Marathwada Institute of Technology

Biodiesel’s application in compression–ignition engines is mostly limited by the type of methyl esters it contains rather than the total amount of feedstocks. In order to modify the fatty acid methyl ester (FAME) profile for better combustion and emissions, cottonseed (CSOME), neem (NOME), and orange peel oil methyl esters (OPOMEs) were carefully mixed. Fuel chemistry was examined using Gas Chromatography–Mass Spectrometry (GC-MS) and Fourier Transform Infrared (FTIR), which confirmed variations in oxygenated functional groups, saturation levels, and volatility. In a single-cylinder CI engine, diesel, single, binary, and ternary biodiesel mixes were tested over 25–100% load at compression ratios of 17 and 18, both with and without 10% EGR. The ester-optimized ternary blend HBO70 delivered the best overall performance at CR 18 with EGR, exhibiting only a 0.61% reduction in BTE while achieving significant reductions in smoke (44%), PM (51%), NOx (30%), HC (11%), CO (10%), and specific fuel consumption (SFC) (6.8%). Regression analysis confirmed a temperature- and oxygen-controlled NOx–PM trade-off, demonstrating that ester-profile optimization is an excellent way to achieve cleaner and more efficient CI engine operation.

Understanding wire-cut electrical discharge machining (WEDM) parameters’ impact on surface roughness (Ra) is crucial for optimizing processes. This study uses artificial neural network (ANN) techniques to estimate the surface roughness of Al/SiC composites during WEDM, examining how process parameters affect the roughness. The experiment used a stir casting aluminum alloy with a 7.5% silicon carbide metal matrix composite (MMC), adjusting parameters like the wire tension (WT), servo voltage (SV), peak current (IP), pulse on time (TON), and pulse off time (TOFF). An ANN model was created to forecast the surface roughness. The study developed an ANN model to forecast surface roughness in Al/SiC composites during WEDM, demonstrating its accuracy in identifying the link between surface finish and input parameters, thereby improving the surface quality. The ANN model accurately predicted the surface roughness based on WEDM parameters, with strong correlations between predictions and actual data, demonstrating its ability to estimate surface quality accurately.

This paper presents a novel approach for historic facial image restoration using Super-Resolution Generative Adversarial Networks (SRGAN). By leveraging adversarial training, SRGAN effectively enhances the resolution and quality of degraded historic facial images, recovering fine details and textures while preserving the authenticity of facial features. The dataset comprises high-resolution and artificially degraded images that mimic the characteristics of aged photographs, enabling the model to learn realistic mappings between low- and high-resolution images. The SRGAN architecture incorporates a generator with residual blocks and a discriminator optimized for adversarial learning, achieving superior visual quality and facial detail restoration. Quantitative evaluations demonstrate significant improvements, with Peak Signal-to-Noise Ratio (PSNR) values ranging from 17.65 dB to 24.18 dB and controlled reconstruction errors (Mean Squared Error: 248.24-1808.87, Mean Absolute Error: 13.01-31.40). The model preserves critical attributes, maintaining brightness and contrast changes within acceptable thresholds (e.g., brightness 23.12, contrast 18.45) while enhancing visual clarity. These results highlight the SRGAN model’s effectiveness in restoring and preserving historic portraits and facial imagery, offering a scalable solution for applications in cultural heritage conservation and digital archiving.

This study examines shifts in shopping activity patterns among grocery and non-grocery purchasers in Nagpur, India, by comparing pre- and post-COVID-19 scenarios. Using descriptive statistics and Structural Equation Modeling (SEM) on 1,546 household survey responses, the study explores the interdependencies between virtual (internet-based) and physical (in-store) purchasing behaviors while accounting for socio-demographics, internet experience, accessibility, and attitudes. The findings reveal a significant increase in virtual grocery and non-grocery purchasing, coupled with a decline in physical store visits for both categories. SEM results indicate strong interconnections between purchasing categories, demonstrating that changes in virtual or physical shopping behavior in one category influence the other. Furthermore, positive attitudes toward e-commerce significantly drive virtual shopping adoption while reducing physical shopping. Factors such as internet experience, household size, and income levels also shape purchasing behavior, with lower- and middle-income groups exhibiting a stronger shift toward virtual grocery shopping. Despite extensive research on virtual shopping adoption, few studies have examined the interdependencies between grocery and non-grocery purchasing behaviors in a post-pandemic context, particularly in heterogeneous urban settings like Indian cities. By addressing this gap, this study offers valuable insights for urban planners, retailers, and policymakers to refine travel demand forecasting models, enhance e-commerce infrastructure, bridge the digital divide, and promote sustainable transportation policies.