Collaborative Research: A Fundamental and Modeling Study of Cluster-Induced Turbulence in Particle-Laden Flows

本研究的目标是建立基于颗粒流动动力学理论的多相湍流模型，并利用高保真度模拟。这些流动的应用领域包括发动机、工业炉、用于化石和可再生能源转换的流化床以及化工厂（化学反应器、裂化装置等），所有这些对美国工业的竞争力都很重要。合作的pi建议将他们的模型和代码全面实施到商业（Ansys Fluent）和公共（OpenFOAM）平台上，并将他们的详细仿真结果的综合数据库提供给研究界。目前可用的多相湍流模型通常基于现象学方法，需要用实验数据进行微调。这些模型的预测能力有限，阻碍了它们在工业上的采用或在研究目的上的推广。然而，这一建议旨在发展和验证一个一致的雷诺平均纳维-斯托克斯（RANS）模型，该模型建立在第一性原理的基础上，因此是可推广的。将从颗粒流动动力学理论得到的精确RANS方程进行积分，得到一组耦合的宏观流体动力学方程，然后可以精确地平均。为了关闭方程组，需要近似的项将通过高保真的欧拉-拉格朗日中尺度模拟得到，这种模拟已经被证明可以准确地预测充满颗粒的湍流流动。本奖项由CBET部流体动力学项目颁发，由CISE/ACI部CIF 21软件重用风险基金项目共同资助。

The goal of the proposed study is to develop multiphase turbulence models based on the kinetic theory for granular flows and utilizing high fidelity simulations. Areas of application of these flows include engines, industrial furnaces, fluidized beds for fossil and renewable energy conversion, and chemical plants (chemical reactors, cracking units, etc.), all of which are important for the competitiveness of the U.S. industry. The collaborating PIs propose to offer a full implementation of their models and code into commercial (Ansys Fluent) and public (OpenFOAM) platforms, as well as to make their comprehensive database of detailed simulation results available to the research community.Currently available models for multiphase turbulence are often based on phenomenological approaches requiring fine-tuning with experimental data. These models have limited predictive capability, preventing their adoption in industry or their generalization for research purposes. This proposal, however, aims at developing and validating a consistent Reynolds-averaged Navier-Stokes (RANS) model that is grounded on first principles, and, thus, generalizable. The exact RANS equations, as obtained from Kinetic Theory for granular flow, will be integrated to obtain a set of coupled macroscale hydrodynamic equations, which can then be averaged exactly. The terms that need to be approximated in order to close the system of equations will then be obtained with high fidelity Euler-Lagrange mesoscale simulations, already shown to predict accurately turbulent particle-laden flows.This award by the Fluid Dynamics Program of the CBET Division is co-funded by CIF 21 Software Reuse Venture Fund Program of the CISE/ACI Division.