Professional Documents
Culture Documents
VerificationSuite17 06 008
VerificationSuite17 06 008
This software and related documentation are proprietary and confidential to Siemens. A list of relevant
Siemens trademarks can be found here. Other trademarks belong to their respective owners.
Simcenter STAR-CCM+ 2210 | Verification Suite
Contents
The Verification Suite provides cases that come from the Siemens Digital Industries Software quality-assurance
process. This extensive process includes an internal test system that is known as STAR-Test, which is used
continuously for the development and release builds of the software. With this suite of cases, you have the
opportunity to verify that the software you have received is able to reproduce the same results on the platform
you are using. Before using the suite, familiarize yourself with how key terms are defined.
The following are provided for each verification case:
Links to the STAR-Test report and simulation files are provided on each case description page. Instructions for
downloading the simulation files are also given on each page. A summary of all the STAR-Test reports can be
found by clicking the following link:
You can also clear the results in the simulation file and rerun the case.
• Clear the existing data using Solution > Clear Solution from the top menu bar. Click OK in the Clear
Solution dialog that appears.
• From the same menu, select Solution > Run.
Definitions
In the STAR-Test report pages, some of the individual tests provide plots on which comparisons are made
between two or three types of data. These data types, which are identified on plot legends, are defined as
follows:
• Baseline data: Simcenter STAR-CCM+ simulation data from the accepted baseline simulation file that is
held within the Simcenter STAR-CCM+ test repository. Subject matter experts certify baseline simulations
as giving acceptable results for the particular scenario being simulated. When a STAR-Test case
completes, data is extracted from it and compared to the same data set from the baseline simulation file.
If the Comparison and Baseline data are the same, the test passes.
• Comparison data: The data that was generated by running the simulation using a release candidate of
Simcenter STAR-CCM+.
• Test data: Reference data, such as experimental or analytical data, which the developer uses to assess
the Comparison data when creating a baseline. An exact match between Comparison data and Test data
is not necessarily expected. Developers decide whether the difference between these data sets is
acceptable for a particular case.
Contents:
Aeroacoustics
Flow and Energy
Time
Radiation
Combustion
Lagrangian Multiphase
Multiphase
Multiphase VOF
Finite Element Solid Stress
Electromagnetism
Compressible Flow
Turbulence
Aeroacoustics
Contents:
Noise Emission from Coaxial Jets
Propagation of a Planar Acoustic Wave
Purpose
This verification case demonstrates solving problems containing acoustical noise sources, using the Broadband
Noise Source model of Goldstein.
Overview
The geometry and flow conditions are based on Juve's configuration [1]. The primary and secondary nozzle
diameters are 30 mm and 100 mm, with corresponding exit velocities of 130 m/s and 52 m/s, respectively.
Physics
An initial velocity profile is prescribed across the inlets, with the velocity set to 130 m/s at the primary jet inlet,
52 m/s at the secondary jet inlet and 1 m/s for the co-flowing region beyond the secondary jet inlet:
The turbulent kinetic energy (k) and dissipation rate (ε) are set to 0.01 at the inlets, and the initial flow field
comes from a steady state converged solution.
The problem is solved as a steady two-dimensional axisymmetric turbulent incompressible flow case with the
Goldstein broadband noise source model active.
The length of the mixing and transition region are responsible for 90% of the acoustic power that is emitted.
Therefore, the quality of computing the acoustic power depends on accurate determination of the
aerodynamic flow field.
A qualitative comparison of this result with the similar acoustic power spectrum per unit volume of the
Goldstein model from Bechara et al. [2] shows that the correct solution is obtained. The general validity of the
Goldstein model for a wide range of Mach numbers means that it can be useful in more complex jet
configurations.
Physics Models:
• Aeroacoustics
• Axisymmetric
• Broadband Noise Sources
• Constant Density
• Gas (Air)
• Goldstein
• Gradients
• K-Epsilon Turbulence
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Standard K-Epsilon Two-Layer
• Steady
• Turbulent
• Two-Layer All y+ Wall Treatment
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[1] Juve D., Bataille J., Comte-Bellot G. (1978) Bruit de jets coaxiaux froids subsoniques, Journal de
Mecanique Appliquee, Vol. 2, N. 3.
[2] Bechara W., Lafon P., Bailly C., Candel S.M. "Application of a K-epsilon Turbulence Model to the
Prediction of Noise for Simple and Coaxial Free Jets", Acoustical Society of America, Vol. 97, Issue 6,
pp. 3518-3531.
Purpose
This case tests acoustic planar wave propagation.
Overview
This three-dimensional case calculates acoustic pressure. Acoustic waves are generated from a wall boundary
and propagate through a square channel. The sinusoidal input is propagated unchanged through the domain
for this planar wave case.
In addition to boundary calculations, data is also extracted through derived parts. A line-probe extends through
the center of the channel, parallel to the symmetry planes. Parallel to the line-probe, a vertical plane section
cuts through the middle of the region, and a point lies at the very center.
Physics
To test acoustics only, the segregated flow and energy solvers are frozen. The acoustic wave is a sinusoidal
input from the wall boundary.
Physics Models:
• Acoustic Wave
• Aeroacoustics
• Cell Quality Remediation
• Gas
• Gradients
• Ideal Gas
• Implicit Unsteady
• Laminar
• Segregated Flow
• Segregated Fluid Isothermal
• Three Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Contents:
Natural Convection in an Eccentric Annulus
Flow Through an Asymmetric Plane Diffuser
Lid-Driven Flow in a Two-Dimensional Square Cavity
Counter-flow Heat Exchanger in a Two-Dimensional Channel Modeled with Periodic Heat Transfer
Porous Dominated Flow
“Darcy’s Law” in One-Dimensional Incompressible Flow Using Porous Media
Buoyancy-Driven Laminar Flow in a Square Cavity (CGD)
Natural Convection in a Laminar Square Cavity with Large Temperature Difference
Karman Vortex Shedding
Porous Dominated Swirling Flow in an Annulus
Porous Baffle
Unsteady Conduction Heat Transfer Through a Solid Slab
Non-Newtonian Generalized Carreau-Yasuda Flow in an Axisymmetric Pipe
Natural Convection between Two Infinite Vertical Walls
Laminar High-Capacity Heat Exchanger Modeled with Fully-Developed Energy
Enthalpy Source to Mimic a Heat Exchanger
Laminar Flow of a Viscoelastic Fluid in Cross Flow
Flow of Viscoelastic Fluid Past a Cylinder
Unsteady 2D Compression Molding of a Newtonian Fluid
Orifice Plate Discharge Compared with ISO-5167 (CGD)
Purpose
This verification case demonstrates solving natural convection problems in Simcenter STAR-CCM+.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences in the
temperatures measured on the symmetry plane between the simulation and the reference, with a threshold
value of 5%, as shown in the following table. The acceptance value is derived from the study of Cho et al. [3].
