Punjab Engineering College

Reinforced concrete (RC) frame-shear wall building systems are widely adopted in India due to their high strength and stiffness, making them effective in resisting lateral seismic forces. However, their performance during extreme seismic events has revealed limitations, particularly in terms of structural resilience. Foundation rocking is a novel approach for enhancing the seismic resilience and has caught the eye of several researchers. However, allowing the foundation to rock will result in the increase of displacement caused by rotations at the base of the foundation which may cause permanent settlements depending on the underlying soil properties. This study investigates the potential of foundation rocking as an innovative strategy to enhance the seismic resilience of high-rise RC frame-shear wall buildings. Four varying degrees of foundation rocking that can be allowed without compromising the integrity of the building system were analyzed to ensure structural integrity while improving performance. The seismic behavior was assessed through key parameters, including maximum superstructure displacement, peak moment at the foundation, settlement of underlying soil, self-centering ratio, and tipping over ratio. Results demonstrate that the displacements induced by rocking remain within the permissible limits outlined in Indian standards. Additionally, rocking foundations effectively reduce peak moments during seismic events, thereby lowering the seismic demand on the superstructure. The calculated self-centering and tipping over ratios confirm adequate safety margins for all considered scenarios. Overall, the study concludes that controlled foundation rocking significantly enhances the seismic resilience of RC frame-shear wall systems, offering a promising direction for earthquake-resistant construction in India.

The adoption of circular economy (CE) practices and artificial intelligence (AI) in the supply chain (SC) has become extremely significant in manufacturing organizations. The CE seeks to facilitate sustainable growth by managing the flow of materials and energy within closed-loop systems. The CE has resulted in the development of sustainable business models. AI capabilities transform work activities, data flows, and organizational processes. Therefore, the present study aims to develop a framework to improve circular supply chain (CSC) adoption in the automobile manufacturing sector by identifying and analyzing CE practices and AI-enabled digital strategies. The proposed framework was analyzed by employing a hybrid approach of Prioritized Weighted Average–Criteria Importance Through Intercriteria Correlation–Preference Ranking Organization Method for Enrichment Evaluations-II (PWA-CRITIC-PROMETHEE-II) under an Interval-Valued Fermatean Fuzzy (IVFF) environment. IVFF-CRITIC was employed to determine the CE practices’ weights, while IVFF-PROMETHEE-II was utilized to establish the relative index of AI-enabled digital strategies to enhance the CSC adoption. The key findings of the current study indicate that “AI-enabled infrastructure configuration for circular economy adoption in the supply chain”, “AI-integrated equipment to facilitate adaptability and mass personalization”, and “Robotics and AI-driven manufacturing and material reclamation” are the most significant AI-based digital strategies that support CE practices to enhance the adoption of a CSC and encourage case example manufacturing organizations to align their operations with AI and CE. Moreover, the outcomes of the study will deliver a comprehensive evaluation of CE practices and AI-enabled digital strategies for SC managers, based on the relative indexing obtained through the implementation of the hybrid approach.

Traditionally, Electric grids are designed to transport the bulk power generated at conventional power plants to load centres. Electricity regulators are tightening the grid codes to improve the performance and efficiency of electric grids. The coordinated planning of reactive power management for voltage profile regulation is an essential aspect of the seamless operation and control of electric grids. The adverse effects of seasonal variation in load profiles and the de-commitment of generating units based on their viability magnify the voltage regulation challenge in the large electric grid. These operational scenarios create low-voltage pockets in the distribution network, and there is a high draw of reactive power from the upstream transmission system. The power transfer capabilities of inter-regional tie lines decrease, and subsequently, this leads to overloading of power equipment and increases active loss in the utility grid. In this research article, reactive power management is formulated as a nonlinear optimisation problem. The objective function is minimised as the sum of reactive power flow over inter-intra-regional tie lines in a large electric grid. The optimisation problem is solved with a set of constraints imposed by placing the local reactive power support devices at critical locations. Selection of critical locations is identified through a new hybrid voltage-grid strength sensitivity index. The proposed index utilises the grid strength in addition to the voltage sensitivity of bus selection of buses for the installation of reactive power support devices. The performance of the proposed algorithm has been tested on real data from the Northern region of the Indian grid, which includes seven power transmission utilities with over 9000 buses. The simulation results show that the proposed reactive power optimisation, based on the hybrid voltage-grid strength sensitivity index, reduces reactive power import from 1592 to 383 MVAr by injecting 9421.8 MVAr at 33kV at only 14.1% of the highly sensitive buses identified through this index. It is observed that injecting compensation devices at 33kV buses improves the average bus voltage to nearly 0.98. The reactive power in the inter-intra tie lines decreases by 76%, and active power losses are reduced by 7.99%. In conclusion, the proposed algorithm can serve as a guiding tool for planning reactive power management in large-scale grid utilities, aiding voltage improvement with the optimal installation of compensating devices at a minimal number of 33kV buses.

