Article
Thermodynamics
Ali Fguiri, Christophe Marvillet, Mohamed Razak Jeday
Summary: This study solved the problem of estimating thermal resistance of fouling in a heat exchanger using an inverse method based on the Levenberg-Marquardt technique. Two parameters characterizing the thermal resistance of fouling were simultaneously estimated through sensitivity analysis. Experimental validation was performed on three Phosphoric Acid/Steam heat exchangers to determine the asymptotic value of the fouling resistance and the characteristic time.
APPLIED THERMAL ENGINEERING
(2021)
Article
Engineering, Chemical
Ehsan Davoudi, Bahareh Kamyab Moghadas
Summary: In this study, artificial neural networks and adaptive neural-based fuzzy inference system were used to predict the fouling factor of hot wire probe based on a large dataset. The cascade feed-forward neural network was found to be the best artificial intelligence model for accurately estimating the fouling factor.
CHEMICAL ENGINEERING RESEARCH & DESIGN
(2022)
Article
Computer Science, Artificial Intelligence
P. S. Febin Sheron, K. P. Sridhar, S. Baskar, P. Mohamed Shakeel
Summary: The article proposes a projection-dependent input processing (PDIP) method to reduce misidentifications in object recognition. This method identifies conjoining indices by projecting the visualizing image in all possible dimensions, preventing errors from occurring.
IMAGE AND VISION COMPUTING
(2021)
Article
Neurosciences
Samuel P. Muscinelli, Mark J. Wagner, Ashok Litwin-Kumar
Summary: The study develops a theory to understand the computational role of compressed sensorimotor inputs in cerebellum-like structures and predicts the function of the pontine relay and glomerular organization. It highlights the difference between clustered and distributed neuronal representations and reconciles recent observations with theories of nonlinear mixing. Overall, the study demonstrates the efficiency of structured compression followed by random expansion for flexible computation.
NATURE NEUROSCIENCE
(2023)
Article
Environmental Sciences
Luisa Marques, Guilherme Teixeira, Ricardo Calado, Ana Isabel Lillebo
Summary: This study surveyed and established a baseline for the ascidian biofouling community in a coastal lagoon using artificial substrate units (ASU) made from oyster shells. The results showed that ASU can be considered an efficient support structure for monitoring the ascidian community and fouling communities in general. Additionally, two ascidian species were recorded for the first time in coastal waters of mainland Portugal.
FRONTIERS IN MARINE SCIENCE
(2022)
Article
Materials Science, Multidisciplinary
V. Mathivet, J. Jouin, M. Parlier, S. Rossignol
Summary: The study focuses on the synthesis of geopolymers from metakaolin and phosphoric acid with varying [P] concentration and the addition of water to improve the viscosity for complex shapes. Measurements of viscosity, density, and consolidation times, as well as studies on microstructure and mechanical properties were conducted to assess the influence of water content on the material's structure. The results showed that the addition of water led to the formation of new amorphous networks in the geopolymers, impacting their properties.
MATERIALS CHEMISTRY AND PHYSICS
(2021)
Article
Engineering, Environmental
Ahmed Benyekhlef, Brahim Mohammedi, Djamel Hassani, Salah Hanini
Summary: An artificial neural network model was developed to predict fouling resistance for heat exchangers. The MLP network with 20 hidden neurons trained using the LM algorithm showed the best performance, with high accuracy in all training, testing, and validation stages. Inputs variables were found to have a strong effect on estimating fouling resistance after conducting a sensitivity analysis.
WATER SCIENCE AND TECHNOLOGY
(2021)
Article
Chemistry, Physical
Taegun Park, Taehyeon Kim, Jongwon Han, Sangwoo Lim
Summary: The improvement of Si3N4 etching kinetics and suppression of oxide regrowth are crucial in the selective Si3N4 etching process for the fabrication of 3D NAND devices on multi-layered Si3N4/SiO2 structures. By reducing the concentration of H3PO4, oxide regrowth was suppressed and 30 wt% H3PO4 showed the highest Si3N4/SiO2 etching selectivity. Additionally, introducing amino acids with hydrophilic side chains into 30 wt% H3PO4 increased the Si3N4 etching rate and suppressed oxide regrowth.
APPLIED SURFACE SCIENCE
(2023)
Article
Green & Sustainable Science & Technology
M. Jahirul, M. G. Rasul, R. J. Brown, W. Senadeera, M. A. Hosen, R. Haque, S. C. Saha, T. M. Mahlia
Summary: This study used principal component analysis and artificial intelligence models to investigate the impact of biodiesel chemical composition on its properties. By utilizing experimental data and literature information, the study established a model to show the complex multi-dimensional correlation between biodiesel properties and chemical composition.
Article
Environmental Sciences
Pedro D. A. Bastos, Claudia F. Galinha, Maria Antonio Santos, Pedro Jorge Carvalho, Joao G. Crespo
Summary: The study focused on identifying the wastewater stream with the highest phenolic impact at a large oil refinery. Through multivariate data analysis and development of predictive models, correlations between process parameters and phenols content in wastewater were established to improve wastewater management system. An optimised system leads to improvement in wastewater quality and reduction in pollutant discharge costs.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2022)
Article
Engineering, Mechanical
Jie Deng, Oriol Guasch, Laurent Maxit, Nansha Gao
Summary: In this paper, a component mode synthesis method using artificial springs to connect subsystems of a built-up structure is developed. The method allows for different approaches in computing the modal behavior of the subsystems, depending on their complexity. The use of artificial springs avoids the difficulty of satisfying the system boundary conditions, and can significantly reduce the computational cost of numerical simulations.
INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES
(2023)
Article
Physics, Multidisciplinary
Jianning Han, Rui Wang
Summary: This study proposes a coding composite structure composed of a perforated plate and a coiling-up cavity with six interdigital rigid walls to achieve switchable acoustic projection displays. By using forward and backward placements, the configuration can serve as bits '0' and '1' for display. The projected image can be switched from 'Off state' to 'On state' by switching control wave P(c).
FRONTIERS IN PHYSICS
(2023)
Article
Materials Science, Multidisciplinary
Rania Jradi, Christophe Marvillet, Mohamed-Razak Jeday
Summary: In this study, an Artificial Neural Network (ANN) is used to predict the fouling resistance in the phosphoric acid concentration process. By selecting the best training algorithm and the optimal number of hidden neurons, the model accurately predicts the fouling resistance.
FDMP-FLUID DYNAMICS & MATERIALS PROCESSING
(2022)
Article
Engineering, Electrical & Electronic
Hamed Karami, Saeed Karami Chemeh, Vahid Azizi, Hooman Sharifnasab, Jose Ramos, Mohammed Kamruzzaman
Summary: Aroma is a significant quality trait for pharmaceutical plants and their products, indicating the quality of the raw material. An electronic nose is an efficient approach for identifying and evaluating the aroma of essential oils. In this study, tarragon was dried at different temperatures and air velocities, and the purity of tarragon essential oil was evaluated using an electronic nose. Multivariate data analysis and artificial neural networks modeling were employed to quantify and classify the obtained essential oils.
SENSORS AND ACTUATORS A-PHYSICAL
(2024)
Article
Computer Science, Interdisciplinary Applications
Mengkai Lu, Liang Zhang, Zhou Yan, Jian Wu
Summary: The paper presents a variational principle for bi-modulus elasticity to model no-tension/compression materials and structures, proving equivalence between the derived complementarity finite element formulation and a variational inequality. The method is demonstrated through numerical examples and verified with existing experimental data, showing good numerical stability and improved computational efficiency.
COMPUTERS & STRUCTURES
(2021)