Overview
An infinitely long pipe of radius 0.0178 m, heated at a constant temperature of 373 K, is placed inside another
larger pipe of radius 0.0463 m at a lower fixed temperature of 327 K. The inner pipe centroid is displaced
below the centroid of the outer pipe as shown:
Physics
The problem is solved as a two-dimensional laminar flow case with natural convection at Rayleigh number of
49500. The ideal gas model is used, with constant material properties, second-order discretization and a
Courant Number of 100.
Comparisons are made between experimental and computed temperature values at points on the vertical
centerline of the annulus. This line corresponds to the symmetry plane in the analysis. The points are
converted from the non-dimensional values from Kuehn and Goldstein [4] to the dimensions used in the
simulation.
Physics Models:
• Coupled Energy
• Coupled Flow
• Gas (Air)
• Gradients
• Gravity
• Ideal Gas
• Laminar
• Steady
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[3] Cho, C., Chang, K. & Park, K., 1982. Numerical Simulation of Natural Convection in Concentric and
Eccentric Horizontal Cylindrical Annuli. ASME. J. Heat Transfer, 104(4), pp. 624-630.
[4] Kuehn, T. H. and Goldstein, R. J. An experimental study of natural convection heat transfer in
concentric and eccentric horizontal cylindrical annuli. J. Heat Trans. (100) pp. 635-640, 1978.
Purpose
This verification case demonstrates predicting flow separation and reattachment through an asymmetric plane
diffuser.
Acceptance Criteria
The acceptance criteria for the comparison are the average of the unsigned relative differences in the U
velocity between the simulation and the reference at various probe locations. The relative differences were
normalized using the largest unsigned values of the U velocity obtained in the reference data. The acceptance
criterion represents a desired accuracy level for CFD simulations regardless of the specific turbulence model
adopted. The 10% criterion is selected as a strict acceptance from the work of Iaccarino [7].
Overview
The geometry is a two-dimensional Buice diffuser [6]. The height H is set to 1 m.
The experimental data consists of the U velocity profiles at a number of planes of constant x/H, as well as the
profile of skin friction coefficient along the bottom wall [6]. Other velocity components and turbulence
quantities are available but have not been used.
Physics
The flow is set to turbulent, with a Reynolds number of 20,000 based on the centerline velocity and the
channel height. The flow is solved using the segregated solver, and five turbulence models are compared:
• A. The two-layer variant of the Realizable K-Epsilon (TLRKE) turbulence model with All-y+ Wall Treatment
• B. The low-Reynolds number variant of the Standard K-Epsilon turbulence model with All-y+ Wall
Treatment
• C. The Elliptic Blending K-Epsilon (EBKE) model with All-y+ Wall Treatment
• D. The Modified Quadratic K-Epsilon (MQKE) low-Reynolds model with All-y+ Wall Treatment (Standard K-
Epsilon Low-Re with constitutive relation set to Quadratic) [5]
• E. The SST K-Omega turbulence model with All-y+ Wall Treatment
The following lists and tables summarize the physics models, material properties, initial and boundary
conditions, solver settings, and stopping criteria. To monitor the convergence of the simulation, the U
velocities are monitored at two different probes, P1 and P2, and the surface-averaged wall shear stress is
monitored at the bottom wall. The simulation is considered converged when the maximum difference in the
values of each of those magnitudes in the last 500 iterations is lower than 10–6 m/s and 10–6 N/m2,
respectively.
Physics Models:
• Constant Density
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[5] Baglietto, E. 2006. Anisotropic Turbulence Modeling for Accurate Rod Bundle Simulations, Miami
USA., Proceedings of ICONE14 Conference.
[6] Buice, C. U. and Eaton, J. K., “Experimental investigation of flow through an asymmetric plane
diffuser”, Report No. TSD-107. Thermosciences Division, Department of Mechanical Engineering,
Stanford University, 1997.
[7] Iaccarino, G. 2001. Predictions of a turbulent separated flow using commercial CFD codes, J. Fluid
Engineering, Volume 123, p. 819-828.
Purpose
This verification case demonstrates solving a traditional square lid-driven cavity flow. The velocity components
in two perpendicular planes passing through the center of the cavity are compared against a selected
reference.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences between the
velocities measured in the simulation at different probes and the reference, with a threshold value of 5%, as
shown in the following table. These differences are converted into relative values by normalizing them using
the maximum unsigned value of the velocity results reported in the reference for the above mentioned planes.
Overview
The problem geometry consists of a two-dimensional square cavity, 1 m in length. The cavity is covered using
an impermeable wall that moves in the x-direction with constant velocity of 1 m/s.
Physics
The Reynolds number that is based on the width of the cavity, top wall velocity, fluid density and fluid dynamic
viscosity is 5000. The problem is solved using the segregated flow model and second order upwind convection
scheme. Both the U- and V-velocity profiles along lines passing through the center of the cavity are compared
with data from the literature, such as [8].
Physics Models:
• Constant Density
• Gradients
• Laminar
• Liquid (H2O)
• Segregated Flow
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[8] Ghia, U., Ghia, K.N., and Shir, C.T. 1982 “High-Re solution for incompressible flow using the Navier-
Stokes equations and a multigrid method,” J. Comp. Phys., 48, p. 387.
Purpose
This verification case demonstrates the use of periodic heat transfer in a channel with both a constant heat flux
wall and a conjugate interface between counter-flowing streams of identical thermal properties.
Acceptance Criteria
The acceptance criteria for the comparison are the overall averages of the unsigned relative differences in the
velocities and temperatures between the simulation and the reference, with a threshold value of 1%, as shown
in the following table. The differences in velocity were normalized using the maximum reference velocity for
each profile, whilst the differences in temperature were normalized using the bulk temperature difference
between the inlets of both fluids. One average is then calculated for the velocity and one for the temperature,
using all the available relative differences in each case.
Overview
The problem geometry consists of two, two-dimensional periodic channels. The top channel is modeled in full,
while a symmetry plane is used for the bottom channel. The channels measure 10 m in length and 2 m in
height. A cold fluid flows from left to right in the bottom channel; a hot fluid flows from right to left in the top
channel. A constant heat flux wall is specified on the top wall of the top channel. A baffle interface is defined
between the two channels. There is no thermal resistance across the baffle.
Physics
The simulation results are compared to the analytical velocity and temperature profile solutions for a unit
channel of half-height, that is given in [9].
Physics Models:
• Constant Density
• Gradients
• Laminar
• Liquid (Fictitious Material)
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Two-Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[9] White, F.M., “Viscous Fluid Flow”, 2nd Ed., McGraw-Hill, 1991.
