Curtis Christopher W, Carretero-González R, Polimeno Matteo
Nonlinear Dynamical Systems Group, Department of Mathematics and Statistics, San Diego State University, San Diego, California 92182-7720, USA.
Computational Sciences Research Center, San Diego State University, San Diego, California 92182-7720, USA.
Phys Rev E. 2019 Jun;99(6-1):062215. doi: 10.1103/PhysRevE.99.062215.
We use the dynamic mode decomposition (DMD) methodology to study weakly turbulent flows in two-dimensional Bose-Einstein condensates modeled by a Gross-Pitaevskii equation subject to band-limited stochastic forcing. The forcing is balanced by the removal of energy at both ends of the energy spectrum through phenomenological hypoviscosity and hyperviscosity terms. Using different combinations of these parameters, we simulate three different regimes corresponding to weak-wave turbulence, and high- and low-frequency saturation. By extracting and ranking the primary DMD modes carrying the bulk of the energy, we are able to characterize the different regimes. In particular, the proposed DMD mode projection is able to seamlessly extract the vortices present in the condensate. This is achieved despite the fact that we do not use any phase information of the condensate as it is usually not directly available in realistic atomic BEC scenarios. Being model independent, this DMD methodology should be portable to other models and experiments involving complex flows. The DMD implementation could be used to elucidate different types of turbulent regimes as well as identifying and pinpointing the existence of delicate and hidden coherent structures within complex flows.
我们使用动态模态分解(DMD)方法来研究二维玻色-爱因斯坦凝聚体中的弱湍流,该凝聚体由受带限随机强迫作用的格罗斯-皮塔耶夫斯基方程建模。通过唯象的亚粘性和超粘性项在能谱两端去除能量,从而平衡强迫作用。利用这些参数的不同组合,我们模拟了对应于弱波湍流、高频饱和和低频饱和的三种不同状态。通过提取并排列携带大部分能量的主要DMD模态,我们能够表征不同的状态。特别地,所提出的DMD模态投影能够无缝地提取凝聚体中存在的涡旋。尽管在实际的原子玻色-爱因斯坦凝聚体情形中通常无法直接获得凝聚体的任何相位信息,但仍能实现这一点。由于该DMD方法与模型无关,它应该能够应用于其他涉及复杂流动的模型和实验。DMD实现可用于阐明不同类型的湍流状态,以及识别和定位复杂流动中微妙和隐藏的相干结构的存在。