TU Darmstadt / ULB / TUbiblio

Modeling the two-point correlation of the vector stream function

Oberlack, Martin and Rogers, M. M. and Reynolds, W. C. (1994):
Modeling the two-point correlation of the vector stream function.
In: Center for Turbulence Research, Proceedings of the Summer Program 1994, Stanford University/NASA Ames, pp. 339-352, [Book Section]

Abstract

A new model for the two-point vector stream function correlation has been developed using tensor invariant arguments and evaluated by the comparison of model predictions with DNS data for incompressible homogeneous turbulent shear flow. This two-point vector stream function model correlation can then be used to calculate the two-point velocity correlation function and other quantities useful in turbulence modeling. The model assumes that the two-point vector stream function correlation can be written in terms of the separation vector and a new tensor function that depends only on the magnitude of the separation vector. The model has a single free model coefficient, which has been chosen by comparison with the DNS data. The relative error of the model predictions of the two-point vector stream function correlation is only a few percent for a broad range of the model coefficient. Predictions of the derivatives of this correlation, which are of interest in turbulence modeling, may not be this accurate.

Item Type: Book Section
Erschienen: 1994
Creators: Oberlack, Martin and Rogers, M. M. and Reynolds, W. C.
Title: Modeling the two-point correlation of the vector stream function
Language: English
Abstract:

A new model for the two-point vector stream function correlation has been developed using tensor invariant arguments and evaluated by the comparison of model predictions with DNS data for incompressible homogeneous turbulent shear flow. This two-point vector stream function model correlation can then be used to calculate the two-point velocity correlation function and other quantities useful in turbulence modeling. The model assumes that the two-point vector stream function correlation can be written in terms of the separation vector and a new tensor function that depends only on the magnitude of the separation vector. The model has a single free model coefficient, which has been chosen by comparison with the DNS data. The relative error of the model predictions of the two-point vector stream function correlation is only a few percent for a broad range of the model coefficient. Predictions of the derivatives of this correlation, which are of interest in turbulence modeling, may not be this accurate.

Title of Book: Center for Turbulence Research, Proceedings of the Summer Program 1994
Series Name: Studying Turbulence Using Numerical Simulation Databases
Number: V
Publisher: Stanford University/NASA Ames
Divisions: 16 Department of Mechanical Engineering > Fluid Dynamics (fdy)
16 Department of Mechanical Engineering
Date Deposited: 30 Aug 2011 14:13
Export:

Optionen (nur für Redakteure)

View Item View Item