[10] Cess, R. D. and Schaffer, E. C., "Heat Transfer to Laminar Flow Between Parallel Plates with a
Prescribed Wall Heat Flux," Appl. Sci. Res., Section A, Vol. 8, p. 339-344, 1959.
Purpose
This verification case demonstrates using a porous media to simulate inviscid flow in which porous body forces
dominate.
Overview
The problem geometry is a channel, split up into three sections: the inlet section, that is skewed at an angle of
26.6 degrees with the horizontal, and two sections that are aligned with the x-axis. The middle section is
porous, with an anisotropic porous media that both induces a pressure drop and turns the flow horizontal.
Physics
The problem is solved using the segregated flow solver. The flow is assumed to be inviscid and have constant
density.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Inviscid
• Segregated Flow
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Porosity 1.0
Viscous Resistance (kg/ Principal Tensor: XX axis = [4, -3], YY axis = [3, 4]
m^3-s) XX = 10, YY = 1000
Inertial Resistance (kg/ Composite Tensor:
m^4) XX = 366.4, XY = 475.2, YX = 475.2, YY = 643.6
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates using porous media to create the desired pressure drop in laminar pipe
flow.
Overview
The pipe is 1 m in diameter and 100 m in length. Modeling the pipe as an axisymmetric domain converts the
geometry to a two-dimensional rectangle, representing one half of the symmetric pipe about the horizontal
axis. The viscous resistance to the flow is the porous resistance. The momentum fluxes in the porous region
are neglected.
Physics
Since the flow is one dimensional, the flow field is uniform in the cross-stream direction. Pressure on the axis is
compared to the theoretical drop computed from
ΔP 4Reμ2
= (1)
ΔL r02
Material Properties:
Initial Conditions
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates buoyancy-induced laminar flow in a square cavity. The buoyancy effect is
created by a temperature difference in the opposing vertical walls. A series of cases with different Rayleigh
numbers were analyzed and the results were compared with reference data.
Acceptance Criteria
The acceptance criteria for the comparison are the average of the relative differences in the results obtained,
for all the Rayleigh numbers analyzed. The results compared are:
• The Nusselt number averaged over the surface of the hot wall
• The maximum and minimum Nusselt numbers on the hot wall
• The Nusselt number at the center of the hot wall
• The maximum horizontal component of the velocity on a vertical plane passing through the center of the
cavity
• The maximum vertical component of the velocity on a horizontal plane passing through the center of the
cavity
The different acceptance criteria and threshold values are shown in the following table and are derived from
the grid convergence work presented by De Vahl Davis [11].
Overview
Buoyancy-driven laminar flow is modeled in a square two-dimensional cavity, as shown in the following
diagram. The buoyancy is created by establishing a temperature difference between both vertical walls, the
left wall being at a higher constant temperature than the right wall. The top and bottom walls are adiabatic.
Two different meshes were employed, as shown in the following image, one polyhedral mesh (a) and one
trimmed mesh (b). In both cases the element base size was set to 1% of the side length L and the prism layer
consisted of two elements with a total thickness equal to 33.33% of the base size and a thickness ratio of 1.5.
Physics
A series of Rayleigh numbers were analyzed, ranging from 1,000 to 100,000, as shown in the following table.
Following [11], the fluid is assumed to have a Prandtl number equal to 0.71.
The reference data [11] was created by means of a series of numerical simulations with different mesh sizes
and then using an extrapolation scheme to generate a solution of greater accuracy.
Rayleigh Numbers
1,000
10,000
100,000
Physics Models:
• Boussinesq Model
• Constant Density
• Gradients
• Gravity
• Laminar
• Liquid
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps Varies for each case. Upon convergence of the following:
• Average Nusselt number on left wall (Nu_avg)
• Horizontal velocity on vertical plane (U_max)
• Vertical velocity on horizontal plane (V_max)
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[11] De Vahl Davis, G. 1983. "Natural convection of air in a square cavity: A benchmark numerical
solution," Int. J. for Numerical Methods in Fluid, 3, pp. 249-264.
Purpose
This verification case demonstrates laminar natural convection with variable properties in a square cavity.
Acceptance Criteria
The acceptance criterion for the comparison is the maximum value of the unsigned relative differences in the
surface averaged Nusselt number on the right wall of the domain, with a threshold value of 5%, as shown in
the following table. The threshold value is derived from the uncertainty in the correlation for Nusselt number
produced by Chenoweth and Paolucci [12] and comparison to the higher accuracy solution by Le Quéré, [13].
Overview
The geometry is a two-dimensional, 1 m x 1 m cavity, discretized using a 6400 quad cell, stretched grid. The
cavity, which is filled with an ideal gas, is heated on one vertical wall, is cooled on the other vertical wall and is
insulated on the top and bottom. The temperature differences are large enough to require temperature-
dependent variable properties.
Physics
This case runs with four different Rayleigh numbers, that are achieved by changing the gravity magnitude. The
reference data that are used are the Nusselt numbers from Chenoweth and Paolucci [12]. They present data
from a Boussinesq calculation (infinitely small value of temperature difference) and provide a correlation for
Nusselt number variation. The correlation formula for a square cavity (for Ra ≥ 10−3) is Nu = 0.1448Ra0.2969.
Physics Models:
• Coupled Energy
• Coupled Flow
• Gas (Air)
• Gradients
• Gravity
• Ideal Gas
• Laminar
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Gravity Conditions :
1.0E5 2.901E-5
1.0E6 2.901E-4
1.0E7 2.901E-3
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Note: For every Ra condition, the solution is cleared and run up to 1000 iterations.
Bibliography
[12] Chenoweth, D. R. and Paolucci, S., 1986. "Natural Convection in an enclosed vertical layer with large
horizontal temperature differences”, J. FluidMech, 169, p. 173-210.
[13] Le Quéré, P., 1991. "Accurate solutions to the square thermally driven cavity at high Rayleigh
number", Computers & Fluids, 20(1), pp. 29-41.
Purpose
This verification case demonstrates solving vortex shedding. The objective is to predict the correct shedding
frequency for Karman vortex street behind a circular cylinder in a uniform flow.
Acceptance Criteria
The acceptance criterion for the comparison is the unsigned relative difference in vortex shedding frequency
between the predictions of the simulation and the reference, as shown in the following table.
Overview
The problem geometry is shown in the diagram below and it consists of a two-dimensional rectangular
channel containing a cylinder of 10 mm in diameter.
The objective is to predict the correct shedding frequency of 2.25 Hz. The simulation shows a period of
approximately 0.44 s, which corresponds to the expected shedding frequency of 2.25 Hz.
