while broader experiments are still required to establish general extrapolation limits.
FPGA transformer accelerators routinely fix their multiply-accumulate (MAC) accumulators at 32 bits, echoing a theoretical 25–27-bit worst case for INT8 BERT-base matrix multiplication that has never been checked against actual per-layer behavior. We provide a transformer-specific layer-level study, profiling 215 linear layers across BERT-base, TinyBERT-4L, DistilBERT-base, ALBERT-base-v2, and ELECTRA-small-discriminator over WikiText-2 samples via PyTorch forward-hook INT8 simulation that tracks the true partial-sum accumulation trajectory (not only the final dot-product value) with a 2-bit margin. Across these 215 profiled layers, the maximum observed accumulator requirement is 20 bits under the specified profiling conditions, and 500 SST-2 samples show zero prediction changes, with δ=2 bits the best-performing margin among those tested. Synthesized on the ZynqTM UltraScale+TM MPSoC ZCU104, a 16-bit MAC cuts flip-flops 39% and LUTs 17% versus a 32-bit baseline while exceeding 400 MHz; a per-layer-instance resource lower bound for a 73-layer BERT-base array projects 34.5% fewer flip-flops and 15.4% fewer LUTs.
Amid rapid urbanization, issues such as traffic congestion, land scarcity, and environmental pollution have made metro systems central to urban public transportation due to their efficiency, capacity, and low emissions. This study involves the development of a spatiotemporal accessibility framework to diagnose supply–demand matching at metro stations, supporting the shift from construction-led expansion to precision governance. Using Nanjing as a case study, it integrates metro network, population heat, and road network data for 217 stations. The Gaussian-based two-step floating catchment area method is employed to measure spatiotemporal accessibility under walking and bike-sharing access, and K-means clustering is performed to classify stations by supply, demand, and accessibility. Results show that concentrated population activity creates a dilution effect. Temporally, accessibility exhibits pronounced diurnal variation, with high-accessibility coverage contracting substantially during commute peaks relative to the early morning maximum; for example, the workday walking mode showed a 72% contraction at the evening peak at 18:00 relative to 6:00. Spatially, this dilution effect is most pronounced in the urban core, where high topological supply coexists with low accessibility. Stations were classified into five types—Balanced and Coordinated (43.8%), Demand-Pressured (18.0%), Low-Competition (17.0%), Potential-Driven (14.7%), and System-Constrained (6.5%). The 2,000-m bike-sharing threshold substantially expanded the high-accessibility coverage relative to the 1,200-m walking threshold across all time periods. Methodologically, the study achieves a temporally integrated supply–demand assessment; practically, it provides a scientific basis for differentiated station strategies, supporting the transition toward quality- and efficiency-oriented metro governance.
Structural characterization of ship hulls becomes challenging when localized changes in mass distribution produce subtle modifications in vibration response that may not be adequately captured by conventional modal-parameter-based assessment. This study develops and experimentally evaluates an integrated RDT–machine learning methodology for characterizing localized mass-induced dynamic perturbations in a scaled stiffened ship-hull model. The methodology combines finite element (FE) modal analysis and experimental modal analysis (EMA) with random decrement technique (RDT)-based dynamic signature extraction and artificial neural network (ANN) modeling. The FE model provides a numerical reference for experimental design, while measured vibration responses are processed using RDT to obtain physically grounded free-decay-type dynamic features. Six perturbation scenarios were investigated by adding 100, 200, and 500 g masses at two distinct hull locations. The RDT-derived displacement and velocity were used as ANN inputs to estimate the perturbation-sensitive dynamic indicator N(μ,μ˙). The ANN achieved testing R2 values of 0.841–0.902 across the six scenarios. Sensitivity analysis identified a normalized RDT trigger level of Ltrig=0.50 as the most suitable common condition for the investigated dataset. Synthetic-noise assessment showed limited performance degradation at moderate noise levels and progressively greater degradation at lower signal-to-noise ratios. The resulting dynamic-indicator signatures exhibited sensitivity to perturbation magnitude and attachment location. These findings demonstrate the feasibility of a physics-informed learning approach based on experimentally derived RDT signatures for vibration-based characterization of localized mass perturbations under controlled laboratory conditions.