ABSTRACT Magnesium (Mg) and its alloys are promising candidates for next-generation biodegradable implants owing to their biocompatibility, suitable mechanical properties, and ability to gradually degrade in vivo, eliminating the need for secondary removal surgery. However, their rapid corrosion in physiological environments remains a challenge. Wire Electrical Discharge Machining (WEDM) is increasingly considered for manufacturing complex Mg-based medical devices, yet its influence on surface integrity and degradation behaviour requires careful evaluation. In the present study, the in vitro corrosion performance of polished (Control), and polished/WEDM-processed AZ91 Mg alloy was systematically investigated. Specimens were immersed in simulated body fluid (SBF) for up to 168 h, following ASTM G31-72, and characterised using SEM/EDX, XRD, and electrochemical impedance spectroscopy (EIS). WEDM processing produced a re-solidified layer with micro-cracks, selective evaporation of Al and Zn, while polished samples retained α-Mg traces. A higher maximum area polarisation resistance for polished samples of 280 Ω cm² was obtained after 72 h immersion in SBF (pH 7.4) maintained at 37 ± 0.5 °C. EIS revealed higher polarisation resistance for polished samples (peak 280 Ω cm² at 72 h) vs. WEDMed (70 Ω cm² at 168 h). Hydrogen evolution rate (HER) and pH monitoring indicated more rapid alkalisation for WEDMed samples, corroborating accelerated corrosion. Post-immersion analysis showed a Ca/P ratio of 1.63 on polished samples, indicating stable hydroxyapatite deposition, compared to 1.29 for WEDMed surfaces, signifying less crystalline Ca – P phases. XRD corroborated these trends, revealing well-crystallised apatite on polished surfaces and predominantly soluble magnesium salts on WEDMed alloys. While WEDM offers manufacturing precision, its surface defects compromise corrosion resistance and bioactivity, highlighting the need for post-processing treatments to optimise implant performance.

Fragmentation warheads are considered critical components of missiles and other air-delivered weapon systems used against ground and aerial targets. The study of the terminal effects of a fragmentation warhead is an essential requirement during the design stage, layout configuration during the experiments, and development of protective system configuration against it. A Monte Carlo simulation-based computational model, named the Munitions Lethality Analysis (MULAS), has been developed to assess the terminal effects of the Pre-Fragmentation (PF) warhead in terms of its lethality and blast. A parametric study was conducted on cylindrical PF warheads with steel balls as a kill mechanism using the MULAS simulation method. The effect of the size of fragment steel balls on both fragment spatial distribution and warhead lethality, and the impact of the initiation point on warhead lethality, have been studied. Experiments with cylindrical model PF warheads were performed to validate the MULAS methodology. The test data analysis revealed that the performance prediction for fragment lethality and blast using the MULAS methodology was in fair agreement with the experimental data within a 10 % variation.

Abstract Scouring around bridge piers and abutments are the most common failure mode for bridges which can lead to infrastructure and economic losses. Thus, it is important to analyze the scour depth around bridge piers and abutments, however, most studies focus on the prediction of scour depth around pier of single geometry. This study focuses on predicting scour depth around vertical wall, 45° wing wall, and semicircular abutments using a comprehensive set of different machine learning (ML) models, i.e., Artificial Neural Network (ANN), Adaptive-Network-Based Fuzzy Inference System (ANFIS), Gene Expression Programming (GEP), Extreme Gradient Boosting (XGBoost), Random Forest (RF), Support Vector Machine (SVM), and Random Subspace KStar (RS-KStar) model. These models are validated using statistical indices like coefficient of determination (R 2 ), root mean square error (RMSE), and mean absolute error (MAE). The models were benchmarked against two existing empirical equations. The results indicate that, XGBoost and Random Forest achieved the highest accuracy for vertical wall and 45° wing wall due to their ability of capturing strong nonlinearities and skewed flow patterns. In case of semicircular abutment, ANN and ANFIS performed better as the flow behavior is smoother and more continuous. All ML models outperform the empirical equations highlighting the limited generalizability of these equations. The comparative analysis indicates that the performance of the model varies with different abutment geometries, highlighting the importance of selection of ML model based on structure geometry.