Comparison of thermodynamic models for ice accretion on airfoils

Abstract

Purpose This paper aims to assess the strengths and weaknesses of four thermodynamic models used in aircraft icing simulations to orient the development or the choice of an improved thermodynamic model. Design/methodology/approach Four models are compared to assess their capabilities: Messinger, iterative Messinger, extended Messinger and shallow water icing models. They have been implemented in the aero-icing framework, NSCODE-ICE, under development at Polytechnique Montreal since 2012. Comparison is performed over typical rime and glaze ice cases. Furthermore, a manufactured geometry with multiple recirculation zones is proposed as a benchmark test to assess the efficiency in runback water modeling and geometry evolution. Findings The comparison shows that one of the main differences is the runback water modeling. Runback modeling based on the location of the stagnation point fails to capture the water film behavior in the presence of recirculation zones on airfoils. However, runback modeling based on air shear stress is more suitable in this situation and can also handle water accumulation while the other models cannot. Also, accounting for the conduction through the ice layer is found to have a great impact on the final ice shape as it increases the overall freezing fraction. Originality/value This paper helps visualize the effect of different thermodynamic models implemented in the same aero-icing framework. Also, the use of a complex manufactured geometry highlights weaknesses not normally noticeable with classic ice accretion simulations. To help with the visualization, the ice shape is presented with the water layer, which is not shown on typical icing results.

A Preliminary Approach towards Rotor Icing Modeling Using the Unsteady Vortex Lattice Method

“…Many methods have been developed for this approach, most notably the Messinger Model, Extended Messinger Model, Iterative Messinger Model, or Shallow Water Icing Model (SWIM). A comparative assessment of these models can be found in [25]. Finally, for the conjugate heat transfer solver, the effect of the internal heat generated by a thermal IPS is transferred to the mass and energy balance equations by a conductive heat transfer term.…”

“…Together with the steady-state icing approach, these assumptions neglect the effects of surface roughness as well as the water and ice thickness on the airfoil surface. While these assumptions have been used in past works and provided acceptable results [25,30], they are known to produce inaccuracies and errors in the predicted ice shape. For the simplified nature of this work, and since the icing analysis is only focused on the stagnation point, the results may be acceptable.…”

Immersed Boundary Methodology for Multistep Ice Accretion Using a Level Set

“…This manufactured geometry was first presented in [21] and is generated from a NACA0012 airfoil with added artificial ice near the leading edge. The three-horn configuration was selected to obtain multiple flow recirculation zones and create a difficult situation for the ice growth solver because of the presence of highly concave and convex features.…”

Sensitivity Study of Ice Accretion Simulation to Roughness Thermal Correction Model

“…Once the heat transfer along the flat plate is known, the ice accretion simulations can be performed. To compute the 2D ice accretions, the Messinger approach is used [6,7]. The basis of this approach is to compute the ice growth rate in a control volume using a mass and an energy balance.…”

“…Most of the icing codes used to predict the ice accretion shape are based on the Messinger model [6]. The model uses a mass and energy balance involving the convective heat transfer to compute the ice growth on the airplane surface [7]. Past studies have shown that the convective heat transfer, modeled by a convective heat transfer coefficient, has a major impact on the final shape of the ice accretion [8,9].…”