Numerical Optimization of Electrothermal Anti-icing and De-icing Systems via Reduced Order Models

Electrothermal ice protection systems (IPS) operating in anti-icing and de-icing modes consume less energy than hot air systems but still need to be optimized for energy usage. The present chapter offers an end-to-end optimization framework in which thermal, structural, aerodynamic and mathematical viewpoints are taken into account in formulating various objective and constraint functions by considering energy consumption, wing skin temperature, thickness, volume, shape and location of the accreted ice on the surface as key parameters affecting the energy usage, the thermal fatigue and the aerodynamic performance. The design variables include the power density, the extent and the activation time of the electric heating blankets. A derivative-free technique called the mesh adaptive direct search (MADS) method is used to carry out the optimization process which would normally need a large number of conjugate heat transfer (CHT) calculations for the IPS simulation. To avoid such prohibitive computations, reduced-order modeling (ROM) is used to construct simplified low-dimensional CHT models. The approach is illustrated through several test cases, in which different combinations of objective and constraint functions, design variables and cycling sequence patterns are examined. In these test cases, the energy consumption is significantly reduced compared to experiments by improving the spatial and temporal distribution of thermal energy usage. The benefits of the approach in combining energy, safety and aerodynamic considerations in designing ice protection systems are demonstrated.