Quasi-Steady Convergence of Multistep Navier–Stokes Icing Simulations

Abstract

A newly developed two-dimensional ice accretion and antiicing simulation code, CANICE2D-NS, is presented. The method is used to predict iced airfoil shapes and performance degradation with a multistep approach. A multiblock Navier-Stokes code, NSMB, has been coupled with the CANICE2D icing framework, supplementing the existing panel method-based flow solver. Attention is paid to the roughness iniplementation within the turbulence model and to the convergence of the steady and quasi-steady iterative procedure. The new conpling allows fully automated multilayer icing simulation, whereas also permitting flow analysis and performance prediction of iced airfoils. Effects of uniform surface roughness in quasi-steady ice accretion simulation are analyzed through different validation test cases. The results demonstrates the benefits and robustness of the new framework in predicting ice shapes and aerodynamic performance parameters, as well as iced airfoil surface pressure coefficients. Finally, the convergence of the quasi-steady algorithm is verified and identifies the need for an order of magnitude increase in the number of multitime steps in icing simulations. Nomenclaturethe nearest wall, m í/() = offset in the wall distance to account for wall roughness, m /(,., = turbulent convective heat transfer coefficient, W/m-•K K¡ = equivalent sand grain roughness height normalized hy chord k = von Kármán constant k^ = equivalent sand grain roughness height, m LWC = liquid water content, kg/m' MED = mean droplet diameter, m Pr, = turbulent Prandtl number 5? = Stanton number Stf. = roughness Stanton number S = transformed vorticity u = longitudinal velocity component, m/s Ug = boundary-layer edge velocity, m/s «J = friction velocity, m/s 1/ = kinematic viscosity, m^/s t/, = turbulent eddy viscosity, m-/s D = transported quantity in the S-A model w = wall value y = distance along the wall normal, m y+ = normalized wall distance •p = density, kg/m-

Experimental characterization of water droplet deformation and breakup in the vicinity of a moving airfoil

“…This has been explicitly argued by Aliaga et al [6] in the introduction section of their article where they write: In this context, a number of numerical (CFD) approaches have been recently published by Aliaga et al [6], Ghenai and Lin [7], Hasanzadeh et al [8], Honsek et al [9], Iuliano et al [10], Jung and Myong [11], Nakakita et al [12], Rendall and Allen [13], and Zeppetelli and Habashi [14]. In practice, these numerical approaches involve an extremely high degree of modeling complexity since multiphase compressible turbulent flows with changing geometries (the ice accretion effect) are considered.…”

“…Ghenai and Lin [7] prescribed the number of particle trajectories, density, size and distribution of water droplets. Hasanzadeh et al [8] solved the equation of motion of the (spherical) droplets using a Runge-Kutta type scheme after estimating their drag from the previously computed flow field solution. Iuliano et al [10] followed a similar approach and neglect the interaction of the particle liquid phase on the gas phase (one way interaction).…”

Water droplet deformation and breakup in the vicinity of the leading edge of an incoming airfoil

“…In this context, a number of numerical (CFD) approaches have been recently published by Aliaga et al [3], Ghenai and Lin [28], Hasanzadeh et al [35], Honsek et al [38], Iuliano et al [45], Jung and Myong [47], Nakakita et al [64], Rendall and Allen [80], and Zeppetelli and Habashi [103].…”

“…For instance, this data could be of interest for researchers developing numerical algorithms that simulate aircraft icing conditions [3,35,45,47,80].…”

“…Ghenai and Lin [28] prescribed the number of particle trajectories, density, size and distribution of water droplets. Hasanzadeh et al [35] solved the equation of motion of the (spherical) droplets using a Runge-Kutta type scheme after estimating their drag from the previously computed flow field solution. Iuliano et al [45] followed a similar approach and neglected the interaction of the particle liquid phase on the gas phase (one way interaction).…”