Abdulghani M. Abdulghani, Mokhles M. Abdulghani, Wilbur L. Walters, Khalid H. Abed^*

Journal on Artificial Intelligence, Vol.5, pp. 1-14, 2023, DOI:10.32604/jai.2023.039786

Abstract Autonomous agents can explore the environment around them when equipped with advanced hardware and software systems that help intelligent agents minimize collisions. These systems are developed under the term Artificial Intelligence (AI) safety. AI safety is essential to provide reliable service to consumers in various fields such as military, education, healthcare, and automotive. This paper presents the design of an AI safety algorithm for safe autonomous navigation using Reinforcement Learning (RL). Machine Learning Agents Toolkit (ML-Agents) was used to train the agent with a proximal policy optimizer algorithm with an intrinsic curiosity module (PPO + ICM). This training aims to improve AI… More >

Mo Yang^a,*, Xiaoming Wang^a, Zhiyun Wang^a, Zheng Li^b , Yuwen Zhang^b

Frontiers in Heat and Mass Transfer, Vol.11, pp. 1-6, 2018, DOI:10.5098/hmt.11.7

Abstract Turbulent convective heat transfer in an elliptical pipe are investigated numerically in this paper. The RSM model is employed in the simulations of elliptical tubes with different aspect ratio a/b and Reynolds numbers within the range of 10,000~120,000. It is found that the maximum deviation between the numerical result and the one from Dittus-Boelter equation contained a hydraulic diameter is 28.4%. Based on simulation results, the correlation between the Nusselt number and Reynolds number in the fully developed fluid section of the elliptical tube is obtained. More >

Zijie Chen^a , Sanat Modak^a, Massoud Kaviany^a,* , Richard Bonner^b

Frontiers in Heat and Mass Transfer, Vol.14, pp. 1-11, 2020, DOI:10.5098/hmt.14.1

Abstract In dropwise condensation on vertical surface, droplets grow at nucleation sites, coalesce and reach the departing diameter. In biphilic surfaces, when the hydrophobic domain is small, the maximum droplet diameter is controlled by the shortest dimension where the droplets merge at the boundary. Through direct numerical simulations this size-effect heat transfer coefficient enhancement is calculated. Then the 1-D biphilic surface is optimized considering the size-dependent hydrophilic domain partial flooding (directly simulated as a liquid rivulet and using the capillary limit), the subcooling (heat flux) and condenser length effects. The predicted performance is in good agreement with the available experiments. More >

Christopher C. Rosser, Wilbur L. Walters, Abdulghani M. Abdulghani, Mokhles M. Abdulghani, Khalid H. Abed^*

Journal on Artificial Intelligence, Vol.4, No.4, pp. 261-277, 2022, DOI:10.32604/jai.2022.039703

Abstract Intrinsic motivation helps autonomous exploring agents traverse a larger portion of their environments. However, simulations of different learning environments in previous research show that after millions of timesteps of successful training, an intrinsically motivated agent may learn to act in ways unintended by the designer. This potential for unintended actions of autonomous exploring agents poses threats to the environment and humans if operated in the real world. We investigated this topic by using Unity’s Machine Learning Agent Toolkit (ML-Agents) implementation of the Proximal Policy Optimization (PPO) algorithm with the Intrinsic Curiosity Module (ICM) to train autonomous exploring agents in three… More >

Fangzhou Wang^a,* , Gennifer A. Riley^b, Munonyedi Egbo^c, Melanie M. Derby^b, Gisuk Hwang^c, Xianglin Li^a,†

Frontiers in Heat and Mass Transfer, Vol.15, pp. 1-8, 2020, DOI:10.5098/hmt.15.1

Abstract This study conducts pore-scale simulations and experiments to estimate the permeability of two different types of porous materials: metal foams and sintered copper particles with porosities of approximately 0.9 and 0.4, respectively. The integration of micro X-ray computed tomography with pore-scale computational fluid dynamics simulations develops a unique tool to capture the pore-scale geometry of porous media and accurately predict non-isotropic permeability of porous media. The pore-scale simulation not only results in improved prediction accuracy but also has the capability to capture non-isotropic properties of heterogeneous materials, which is a huge challenge for empirical correlations, volume averaged simulations, and simulations… More >

