Browse Articles

Machine learning models based wear performance prediction of AZ31/TiC composites

Abstract This study presents the fabrication of AZ31 magnesium matrix composites reinforced with 5, 10 and 15 vol% TiC particles using the Friction Stir Processing (FSP) technique and evaluates their wear behavior under varying loads (10–50 N) and sliding speeds (75–225 mm/s). The incorporation of TiC significantly enhanced the microstructural and mechanical properties of the composites. In particular, the AZ31/15 vol% TiC composite exhibited a refined grain structure with an average grain size of ~ 8 μm, compared to ~ 60 μm in the unreinforced AZ31 alloy. The same composite also demonstrated a substantial increase in hardness from 62 HV (base alloy) to 116 HV, highlighting the effectiveness of TiC reinforcement in improving strength. A key innovation of this work is the application of five machine learning (ML) algorithms, trained on experimental data using input features such as load, sliding speed and reinforcement content, to model and predict wear performance. After rigorous hyperparameter optimization, the Gradient boost algorithm achieved the highest predictive accuracy (R² = 0.9987), with errors falling within the range of experimental uncertainty. The study further includes residual analysis and computational efficiency assessment, supporting interpretable and robust AI-driven modeling. This integrated experimental-ML approach establishes a new benchmark for predictive modeling and data-driven material design in magnesium-based metal matrix composites.

Digital twin-based machine learning framework for predicting nonlinear seismic response of reinforced concrete shear walls using analytical data

Effects of environmental factors on arsenic accumulation in rice under field conditions using a Bayesian state space model

Development of ascorbyl palmitate based hydrophobic gold nanoparticles as a nanocarrier system for gemcitabine delivery

Research on ambidextrous digital innovation strategies of SMEs embedded in industrial internet platforms based on evolutionary game theory

In vitro evaluation of the anthelmintic activity of citrus flavonoids against free-living and parasitic nematodes

Dual optical frequency comb using a multi-functional PIC based on fully integrated injection locked gain-switched lasers

Reviving expired pharmaceuticals using 2-[(butylamino)propan-2-yl]phosphinic acid for corrosion protection of carbon steel in 1.0 M HCl

Urbanization accelerates soil degradation in peri-urban compared to rural farms

Parametric analysis of electromagnetic wave interactions with layered biological tissues for varying frequency, polarization, and fat thickness

Abstract Electromagnetic wave interaction with biological tissue is frequency-, angle-, and polarization-dependent, influencing both dosimetric parameters and resultant thermal effects. This work presents a comprehensive analysis across the major ISM bands (433, 915, 2450, and 5800 MHz) for transverse electric (TE) and transverse magnetic (TM) polarizations incident on a three-layer tissue model (skin–fat–muscle). A custom MATLAB code was developed to integrate the multilayer transmission line formalism, polarization-specific wave impedance modeling, Cole–Cole dielectric parameterization, and a finite difference method (FDM) solution of the Pennes bioheat equation. Simulations were performed for incident power density 50 W/ $$\hbox {m}^2$$ and fat thicknesses from 0.005 m to 0.03 m, over incidence angles 0 °C– 80 °C. Throughout the manuscript, reflection is reported strictly as a power quantity $$R=|\Gamma |^2$$ rather than a field-amplitude coefficient. The thermal pipeline solves the steady-state Pennes equation in its direct $$\Delta T$$ form with consistent surface (Robin) and deep (Dirichlet) boundary conditions, and simulations are audited by an energy-conservation budget. Results indicate that while temperature increases remain below 0.4 °C at lower frequencies (433–915 MHz), significant superficial heating (up to 3.5 °C) occurs at 5.8 GHz due to reduced penetration depth, even at moderate exposure levels. The results demonstrate that subcutaneous fat acts as a low-loss impedance transformer whose thickness strongly modulates the balance between reflection and internal absorption, while polarization and angle primarily tune the detailed shape of angular reflection curves (including TM Brewster-like minima) at a given incident power. The analytical framework therefore complements voxel-based full-wave numerical models by providing fast, physically transparent trends across ISM bands that are directly relevant for preliminary assessment of wearable devices, implanted sensors, and compliance with radiofrequency safety limits.

