
Two different types of perturbations of the Lorenz 63 dynamical system for
RayleighBenard convection by multiplicative noise  called stochastic
advection by Lie transport (SALT) noise and fluctuationdissipation (FD) noise
 are found to produce qualitatively different effects, possibly because the
total phasespace volume contraction rates are different. In the process of
making this comparison between effects of SALT and FD noise on the Lorenz 63
system, a stochastic version of a robust deterministic numerical algorithm for
obtaining the individual numerical Lyapunov exponents was developed. With this
stochastic version of the algorithm, the value of the sum of the Lyapunov
exponents for the FD noise was found to differ significantly from the value of
the deterministic Lorenz 63 system, whereas the SALT noise preserves the Lorenz
63 value with high accuracy. The Lagrangian averaged version of the SALT
equations (LA SALT) is found to yield a closed deterministic subsystem for the
expected solutions which is found to be isomorphic to the original Lorenz 63
dynamical system. The solutions of the closed chaotic subsystem, in turn, drive
a linear stochastic system for the fluctuations of the LA SALT solutions around
their expected values.

Computational codes for direct numerical simulations of RayleighB\'enard
(RB) convection are compared in terms of computational cost and quality of the
solution. As a benchmark case, RB convection at $Ra=10^8$ and $Pr=1$ in a
periodic domain, in cubic and cylindrical containers is considered. A dedicated
secondorder finitedifference code (AFID/RBflow) and a specialized
fourthorder finitevolume code (Goldfish) are compared with a general purpose
finitevolume approach (OpenFOAM) and a general purpose spectralelement code
(Nek5000). Reassuringly, all codes provide predictions of the average heat
transfer that converge to the same values. The computational costs, however,
are found to differ considerably. The specialized codes AFID/RBflow and
Goldfish are found to excel in efficiency, outperforming the general purpose
flow solvers Nek5000 and OpenFOAM by an order of magnitude with an error on the
Nusselt number $Nu$ below $5\%$. However, we find that $Nu$ alone is not
sufficient to assess the quality of the numerical results: in fact,
instantaneous snapshots of the temperature field from a near wall region
obtained for deliberately underresolved simulations using Nek5000 clearly
indicate inadequate flow resolution even when $Nu$ is converged. Overall,
dedicated special purpose codes for RB convection are found to be more
efficient than general purpose codes.

Rotation modulates turbulence causing columnar structuring of a turbulent
flow in case of sufficiently strong rotation. This yields significant changes
in the flow characteristics and dispersion properties, which makes rotational
turbulence modulation particularly relevant in the context of atmospheric and
oceanic flows. Here we investigate the canonical flow of turbulence in a
periodic box, subjected to rotation about a fixed vertical axis. As point of
reference we consider direct numerical simulations of homogeneous isotropic
turbulence. Modulation due to rotation at various rotation rates (i.e.,
different Rossby numbers) is investigated. Special attention is paid to the
alteration of intermittency, which is measured in terms of changes in the
scaling of the structure functions. A reduction of intermittency quantified
with the longitudinal structure functions in the direction perpendicular to the
rotation axes will be presented. These numerical findings correspond well to
recent results obtained in experiments by Seiwert et al. (2008).

A new computational framework for the simulation of turbulent flow through
complex objects and along irregular boundaries is presented. This is motivated
by the application of metal foams in compact heattransfer devices, or as
catalyst substrates in processengineering. The flowconsequences of such
complicated objects are incorporated by adding explicit multiscale forcing to
the NavierStokes equations. The forcing represents the simultaneous agitation
of a wide spectrum of lengthscales when flow passes through the complex
object. It is found that a considerable modulation of the traditional energy
cascading can be introduced with a specific forcing strategy. In spectral
space, forcing yields strongly localized deviations from the common Kolmogorov
scaling law, directly associated with the explicitly forced scales. In
addition, the accumulated effect of forcing induces a significant nonlocal
alteration of the kinetic energy including the spectrum for the large scales.
Consequently, a manipulation of turbulent flow can be achieved over an extended
range, well beyond the directly forced scales. The turbulent mixing of a
passive scalar field is also investigated, in order to quantify the
physicalspace modifications of transport processes in multiscale forced
turbulence. The surfacearea and wrinkling of levelsets of the scalar field
are monitored as measures of the influence of explicit forcing on the local and
global mixing efficiency.

The response of turbulent flow to timemodulated forcing is studied by direct
numerical simulations of the NavierStokes equations. The largescale forcing
is modulated via periodic energy input variations at frequency $\omega$. The
response is maximal for frequencies in the range of the inverse of the large
eddy turnover time, confirming the meanfield predictions of von der Heydt,
Grossmann and Lohse (Phys. Rev. E 67, 046308 (2003)). In accordance with the
theory the response maximum shows only a small dependence on the Reynolds
number and is also quite insensitive to the particular flowquantity that is
monitored, e.g., kinetic energy, dissipationrate, or TaylorReynolds number.
At sufficiently high frequencies the amplitude of the kinetic energy response
decreases as $1/\omega$. For frequencies beyond the range of maximal response,
a significant change in phaseshift relative to the timemodulated forcing is
observed.