Physics
This test is set up to generate a Karman vortex street behind a circular cylinder of diameter d in uniform cross
flow of velocity u. The vortex shedding frequency f is characterized by the Strouhal number, St = fd/v, which
for a Reynolds number Red = ρud /μ of 75 is given as 0.15 [14]. Water at a constant velocity of 0.15 m/s
enters the domain through the velocity inlet.
Physics Models:
• Constant Density
• Coupled Flow
• Gradients
• Implicit Unsteady
• Laminar
• Liquid (H2O)
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[14] Daily, J.W., and Harleman, D.R.F. 1966. Fluid Dynamics, Addison-Wesley, MA.
Purpose
This verification case demonstrates the use of porous dominated flow with defined resistances in a cylindrical
coordinate system.
Overview
The problem geometry consists of a two-dimensional annulus of inner radius 1.0 m and outer radius 2.0 m.
Swirling flow enter at the inner radius of the annulus and exits out of the outer radius. Within the annulus, an
anisotropic porous medium is defined, which tends to remove the swirl component and induce a radial
pressure drop. The radial and tangential components in this case are set using a cylindrical coordinate system
at the center of the annulus.
Physics
The simulation results are compared to reference data for velocity and pressure. The analytical solution is:
riuri
ur r = (2)
r
riuθ1 Cθ ri2 − r2
uθ r = exp (3)
r 2ρriuri
ro
ρu2 − Crriuri
p r = dr′ (4)
r′
r
where Cr and Cθ are the radial and tangential viscous resistances, respectively.
Physics Models:
• Constant Density
• Coupled Flow
• Gradients
• Inviscid
• Liquid (H2O)
• Steady
• Two-Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Region Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Porous Baffle
Purpose
This verification case demonstrates the pressure drop of a fluid passing through porous baffles.
Overview
The problem geometry consists of a two-dimensional, 0.1 m by 1.0 m channel split into three sections. The
three sections belong to the same fluid continuum and are separated by porous baffles. The pressure drop at
each baffle is set to unity for the given fluid properties and uniform velocity.
Physics
The simulation results are compared to reference data for pressure distribution along the length of the channel
which was determined through trivial analysis.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Laminar
• Segregated Flow
• Steady
• Two-Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Interface Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates solving an unsteady conduction heat transfer problem.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences in the
temperatures measured at different probes, when comparing simulated and analytical values, as shown in the
following table. The relative differences were normalized using the largest reference temperature value.
Overview
The problem geometry consists of a two-dimensional slab, 1 m in length, at a uniform initial temperature T_i =
200 K. At time t = 0 s, the temperature at both ends of the geometry is lowered to and maintained at T_inf =
100 K for all time t > 0.
Physics
The problem is solved using the coupled solid energy and implicit unsteady models. The temperature
distribution along the length of the slab is plotted at time t = 0.05 s. The temperature profile corresponds to an
analytical solution given in [15]. The solution is:
4 ∞ 1 −α 2
T x, t = T∞ + T − T∞ e nπ/L tsin nπ x (5)
π i n = 1, 3, 5, … n L
k
where α = ρCp
.
Physics Models:
• Constant Density
• Coupled Solid Energy
• Gradients
• Implicit Unsteady
• Solid (Al)
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[15] Ozisik, M. M. 1985. “Heat Transfer, A Basic Approach”, McGraw-Hill.
Purpose
This verification case demonstrates that viscosity shear-thinning is correctly captured in a Carreau-Yasuda fluid
flow inside an axisymmetric pipe. Results are compared against experimental data for 0.125% polyacrylamide
solution at a Reynolds number of 676.
Overview
This case uses axisymmetric pipe geometry with no-slip wall boundary conditions on the top wall, uniform
velocity at the inlet of 0.256 m/s, and a pressure outlet boundary condition at the outlet.
Physics
The Laminar model is used with steady segregated flow in a two-dimensional mesh.
Physics Models:
• Axisymmetric
• Coupled Energy
• Constant Density
• Gradients
• Laminar
• Liquid
• Segregated Flow
• Steady
Material Properties:
Initial Conditions:
Pressure (Pa) 0
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[16] Escudier, M.P., Poole, R.J., Presti, F., Dales, C., Nouar, C., Desaubry, C., Graham, L., Pullum, L. 2005.
“Observations of asymmetrical flow behaviour in transitional pipe flow of yield-stress and other
shear-thinning liquids”, J. Non-Newtonian Fluid Mech. 127 pp. 143155.
Purpose
This verification case demonstrates natural convection using a two-layer model for turbulence.
Acceptance Criteria
The acceptance criteria for the comparison are the overall averages of the unsigned relative differences in the
velocities and temperatures between the simulation and the reference. The acceptance thresholds are selected
from the comparison between the numerical solution of Versteegh [19] and the experiments of Dafa'Alla &
Betts [17] as shown in the following table. These differences are converted into relative values by normalizing
them using the maximum unsigned value found in the reference for each compared profile. Overall average
values are then calculated for the different results, using all the available relative differences in each case.
Overview
In this case, two infinitely long vertical walls are differentially heated in the presence of a gravitational field.
Physics
The Rayleigh number associated with this case is 5 x 10^6. The Standard K-Epsilon Two-Layer model is used
with steady coupled flow in a two-dimensional mesh.
Physics Models:
• Boussinesq Model
• Constant Density
• Coupled Energy
• Coupled Flow
• Gas
• Gradients
• Gravity
• K-Epsilon Turbulence
• Reynolds-Averaged Navier-Stokes
• Standard K-Epsilon Two-Layer
• Steady
• Turbulent
• Two-Dimensional
• Two-Layer All y+ Wall Treatment
Material Properties:
Initial Conditions:
Pressure (Pa) 0
Static Temperature (K) 1.0
Turbulent Dissipation Rate (m2/s3) 0.1
Turbulent Kinetic Energy (J/Kg) 0.2
Velocity (m/s) U=V=0
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[17] Dafa'Alla, A. & Betts, P., 1996. "Experimental study of turbulent natural convection in a tall cavity",
Exp. Heat Transfer, 9, pp. 165-194.
[18] Nieuwstadt, F.T.M and Versteegh, T.A.M. 1997. “DNS of Natural Convection Between Two Vertical
Differentially Heated Walls”, Proc. 11th Turblent Shear Flows Symposium, Grenoble.
[20] Xu, W., Chen, Q., and Nieuwstadt, F.T.M. 1998. “A new turbulence model for near-wall natural
convection”, Int. J. Heat and Mass Transf., 41, pp. 3161-3176.
Purpose
This verification case demonstrates a high-capacity heat exchanger in a two-dimensional channel modeled
with periodic heat transfer.
Overview
The two-dimensional channel has both a constant temperature wall and a conjugate interface between a fluid
stream and a solid stream. The solid stream has a much larger thermal capacity, and therefore is at an
approximately constant temperature.
