Searcharxiv⌕ Search

arXiv subjects

Jia-Jun Zou

Publications and source records attributed to Jia-Jun Zou.

2 recordsLinked to original sources

A bound-preserving oscillation-eliminating discontinuous Galerkin method with operator splitting for solving Kapila's five-equation model

This paper proposes a robust operator-splitting discontinuous Galerkin (DG) framework to overcome the severe stiffness-induced instabilities in simulating compressible two-phase flows governed by Kapila's five-equation model with the Tammann equation of state. Specifically, the system is decoupled into a five-equation transport model and a stiff $Îș$-source term. The former is discretized via a quasi-conservative DG method \cite{cheng2020quasi}, while the latter is resolved by the local DG method combined with a novel adaptive implicit strategy that hybridizes the backward Euler and second-order singly diagonally implicit Runge-Kutta schemes. This implicit strategy possesses the unconditionally bound-preserving property, and thus effectively circumvents the severe stability constraints and time-step penalties inherent in traditional explicit schemes. Furthermore, to enhance computational robustness, we integrate an oscillation-eliminating DG (OEDG) procedure to suppresses spurious oscillations without characteristic decomposition, complemented by a bound-preserving limiter to maintain physically admissible numerical solutions. We also prove that the proposed operator-splitting DG framework, integrated with the oscillation-eliminating limiter, and the bound-preserving limiter, strictly satisfies the Abgrall condition. Finally, extensive numerical experiments are conducted to demonstrate the superior robustness and efficiency of the method.

math.NA↗

Moving mesh FSI approach for VIV simulation based on DG method with AMR technique

Vortex-induced vibration (VIV) remains a fundamental yet computationally challenging problem in computational fluid dynamics (CFD). This study develops a moving mesh Fluid-structure interaction (FSI) algorithm within a Runge-Kutta Discontinuous Galerkin (RKDG) adaptive mesh refinement (AMR) framework. The viscous term in the compressible Navier-Stokes (NS) equations is discretized using the high-order Interior Penalty Discontinuous Galerkin (IPDG) method. In addition to the above, key numerical advancements encompass the rigorous derivation of the Lax-Friedrichs (L-F) numerical flux formulation tailored for moving meshes, an enhanced AMR-driven nodal correction methodology designed for curved surface geometries, and the implementation of a ghost-node boundary condition treatment scheme to address dynamic mesh motion. Numerical validation proceeds through three phases: First, Couette flow simulations confirm the IPDG method's spatial convergence order. Subsequent analysis of unsteady flow past a cylinder demonstrate the AMR framework's efficacy in resolving vortex-dominated flow. Finally, six VIV benchmark cases are simulated using third-order IPDG discretization, establishing the proposed FSI algorithm's accuracy. Furthermore, synthetic jets (SJs) flow control is investigated through four frequency-variant SJs configurations. The results reveal that SJs can achieve completely VIV suppression at a low actuation frequency, while higher actuation frequencies reduce suppression efficiency due to the energy of the SJs is more in the form of acoustic wave.

physics.flu-dyn↗