9.6 Very Large Eddy Simulation
307
is no reason t o use LES and DNS. On the other hand, when RANS does not
work, it may be a good idea to try LES.
One way to use simulation methods in the near term is t o perform LES
and/or DNS of 'building block' flows, ones that are structurally similar t o
those of actual interest. From the results, RANS models that can be applied
to more complex flows can be validated and improved. RANS computations
can then be the everyday tool. LES need be performed only when there are
significant changes in the design.
There have been large eddy and direct numerical simulations of complex
flows. Some have been spectacularly successful while others met with more
limited success. It is important t o determine what kinds of flows LES is
good at and which ones give it problems. Large parts of the roadmap await
completion.
It appears that we have to either use RANS, which is affordable, or LES,
which is more accurate but rather expensive. It is natural t o ask whether
there is a method that provides the advantages of both RANS and LES
while avoiding the disadvantages?
Flows over bluff bodies usually produce strong vortices in their wakes.
These vortices produce fluctuating forces on the body in both the streamwise
and spanwise directions whose prediction is very important. These include
flows over buildings (wind engineering), ocean platforms, and vehicles, among
others. If the vortices are sufficiently larger than the bulk of the motions that
constitute the 'turbulence', it should be possible to construct a filter that
retains the vortices while removing the smaller-scale motions. In so doing, one
may convert an aperiodic flow into a periodic one, which may have significant
consequences.
A method that accomplishes this is called either very large eddy simulation (VLES) or unsteady RANS. In this method, one uses a RANS model but
computes an unsteady flow. The results often contain periodic vortex shedding. When the results of such a simulation are time-averaged, they often
agree better with experiments than steady RANS computations. While there
are questions about the quantitative accuracy of this approach, it certainly
has some merit, at least for the near future. An example of this approach is
the prediction of buoyancy-driven flows by KenjereS (1998).
Finally, we mention a method called detached eddy simulation (DES)
which has been suggested for separated flows (Travin et al., 2000). In this
approach, RANS is used for the attached boundary layer and LES is applied
to the free shear flow resulting from separation. This requires some means of
producing the initial conditions for the LES in the separation region and this
is a difficulty. Only a few simulations of this kind have been made to date
and the results are not yet conclusive.
307
is no reason t o use LES and DNS. On the other hand, when RANS does not
work, it may be a good idea to try LES.
One way to use simulation methods in the near term is t o perform LES
and/or DNS of 'building block' flows, ones that are structurally similar t o
those of actual interest. From the results, RANS models that can be applied
to more complex flows can be validated and improved. RANS computations
can then be the everyday tool. LES need be performed only when there are
significant changes in the design.
There have been large eddy and direct numerical simulations of complex
flows. Some have been spectacularly successful while others met with more
limited success. It is important t o determine what kinds of flows LES is
good at and which ones give it problems. Large parts of the roadmap await
completion.
It appears that we have to either use RANS, which is affordable, or LES,
which is more accurate but rather expensive. It is natural t o ask whether
there is a method that provides the advantages of both RANS and LES
while avoiding the disadvantages?
Flows over bluff bodies usually produce strong vortices in their wakes.
These vortices produce fluctuating forces on the body in both the streamwise
and spanwise directions whose prediction is very important. These include
flows over buildings (wind engineering), ocean platforms, and vehicles, among
others. If the vortices are sufficiently larger than the bulk of the motions that
constitute the 'turbulence', it should be possible to construct a filter that
retains the vortices while removing the smaller-scale motions. In so doing, one
may convert an aperiodic flow into a periodic one, which may have significant
consequences.
A method that accomplishes this is called either very large eddy simulation (VLES) or unsteady RANS. In this method, one uses a RANS model but
computes an unsteady flow. The results often contain periodic vortex shedding. When the results of such a simulation are time-averaged, they often
agree better with experiments than steady RANS computations. While there
are questions about the quantitative accuracy of this approach, it certainly
has some merit, at least for the near future. An example of this approach is
the prediction of buoyancy-driven flows by KenjereS (1998).
Finally, we mention a method called detached eddy simulation (DES)
which has been suggested for separated flows (Travin et al., 2000). In this
approach, RANS is used for the attached boundary layer and LES is applied
to the free shear flow resulting from separation. This requires some means of
producing the initial conditions for the LES in the separation region and this
is a difficulty. Only a few simulations of this kind have been made to date
and the results are not yet conclusive.