No description available.

In this thesis, large-eddy simulation has been performed to investigate a heated plane channel flow subjected to streamwise system rotations. A variety of rotation numbers ranging from Roτ = 0 to 15 have been tested in conjunction with two fixed low Reynolds numbers Reτ = 150 and 300. The fundamental characteristics of the resolved velocity and temperature fields in terms of their mean and root-mean-square (RMS) values are investigated. Advanced physical features in terms of the transport of turbulent stresses, turbulent kinetic energy (TKE), heat fluxes and forward and backward scatter of local kinetic energy (KE) fluxes between the resolved and subgrid scales are also studied. Numerical simulations were performed using the conventional dynamic model (DM) and an advanced dynamic nonlinear model (DNM) for closure of the filter momentum equation, and an advanced dynamic full linear tensor thermal diffusivity model (DFLTDM) for closure of the filtered thermal energy equation.

No description available.

No description available.

No description available.

No description available.

Die Zerlegung von Beobachtungsgrößen in sogenannte Mittelwerte und Fluktuationen führt zur Parametrisierung des turbulenten Flusses aber auch zu Problemen. Der Erwartungswert der turbulenten Größe ψ ist das Ensemble Mittel über eine große Zahl von Realisierungen, falls ψ normalverteilt ist. Geophysikalische Daten bestehen jedoch aus Zeitreihen und/oder räumlichen Daten. Daher muß jeder vernünftige Mittelungsprozeß von ψ in der Zeit und/oder im Raum durchgeführt werden. Um die Fluktuationen ψ' von ψ zu trennen, müssen wir den Erwartungswert von ψ bestimmen, d.h. die Fluktuationen in den langen Zeit- und/oder Raumskalen. Für dieses Problem ist der Mittelwert über das Meßintervall eine schlechte Approximation, das gleitende Mittel eine bessere und der numerisch tiefpassgefilterte Wert die bestmögliche Approximation. Eine Fluktuationsmessung (surface flux) im Bereich niedriger Flüsse wurde ausgewertet 1) nach der gewöhnlichen Methode und 2) mit einem numerischen Tiefpass Lanczos-Filter. Mit 2) erhielten wir bessere Ergebnisse. / Decomposition of some observables into so-called mean parts and fluctuations leads to parameterisation of turbulent flow but is also the cause of different problems. The expectation of the turbulent field ψ is, the ensemble mean over a large number of realizations if ψ follows a normal distribution. Geophysical data, however, consist of time- and/or space series. Thus every reasonable averaging process of ψ must be over time and/or space. To separate fluctuations ψ' from ψ we must estimate the expectation value of ψ, i.e. fluctuations on long time and/or space scales. For this problem the mean over the measuring interval is an inexact approximation, the moving mean is better but the numerically low-pass filtered value probably the best possible approximation. A surface flux measurement in low flux regime is evaluated with 1) a usual procedure and 2) with a numerical low-pass Lanczos-filter. With 2) we obtain better results.

No description available.

The purpose of this research is to numerically simulate, analyze, and visualize turbulent flow around rotating aerodynamic shaped 3-dimentional geometries using a custom-made software suite. The computational fluid dynamic program used for this research is called Numerical Wind Tunnel, NWT, which was developed by Dr. J. Militzer and his students over the last 15 years. In order to meet various simulation and prediction requirements of this research, the NWT was modified and improved by implementing many new features; in addition, many bugs have been fixed. Key features added to the NWT include improved boundary layer handling for Detached Eddy Simulation method, new implementations of Surrounding Cell Method and rewritten Lift and Drag Coefficients calculation algorithms, and new approaches to Mesh Refinement and Adaptation Criteria. The improved software is tested extensively by simulating turbulent flows around a rotating National Advisory Committee for Aeronautics (NACA) 0009 airfoil, and test results are compared with both experimental data and previous simulation data. The research was successful mainly because of the much-improved accuracy in predicting static lift and drag coefficients. Another achievement of this research is that the software also successfully predicted various events during an airfoil dynamic stall condition, which is a result of both accurate flow prediction and a NWT feature called Automatic Anisotropic Grid Adaptation.

No description available.