微尺度颗粒（包括细胞、囊泡、液滴等）的精确操控在生物医学分析和临床诊断等领域有着至关重要的应用。惯性微流动操控技术自2007年被首次提出以来，在对颗粒的高通量聚焦、过滤、富集和分离等方面表现出了巨大潜力。惯性微流动操控技术主要基于惯性迁移效应，微通道中的颗粒会由于受到流体惯性影响而穿过流线进行侧向迁移。颗粒的精确操控赖于对惯性迁移影响因素的深入认识，本文通过数值模拟分别对颗粒形状，通道结构以及流体性质的影响进行了系统研究。

微尺度流动中颗粒的惯性迁移效应已成为操控生物颗粒和工程颗粒的有效工具，这些颗粒的形状普遍是非球形的。科学界已经对球形颗粒的惯性迁移效应进行了大量研究，而非球形颗粒的研究依旧还有很大的探索空间。在这里，我们利用三维直接数值模拟研究了柱状颗粒在矩形截面通道内的惯性迁移。我们系统地研究了柱状颗粒在惯性迁移过程中所受表面应力以及惯性升力的动态变化，揭示了典型迁移行为并讨论了与球形颗粒迁移行为的关系和差异，这些发现将有助于精确控制柱状颗粒。

基于颗粒的惯性迁移效应，研究者们通过组合多段通道或引入二次流设计了大量微流动颗粒操控装置，可以灵活地实现各种操控目的。惯性微流动操控装置的设计在很大程度上取决于对颗粒轨迹的精确预测，该预测主要通过颗粒所受惯性升力确定。作为能够准确获得升力的唯一方法，直接数值模拟通常会消耗大量计算成本，具有复杂结构的微通道的计算成本将变得非常高以至于难以实现。我们结合机器学习技术提出了一种数值算法，可以实现三种常用通道截面内惯性升力数值解的快速获取。该惯性升力数值解能够显式地代入到拉格朗日追踪公式中，快速有效地预测各种颗粒惯性操控装置的功能，从而帮助研究人员加快微流动颗粒操控装置的开发。

基于流体惯性和流体弹性的综合作用，粘弹性流中颗粒的弹性-惯性迁移可以诱导高质量颗粒聚焦。由于计算流体动力学方法存在无法处理相对较大Weissenberg数的缺点，阐明大Weissenberg数流动中颗粒的弹性-惯性迁移机制仍然是一个挑战。在这里，我们使用耗散粒子动力学方法研究了矩形微通道中球形颗粒的弹性-惯性迁移。基于聚合物溶液模型，我们成功地再现了具有聚合物链拉伸行为的大Weissenberg数流动。我们发现随着Weissenberg数的增大，颗粒聚焦位置由通道中心处转移至长中线上的偏心位置。有趣的是，作用在颗粒上的弹性升力和惯性升力是相互合作而非竞争将颗粒聚焦到偏心位置。通过分析法向应力差的分布和聚合物链拉伸行为，我们发现大Weissenberg数加剧了颗粒扰动影响，从而改变了通道中心附近弹性升力方向。上述研究对基于粘弹性效应的颗粒微流动操控器件设计将起到指导作用。

The precise manipulation of microparticles, including cells, vesicles, and droplets, has important applications in biomedicine analysis and clinical diagnosis. Since the inertial microfluidic technology first proposed in 2007, it has shown great potential in high-throughput focusing, filtering, concentrating, and separation of particles. The inertial microfluidic technology is mainly based on inertial migration, that is, the particles in the microchannel laterally move cross the streamline due to the fluid inertia. The realization of precise particle manipulation depends on the in-depth understanding of inertial migration. In this paper, we systematically studied the influence of particle shape, channel structure, and fluid properties using direct numerical simulation.

Inertial migration has emerged as an efficient tool for manipulating both biological and engineered particles that commonly exist with non-spherical shapes in microfluidic devices. There have been numerous studies on the inertial migration of spherical particles, whereas the non-spherical particles are still largely unexplored. Here, we conducted three-dimensional direct numerical simulations to study the inertial migration of rigid cylindrical particles in rectangular microchannels. We systematically studied the dynamic changes of the surface stress and inertial lift acting on cylindrical particles, revealed the typical behavior of particle migration, and discussed the relationship and difference between cylindrical particles and spherical particles, which can be useful for the precise manipulation of cylinder-like particles.

Based on the inertial effect, researchers have designed many microfluidic devices by combining multiple microchannels or introducing secondary flow, which can flexibly achieve various manipulation purposes. The design of inertial microfluidic devices largely relies on the precise prediction of particle migration that is determined by the inertial lift acting on the particles. Despite being the only means to accurately obtain the lift forces, direct numerical simulation (DNS) often consumes high computational cost and even becomes impractical when applied to microchannels with complex geometries. Herein, we proposed a fast numerical algorithm in conjunction with machine learning techniques to quickly obtain the numerical solution of the inertial lift in the microchannels with three types of commonly used cross-sectional shapes. The numerical solution of the inertial lift force can be explicitly integrated into the Lagrangian tracking method to effectively predict the functions of various particle manipulation devices, thereby helping researchers to expedite the development of inertial microfluidic devices for particle manipulation.

Elasto-inertial migration of particles in a viscoelastic flow can result in high-quality particle focusing due to the combined effect of fluid elasticity and flow inertia. Owing to the drawback of computational fluid dynamics methods being incapable of dealing with relatively large Weissenberg numbers, clarifying the mechanism of elasto-inertial migration in large Weissenberg number flow remains a challenge. Here, we investigated the elasto-inertial migration of spherical particles in a rectangular microchannel, using a dissipative particle dynamics method. Based on the polymer solution model, we successfully reproduced large Weissenberg number flow with polymer stretching behavior. We found that with the increase of the Weissenberg number the focusing position of particles shifts from the channel center to the off-center position on the long midline. Interestingly, the elastic lift force and inertial lift force acting on the particles do not compete but cooperate to focus the particles to the off-center position. By analyzing the distribution of the normal stress differences and the stretching behavior of the polymer chains, we found that the large Weissenberg number aggravates the impact of particle disturbance, which changes the direction of the elastic lift force near the channel center. The above research provides useful guidelines for the design of particle microfluidic devices based on the viscoelastic effects.