Assignment 5 Faiz

UNIVERSITI TEKNOLOGI MALAYSIA (UTMSPACE)
2017
BACHELOR IN MECHANICAL ENGINEERING
SME 4353 – COMPUTER FLUID DYNAMICS
ASSIGNMENT 5
LECTURER: PROF MADYA DR. KAHAR BIN OSMAN
NAMA NO. MATRIK
MOHD FAIZ BIN AHMAD PHUAT SX111176MMS04

Two Model – Reynolds Stress Turbulence Model (RSM&RNG)
Assumption: Jet flow simulation
The flow considered in this study is weakly turbulent, and a turbulence model is thus
required for the computation of the flow.
In the present study, we employed six different two-equation models in their low-
Reynolds number forms. Among them, four linear eddy viscosity models are considered: the
LR k–ε model of Lien et al., the RNG k–ε model, the standard k–ω model and the SST k–
ω model.
The standard k–ε model is the most popular turbulence model, in which transport
equations are solved of the turbulent kinetic energy, k, and its dissipation rate, ε. The LR k–
ε model of Lien et al. Is similar to the standard k–ε model except that uses damping function
instead of the wall-function and contains extra source terms in its turbulent kinetic energy and
dissipation rate equations.
The RNG k–ε based algebraic model is derived from the instantaneous Navier–Stokes
equations using a mathematical technique known as the “renormalization group” methods. The
major difference between the RNG k–ε model and the standard k–ε model is that the RNG model
has an additional term that significantly improves the accuracy for rapidly strained flows. The
RNG model has been shown to perform better than the standard model for flows with high
streamline curvature, vortex shedding, etc. . The standard k–ω model is one of the most common
turbulence models. The ω equation has several advantages compared with k–ε. One of them is
that the equation can be used without additional terms through the viscous sub layer. The SST k–
ω model was developed by Menter which combined the robustness of k–ω turbulence model near
walls with capabilities of the k–ε model away from the walls.
Two nonlinear low Reynolds number k–ε turbulence models, i.e., quadratic and cubic,
according to the order of the characteristic time scale τ=k/ε in the anisotropy tensor equation, are
also considered.
We also considered the Reynolds stress turbulence model (RSM) proposed by Gibson and
Launder with a linear correlation for the pressure–strain term of the Reynolds-stress transport
equation. The model was associated with a two-layer approach for modelling the near-wall
region. RSM is the most elaborate turbulence model among the RANS based models. This
turbulence model should be considered whenever non-isotropic effects are important, for
example, in flows with strong curvature, swirling flows, and flows with strong
acceleration/retardation. The fidelity of RSM predictions is still limited by the closure
assumptions employed to model various terms in the exact transport equations for the Reynolds
stresses. The modelling of the pressure–strain and dissipation-rate terms is particularly
challenging, and often considered to be responsible for compromising the accuracy of RSM
predictions. Among some of the disadvantages of the RSM, Ranade mentions that the model is
computationally expensive, that it performs as poor as the k–ε model in some flows due to
problems with the dissipation rate equation, and that it has not been widely validated yet. The
detailed description of RSM model can be found in textbooks, such as Pope.
The RSM model which directly solves transport equations for each element of the
Reynolds stress tensor was associated with a two-layer approach, using the Wolf stein model for
the near-wall region. This model smoothly blends a one-equation model in the wall-adjacent
region with the values computed from solving the transport equation far from the wall. The Wolf
stein model solves turbulent kinetic energy, but prescribes the turbulence dissipation rate and
turbulent viscosity algebraically as functions of the wall distance, and the used turbulence model
away from the wall. This choice is appropriate for flows that are not dominated by buoyancy
forces. Our y+ values are less than 4 and are also satisfactory in this case. For k–ε based
turbulence models, the flow was resolved down to the viscous sub layer and the same two-layer
approach for modeling the near-wall region was applied. The use of a k–ω formulation for the k–
ω based turbulence models in the inner parts of the boundary layer makes the models directly
usable all the way down to the wall through the viscous sub-layer.
Results were compared with the experimental data of the turbulent cross-shaped jet at
moderate Reynolds number and the results of the comparison was satisfactory convincing us that
we have a good quality mesh
Conclusion
The article investigates the three-dimensional mean flow issued from a cross-shaped
orifice at moderate Reynolds number using seven turbulence models, i.e. linear (LR and RNG)
and nonlinear (quadratic and cubic) k–ε turbulence models, k–ω and shear stress transport
(SST) k–ω turbulence models and Reynolds stress turbulence model (RSM). It is revealed that
none of the turbulence models is able to predict well all jet characteristics in the same time. From
the point of view of the aimed applications (air diffusion for heating ventilating and air
conditioning systems) we found that the most suitable turbulence model is SST k–ω.
The selected model differs from the other six models on several points that can be
summarized as follows:
(1) The vena contracta and mean exit velocity profiles at the jet exit and at farther downstream
are best predicted by this model which is correctly handling the transition of the flow through the
diffuser,
(2) SST k–ω is superior to the other models in predicting mean flow characteristics in the jet near
field (axis-switching phenomenon, the secondary corner-vortices, stream wise velocity and
vortices fields in this region are best predicted by this model),
(3) The inflection point shown by the experiments on jet centreline velocity is better predicted by
this model,
(4) SST k–ω is able to predict reasonably well axial jet changes in the full observed axial range,
(5) In the context of the present study the most important parameters allowing orifice geometry
optimization is undoubtedly the mean flow integral parameters, particularly the axial evolution
of jet volumetric flow rate and global flow expansion which measure the ambient air induction
which are best predicted by the SST k–ω model.

Assignment 5 Faiz

Uploaded by

Copyright:

Available Formats

Assignment 5 Faiz

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Assignment 5 Faiz

Uploaded by

Copyright:

Available Formats

UNIVERSITI TEKNOLOGI MALAYSIA (UTMSPACE)

BACHELOR IN MECHANICAL ENGINEERING

SME 4353 – COMPUTER FLUID DYNAMICS

LECTURER: PROF MADYA DR. KAHAR BIN OSMAN

NAMA NO. MATRIK

MOHD FAIZ BIN AHMAD PHUAT SX111176MMS04

Assumption: Jet flow simulation

You might also like