Two regions with periodic interfaces are used. The top region is a channel that is modeled completely. The
bottom region is a solid, with a constant temperature wall on the bottom. The hot fluid on the top flows from
right to left. A constant heat temperature wall is specified on the top wall of the channel.
Physics
This case involves laminar flow and fully developed energy with constant wall temperature scaling. Conjugate
heat transfer occurs between fluid streams through a baffle interface.
The exact velocity solution [21] for a unit channel half-height is given by:
3 y 2
u= V 1− (6)
2 y0
where y0 is the channel half-height, and y is measured from the symmetry plane. The mean velocity V is given
in terms of the streamwise pressure gradient as:
1 dP
V= − (7)
3μ dx
The temperature profile for a constant wall temperature heating condition is:
T − Tw ∞ 3Nu
= C − Cn − 1 (8)
Tb − Tw n=0 2 2n − 1 n − 2
3
where Nu = 11.885, C0 = 132908, and C1 = − 4 C0Nu.
Physics Models:
• Constant Density
• Gradients
• Steady
• Two Dimensional
Fluid Continuum:
• Coupled Energy
• Coupled Flow
• Gas
• Laminar
Solid Continuum:
Material Properties:
Initial Conditions:
Description Type/Conditions
inflow [Periodic 2] Internal Interface
interface 1 [In-place 1] Contact Interface
outflow [Periodic 2] Internal Interface
wall Wall
Static Temperature (K): 400.0
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[21] White, F. M. 1991. Viscous Fluid Flow, 2nd ed. McGraw-Hill.
Purpose
This verification case tests the Heat Exchanger option of the Energy Source Option region physics condition in
Simcenter STAR-CCM+. In this case, it simulates local heat passing to and from the fluid, distributed by the
local temperature difference with Tref.
Overview
This simulation is a two-dimensional flow continuum with a heat exchanger energy source region.
The energy source Q and a reference temperature Tref are given as input parameters for the “Heat Exchanger
Energy Source”. The local heat transfer at each cell is scaled by the local temperature difference between the
cell and Tref using a local heat transfer coefficient ul, computed such that the volume sum of all local heat
sources adds to Q.
Physics
This two-dimensional case models the steady inviscid flow of a constant-density gas.
dQ = ual Tref − T dV
ρuACp + dT = dQ (9)
= ual Tref − T dV
where ual represents a local overall heat transfer coefficient times the heat transfer area per unit volume of the
device. For this verification problem, the flow area, A, is a constant 1.0m x 1.0m, so:
dV = dx 1.0m 1.0m = dx
(10)
m′V = m′dV = ρu 1.0m 1.0m
T x
dT ual
= dx
Ti Tref − T 0 m′dVCp
(11)
Tref − T ual
ln = x
Tref − Ti m′VCp
gives:
ual
x
T x = Tref − Tref − Ti e m′VCp (12)
ual can be calculated given Q, the total heat transfer of the region (device) by evaluating the temperature at
x − L as follows:
ual
L
T L = Tref − Tref − Ti e m′VCp (13)
Q
TL = Ti + (14)
m′VCp
giving:
m′VCp Q
ual = ln 1 − (15)
L m′VCp Tref − Ti
Physics Models:
• Constant Density
• Gas
• Gradients
• Inviscid
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Region Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case tests solution accuracy for two fluids that approach each other in opposite directions.
Overview
Viscoelastic flows are common in polymer melt applications. In this case, comparison is made between
Simcenter STAR-CCM+ results and the work of Verbeeten et al. [23] for the velocity profiles and stress
components along a symmetry boundary. The comparison gives good agreement.
Physics
The flow is assumed to be laminar. For such flows two components are modeled: the solvent component
(through dynamic viscosity, which is constant in this case) and a viscoelastic component. To simulate the
inertial mass of the viscous fluid, the Viscous Flow model includes inertia terms in the momentum equation.
Four modes of the Giesekus-Leonov method are used in the material properties.
Physics Models:
• Constant Density
• Laminar
• Liquid
• Steady
• Two Dimensional
• Viscoelastic
• Viscous Flow
Material Properties:
As per the work by Peters et al. [22], viscoelastic rheology parameters (Pa-s) are estimated from aT and bT shift
factors. These shift factors are based on low-density polyethylene (LDPE) at reference temperatures of 190°C
and 150°C.
Initial Conditions:
Velocity U = V = W = 0.0
Boundary Conditions:
Maximum Steps 15
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[22] Peters, G.W.M., Schoonen, J.F.M., Baaijens, F.P.T., and Meijer, H.E.H. 1999. “On the performance of
enhanced constitutive models for polymer melts in a cross-slot flow”, J. Non-Newtonian Fluid Mech.,
82 (2), pp. 387-427.
[23] Verbeeten, W.M.H., Bogaerds, A.C.B., Peters, G.W.M., and Baaijens, F.P.T. 1999. “Visco-elastic
analysis of polymer melts in complex flows”, Eindhoven University of Technology, Materials
Technology, the Netherlands.
Purpose
This verification case checks the accuracy of the computed first normal stress difference.
Overview
This simulation applies the following:
Comparison is made between Simcenter STAR-CCM+ results and the work of Verbeeten et al. [24] for the stress
component over the cylinder wall. The comparison gives good agreement.
Physics
The flow is assumed to be laminar. To simulate the inertial mass of the viscous fluid, the Viscous Flow model
includes inertia terms in the momentum equation.
Four modes of the Phan-Thien Tanner/Johnson-Segalman method are used in the material properties.
Physics Models:
• Constant Density
• Laminar
• Liquid
• Steady
• Two Dimensional
• Viscoelastic
• Viscous Flow
Material Properties:
Initial Conditions:
Velocity U = V = W = 0.0
Boundary Conditions:
Maximum Steps 20
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[24] Verbeeten, W.M.H., Peters, G.W.M., and Baaijens, F.P.T. 2002. “Viscoelastic analysis of complex
polymer melt flows using the eXtended Pom–Pom model”, J. Non-Newtonian Fluid Mech., 108, pp.
301–326.
Purpose
This verification case tests the accuracy of compression molding of a Newtonian fluid.
Overview
This verification case examines a blank of fluid which is compressed in a mold. The mold is of width W. The
blank is of width W, initial thickness of ℎ0, and initial length of 2l0:
The rheology of fluid is assumed to be Newtonian with a viscosity of μ, and the flow is assumed to be
isothermal and quasi-steady. The objective is to find an analytical solution for the velocity and pressure field
using lubrication approximation. It is considered that ℎ ≪ 1 and 1 ≪ W, and hence vy ∼ 0.
