Abstract-
First, a brief discussion on the need for bridging models is presented.
Some thoughts on the whats and whatnots of the seemingly ad hoc hybrid
approaches are discussed. Then I present the development of a new
bridging model called the Partially-averaged Navier Stokes (PANS)
method. This model is derived by asking several interesting questions. (i) What is the true velocity decomposition in a hybrid model?
(ii) If RANS represents the closure of fully-averaged statistics, what
is the implied closure of partially averaged statistics? (iii) How does one quantify the level of bridging?
(iv)Where do the bridging and hybrid models belong in the traditional
modeling hierarchy?
Finally, several PANS results from flow past square and circular
cylinders, backward-facing step and three-dimensional cavity are
presented. Exhaustive comparison against experimental data and other
methods are made.
GALCIT Home Page
|
|