1Concordia University, Montreal, Quebec, Canada
2Universita degli Studi di Udine, Udine, Italy
Aircraft taking off from or landing at airports and/or flying through clouds under adverse conditions may encounter the problem of ``in-flight'' ice formation on aircraft surfaces and critical engine components. Aircraft and system design and operational procedures still have not totally conquered the problem: aircraft operations in icing conditions continue to result in incidents and accidents with no aircraft type, size or configuration being immune.
To optimally design and protect aerodynamic systems prone to icing requires techniques that can accurately predict ice accretion on the entire aircraft under all atmospheric conditions, enabling designers to predict droplet trajectories, limits of impingement, ice accretion, anti- and de-icing heat loads, melted ice runback, as well as accurately assess how the iced aircraft's performance characteristics are affected.
In the above process, experimental and empirical techniques are constantly complemented with computational methods. An assessment of the state-of-the-art in this field rapidly shows that the CFD techniques used are disparate. For example, a Panel method is usually used to calculate the flow, followed by a Lagrangian approach to calculate droplets trajectories, making ice impingement calculations on multi-component airfoils a tedious sequential chore and limited to low Mach numbers. Ice accretion is carried out via control volume methods, with heat transfer calculated separately for the ice accretion and for the internal flow inside hot wings. Performance deterioration is then calculated using viscous-inviscid interaction codes or full Navier-Stokes codes. It goes without saying that a fresh approach is needed since advances in CFD make it now possible to contemplate more unified approaches, not feasible or imaginable only a decade ago.
The Concordia CFD Lab is leading an initiative intended to regroup in an integrated Consortium, right from the conceptual level, most entities concerned with the icing problem such as meteorologists, aircraft and engine manufacturers, flight simulation companies, airline operators, government entities and airline pilots. This Consortium acts as an advisory body for the development of a single-code ``numerical simulator'' to improve and optimize designs, reduce testing, accelerate certification, investigate situations difficult to reproduce and provide a more realistic training tool for pilots. This co-operative effort has thus launched the development of a three-dimensional Full Navier-Stokes-based icing code, FENSAP-ICE (Finite Element Navier-Stokes Analysis Package), with applications ranging from wings, to multi-component wings, to nacelle cowl sections, providing the following capabilities:
1. An integrated Eulerian approach to simultaneously solve the flow field and the droplets concentration (rather than trajectories). This eliminates the need for Panel methods and Lagrangian tracking and removes their limitations.}
2. A full-blown prediction of ice accretion, followed by modeling of the airflow over the shapes resulting from this accretion. This necessitates the development of moving grids techniques using an Arbitrary Lagrangian Eulerian method, in which surface grids distort with the evolving body shape, and field grid points move with a velocity dictated by the speed of body shape change, without the need for remeshing or interpolation,
3. Extend these capabilities to multi-element wings with deployed flaps and slats, with ice accretion on all these components calculated in one shot. This is greatly enhanced by a new mesh adaptation technology that captures the details of the recirculation regions and wakes of such flows,
4. Development of transition and turbulence models for small roughness (glaze ice), large roughness (rime ice and runback) and for components operating in the wake of other components (multi-element airfoils),
5. Development of a conjugate heat transfer model to simultaneously solve the heat transfer between the flow inside a hot wing (piccolo tube) and the external flow, through the aircraft skin. This would lead to the optimization of bleed air to prevent ice accretion can then be determined;
6. Development of an aerodynamics database usable for a more realistic training simulator.
The presentation will highlight the progress in the development of FENSAP-ICE such as an Eulerian droplets impingement algorithm coupled to an Euler or Navier-Stokes solver, flow in piccolo tubes, advances in the turbulence modeling of roughness, anisotropic adaptive and moving grids for multi-element airfoils, and prediction of performance degradation. The important generic implications of mesh adaptation on CFD will be highlighted as they are leading to the ultimate goal of mesh-independent, user-independent and solver-independent CFD.