Having performed the conventional lubrication approximation analysis, one can find the velocity and pressure
dℎ
profiles as well as the force F required to maintain a uniform closing rate of dt according to:
dℎ 6μ 2
p x, t = l − x2 (16)
dt ℎ3
6x dℎ z z 2
vx = − (17)
ℎ dt ℎ ℎ
dℎ 1 z 2 1 z 3
vz = − 6 − (18)
dt 2 ℎ 3 ℎ
8Wμl3 dℎ
F= (19)
ℎ3 dt
For quantitative comparison with Simcenter STAR-CCM+, a 2D compression molding (initial height
ℎ0 = 0.02 m and length of l = 0.2 m) was simulated on a rectangular domain as shown schematically above.
The geometry was discretized using rectangular cells (20 x 200 = 4000 cells).
kg
A Newtonian model is assumed with constant viscosity μ = 1000 Pa.s and density ρ = 1000 . Furthermore,
m3
dℎ
the moving plate on top is considered to have a velocity of dt
= − 0.001 m/s and the initial load of fluid is
occupying the region where −0.03 < x < 0.03. The simulation runs in transient with a sufficiently small time-
step of Δt = 0.05 s to make sure the fluid front does not move significantly within the time interval. In the
Simcenter STAR-CCM+ simulation, 10 different line probes were introduced in the channel cross-section at
different locations (that is, different values of x).
For quantitative comparison, both velocity fields from the analytical solution (mentioned previously), vz and
vx, were plotted on the same XY plot as the Simcenter STAR-CCM+ prediction. The agreement is remarkable in
all line probes, indicating the compression molding feature has been implemented correctly in
Simcenter STAR-CCM+.
Physics
The Generalized Newtonian model is used in a 2D transient case.
Physics Models:
• Constant Density
• Free Surface
• Generalized Newtonian
• Implicit Unsteady
• Laminar
• Liquid (H2O)
• Partial Fill
• Two Dimensional
• Viscous Flow
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates the flow through an orifice plate at several pressure drops. The mass flow
rates obtained are compared with the values prescribed by the ISO-5167 (2003) Standard [26]. Relevant
discussions are also found in Otgonbaatar, et al. [27].
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences in the mass
flow rate, between the simulation values and the reference values. Several Reynolds numbers were imposed in
the simulations. The reference mass flow rate values were reconstructed from the pressure drops obtained in
the simulations using the ISO-5167 (2003) Standard, to allow for direct comparison. The average was
calculated from the Reynolds numbers analyzed. A threshold value of 5% was used, as shown in the following
table. The acceptance criterion represents a desired accuracy level for CFD simulations regardless of the
specific turbulence model adopted. The performance of a specific model allows much improved accuracy from
the acceptance level.
Overview
Turbulent flow is modeled through an orifice plate inside a circular pipe, as shown in the following image. The
flow is driven by imposing several values of Reynolds number on the pipe. The pressure drops across the orifice
plate are measured from the probes p1 and p2. The probes are located on the pipe wall, at a distance from the
perforated plate equal to D and D/2 upstream and downstream, respectively. The pipe and perforated plate
have diameters D = 252 mm and d = 173 mm, respectively.
As shown in the following illustration of the mesh employed around the orifice plate, the perforated plate has
been modeled as a baffle with zero thickness. The reason is that according to the ISO-5167 standard, the
perforated plate should be as thin as possible without compromising the structural integrity of the plate. In the
simulation, it is possible to take this statement to the limit and create a plate with zero thickness. The Trimmer
mesher has been employed to generate a wall-function mesh, with a prism layer comprising 2 cells.
The mesh has a base size of 5 mm and has been refined to 1.25 mm in a volume extending 1D upstream and
5D downstream from the perforated plate. The table below summarizes the dimensions and mesh types
employed.
Ldownstream = 56D
Perforated plate d = 173 mm
Mesh Type Trimmer mesh
Physics
ISO-5167 specifies mass flow rate values based on measurements of the pressure drop between the probes p1
and p2. In the simulations, the Reynolds number is specified. A series of simulations were run, with the
Reynolds numbers as follows:
• 50,000
• 100,000
• 250,000
• 500,000
By specifying the Reynolds number, the mass flow rate in the simulation is also specified as the dimensions
and material properties are fixed beforehand. Then, the reference mass flow rate values are calculated from
the pressure drops obtained in the simulations with the formula published in the Standard. The pressure drops
in the simulations are calculated as specified in the Standard.
• A. The two-layer variant of the Standard K-Epsilon (TLSKE) turbulence model with All-y+ Wall Treatment
• B. The two-layer variant of the Realizable K-Epsilon (TLRKE) turbulence model with All-y+ Wall Treatment
• C. The Elliptic Blending K-Epsilon (EBKE) model with All-y+ Wall Treatment
• D. The Modified Quadratic K-Epsilon (MQKE) high-Reynolds model with High-y+ Wall Treatment [25]
Physics Models:
• Axisymmetric
• Steady
• Liquid (H2O)
• Segregated Flow
• Constant Density
• Turbulent
• Reynolds-Averaged Navier-Stokes
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[25] Baglietto, E., 2006. Anisotropic Turbulence Modeling for Accurate Rod Bundle Simulations. Miami
USA., Proceedings of ICONE14 Conference.
[26] ISO 5167-2:2003, 2003. Measurement of fluid flow by means of pressure differential devices
inserted in circular cross-section conduits running full — Part 2: Orifice plates, s.l.: s.n.
[27] Otgonbaatar, U., Baglietto, E. & Todreas, N., 2016. Methodology for Characterizing
Representativeness Uncertainty in Orifice Plate Mass Flow Rate Measurements Using CFD
Simulations. Nuclear Science and Engineering· Volume 184, pp. 430–440.
Time
Contents:
Stokes’ First Problem: The Suddenly Accelerated Wall
Purpose
This verification case compares the velocity distribution against theory [28] for a given time period. Essentially
Stokes’ First Problem considers the effect on an infinite body of fluid when an infinite wall adjacent to that fluid
instantly accelerates from zero to a constant velocity in the direction of the fluid.
Overview
A vertical column with a 2D mesh of 80 quadrilateral cells is used for this test.
Physics
The Laminar model is used with unsteady segregated flow in a two-dimensional mesh.
Physics Models:
• Constant Density
• Gas
• Gradients
• Implicit Unsteady
• Laminar
• Segregated Flow
• Two-Dimensional
Material Properties:
Initial Conditions:
Pressure (Pa) 0
Velocity (m/s) U=V=0
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[28] Schlichting, H. 1979. Boundary-Layer Theory, McGraw-Hill.
Radiation
Contents:
Radiative Heat Transfer in Scattering Media Using the Discrete Ordinate Method
Surface-To-Surface Radiation in an Empty Rectangular Box
Surface-To-Surface Radiation in a Cylindrical Hole
Surface-To-Surface Radiation in a Hollow Cylinder
Surface-to-Surface Radiation Heat Transfer Reduction Across a Single Shield
Gray Thermal Radiation Using a Hexahedral Mesh
Surface Photon Monte Carlo Modeling of Solar Loads Focusing through a Lens
Purpose
This verification case demonstrates solving radiative heat transfer in scattering media using the discrete
ordinate method (DOM).
