Centroidal Voronoi tessellation, leave-one-out cross validation, proper orthogonal decomposition, and multidimensional interpolation are integrated to define a reduced-order modeling approach for the parametric evaluation of steady aerodynamic loads. The proper orthogonal decomposition-based methodology allows reducing the number of degrees of freedom of the problem while maintaining good accuracy for the solution of complex three-dimensional viscous turbulent flows. As a result, it yields fairly accurate solutions at a fraction of the time required by standard computational fluid dynamics approaches. Three-dimensional examples for fixed- and rotary-wing cases of industrial relevance are used to assess the method in the cases of subsonic and transonic flow conditions.

Steady and unsteady three-dimensional viscous turbulent aero-icing simulations are computationally expensive, especially for certification campaigns when broad parametric studies are needed. To overcome the computational effort of such investigations, a Reduced Order Modeling approach, based on Proper Orthogonal Decomposition and Kriging interpolation, is proposed. Using a database of high-fidelity numerical simulations, experimental data, or combinations of both, the proposed technique allows approximating solutions by linear combination of a limited number of eigenfunctions. Bayesian Kriging, a recent variant, is used to obtain the scalar coefficients for the expansion. The accuracy of the proposed method is assessed against reference solutions from two- and three-dimensional aero-icing simulations as well as against experimental data.

SU2-NEMO, a recent extension of the open-source SU2 multiphysics suite’s set of physical models and code architecture, is presented with the aim of introducing its enhanced capabilities in addressing high-enthalpy and high-Mach number flows. This paper discusses the thermal nonequilibrium and finite-rate chemistry models adopted, including a link to the Mutation++ physio-chemical library. Further, the paper discusses how the software architecture has been designed to ensure modularity, incorporating the ability to introduce additional models in an efficient manner. A review of the numerical formulation and the discretization schemes utilized for the convective fluxes is also presented. Several test cases in two- and three-dimensions are examined for validation purposes and to illustrate the performance of the solver in addressing complex nonequilibrium flows.

A conservative Eulerian numerical approach for modeling postimpact Supercooled Large Droplets undergoing splashing and bouncing on aircraft surfaces is presented. The approach introduces the effect of the postimpact droplets by successive solutions of the conservation equations. Two models have been selected to identify the droplet splashing and bouncing conditions, and to provide initial conditions for the reinjected water. The method has been applied to droplet impingement in Supercooled Large Droplet conditions on clean and iced NACA 23012 geometries, as well as the MS(1)-0317 airfoil, and the results have been compared to experimental data. Good agreement is observed for both impingement limits and collection efficiency. Additionally, the method has been applied to a three-element high-lift configuration operating in one of the proposed Appendix O Supercooled Large Droplet environments to demonstrate the danger posed by the re-impingement of splashing and bouncing droplets on complex interacting aerodynamic components.

A framework of local reduced-order modeling using machine learning algorithms is presented together with an approach to optimally select the snapshots for strongly nonlinear problems. By using an unsupervised learning algorithm, solutions are grouped into clusters of similar features. The input parameter space is divided into subregions by decision boundaries based on a supervised learning algorithm. Local reduced-order bases are extracted on each cluster, for which the solutions are represented as a linear combination of the basis vectors from their corresponding subregion. The proposed methodology is employed to conduct a comprehensive, exploration of the in-flight icing certification envelopes.

Numerically predicted roughness distributions obtained in in-flight icing simulations with a beading model are used in a quasi-steady manner to compute ice shapes. This approach, called "Multishot," uses a number of steady flow and droplet solutions for computing short intervals (shots) of the total ice accretion time. The iced geometry, the grid, and the surface roughness distribution are updated after each shot, producing a better match with the unsteady ice accretion phenomena. Comparisons to multishot results with uniform roughness show that the evolution of the surface roughness distribution has a strong influence on the final ice shape. The ice horns that form are longer and thinner compared to constant roughness results. The constant roughness approach usually fails to capture the formation of the pressure side horns and under-predicts the thickness of the ice in this region. With the variable roughness produced by the beading model, the ice shape in general is captured more accurately, better matching the experimental ice shapes. Results are provided for NACA 0012 and SC 1095 airfoils for 2D icing. The flow conditions range between Mach numbers of 0.3 to 0.6, free stream static temperatures of -27°C to -6°C, free stream liquid water contents of 0.5 g/m₃ to 1.4 g/m₃, Reynolds numbers of 2.6 to 4.4 million, and droplet diameter of 20 microns. Total ice accretion times range between 45 seconds to 7 minutes. The FENSAP-ICE icing simulation system is utilized to carry out the CFD-Icing calculations in 2D and 3D.