Numerical Simulation of In-Flight Icing via a Particle-Based Morphogenetic Method

Abstract

Most current numerical ice accretion models are constrained by their reliance on solving continuous partial differential equations for energy, momentum, and mass conservation. This procedure naturally leads to the prediction of continuous, relatively smooth, and compact ice shapes. Unfortunately, natural ice accretion structures are often complex, rough, and disjointed. This limitation of continuous methods can be circumvented by particle-based methods, where the particles mimic the behavior of large ensembles of water molecules as they impact, move, and freeze. This paper describes a computationally efficient particle-based method called morphogenetic modelling. The numerical particle size can be varied for both computational and physical reasons, depending on the need for numerical efficiency and spatial resolution. The particle behavior is driven by physically based, stochastic rules. Since morphogenetic models emulate the time evolution of the accretion shape in a way that reasonably mimics the real physical processes, they can improve on continuous ice accretion models. This improvement is dramatic when simulating rough, non-contiguous, and highly three-dimensional ice structures, which incorporate substructures spanning a wide range of scales. For example, morphogenetic models can predict simultaneous rime and glaze ice accretions, ice accretions with variable density, and discontinuous ice accretions such as rime feathers and lobster tails. Numerous examples are presented here.