Overview
A square two-dimensional domain is used. The bottom wall has a fixed temperature of 400 K while the other
three walls have a fixed temperature of 300 K. The media is scattering (scattering coefficient of 1) and non-
absorbing (absorption coefficient of 0). The problem involves pure radiative heat transfer; advection and
conduction are not considered.
Physics
The problem is solved as a two-dimensional, steady state, laminar flow case with the gray thermal radiation
option. The gas is assumed to be air with constant material properties. Since fixed temperature walls are used
and the media is non-absorbing, the fluid energy equation is decoupled from the radiative transfer. Thus, the
RTE solution is independent of the fluid solution and can be finalized in only one global iteration.
q
q* = (20)
σ Tℎ4 − Tc4
G − 4σTc4
G* = (21)
4σ Tℎ4 − Tc4
where q is the boundary radiation heat flux, G is the incident radiation, σ is the Stefan-Boltzmann constant
(5.67e − 8Wm−2K−4), Tℎ is the hot wall temperature and Tc is the cold wall temperature.
The values are highest near the hot bottom wall. The rays with higher intensities that are emitted from the
bottom wall are attenuated as they move up through the media and are scattered.
Both analytical [29] and numerical [30] reference data is available for comparison.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Gray Thermal Radiation
• Laminar
• Participating Media Radiation (DOM)
• Radiation
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps 1
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Reference
[29] Crosbie, A.L., and Schrenker, R.G., 1984, “Radiative Transfer in a Two-Dimensional rectangular
Medium Exposed to Diffuse Radiation,” J. Quant. Apectrosc. Radiat. Transfer, Vol. 31, pp. 339-372.
[30] Raithby, G.D., and Chui, E.H., 1990, “A Finite-Volume Method for Predicting Radiant Heat Transfer in
Enclosures with Participating Media,” J. Heat Transfer, Vol. 112, pp. 415-423.
Purpose
This verification case demonstrates surface-to-surface radiation in an empty rectangular box and compares the
simulation prediction with analytical data.
Overview
The case tests surface-to-surface radiation in an empty rectangular box for four different conditions:
Physics
Aside from Condition 4, all cases are run using the steady segregated solver. The implicit unsteady segregated
solver is required to run Condition 4 as it has zero gas conductivity.
All four tests are compared to analytical data:
• Condition 1: the total radiation heat transfer is calculated for each fixed temperature surface using the
network theory found in [31] and [32] and exact view factors.
• Condition 2: at equilibrium, next to the fixed-temperature wall, the temperature of the insulated walls
and the gas is equal to the imposed temperature.
• Condition 3: the wall temperature has to be equivalent to the temperature of the fixed-temperature wall.
At equilibrium, the radiation heat flux is null everywhere. The gas temperature is the same as the wall
temperature.
• Condition 4: same as above, except that the gas temperature is at the initial gas temperature since the
gas is thermally decoupled from the wall.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Gray Thermal Radiation
• Laminar
• Radiation
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Surface-to-Surface Radiation
• Three Dimensional
• View Factors Calculator
Material Properties:
Initial Conditions:
Boundary Conditions:
Conditions Conditions 4
1-3
Implicit Unsteady Timestep N/A 0.001 s
Segregated Flow Velocity URF 0.7 0.8
Segregated Flow Pressure URF 0.3 0.2
Segregated Energy Fluid URF 0.9 0.9
Segregated Energy Solid URF 0.99 0.99
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[31] Evans, S., Validation of Radiation Model in STAR-CD, CD-adapco Internal Report CDH/101103/R1, 11
November 2003, London.
[32] Holman, J.P., Heat Transfer, McGraw Hill, Eighth (SI) Edition, 2001.
Purpose
This verification case demonstrates using the view factor generator and the symmetry boundary conditions for
problems involving surface-to-surface radiation.
Overview
The case corresponds to example 8-17 in [33], which has an analytical solution for a case having five patches.
The geometry is comprised of a 45 degree wedge, 10 mm in radius and 30 mm in length, with symmetry
planes as shown in the following diagram:
Physics
The problem is solved using the segregated flow and energy solver. The radiation heat flux for five patches
from the hole is compared to analytical data. Five patches are fewer than the number that are used in the
simulation, but already accurate, as Holman [33] showed. Further reducing the number of patches to two in
the analysis, by assuming uniform radiosity over the hole and the opening, only changes the final heat fluxes
by 3.9%.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Gray Thermal Radiation
• Laminar
• Radiation
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Surface-to-Surface Radiation
• Three Dimensional
• Viewfactors calculator
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps 20
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Reference
[33] J. P. Holman, J. P. (2001) Heat Transfer, McGraw Hill, Eighth (SI) Edition.
Purpose
This verification case demonstrates using the view factor generator and surface-to-surface solver with periodic
interfaces. The objective is to predict the radiative heat flux on the inner and outer walls of a cylinder.
Acceptance Criteria
The acceptance criterion for the comparison is the maximum unsigned relative difference in radiation heat flux
between the predictions of the simulation and the analytical calculations, as shown in the following table.
Overview
The problem geometry consists of two concentric cylinders of inner radius 0.3 m and outer radius 1.0 m. The
cylindrical surfaces radiate with prescribed temperatures, and periodic conditions are prescribed on the ends of
the geometry.
Physics
The problem is solved using the coupled flow and energy solver. The radiation heat flux on the inner and outer
walls of the cylinder are compared to the reference solution for infinite concentric cylinders in Holman [34],
Fig. 8-30. For this model, the analytical solution is -40110.8 W/m^2 on the inner wall, and 12033.3 W/m^2 on
the outer wall.
Physics Models:
• Constant Density
• Coupled Energy
• Coupled Flow
• Gas (Air)
• Gradients
• Gray Thermal Radiation
• Laminar
• Steady
• Surface-to-Surface Radiation
• Three-Dimensional
• Viewfactors calculator
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps 40
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[34] Holman, J.P. 2001. “Heat Transfer”, Eighth (SI) Edition., McGraw-Hill.
Purpose
These two cases demonstrate the use of baffles for surface-to-surface radiation across a single shield. In one
case the shield is opaque, and in the other the shield is fully transmissive.
Acceptance Criteria
The acceptance criterion for the comparison is the maximum unsigned relative difference in total radiation
power between the predictions of the simulation and the analytical calculations, as shown in the following
table.
