Abstract The hydrodynamics around a Brownian particle has a noticeable impact on its hindered diffusion in arbitrary geometries (such as channels/pores) due to reduced mobility close to walls. These effects are difficult to describe at sub-pore scales, wherein a complete analytical solution of the underlying hydrodynamics is challenging to obtain. Here, we propose a coupled Langevin-multiphase direct numerical simulation (DNS) framework, that fully resolves the hydrodynamics in such systems and consequently provides an on-the-fly capability to probe local instantaneous particle diffusivities. We validate and establish the capabilities of this framework in square micro-channels (under varying degrees of hydrodynamic confinement) and in an arbitrary pore. Our results show that directional variations in mean-squared displacements, velocity auto-correlation functions and diffusivities of the Brownian particle, due to inherent asymmetries in the geometry are adequately captured. Further, a local anisotropy in the hydrodynamic resistances along the co-axial direction of the channel is also noted.

中文翻译：

---

使用多相 DNS 框架对任意几何形状中的受阻扩散进行大量评估

摘要 由于靠近壁的流动性降低，布朗粒子周围的流体动力学对其在任意几何形状（例如通道/孔）中的受阻扩散具有显着影响。这些影响很难在亚孔隙尺度上描述，其中潜在流体动力学的完整解析解很难获得。在这里，我们提出了一个耦合的朗之万-多相直接数值模拟 (DNS) 框架，它完全解决了此类系统中的流体动力学问题，因此提供了一种动态探测局部瞬时粒子扩散率的能力。我们验证并建立了该框架在方形微通道（在不同程度的流体动力学限制下）和任意孔中的能力。我们的结果表明，均方位移的方向变化，由于几何中固有的不对称性，布朗粒子的速度自相关函数和扩散率被充分捕获。此外，还注意到沿通道同轴方向的流体动力学阻力的局部各向异性。