Search

The real power of artificial intelligence appears in reinforcement learning, which is computationally and physically more sophisticated due to its dynamic nature. Rotation and injection are some of the proven ways in active flow control for drag reduction on blunt bodies. In this paper, rotation will be added to the cylinder alongside the deep reinforcement learning (DRL) algorithm, which uses multiple controlled jets to reach the maximum possible drag suppression. Characteristics of the DRL code, including controlling parameters, their limitations, and optimization of the DRL network for use with rotation will be presented. This work will focus on optimizing the number and positions of the jets, the sensors location, and the maximum allowed flow rate to jets in the form of the maximum allowed flow rate of each actuation and the total number of them per episode. It is found that combining the rotation and DRL is promising since it suppresses the vortex shedding, stabilizes the Karman vortex street, and reduces the drag coefficient by up to 49.75%. Also, it will be shown that having more sensors at more locations is not always a good choice and the sensor number and location should be determined based on the need of the user and corresponding configuration. Also, allowing the agent to have access to higher flow rates, mostly reduces the performance, except when the cylinder rotates. In all cases, the agent can keep the lift coefficient at a value near zero, or stabilize it at a smaller number., Comment: 1st Edit: Parts of the introduction, simulation environment, and the network and reinforcement learning framework have been revised. --- 2nd Edit: Added real-world scenario, possible design of rotating cylinder, sensors limited to the body of the cylinder, Re = 200, compared the results and runtime of a shallow network with a higher number of neurons vs deeper network but lower number of neurons

A fully-coupled-fluid-structure-piezoelectric model is presented based on the finite element method that is capable of modeling piezoelectric harvesters in the presence of free-surface flow and floating lightweight harvesters with arbitrary movements. The Navier-Stokes equations and the phase-field method are employed to describe the free-surface waves. Equations of the conservation of linear momentum in company with the piezoelectric constitutive relations in the strain-charge form are utilized to obtain solid deformation and the electric field intensity. According to the results, attaching mass to the tip of the beam leads to 13.5% rise in the output voltage compared to the state without the attached mass. Another studied factor was the influence of the load resistance on voltage and the output power. The generated voltage grows along with the load resistance until it reaches a constant value. However, the power has an optimum load resistance that is 2.61 times higher than the reference state. The beam's inclination is significant in effectively exploiting water waves due to raising the root mean square (RMS) value of the voltage by 89.53% at an angle of 40 degrees relative to the vertical state. By altering the thickness of the beam from 1 mm to the value of 1.5 mm, the RMS voltage exhibits a considerable upward change of 66%. By increasing the length of the cantilever beam connected to the buoyant structure, and therefore, the indentation of the beam in the water, the output voltage grows, such that a beam with the length of 25 cm shows a 2.92 times increase in the output voltage relative to the beam with a length of 15 cm.

Numerical simulation of convective heat transfer over a stationary and transversely oscillating partial super-hydrophobic cylinder has been performed using OpenFOAM libraries. Superhydrophobicity of the cylinder surface has been addressed by means of a partial slip boundary condition. Applying the slip condition to the surface of the stationary cylinder causes the drag and the rms lift coefficients to reduce by 46 and 75 percent, respectively. It also augments the average Nusselt number by 55 percent accompanied by a 21 percent increase of the natural shedding frequency. The partially superhydrophobic cylinder has also been investigated and the effects of slip on different sections of the cylinder surface have been analyzed. Considering the reduction of force coefficients, it is shown that the application of slip over a 135 segment of the surface is an optimum case, resulting in a 47 and 85 percent decrease of the drag and the rms lift coefficients, respectively. However, the fully superhydrophobic cylinder provides higher heat transfer rates. Regarding the transversely oscillating cylinder, superhydrophobicity extends the primary synchronization region, and also exhibits different wake dynamics behavior compared to the no-slip case. The slip over surfaces also causes the average Nusselt number to become nearly 6 times greater than the no-slip oscillating cylinder at the lock-in condition. Further analysis based on thermal performance index (TPI) proves that a high value of TPI = 6 can be reached for the superhydrophobic cylinder.

