The adjoint method allows the computation of sensitivities (derivatives) of flow variables or quantities derived from them with respect to geometrical parameters. This is more efficient than the direct computation of derivatives from flow solutions, e.g. by finite differences, if the number of parameters is large compared to the number of objective functionals.
The increasing model accuracy to describe physical processes like aerodynamics, aeroacoustics, aerothermodynamics, aeroelastics or turbulence as well as their interaction leads to an ever growing complexity of the simulated configurations. For reduction of the required simulation duration different acceleration methods are used.
Essential physical phenomena occur in the simulation of flow in turbomachinery often with specific frequencies, e.g. at multiples of the rotation frequency or at the eigenfrequencies of blades. In these cases, it is sufficient to calculate the solution of the flow equations for some frequencies only. This is achieved by transforming the flow equations into the frequency domain and solving these equations for selected frequencies. Thus the computation time is reduced by one to two magnitudes compared to a nonlinear, unsteady calculation. In TRACE two types of frequency domain methods are implemented: in a classical time-linearized Navier-Stokes method and a harmonic-balance method which takes into account a set of frequency and their non-linear coupling.
The TRACE suite provides an in-house, industrial prove process from the grid generation to the post-processing. Amongst others this includes a tool for the mesh generation, a pre- and post-processing-tool. The in-house tool G3DHEXA generates structured meshes for turbomachine components. PREP is an aeroelastic pre-processor. After the computation of the flow solution the post-processing tool POST is used to analyze the flow field e. g. with respect to turbomachine performance data.
Virtually all turbomachinery flows involve turbulent phenomena, including transition from laminar to turbulent states. Thus, the adequate representation of transition and turbulence is crucial for the predictive accuracy of thecomputational method used in the turbomachinery design process.
The Combustion Group has been created in 2009 in the Numerical Methods Department with the objective of expanding the multi-physics capabilities of TRACE and carry out advanced research in the development and implementation of state-of-the-art combustion models for RANS/LES simulations in multi-component configurations, in close cooperation with industrial and academic partners.
In turbomachinery powerful numerical methods have been developed and used successfully to simulate the flow. However, these classical first- and second-order accurate algorithms for spatial discretization are inefficient for flow problems with complex physics and geometry. In particular, such applications as computational aeroacoustics (CAA) or turbulent combustion problems require a more accurate prediction of turbulent phenomena than what is attainable with second-order RANS simulations.
The understanding of aero- and thermodynamics of the turbine is of essential relevance for the design of blade cooling methods. A state the art heat transfer modelling is under development in TRACE to improve the prediction accuracy of the behaviour of thermal and momentum boundary layers.