Flextensional transducers (FTs) are highly efficient devices for converting mechanical vibration into electrical power, offering compactness, scalability, and superior mechanical amplification compared to traditional bulk piezoelectric elements. Among them, the class V cymbal transducers are widely employed for vibration energy harvesting due to their simple geometry and broadband response. However, they suffer from high stress concentrations, limited load-bearing capacity, and reduced durability under dynamic excitation, constraining the long-term performance in practical energy harvesting systems.This study explores wagon wheel-shaped end cap architectures in class V FTs to achieve dynamic optimisation. By varying the number of spokes (4, 6, and 8), these designs enhance flexural compliance, aiming to improve vibration amplification, mechanical stability, and energy conversion efficiency. Finite element simulations and experimental measurements were conducted to characterise the dynamic responses, stress distributions, and electromechanical behaviours of the FTs, including additional tests with attached metal masses to evaluate the load-bearing capability.Results show that, compared with the conventional solid cymbal transducer, wagon wheel FTs reduce resonant frequency by up to 28% and enhance peak displacement amplitude by three times at high excitations. Among all configurations, the 6-spoke wagon wheel FT exhibits the best overall performance, achieving a maximum displacement amplitude of approximately 17.5 µm without a rod and nearly 20 µm with a rod, more than three times that of solid cymbal FT. This improvement is attributed to its balanced combination of moderate flexural compliance, high vibration amplification, and reduced stress concentration, demonstrating a strong potential for robust and efficient vibration energy harvesting applications.
Chronic diseases such as diabetes, hypertension, chronic kidney disease (CKD), and heart disease impose a major global health burden. Machine-learning prediction systems offer valuable early-diagnostic support, but their performance is often degraded by outliers and class imbalance in real-world clinical tabular data.The standard remedy, DBSCAN, requires manual ε-tuning and assumes globally uniform density — assumptions that rarely hold on heterogeneous medical data. The recent AutoSCAN method removes the manual ε but has been validated only on synthetic clustering benchmarks, not on downstream medical classification.We propose YadroSeg, an extension of the graph-based coring framework of Le et al. [1], originally introduced for image segmentation, and adapted here to tabular medical data. YadroSeg builds a k-nearest-neighbour weighted graph and analyses a heap-based density variation sequence to identify noisy samples in an adaptive, parameter-free manner, using a novel geometric elbow criterion for automatic threshold selection. The cleaned data are then balanced with SMOTE-ENN and classified by a Random Forest.On five public chronic-disease datasets — diabetes, hypertension, CKD, Cleveland and Statlog heart disease — under 10-fold cross-validation, YadroSeg attains the best F1 on three datasets, reaching 98.87%. On the remaining two, YadroSeg-Grid and AutoSCAN are best, the latter benefiting from full data retention combined with SMOTE-ENN.These results confirm that graph-theoretic density analysis provides a mathematically principled and practically effective approach to noise filtering in imbalanced clinical tabular data.
Engineered public school infrastructure in emerging economies is often delivered through standardized, cost-driven designs, while the life-cycle consequences of material choices remain poorly quantified. This study applies a BIM-assisted whole-building life-cycle assessment (WBLCA) to a 3,982 m², four-story engineered public school in Nepal and evaluates three alternatives for its non-structural envelope and interior partitions. The as-built reinforced-concrete frame with fired-brick masonry was compared with rammed earth, cross-laminated timber (CLT), and concrete-block alternatives while maintaining the same structural frame, foundations, floor systems, service life, and operational-energy assumptions. BIM-derived material quantities were verified and linked to local, regional, and proxy life-cycle inventory datasets within a cradle-to-grave assessment covering eight environmental impact categories. Material production (A1–A3) emerged as the dominant environmental hotspot across most impact categories. The baseline building generated an A1–A3 GWP of 952 kg CO₂e/m² and a total carbon GWP of 1,209 kg CO₂e/m². Rammed earth provided the greatest reduction, followed by CLT, while concrete block produced comparatively modest improvements. CLT additionally provided temporary biogenic carbon storage, reported separately from GWP. Sensitivity analysis showed that the comparative ranking remained stable across variations in CLT datasets, transportation, grid emission factors, study period, and end-of-life assumptions. The findings demonstrate that targeted substitution of envelope and partition materials can reduce life-cycle impacts without complete structural redesign and provide preliminary case-study reference values to support carbon-informed material selection and procurement for public buildings in data-limited regions.