Diagnosing the innovation atmosphere of industrial parks through urban spatial perception: a multimodal large language model approach

Ethnobotanical study of medicinal plants in Meketewa District, northwestern Ethiopia

Investigating the mechanisms of lactulose on gut health and uric acid metabolism in geese via transcriptomics

Integration of human ERKs and DUSPs into the yeast cell wall integrity pathway

Structural inequalities and dietary diversity in Odisha: evidence from NSSO 68th and 79th rounds

Effect of incorporating bone char with sulfur or humic acid on phosphorus availability and spinach growth in calcareous sandy soil

Abstract This study investigated the effects of applying modified bone char by sulfur (MBC) with humic acid and co-applying bone char (BC) with sulfur (S) or humic acid (HA) on chemical properties, phosphorus (P) availability, and spinach growth in calcareous sandy soil. This pot experiment has twelve treatments: Control (CK), bone + S (BS), bone + HA (BHA), BC + S (BCS), BC + HA (BCHA), MBC, MBC + HA (MBCHA), acidified BC with 0.1 N H 2 SO 4 (0.1ABC), acidified BC with 1 N H 2 SO 4 (1ABC), rock phosphate (RP), RP + S (RPS), and RP + HA (RPHA). The B, BC, MBC, 0.1ABC, 1ABC, and RP were added at 300 mg P kg − 1 soil doses. Spinach was grown in this experiment. Applying all treatments significantly increased soil phosphorus availability. Available phosphorus increased from 11.61 mg kg − 1 (CK) to 19.70, 19.76, 21.82, 22.25, 22.45, 26.09, 19.58, 21.01, 15.26, 18.95, and 17.77 mg kg − 1 for BS, BHA, BCS, BCHA, MBC, MBCHA, 0.1ABC, 1ABC, RP, RPS, and RPHA, respectively. The effectiveness of the treatments in this study on the available phosphorus improvement was in the order of MBCHA > MBC > BCHA > BCS > 1ABC > BHA > BS > 0.1ABC > RPS > RPHA > RP > control. Compared to the control treatment, applying BHA, BCS, BCHA, MBC, MBCHA, 1ABC, RPS, and RPHA to the soil significantly increased the fresh shoot of the spinach plant. Fresh shoot of spinach increased from 46.02 g pot − 1 for CK to 54.41, 54.36, 56.94, 50.39, 51.91, 48.83, 54.24, and 49.52 g pot − 1 for BHA, BCS, BCHA, MBC, MBCHA, 1ABC, RPS, and RPHA, respectively. The effectiveness of treatments in improving the fresh weight of spinach was in the order of BCHA > BHA ≈ BCS > RPS > MBCHA > MBC > RPHA > 1ABC > control > RP > BS > 0.1ABC. Our results concluded that co-applying bone char with sulfur is optimal for enhancing soil quality indicators and improving fresh and dry shoots of spinach. Due to its cheaper price, it is preferable to add sulfur with bone char rather than humic acid.

Evaluation of arsenic-Tolerant plant growth-promoting rhizobacteria from Manipur for mitigating arsenic translocation and enhancing growth in rice (Oryza sativa)

MC-LBTO: secure and resilient state-aware multi-controller framework with adaptive load balancing for SD-IoT performance optimization

HIV case management using agent-based modeling approach subject to antiretroviral therapy and lifestyle treatment plan

High-fidelity 3D mesh generation from a single sketch using shape constraints

Abstract The research on 3D model reconstruction from a single image using deep learning technology has achieved remarkable progress. However, compared with images, sketches lack sufficient visual information, which challenges the reconstruction algorithm’s ability to correctly interpret sketches. Herein, we introduce a streamlined network architecture for sketch-to-3D mesh generation, designed to address the challenge of reconstructing high-fidelity 3D models from single-hand sketches. Our approach deploys the expressive PowerMLP architecture within an encoder-decoder framework, surpassing traditional MLP implementations in representation capability. By integrating 3D shape constraints instead of relying on conventional discriminators, we achieve geometric fidelity in a collaborative generation process. Experimental results demonstrate state-of-the-art (SOTA) performance on both synthetic stylized sketches and real-world handwritten inputs, validating the method’s robustness and adaptability.