Coimbatore Institute of Engineering and Technology

This study evaluates the high-temperature oxidation and corrosion resistance of Inconel 617 (IN617) fabricated using the Wire Arc Additive Manufacturing (WAAM) process, with a focus on its performance under hot air and molten salt environments at 700 degrees C. The as-built microstructure exhibited columnar dendritic grains, with chromium- and molybdenum-rich interdendritic precipitates. Electron Backscatter Diffraction analysis revealed a strong < 001 > fiber texture with localized strain. WAAM-processed IN617 exhibited higher weight gain in molten salt (15.34 mg/cm2) compared to air (1.72 mg/cm2), attributed to salt-induced oxide growth at high temperatures. Oxidation in hot air formed a protective Cr2O3 and NiCr2O4 spinel phase. In contrast, exposure to a Na2SO4-60%V2O5 molten salt produced a porous, brittle oxide layer containing Ni3V2O8, Cr-V-O compounds, and sulfates, which caused severe scale breakdown, as confirmed by X-ray Photoelectron Spectroscopy (XPS). Parabolic rate constants indicated faster corrosion in molten salt (Kp = 5.16 x 10- 10 g2.cm- 4.s- 1) than in air (Kp = 1.26 x 10- 11 g2.cm- 4.s- 1). WAAM IN617 exhibits good oxidation resistance in air but is severely degraded in environments containing sodium, vanadium, and sulfur, particularly at high temperatures.

Health determines life quality and economic productivity. Internet of Medical Things (IoMT) based monitoring using fog computing with local servers and computers offers an efficient solution for real time healthcare management. Traditional healthcare systems often face challenges such as delayed diagnoses, inefficient monitoring and limited access to real-time data especially in remote areas. These issues hinder timely medical intervention and reduce overall healthcare efficiency. Therefore, a Hybrid Optimized Dynamic Graph Convolutional Recurrent Imputation Network for Fog Computing in Health Monitoring Using Internet of Medical Things (HYB-DGCRIN-FCHM-IoMT) is proposed in this paper. Initially, the proposed method is validated on both the Sleep-EDF-2018 and MIT-BIH Polysomnographic datasets to assess its robustness across diverse input sources. Then the input signals are pre-processed utilizing Maximum Correntropy Quaternion Kalman filter (MCQKF) for enhancing signal clarity and reducing noise. The pre-processed signal is given into the Dynamic Graph Convolutional Recurrent Imputation Network (DGCRIN) that accurately monitor and classifies the Sleep Apnea. The Hybrid Bitterling Fish Optimization Algorithm and Bitterling Fish Optimization Algorithm (HBFOA-PEOA) is employed to enhance DGCRIN, which effectively classify input signals by improving accuracy and reducing the computational time. Finally, the categorized signal is stored in fog nodes using Precise Elliptical Curve Cryptography (PECC) technique. The performance of the proposed HYB-DGCRIN-FCHM-IoMT approach achieves 20.28 %, 28.22 % and 29.27 % higher accuracy compared with existing FCA-IOMT-HM, EEC-IoT-FCand FCSEED-IoHTmethods.

The growing carbon footprint of the construction industry, which is largely due to Ordinary Portland Cement production, demands for the development of new sustainable, high-performance alternatives. This research assesses a ternary geopolymer concretesystem using ground granulated blast furnace slag (GGBS), fly ash, and sugarcane bagasse ash (SCBA). The studies examined how different alkaline activator molarities (8 M, 10 M, and 12 M) and different SCBA replacement percentages (0–20

This research introduces Time-Series AI Model to explain Macroeconomic Policy Outcomes based on Hybrid Deep Learning with Causal Inference which combines long short-term memory networks (LSTM) with Granger Causality testing. The model is formulated in a manner that enhances accuracy and interpretability of the forecast of key indicators of the macroeconomic conditions, including GDP growth, inflation, and unemployment in response to changes in the policy. The model development is done on MATLAB platform using its powerful tools in deep learning and econometrics. The findings show that the hybrid model attains a better prediction accuracy (89) than the conventional econometric models (ARIMA) and machine learning models (SVM, Random Forest). Also, the use of causal analysis enables the model to determine and measure the effect of policy decisions, which is important information to policy makers. This proposed methodology is better than the current methods in predictive strength and in computer efficiency and hence provides a powerful macroeconomic forecasting and decision-making tool.

Implementation of Cloud based Recruitment System in Human Resources (HR) will increase the efficiency of the organization by simplifying the recruitment process. A cloud-based Applicant Tracking System (ATS) allows the HR teams to automate the most important areas, including the screening of resumes, communication with the candidates, and the scheduling of the interviews. Such automation does not only minimize the human resources but it also speeds up the recruitment process enabling qualified candidates to be placed more quickly. The cloud characteristics of the system enables real-time cooperation between the HR managers and the hiring teams regardless of the location, and this fosters easy communication and decision-making. In addition, analytics tools in the ATS provide significant information on the performance of the candidates, the tendencies of the recruitment process, and the efficiency of the different methods of hiring. Consequently, there will be more informed hiring choices, a decrease in the costs of hiring, and improved candidates experience. Cloud-based recruitment system is vital to the contemporary HR departments, and it enhances the scalability, flexibility, and general effectiveness of recruitment.