In this research, Elman least mean square (ELMS) based inductively coupled distributed static compensator (IC-DSTATCOM) is studied. The implementation procedure is presented in two sections. First, the source, load and filter current are determined by utilizing the inductive filtering transformer (IFT) on the basis of impedance matching principle, second, the six ELMS subnet structures are employed for both direct and quadrature components of three phase to improve the dynamic response. The comparative performances between IC-DSTATCOM and direct coupled distributed static compensator (DC-DSTATCOM) are demonstrated at various dynamic scenarios such as source current harmonic reduction, power factor (PF) improvement, voltage regulation, voltage balancing as well as controlled DC link voltage. A 10 kVA, 230 V, 50 Hz prototype IC-DSTATCOM and DC-DSTATCOM are built to verify the experimental results as per the prescribed guidelines.
This study investigates the multiscale degradation of a graphite-supported CVD-SiC upper electrode protection ring after 1000 h of service in an 8-inch production plasma etcher. XRD and Raman results indicate that the residual β-SiC coating retains its main crystallographic structure, whereas the surface exhibits a pronounced radial degradation gradient consistent with coupled physical sputtering and fluorine-assisted chemical erosion. In the mildly eroded regions, shallow craters and an increase in the mean Sa to 11.16 ± 1.15 μm suggest that physical ion sputtering is the main contributor. With increasing damage severity, these surface defects may act as preferential sites for fluorine-containing reactive species. EDS analyses reveal localized Si depletion, C enrichment, and F accumulation within deep pits. The Si content decreases to 11.9 at.%, which is consistent with preferential Si removal and the possible formation of volatile SiF4 species. These observations support a positive-feedback degradation mechanism that may promote pit growth and transform the coating into a highly porous structure with a mean Sa value of 208.2 ± 11.6 μm, approximately 460 times that of the pristine surface. The resulting fragile surface features may perturb the local plasma environment and increase the risk of particle release. These findings provide guidance for optimizing component geometry and developing erosion-resistant coatings.
The present work presents the results of investigations into the composition of lithium-bearing rare-metal tailings and the development of methods for their processing. The treatment of rare-metal tailings for lithium recovery, considering that lithium is one of the most strategically important and highly demanded materials, represents a relevant and high-priority research area. The composition of the tailings was investigated using a range of physicochemical characterization techniques. The major mineral phases were identified as albite (Na[AlSi3O8]), muscovite (KAl2[AlSi3O10](OH)2), quartz (SiO2), microcline (K[AlSi3O8]), and biotite (K(Mg,Fe2+)3(Si3Al)О10(OH,F)2). Lithium is predominantly present as an impurity and in isomorphically substituted forms within the crystal structures of muscovite and biotite. Among the lithium-bearing minerals, spodumene (LiAl[Si2O6]) and lepidolite (KLi2Al[(AlSi)4O10](F,OH)2) were identified. Two sulfuric acid processing routes were investigated: with and without preliminary roasting of the tailings. The experimental results demonstrated that direct sulfuric acid treatment without prior roasting is the more effective approach. Direct sulfation of the tailings using 93 wt.% H2SO4 at 300°C, with a solid-to-liquid ratio of 6-10:1 and a reaction time of 1-2 h, followed by water leaching at 90-98°C with a solid-to-liquid ratio of 1:6-10 for 1-2 h, enabled lithium extraction into the aqueous solution of approximately 72-98%. Lithium was subsequently recovered from the obtained lithium-bearing leach solution by chemisorption onto freshly precipitated aluminum hydroxide. The sorbent was synthesized using the aluminum already present in the solution as the major impurity, with the addition of ammonium hydroxide, which enabled the recovery of 99.78% of lithium from the solution into the sorbent phase. The distribution and separation coefficients for the chemisorption process were calculated for lithium, magnesium, and aluminum. The distribution coefficient for lithium was Kd(Li) = 31,840, indicating a high degree of lithium uptake by the freshly precipitated sorbent and high selectivity. The separation coefficients of lithium from magnesium and calcium were Ks(Li/Mg) = 190–439 and Ks(Li/Ca) = 56–135, respectively, demonstrating the potential for effective separation of lithium from most of the accompanying impurities.
