GALCIT Home Page
| 2001-2002 Fluids Seminar Page |
Department of Mechanical and Aerospace Engineering
University of California, San Diego
Abstract-
With the abundance of computational power currently available, and
relatively mature numerical techniques to leverage this power, many
complex, unsteady flow phenomena of engineering interest may now be
modeled accurately by simulation of the Navier-Stokes equation or
filtered versions thereof. Further, a significant expansion of this
capability might be projected over the next several years. Leveraging
this remarkable capability, the time is ripe to make the next major
advance: the control and optimization of flow systems dominated by the
effects of transition and turbulence for desired engineering
objectives, such as drag reduction, mixing enhancement, noise
mitigation, flow stabilization, etc. Such a goal is now feasible
because it is possible, with care, to utilize the fundamental
equations governing fluid flow in a computational setting to compute
directly, or to optimize iteratively, effective control strategies.
The application of model-based control theories to fluid systems,
however, is an extremely delicate matter. Many of these theories have
been developed for low-dimensional ODE system models which simply can
not capture the multi-scale complexity of transitional and turbulent
flows. The present talk will survey recent developments of two
closely related analysis tools at the heart of the noncooperative
control framework we are developing for this class of problems. Both
tools have been benchmarked by our lab on the difficult yet canonical
problem of stabilization of turbulent channel flow at Re_tau=100
by coordinated control of small amounts of zero-net blowing/suction
distributed over the channel walls. Both tools have proven capabable
of completely relaminarizing this highly unstable flow. Extensions of
these proven tools to a variety of other flow control problems are
currently underway, and will be surveyed in this talk, highlighting
the necessary extensions of the control framework currently under
development.
|
GALCIT Home Page
|
|