O. M. Oyewola^a,b,*, O. S. Ismail^b, J. O. Bosomo^b

Frontiers in Heat and Mass Transfer, Vol.17, pp. 1-5, 2021, DOI:10.5098/hmt.17.8

Abstract This paper investigates possible inherent modifications of the radiative heat source term due to the influence of turbulence in the thermal radiation field of a gas turbine combustor flame. Adapting a flame temperature of 2000[K], COMSOL Multiphysics software was utilized to numerically simulate the process, assuming a gray gas participating medium with absorption coefficient of 0.03[m^-1]. The analysis of the results for five (5) different radial cut sections of the simulated combustor chamber shows that turbulence-radiation interactions cause radiative heat losses from the flame, with the divergence of the radiative heat flux having a deviation factor of 3.48, and a… More >

Farhan Lafta Rashid^a, Sarmad Kamal Fakhrulddin^b, Muhammad Asmail Eleiwi^c, Ahmed Kadhim Hussein^d,e, Tahseen Ahmad Tahseen^f , ObaiYounis^g,h,*, Mohammed Ibrahim Ahmedⁱ

Frontiers in Heat and Mass Transfer, Vol.19, pp. 1-8, 2022, DOI:10.5098/hmt.19.6

Abstract The current study numerically analyzes the heat transfer enhancement and laminar fluid flow characteristics of four flat tubes with varying attack front airflow. The heat transfer characteristics of flat tubes are investigated in terms of Reynolds number, heat fluxes, and inclination angle. Four Reynolds numbers are studied (100, 200, 300, and 400), and three heat fluxes on the surface of the tubes are (1000, 2500, 3800 W/m2 ), the inclination angle of the four flat tubes are (30o , 45o , 90o). ANSYS Fluent software v.18 discretizes and solves the governing equations using the finite volume approach across a specified… More >

Pushpendra Kumar^1,*, Vedat Suat Erturk², V. Govindaraj¹, Dumitru Baleanu^3,4,5

CMES-Computer Modeling in Engineering & Sciences, Vol.136, No.3, pp. 2487-2506, 2023, DOI:10.32604/cmes.2023.026009

Abstract In this article, we introduce a nonlinear Caputo-type snakebite envenoming model with memory. The well-known Caputo fractional derivative is used to generalize the previously presented integer-order model into a fractional-order sense. The numerical solution of the model is derived from a novel implementation of a finite-difference predictor-corrector (L1-PC) scheme with error estimation and stability analysis. The proof of the existence and positivity of the solution is given by using the fixed point theory. From the necessary simulations, we justify that the first-time implementation of the proposed method on an epidemic model shows that the scheme is fully suitable and time-efficient… More >

Yiguo Chen^1,2,3,*, Congjun Feng^1,2, Yonghong He³, Zhijun Chen³, Xiaowei Fan³, Chao Wang³, Xinmin Ge⁴

CMES-Computer Modeling in Engineering & Sciences, Vol.136, No.3, pp. 2451-2463, 2023, DOI:10.32604/cmes.2023.021145

Abstract The low-field nuclear magnetic resonance (NMR) technique has been used to probe the pore size distribution and the fluid composition in geophysical prospecting and related fields. However, the speed and accuracy of the existing numerical inversion methods are still challenging due to the ill-posed nature of the first kind Fredholm integral equation and the contamination of the noises. This paper proposes a novel inversion algorithm to accelerate the convergence and enhance the precision using empirical truncated singular value decompositions (TSVD) and the linearized Bregman iteration. The L1 penalty term is applied to construct the objective function, and then the linearized… More > Graphic Abstract

Yiming Zhang^1,*, Zhiran Gao¹, Xueya Wang¹, Qi Liu²

CMES-Computer Modeling in Engineering & Sciences, Vol.134, No.2, pp. 821-833, 2023, DOI:10.32604/cmes.2022.022088

Abstract A large amount of data can partly assure good fitting quality for the trained neural networks. When the quantity of experimental or on-site monitoring data is commonly insufficient and the quality is difficult to control in engineering practice, numerical simulations can provide a large amount of controlled high quality data. Once the neural networks are trained by such data, they can be used for predicting the properties/responses of the engineering objects instantly, saving the further computing efforts of simulation tools. Correspondingly, a strategy for efficiently transferring the input and output data used and obtained in numerical simulations to neural networks… More >