Modeling of Ice Accretion over Aircraft Wings Using a Compressible OpenFOAM Solver

Abstract

A method to simulate ice accretion on an aircraft wing using a three-dimensional compressible Navier-Stokes solver, a Eulerian droplet flow field model, a mesh morphing model, and a thermodynamic model, is presented in this paper. The above models are combined together into one solver and implemented in OpenFOAM. Two-way coupling is achieved between airflow field calculation and ice simulation. The density-based solver rhoEnergyFoam is used to calculate the airflow field. The roughness wall function is proposed to simulate the roughness effect caused by ice accretion. For droplet flow field calculation, the Eulerian model is applied and the permeable wall boundary condition is used on the wing to simulate the droplet impingement. The icing thermodynamic model is built based on the Messinger model. The mesh morphing model adjusts the wing’s shape every time step based on the amount of accreted ice so that the airflow field is updated during the simulation. The effect of the ice accretion on the airflow is studied by comparing the aerodynamic performance—with and without ice. The ice accretion on the ONERA M6 wing model under a specific condition has been simulated to validate the solver’s performance and investigate the effect of the accreted ice on the aerodynamic performance.

Fast Evaluation of Aircraft Icing Severity Using Machine Learning Based on XGBoost

“…Cao et al [10,11] developed mathematical models to simulate the ice accretion on three-dimensional bodies directly. Li and Paoli [12] developed the Icing solver based on OpenFOAM framework [13] with two-way coupling between air flow and ice accretion. A simulation tool was developed by Lampton and Valasek [14] to evaluate icing severity and its effect on aircraft flight performance.…”

“…The development of computational methods for aircraft icing was motivated on one hand by the progress of theoretical icing models incorporating richer and/or more complex physics, and on the other hand by the need of limiting the high cost of carrying out expensive experimental campaigns. From a computational perspective, however, in order to obtain qualified icing results numerically, fine grids and regularly remeshing to update the air flow field [12] are necessary, which requires tremendous computational resources.…”

“…The present study is a first step toward the development of a data-driven approach to predict aircraft icing severity by exploring the complex pattern between LWC, droplet diameter and exposure time. This work is motivated by the fact that numerical simulations generally require computationally expensive and/or cumbersome treatments to calculate the ice displacement and accretion along the wing, such as remeshing [12,20] or other sophisticated techniques. Data-driven methods can help alleviate this constraints once trained, especially if the objective is to extract general yet critical ice features such as the ice global coverage or maximum ice thickness rather than determining the full flow-field and ice distribution as in CFD simulations.…”

“…To the best of our knowledge, this current study presents the first application of XGBoost in aircraft icing prediction. The model is trained and evaluated on a database of available flow data that include simulation data from Reynolds-averaged Navier-Stokes (RANS) models, partly obtained from previous simulations [12] and partly generated here; and new data from high-fidelity delayed detached eddy simulations (DDES) [25,26].…”

Scalability of OpenFOAM Density-Based Solver with Runge–Kutta Temporal Discretization Scheme

“…erefore, there is a growing interest in the CFD community to develop and implement higher-order methods in OpenFOAM for transient flow calculations. ese include, for example, numerical algorithms for DNS/ LES of incompressible flows [2], compressible flows [3][4][5], and reacting flows [6].…”