Enhancing the fatigue performance of M50NiL steel is critical for its application in aircraft engine bearings and gears, and a recent microstructure-property framework links fatigue behavior to the constituent phases, namely martensitic matrix, δ ferrite, retained austenite, and carbides. Tailoring the quantity, size, and distribution of these phases effectively relieves interfacial stress concentration and suppresses crack propagation, especially in post-carburization materials where increasing carbide count and ensuring uniform dispersion become essential. To achieve this optimization, chemical heat treatments including carburizing, nitriding, and carbonitriding along with surface modification techniques such as shot peening and ultrasonic rolling are applied, which generate gradient structures that balance high surface hardness with enhanced core toughness while also introducing compressive residual stresses. These synergistic measures substantially improve fatigue resistance and extend the service limits under extreme conditions. Nevertheless, current fatigue life prediction still relies on conventional mechanism-driven models, and a gradual shift toward a modern multi-scale framework that incorporates machine learning algorithms and mechanistic understanding is urgently needed to enable quantitative life forecasts under complex service environments.
Perovskites metal halide materials such as halide-modified methylammonium lead perovskites, MAPbX3 (X = Cl, Br, I) represents the most promising semiconductor materials for next-generation optoelectronics and photovoltaics. The present work is a comprehensive study based on the Vienna ab initio simulation package (VASP) with the generalized gradient approximation (GGA-PBE) of their structural, optical and electronic properties from first principles. Structural optimization shows that both orthorhombic lattice parameters and orthorhombic unit cell volumes decrease with the decrease in the halide ionic radius (from iodide to chloride).All three compounds have the optimum direct band gaps at the Γ point electronically. Through halide substitution, precise bandgap tuning is achieved, with the MAPbI3, MAPbBr3, and MAPbCl3 having band gaps of 1.6 eV, 2.04 eV and 2.4 eV, respectively. These calculations are in very good agreement with the experimental data, using a variety of parameters such as density of states, Tauc analysis and absorption edges. PV evaluations have been carried out in detail and have confirmed that MAPbI3 is highly promising for solar applications. It has the highest electronic polarizability, the best visible light absorption and the highest short-circuit current density (Jsc), and finally the highest theoretical photovoltaic efficiency (PCE). On the other hand, MAPbCl3 has a large open-circuit voltage (Voc), but a wider band-gap that limits its solar efficiency, thus opening up a new approach for UV optoelectronic applications. Under the idealized assumptions of the SLME model, MAPbI3 exhibits the most favorable theoretical photovoltaic performance among the investigated compounds, with a calculated SLME of approximately 31%.
This paper presents a comprehensive experimental, numerical and theoretical study on the mechanical behavior of CRB600H high-strength stirrup reinforced concrete columns under eccentric compression. Six eccentrically compressed specimens were designed and tested, with stirrup grade, viz. CRB600H and HRB400, and substitution method, viz. equal-volume and equal-strength, as key variables, to explore their failure modes and mechanical properties under both small and large-eccentricity compression. A finite element model was developed using ABAQUS software and verified by comparing test and simulation results of load-deflection curves and load-carrying capacities. Parametric analyses were carried out to investigate the effects of concrete compressive strength, eccentricity, longitudinal reinforcement ratio, stirrup ratio and longitudinal reinforcement strength. The results indicate that CRB600H stirrups effectively improve the ductility of columns under equal-volume substitution; increasing longitudinal and stirrup ratios enhances both load-carrying capacity and deformation performance. Finally, the applicability of GB 50010-2010 and ACI 318 formulas was evaluated, and a modified method based on the Mander confinement model was proposed for small-eccentricity compression, providing more reliable predictions for engineering design.
