A numerical study of propagation of cohesive fluid mud gravity currents in the form of lock-exchange was done using the OpenFOAM open source toolbox. An Eulerian approach solution for three separate phases was developed by incorporating a rheological model to predict the front position of cohesive fluid mud gravity currents. The model also simulates features in the complete movement phases including slumping, self-similar, and viscous in which the dynamics of propagation are affected by the balance of viscous and buoyancy forces, and the inertia force is negligible. The influence of using different turbulence models containing sub-grid scale (SGS), modified SGS, detached eddy simulation (DES), delayed-detached eddy simulation (DDES), Launder-R eece-Rodi (LRR), and k-ɛ models on the accuracy of simulation results was evaluated by comparing with available experimental data. The results show that the selection of the proper turbulence model is one of the most important issues for this type of the numerical modeling. The more efficient turbulence model was suggested and tabulated for each stage of propagation and different selected concentrations of 1,045, 1,140, and 1,214 g/L. Although different turbulence models (except k-ɛ) lead to front propagation dynamic simulation results that are in good agreement with the experimental measurements in the early stage of propagation for low concentrations, only DES, SGS, and modified SGS are able to capture the Kelvin-Helmholtz instability vortex shapes at the dense fluid interface, which is the main characteristic of the gravity current through the slumping phase. The calculated accuracies of SGS and modified SGS in predicting gravity current propagation for the both self-similar and viscous phases also are slightly better than DES, DDES, and LRR model results. The results of this study confirmed the performance and efficiency of the modified SGS model in which the interaction coefficients between phases are calibrated for the numerical modeling of fluid mud gravity current propagation.

使用OpenFOAM开源工具箱对粘性流体泥浆重力流以闭锁交换的形式进行了数值研究。通过合并流变模型来预测粘性流体泥浆重力流的前部位置，开发出了三个独立阶段的欧拉方法解决方案。该模型还模拟了整个运动阶段的特征，包括下陷，自相似和粘滞，其中传播的动力学受粘滞力和浮力的平衡影响，而惯性力可以忽略不计。使用不同湍流模型的影响包括子网格比例（SGS），改进的SGS，分离涡模拟（DES），延迟分离涡模拟（DDES），Launder-R eece-Rodi（LRR），通过与现有实验数据进行比较，评估了k- and模型对模拟结果的准确性。结果表明，正确的湍流模型的选择是这类数值模型最重要的问题之一。提出了更有效的湍流模型，并针对传播的每个阶段以及不同的选定浓度1,045、1,140和1,214 g / L列出了湍流模型。尽管不同的湍流模型（k-ɛ除外）导致前传播动态模拟结果与低浓度传播早期阶段的实验测量结果非常吻合，但只有DES，SGS和改良的SGS能够捕获开尔文-致密流体界面处的亥姆霍兹不稳定性涡旋形状，这是流经塌陷相的重力流的主要特征。在预测自相似相和粘性相的重力电流传播时，SGS和修正SGS的计算精度也略好于DES，DDES和LRR模型结果。这项研究的结果证实了改进的SGS模型的性能和效率，其中校正了相之间的相互作用系数以用于流体泥浆重力流传播的数值模拟。