
technically an anisotropic turbulence model throu gh UDF C functions in the
ANSYS-FLUENT environment, the objective of Chap. 3 is to fi ll the knowledge gap
in this subject area.
In Chap. 4, two-dimensional computational examples are presented using the
anisotropic hybrid k-x Shear-Stress Transport/Stochastic Turbulence Model
(SST/STM) [10]. It is important to highlight that the physics of turbulence is always
a three-dimensional phenomenon and strictly speaking two-dimensional turbulence
does not exist in nature at all. However, when a Reynolds-Averaged Navier–Stokes
(RANS) engineering turbulence model is under development, its validation for
two-dimensional well-established classical benchmark problems is still recom-
mended and is in the centre of interest. Therefore, numerical simulations have been
performed for turbulent flows (a) over a flat plate with zero pressure gradient
(Klebanoff problem), (b) over a NACA 0012 airfoil (NASA test problem), (c) in
axisymmetric straight smooth circular pipes at eight different Reynolds numbe rs,
(d) in a rotationally symmetric coaxial curved duct and (e) over a plane
backward-facing step, respectively. The description of these two-dimensional
numerical simulations can be considered as tutorials for solving simple anisotropic
turbulent flow benchmark problems with the use of the ANSYS-FLUENT R19.1
software package. In Chap. 4, all simulation results have bee n compared to
experimental data taken from the literature. For two-dimensional flows, the
dimensionless scalar weight parameter l
H
ð0\ l
H
jj
1Þ and the anisotropic scale
factors k
H
11
and k
H
22
in the modified deviatoric similarity tensor H
H
(3.99) of the
anisotropic hybrid k-x SST/STM closure model [10] are calibrated and validated
related to meas urements. The sim ulation results indicate that the physical behaviour
of the turbulent mean flow is captured correctly with the isotropic k-x SST model
of Menter [12, 13, 14] and the anisotropic hybrid k-x SST/STM closure model [10]
as well. However, further improvements can be achieved wi th the anisotropic
approach in terms of predicting the physically correct Reynolds stress distributions
in the framework of RANS turbulence modelling. The objective of the fourth
chapter is to show the capabilities of the anisotropic approach. Therefore, 156
simulation results are presented and analysed in Chap. 4.
In the short closing Chap. 5, three-dimensional simulation resul ts have been
presented with the anisotropic hybrid k-x Shear-Stress Transport/Stochastic
Turbulence Model (SST/STM) [10] to make an attempt and a first step towards
more complex real-world engineering applications. Therefore, the objective of the
fifth chapter is to shed light on possible research areas where further improvements
can be made as a future work on the new anisotropic turbulence modelling
approach investigated in this book. Numerical simulations have been performed for
three-dimensional turbulent flows (a) in a horizontal smooth cylindrical pipe,
(b) over a NACA 0013 wing and (c) over the Jetstream 31 aircraft, respectively. In
Chap. 5, all simulation results have been compared to the k-x SST model [12, 13,
14] and experimental data taken from the literature. The dimensionless scalar
weight parameter l
H
ð0\ l
H
jj
1Þ and the anisotropic scale factors k
H
11
, k
H
22
and k
H
33
in the modified deviatoric similarity tensor H
H
(3.99) have been calibrated and
x Preface