Simulation Of Turbulent Boundary Layer Research Articles

Investigation of the turbulent model for pressure fluctuations with spectral theory

The pressure fluctuations in turbulent shear flows are investigated with the theory of spectral analysis. An expression for pressure spectra is analytically derived in terms of velocity spectra. This derivation is based on a formal solution of the Navier-Stokes equation and quasi-normal assumption to express the third and fourth order velocity correlations in terms of double velocity correlation. Then, a turbulent model for the computation of pressure fluctuation intensity with Renolds stress and mean flow velocity gradients is established. The turbulent constants in the model are calculated from the assumptions about the general behaviour of velocity spectra in high Renolds number flows. Comparison with direct simulation of turbulent boundary layer is made. It is found that the turbulent-turbulent, cross correlation, and turbulent-shear source terms for mean square value of pressure fluctuation are about the same magnitude.

Numerical simulation of the dynamics of turbulent boundary layers: perspectives of a transition simulator

The current and prospective capabilities of numerical simulations of turbulent boundary layers are discussed. The stringent resolution requirements for resolving instantaneous structures and dynamics (rather than just for producing statistics) are emphasized. Improvements and alternatives to the prevailing simulation methodology are proposed.

The simulation of offshore turbulent dispersion using seeded eddies

A way of representing turbulence in a two-dimensional situation is introduced appropriate to depth-independent offshore fluid mechanics. The turbulence is simulated by a collection of eddies, each of which has an analytically simple form but whose size, strength and position is governed by stochastically assigned variables. The problem addressed here is how contaminant is dispersed in such an eddy field. A number of experiments are performed whereby the eddies are seeded with marked particles that move with the fluid. The variance of these particles is monitored as time varies, and the results are compared with an assumed power law distribution. Although not a perfect fit, the results are in general accord with a power law with index between 1.5 and 2.5, which is in agreement with the observed power law of 2.34 due to Okubo, and a marked improvement on random walk models which give a variance directly proportional to time. Some further applications of this technique are discussed, namely the simulation of turbulent boundary layers and the simulation of the cascade of energy up turbulent length scales.

Wind loads on an ancient church spire

Wind tunnel tests have been carried out on models of an ancient church spire which had deteriorated. Dynamic pressures on the surface of the octagonal spire were measured on two models, 1:250 scaled full model of the church with turbulent boundary layer simulation, 1:50 scaled spire sectional model exposed to uniform turbulent flow. Total base overturning moment of the spire was also measured by an aerodynamic balance. The results were used for the reinforcement calculations for the church spire during restoration.

Simulation of flat-plate turbulent boundary layers in cryogenic tunnels

The magnitudes of real-gas effects on flat-plate turbulent boundary layer simulations in a cryogenic nitrogen wind tunnel are investigated in order to determine the validity of the method used by Inger (1979) to estimate real-gas effects. Boundary layer solutions for real gases, ideal gases with a specific heat ratio of 1.6 and ideal diatomic gases (specific heat ratio 1.4) were obtained for the worst case conditions of maximum stagnation pressure (9 atm), minimum stagnation temperature (120 K) and Mach number of 1.2. Calculated boundary layer parameters such as friction coefficient and displacement thickness are shown to agree closely for the real gas and the ideal diatomic gas (specific heat ratio 1.4), while the ideal gas solution used by Inger is shown to differ from the real-gas values considerably. Results indicate that real-gas effects on a flat-plate turbulent boundary layer simulation in a cryogenic nitrogen tunnel are insignificant, and suggest the unlikelihood of the large real-gas effects reported by Inger for turbulent boundary layer shock interactions.