Engineering Reports

The Borassus flabellifer husk, with its cross-linked, hollow cylindrical structure and a density of 0.74 g/cm³, is lightweight and exhibits low water absorption (47.25%) and moisture regain (4.41%), enhancing weathering resistance. These properties make it a competitive alternative to bio-fibers like jute and sisal for lightweight engineering applications.

Composite (DMA): This study examines the alkali treatment effect on Borassus husk fiber-reinforced epoxy composites. Alkali treatment improves mechanical properties, including increased storage modulus, glass transition temperature, and damping factor, especially for 0.75-h treated fiber/epoxy composite. Alkali-treated/epoxy composites show enhanced thermos-mechanical stability and vibration damping, making them suitable for aerospace and automotive applications.

The service provision method is introduced based on community detection in the SIoT environment. Using a community detection algorithm, the IoT network is divided into smaller communities. Then, the proposed method searches for the services users request within the detected communities, selects and combines these, and runs on an application.

The centrifugal pump achieved optimal performance at lower viscosity with a minimal number of blades, as evidenced by reduced turbulent kinetic energy, leading to minimal energy loss and maximum efficiency, with the best operation observed using water and five blades.

In ASEB, two key techniques are employed for synthetic sample generation: ADASYN and Generative Adversarial Networks (GANs). ADASYN focuses on generating synthetic samples near the decision boundary of minority class instances, addressing challenging and often misclassified examples. This targeted generation helps the model become more sensitive to subtle patterns and nuances in the minority class, leading to better performance on these critical cases. GANs, on the other hand, utilize a generative adversarial framework where a generator creates synthetic samples and a discriminator evaluates their authenticity. This method produces high-quality synthetic data that closely mirrors real data, enhancing the diversity and complexity of the training set. The use of GANs in ASEB ensures that the synthetic samples not only balance the class distribution but also maintain the integrity and variability of the original data. The impact of these synthetic samples is evaluated through various performance metrics. Enhanced precision and recall for the minority class, improved F1 scores, and higher area under the Precision-Recall curve (PR AUC) are observed, indicating that the synthetic samples effectively contribute to more accurate and reliable predictions. Furthermore, these improvements are accompanied by a better balance in the Matthews Correlation Coefficient (MCC), which reflects the overall quality of the model in handling both classes. Overall, the analysis demonstrates that the synthetic samples generated by ASEB significantly bolster the model's ability to learn from and predict minority class instances. By addressing the class imbalance through sophisticated synthetic data generation techniques, ASEB not only enhances model accuracy but also ensures a more equitable performance across all classes, paving the way for more robust and fair predictive models in heart disease diagnosis and beyond.

A wireless detection wall-climbing robot was developed to solve the problems of high labor intensity and safety hazards for plane steel gate panel coating thickness detection. Through structural design and control system design, the robot integrates the functions of steel gate panel cleaning and coating thickness detection.

Deep-learning techniques were applied in this study. EfficientNetB3, MobileNetV2, and InceptionV3 were employed to assess their effectiveness in detecting melanoma. Among these architectures, EfficientNetB3 emerged as the standout performer, achieving an exceptional accuracy rate of 90.7% and an impressive area under the curve (AUC) score of 97%. A cascading combination technique was utilized to develop a multi-architecture model by combining the output of all the individual models. This reduced the computational cost of the model and enhanced the efficiency of the model for melanoma detection.

The paper summarizes the published studies on additive manufacturing in the powder bed process of CuSn10, covering manufacturing parameters, material properties, microstructure, corrosion behavior, and heat treatments. Multi-material compounds such as 316L/CuSn10 or similar are also considered. In addition, future developments are derived.

This study aims to analyze the age related characteristics of malaria in human host by exploring Caputo fractional order models with temperature variability, that is looked into the combined effects of fractional order and temperature variability on malaria dynamics. MATLAB was used to simulate Caputo fractional order with temperature variability and to apply the Adams–Bashforth–Moulton numerical approach.

This study presents the design of a novel secondary controller for Automatic Generation Control (AGC) in multi-area; multi-source power systems incorporating an appropriate generating rate constraint (GRC) has been considered for the hydro and thermal power plants. The proposed controller, optimized using the Grey Wolf Optimizer (GWO) algorithm, improves AGC performance in systems integrating hydro-thermal. The controller's effectiveness is demonstrated through enhanced dynamic response, reduced frequency deviations, and better load handling. The optimization process fine-tunes the controller parameters to ensure superior performance and robustness under varying operational conditions.

This study compares one-step and multi-step numerical methods for draining a water tank, assessing accuracy, stability, computational time, and memory. Multi-step methods offer greater efficiency but require careful initialization and are more memory-intensive. Higher-order Runge-Kutta methods enhance stability, but multi-step methods may reduce stability with increased order.

Numerical simulation to study the extraction and storage mechanism of three types of gas reservoirs: (1) Low-permeability gas reservoirs: Competing adsorption and pressurization to replenish energy. (2) Mesophilic depleted gas reservoirs: Pressurization to replenish energy, gas injection rate, and gas injection components. (3) Medium-permeability water–intrusive gas reservoirs: Injection/extraction ratios and gas injection fractions.

Based on these insights, mold design is optimized through strategic gate placement, weld line management, elimination of gas traps, pressure drop equalization, and stress reduction. Implementing these optimized parameters in the manufacturing phase leads to improved part quality, characterized by fewer defects and better dimensional accuracy. Furthermore, this research aids in minimizing production costs by decreasing rejections and rework.

This study was performed to summarize the current research status of photodynamic therapy (PDT) for bladder tumors in China and provide a comprehensive analysis of global research frontiers and emerging trends in PDT for bladder tumors.

At United Steel, Kuwait, replacing lime fines with dololime fines in the electric arc furnace optimizes material use and resolves size disparity issues. This approach reduces lime and dololime consumption, eliminates the need for external briquetting, and saves approximately USD 350,000 annually, enhancing both efficiency and cost-effectiveness.