Mathematical regularisation of the nonlinear terms in the NavierStokes
equations provides a systematic approach to deriving subgrid closures for
numerical simulations of turbulent flow. By construction, these subgrid
closures imply existence and uniqueness of strong solutions to the
corresponding modelled system of equations. We will consider the large eddy
interpretation of two such mathematical regularisation principles, i.e., Leray
and LANS$\alpha$ regularisation. The Leray principle introduces a {\bfi
smoothed transport velocity} as part of the regularised convective
nonlinearity. The LANS$\alpha$ principle extends the Leray formulation in a
natural way in which a {\bfi filtered Kelvin circulation theorem},
incorporating the smoothed transport velocity, is explicitly satisfied. These
regularisation principles give rise to implied subgrid closures which will be
applied in large eddy simulation of turbulent mixing. Comparison with filtered
direct numerical simulation data, and with predictions obtained from popular
dynamic eddyviscosity modelling, shows that these mathematical regularisation
models are considerably more accurate, at a lower computational cost.

A new formulation is derived for the commutatorerrors in largeeddy
simulation of incompressible flow. These commutatorerrors arise from the
application of nonuniform filters to the NavierStokes equations. As a
consequence, the filtered velocity field is no longer solenoidal. The order of
magnitude of the commutatorerrors is compared with the divergence of the
turbulent stresses. This shows that one can not reduce the size of the
commutatorerrors independently of the turbulent stress terms by some judicious
construction of the filter operator. Similarity modeling for the
commutatorerrors is presented, including an extension of Bardina's approach
and the application of Leray regularization. The performance of the
commutatorerror parameterization is illustrated with the onedimensional
Burgers equation. For large filterwidth variations the Leray approach is shown
to capture the filtered flow with better accuracy than is possible with
Bardina's approach.

A new modeling approach for largeeddy simulation (LES) is obtained by
combining a `regularization principle' with an explicit filter and its
inversion. This regularization approach allows a systematic derivation of the
implied subgridmodel, which resolves the closure problem. The central role of
the filter in LES is fully restored, i.e., both the interpretation of LES
predictions in terms of direct simulation results as well as the corresponding
subgrid closure are specified by the filter. The regularization approach is
illustrated with `Leraysmoothing' of the nonlinear convective terms. In
turbulent mixing the new, implied subgrid model performs favorably compared to
the dynamic eddyviscosity procedure. The model is robust at arbitrarily high
Reynolds numbers and correctly predicts selfsimilar turbulent flow
development.

Buoyancy effects in unstably stratified mixing layers express themselves
through gravity currents of heavy fluid which propagate in an ambient lighter
fluid. These currents are encountered in numerous geophysical flows, industrial
safety and environmental protection issues. During transition to turbulence a
strong distortion of the separating interface between regions containing
`heavy' or `light' fluid arises. The complexity of this interface will be used
to monitor the progress of the mixing. We concentrate on the enhancement of
surfacearea and `surfacewrinkling' of the separating interface as a result of
gravityeffects. We also show that this process can be simulated quite
accurately using largeeddy simulation with dynamic subgrid modeling. However,
the subgrid resolution, defined as the ratio between filterwidth Delta and
gridspacing h, should be sufficiently high to avoid contamination due to
spatial discretization error effects.

We consider socalled Leray regularization of the convective contributions.
This gives rise to a subgrid parameterization which involves both explicit
filtering and (approximate) inversion. The Leray model also arises from the
alphamodeling strategy derived via Kelvin's circulation theorem. We study the
dynamics associated with the Leray model in a turbulent mixing layer and
compare predictions with filtered DNS results and findings due to dynamic
(mixed) models. In particular, the kinetic energy, momentum thickness and
energyspectra are analyzed, establishing favorable performance of the Leray
model and robustness at arbitrarily high Reynolds number. This is unique for a
similaritytype model that does not contain an explicit eddyviscosity term.

The $\alpha$modeling strategy is followed to derive a new subgrid
parameterization of the turbulent stress tensor in largeeddy simulation (LES).
The LES$\alpha$ modeling yields an explicitly filtered subgrid
parameterization which contains the filtered nonlinear gradient model as well
as a model which represents `Lerayregularization'. The LES$\alpha$ model is
compared with similarity and eddyviscosity models that also use the dynamic
procedure. Numerical simulations of a turbulent mixing layer are performed
using both a second order, and a fourth order accurate finite volume
discretization. The Leray model emerges as the most accurate, robust and
computationally efficient among the three LES$\alpha$ subgrid
parameterizations for the turbulent mixing layer. The evolution of the resolved
kinetic energy is analyzed and the various subgridmodel contributions to it
are identified. By comparing LES$\alpha$ at different subgrid resolutions, an
impression of finite volume discretization error dynamics is obtained.