Search

This paper presents a proposition of a dual grid (DG)-based coupled deep reinforcement learning (DRL) and computational fluid dynamics (CFD) method for active flow control. The DG-DRL-CFD method uses a dual-resolution grid, with a coarser grid for the training phase and a finer grid for the testing phase of the DRL-CFD method. Further, after a validation study for our DRL-CFD-based parallelized in-house solvers, this paper presents a performance study for a periodic suction/ejection-based drag reduction for a cylinder in a channel-confined flow. Using an immersed boundary method for CFD on a Cartesian grid, this novel DG-DRL-CFD method is shown to result in almost same accuracy (within 1%) in a substantially reduced computational time as compared to the traditional DRL-CFD method (on the finer uniform grid). Finally, using a sharp interface level set method for an axisymmetric two-phase flow, this paper presents an application of the DR-DRL-CFD method for an oscillating base plate-based enhancement of heat transfer during nucleate pool boiling. As compared to our recent CFD study-based parametric study for the nucleate pool boiling problem, the present DG-DRL-CFD method leads to a profile and frequency of plate-oscillation that results in a larger value of average Nusselt number Nu avg for the plate. The coupled DRL-CFD method leads to plate oscillation profile giving higher enhancement in Nu avg , compared to an open-loop control strategy that involves a parametric sweep involving multiple simulations. • A dual grid (DG) based novel method is proposed which couples deep reinforcement learning and CFD for active flow control. • The method is validated on the problem of flow past cylinder with suction/ejection based drag reduction. • The method is applied to nucleate boiling with oscillating base plate where plate position is controlled using DRL. • The enhancement of Nusselt number is found to be higher than open-loop control strategy that involves a parametric sweep. [ABSTRACT FROM AUTHOR]

• Single bubble nucleate boiling in zero gravity studied using level-set based numerical simulations. • Asymmetric sinusoidal motion of base plate leads to sustenance of bubble departure and nucleate boiling. • Nusselt number enhancement is observed due to counter-clockwise vortices impinging on base plate. • Coalascence of departing bubbles may be avoided within a certain parameter range for plate oscillations. This work presents Direct Numerical Simulations to study the effect of base plate oscillations on nucleate boiling heat transfer at low superheat, in the isolated bubble regime, under zero gravity conditions. An in-house solver, based on sharp interface dual grid level set method (SI-DGLSM), is used to carry out the numerical simulations. To enable nucleate boiling in zero gravity, an asymmetric periodic motion of the base plate is proposed, in which the average acceleration of the plate is different along the two plate-normal directions. This time-asymmetry in plate motion imparts net motion of bubbles away from the plate, which in turn sustains nucleate boiling. The departure of the bubble is presented first for cases with no phase change, where the effect of varying plate oscillation parameters on the departure of single bubble over the plate is studied for zero gravity conditions. Thereafter, the effect of the plate motion on nucleate boiling characteristics is presented, and the regime in which nucleate boiling is sustained is identified. We report the effect of plate motion on time series of bubble departure radius, Nusselt number and overall flow dynamics during nucleate boiling in zero gravity. The proposed technique for the sustenance of nucleate boiling in zero gravity conditions has numerous space-related applications for heat transfer enhancement. [ABSTRACT FROM AUTHOR]