Numerical investigation of a wing hot air ice protection system

Abstract

Aircraft icing is a recurrent aviation safety concern. In the past eight years alone, eight icing accidents involving business jets and other aircraft have occurred. The accumulation of ice on critical aerodynamic surfaces, the primary cause of these accidents, leads to considerable performance degradation that compromises the safety of the passengers, the crew, and the vehicle. A variety of surface-deformation and thermal systems provide icing protection for aircraft. Hot air anti-icing systems are the most common for airplanes with aluminum leading edges on wing and tail surfaces, and engine inlets. These surfaces are heated using bleed air redirected from the jet engine compressor and channeled through a piccolo tube located inside the leading edge. A series of hot air jets emanate from small holes on the piccolo tube (piccolo holes) and impinge on the internal surface of the leading edge skin, transferring heat, and increasing the skin temperature to prevent ice accumulation. The design and optimization of hot air anti-icing systems involve both experimental and numerical studies. Computational Fluid Dynamics (CFD) is a cost-effective analysis tool for bleed air ice protection system design and evaluation. CFD analysis tools, however, require validation against experimental data to determine the accuracy of the numerical schemes, turbulence models, boundary conditions, and results obtained. The present thesis details a CFD methodology developed to simulate the performance of a wing hot air anti-icing system under dry air conditions (no water impingement). Computational simulations were conducted with the commercial CFD code FLUENT to investigate the performance of a hot air anti-icing system installed in the leading edge of a 72-inch span, 60-inch chord business jet wing model. The analysis was performed with a full-span model (FSM) and a partial-span model (PSM). The FSM was used to model the entire length of the piccolo tube to investigate the development of spanwise flow inside the piccolo tube. The PSM was used to model a 2.44-in spanwise section of the wing in order to investigate the internal and external flow properties about the wing with the bleed air system in operation. Computational results obtained with the PSM model were compared with experimental data obtained from icing tests performed at the NASA Glenn Icing Research Tunnel (IRT) facility. The work presented in this thesis includes extensive 2D axisymmetric computational studies performed with a subsonic, heated, turbulent jet impinging on a flat plate to evaluate the performance of five eddy-viscosity turbulence models available in the FLUENT code. The turbulence model studies showed that the Shear Stress Transport (SST) ? -? formulation provided the most consistent prediction of recovery temperatures at the impingement wall. Grid resolution and spatial discretization studies were completed with a three-dimensional version of the jet impingement scenario employed in the turbulence study, and first- and second-order upwind schemes. Three grid resolution levels were considered based on the number of nodes distributed around the nozzle exit circumference in order to apply the same distribution around the piccolo holes circumferences in the anti-icing system PSM. A boundary condition study was performed with the anti-icing models (FSM and PSM). The PSM did not model the piccolo tube internal flow and, consequently, required inflow boundary conditions to be specified at the piccolo holes' exits. The FSM was employed to analyze the flow inside the piccolo tube and to obtain the inflow boundary conditions for the PSM. The approaches applied to extract the boundary conditions were centerline and cell-averaged. Skin temperature results from the PSM were compared with available experimental data and showed that the cell-averaged approach provided the most accurate simulation. Finally, a parametric study was conducted with the anti-icing models (FSM and PSM) to validate the computational methodology with a broad range of cases with variable internal and external flow parameters for which experimental data was available. The results for leading-edge skin temperature as well as piccolo flow properties demonstrated in all cases high-fidelity agreement with experimental data.