In this paper, the pore network model to investigate the reaction-diffusion process in the hierarchically structured porous catalyst particle is extended to consider the phenomenon of deactivation by coking. In this framework, the interaction of internal particle pore structure and mass transfer under the condition of coke deposition are examined. A primitive experimental investigation has been performed as an introduction to the development of the model. Then, the effect of structural features namely macroporosity and pore size ratio, the deactivation properties, the maximum loading of coke as well as the transport properties, the pore Damkohler number on the net reaction rate and deactivation of the particle have been investigated. Three deactivation mechanisms are accounted for, namely, the site coverage, the pore narrowing, and the pore blockage. It is found that the deactivation of the catalyst particle can be divided into two conditions: the kineticsal deactivation and the structural deactivation. It is shown that depending on the Damkohler number, increasing the macroporosity does not necessarily improve the reactivity and deactivation resistance of the catalyst. The key finding of this work is to demonstrate and quantify how changing the typical fresh catalyst microstructure observed in the experimental characterization into a hierarchical one influences the reactivity and deactivation.

In heat transfer augmenting methods such as radial fins, the heat transfer enhancement commonly leads to the drag force increment. In the present paper, slots are inserted over the fins to simultaneously reduce the drag coefficient. Turbulent convection heat transfer around a cylinder, as well as oscillating bundle of tubes including the slotted radial fins have been investigated. The governing equations are solved in two-dimension utilizing OpenFOAM software based on k-{\omega} SST closure model. The cases with various slot location, slot width, the number of the slots, the fin height, and oscillation frequencies are examined. In all cases, the Reynolds number is taken to be equal to 5000. Presence of three slots on the fins reduces the drag coefficient by 23% and augments the Nusselt number by 76%. In order to enhance the heat transfer from the bundle of tube, oscillation of tubes and utilization of the slotted fins are applied. The optimum situation occurs for a sample with the oscillating third column that shows 3% increment in heat transfer relative to that of the fixed case. This is while adding the slotted fins to the oscillating tube bank increases the heat transfer up to 2.5 times., Comment: 25 pages, 25 figures

The accuracy of the classical heat conduction model, known as Fourier's law, is highly questioned, dealing with the micro and nanosystems and biological tissues. In other words, the results obtained from the classical equations deviate from the available experimental data. It means that the continuum heat diffusion equation is insufficient and inappropriate for modeling heat transport in these cases. There are several techniques for modeling non-Fourier heat conduction. In the present paper, we place our focus on the dual-phase-lag (DPL) approach. The DPL model, as a popular modification of Fourier's law, has already been utilized in numerous situations, such as simulating ultrafast laser heating and heat conduction in carbon nanotubes. There has been a sharp increase in research on non-Fourier heat conduction in recent years. Several studies have been performed in the fields of thermoelasticity, thermodynamics, transistor modeling, and bioheat transport. This review presents the most recent non-Fourier bioheat conduction works and the related thermodynamics background. The various mathematical tools, modeling different thermal therapies, and relevant criticisms and disputes are discussed. Finally, the novel and other possible studies are also presented to provide a better overview, and the roadmap to the future research and challenges ahead is drawn up., Comment: 81 pages, 4 figures

In the last decade, an explosive attention has been paid to piezoelectric harvesters due to their flexibility in design and increasing need to small-scale energy generation. As a result, various energy review papers have been presented by many researchers to cover different aspects of piezoelectric-based energy harvesting, including piezo-materials, modeling approaches, and design points for various applications. Most of such papers tried to shed light on recent progresses in related interdisciplinary fields, and to pave the road for future prospects of development of such technologies. However, there are some missing parts, overlaps, or even some contradictions in the review papers. In the present review of review articles, recommendations for future research directions suggested by the review papers have been systematically summed up under one umbrella. In the final section, topics for missing review papers, concluding remarks on outlooks and possible research topics, and strategy-misleading contents have been presented. The review papers have been evaluated based on merits and subcategories and authors' choice papers have been presented for each section based on clear classification criteria.