The objective of this work is to evaluate the capability of different combinations of a turbulence model and a Lagrangian particle tracking (LPT) model integrating a particle-wall interaction (PWI) model to predict particle-laden flow in 90-deg bends, as well as the impact of the PWI model on the prediction of the referred flow. The experimental data from Kliafas and Holt (1987) (LDV measurements of a turbulent air-solid two-phase flow in a 90° bend. Experiments in Fluids, 5: 73-85) concerning a vertical to horizontal square-sectioned duct with a hydraulic diameter of 0.1 m that are connected by a 90-deg bend with a curvature ratio of 3.52, served as the benchmark for the aimed analysis. Air with glass spheres of 50 μm diameter flows in the experimental duct system with a Reynolds number of 3.47×10⁵. The airflow was modelled by four different turbulence models: a low Reynolds number k-ε model, the SST k-ω model, the v²-f model, and the RSM SSG model. The particle-phase was modelled by a LPT formulation, and the particle-wall interaction was calculated using four different models: Brauer, Grant & Tabakoff, Matsumoto & Saito and Brach & Dunn PWI models. The 3D simulation results of mean streamwise velocities from the sixteen RANS-LPT/PWI combinations were compared qualitatively and quantitatively to experimental and numerical data available in the literature. The four turbulence models produced errors for the gas-phase in the order of 8%. Concerning the particle-phase, the errors produced by all RANS-LPT/PWI combinations were below 4% for bend angles up to 15° and up to 18% for bend angles higher than 30°. The best results for the particle-phase were obtained with the SST k-ω and v²-f model combined with the LPT/Brauer or LPT/Brach & Dunn PWI models, which produced errors inferior to 14%.

中文翻译：

---

关于使用粒子-壁相互作用模型预测90度弯头中充满粒子的流动

这项工作的目的是评估湍流模型和拉格朗日粒子跟踪（LPT）模型的不同组合的能力，该模型结合了粒子-壁相互作用（PWI）模型来预测90度弯头中充满粒子的流动PWI模型对参考流量预测的影响。来自Kliafas和Holt（1987）的实验数据（LDV测量90°弯角处的湍流气固两相流动。在Fluids中的实验，5：73-85）涉及具有水力直径为0.1 m，通过90度弯曲（曲率比为3.52）连接起来，作为目标分析的基准。直径为50μm的玻璃球的空气在实验风道系统中流动，雷诺数为3.47×10⁵。通过四种不同的湍流模型对气流进行建模：低雷诺数k - ε模型，SST k - ω模型，v ² - f模型，以及RSM SSG模型。通过LPT公式对颗粒相进行建模，并使用四种不同模型（Brauer，Grant和Tabakoff，Matsumoto＆Saito和Brach＆Dunn PWI模型）计算颗粒-壁相互作用。将16种RANS-LPT / PWI组合的平均水流速度的3D模拟结果定性和定量地与文献中的实验和数值数据进行了比较。四个湍流模型产生的气相误差约为8％。关于颗粒相，所有RANS-LPT / PWI组合所产生的误差在最大15°的弯曲角度下均低于4％，而在高于30°的弯曲角度下则高达18％。使用SST k - ω和v可获得最佳的颗粒相结果² - f模型与LPT / Brauer或LPT / Brach＆Dunn PWI模型结合使用，产生的误差低于14％。