"The estimation of integral time scales in unsteady flows"
by
Lewalle, J., Ashpis, D.E., Halstead, D.E. and Wisler, D.C.
ABSTRACT
Time scales in a turbulent flow provide a link between
experimental data and modeling, both in terms of physical
content and for quantitative assessment. The problem of
interest here is the definition of time scales in an unsteady flow.
Using a GE turbomachinery dataset with wake-induced transition,
we document several methods to extract dominant frequencies,
and compare the results. We show that
conventional methods of time scale evaluation
(based on autocorrelation functions and
on Fourier spectra) and wavelet-based methods provide similar
information when applied to stationary signals. We also
show the greater flexibility of the wavelet-based methods when
dealing with intermittent or strongly modulated data, as are
encountered in transitioning boundary layers and in flows with
unsteady forcing associated with wake passing. We define phase-averaged
dominant frequencies that characterize the turbulence associated
with freestream conditions and with the passing wakes
downstream of a rotor.
The relevance of these results for modeling is discussed in
the paper.
to appear, NASA-TM-2000-209452, 2000.
Jacques Lewalle, jlewalle@syr.edu