Overview
These cases work with Test Example 8-9 from Holman [35]. In a cube, two parallel facing walls radiate with a
prescribed temperature. The remaining outside surfaces are specified as symmetry boundaries. The two
radiating walls are separated by one baffle which acts as a shield for radiation -- in one case opaque and in the
other case fully transmissive. The radiation flux reduction by the shield can be expressed by an analytical
function.
This case validates radiative parameters such as heat flux due to radiation on boundaries, and the effect of a
shield on incoming and outgoing radiation in the presence or absence of the shield.
The grid consists of 216 hexahedral cells, 504 quadrilateral interior faces, and 343 vertices.
Physics
This three-dimensional stationary case is single-phase and single-component. It involves surface-to-surface
radiation using the Gray thermal radiation model with internally-calculated viewfactors.
The simulation results are compared to analytical data for total radiation power.
T14 − T24
qno shield = A Σ 1 1
(22)
ε1
+ ε2
−1
T14 − T24
qshield = A Σ 1 1 1 1
(23)
ε1
+ ε3
−1 + ε3
+ ε2
−1
2
qno shield − qshield ε3
−1
= (24)
qno shield 1
+
1
−1 +
1
+
1
−1
ε1 ε3 ε3 ε2
Σ = 5.67 × 10−8 W/m2K4
qno shield = 1360.4 W (25)
qshield = 92.706 W
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Gray Thermal Radiation
• Laminar
• Radiation
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Surface-to-Surface Radiation
• Three Dimensional
• View Factors Calculator
Material Properties:
Initial Conditions:
Boundary Conditions:
Opaque:
Fully Transmissive:
Interface Conditions:
Maximum Steps 30
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[35] Holman, J.P. 2001. Heat Transfer, Eighth SI Metric Edition, McGraw Hill.
Purpose
This case demonstrates surface-to-surface radiation for the situation of semi-transparent external walls.
Overview
This case uses a cube with the top, sides, and bottom semi-transparent. Temperatures are fixed on the top and
sides, and are adiabatic on the bottom. The viewfactors and fluxes may all be calculated analytically.
Physics
This three-dimensional stationary case is single-phase and single-component. It involves surface-to-surface
radiation using the Gray thermal radiation model with internally-calculated viewfactors.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• Gray Thermal Radiation
• Laminar
• Radiation
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Surface-to-Surface Radiation
• Three Dimensional
• View Factors Calculator
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps 40
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[36] Fiala, A., 2006. "Solar & thermal radiation FASTRAC model validation", CD-adapco, Applications
Proving Group, CD/130406/R1.
Purpose
This verification case simulates the focusing of light through a lens. Given the lens shape/curvature, the focal
point is known using an analytical expression. The test checks whether the light is being focused appropriately
around the focal point in Simcenter STAR-CCM+.
Overview
There is a small circular disc, the collector, at the focal plane of the lens, and the light should focus entirely
within that circular disc. The integrated boundary irradiation on the circular disc is used as the correctness
criterion. This integrated quantity, along with the scalar scene for the boundary irradiation on the circular disc,
indicates the result.
Physics
The radiation model for this case is Surface Photon Monte Carlo (PMC).
The focal length for the lens is based on the lensmaker’s equation ([37], [38]):
1 1 1 n−1 d
= n−1 − + (26)
f R1 R2 nR1R2
where:
• d is lens thickness
• f is lens focal length
• n is refractive index of the lens material
A lens geometry was considered with R2 as infinity (that is, a flat surface) and d > 0. R1 was set to 0.1562 m,
and n was set to 1.5 for the lens (the approximate refractive index of glass). From the above equation the focal
length f was calculated to be 0.3124 m.
Physics Models:
• Constant Density
• Gradients
• Gray Thermal Radiation
• Radiation
• Refraction (Gray)
• Segregated Solid Energy
• Solar Loads
• Solid
• Steady
• Surface Photon Monte Carlo
• Three Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[37] Greivenkamp, J. E. 2004. "Field Guide to Geometrical Optics", SPIE Field Guides v. FG01. SPIE.
[38] Hecht, E. 1987. "Optics", (2nd ed.). Addison Wesley. Chapters 5 & 6.
Combustion
Contents:
Sandia Piloted CH4/Air Jet Flame D
Turbulent Lifted Hydrogen Flame
C2H4/Air Coppalle Jet Flame
Purpose
The Sandia jet flame appears in multiple verification cases to simulate a turbulent, non-premixed flame using a
variety of combustion modeling techniques.
Overview
The Sandia piloted methane-air jet flame geometry was developed by Sydney University [39].
The two-dimensional axisymmetric geometry consists of the main jet with a nozzle diameter of 7.2 mm,
burning a premixture of 25% methane and 75% dry air by volume. The pilot inlet surrounds the nozzle. The
inlet has a diameter of 18.2 mm. The pipe walls are modeled with zero thickness.
• temperature
• mass fractions of CH4, O2, and CO
• the centerline
• values for z/D of 15, 30, and 45, where z is the axial location and D is the nozzle diameter of 0.0072 m.
Physics
For the Sandia piloted methane-air jet flame, the pilot is operated lean, φ = 0.77, and the flow rate is scaled in
the turbulent flame to maintain the pilot at approximately 6% of the power of the main flame.
Five combustion models are tested on an axisymmetric two-dimensional mesh, with steady segregated
turbulent flow and a non-premixed flame:
• The hybrid Eddy Break-up (EBU) model is used with NOx thermal emission and a NOx Flamelet library.
Turbulence is modeled with the Standard K-Epsilon model and high-y+ wall treatment.
• The Non-Adiabatic PPDF model is used with NOx thermal emission and PPDF Equilibrium. Turbulence is
modeled with the Standard K-Epsilon model and two-layer all-y+ wall treatment.
• The Adiabatic PPDF model is used with NOx thermal emission and PPDF Flamelet. Turbulence is modeled
with the Realizable K-Epsilon Two-Layer model and two-layer all-y+ wall treatment.
• The Flamelet Generated Manifold (FGM) is used, taking into account participating media radiation using
the Discrete Ordinate Method (DOM). Turbulence is modeled with the Realizable K-Epsilon Two-Layer
model and two-layer all-y+ wall treatment.
• The Complex Chemistry model is used with the Eddy Dissipation Concept (EDC). Turbulence is modeled
with the Realizable K-Epsilon Two-Layer model and two-layer all-y+ wall treatment.
Physics Models:
• Axisymmetric
• Exact Wall Distance
• Gradients
• Ideal Gas
• K-Epsilon Turbulence
• Multi-Component Gas
• Non-Premixed Combustion
• Reacting
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Steady
• Turbulent
• Emissions PPDF
• High-y+ Wall Treatment
• NOx Emission/NOx Thermal/NOx PPDF Equilibrium
• Segregated Fluid Enthalpy
• Segregated Species
• Standard K-Epsilon
Material Properties:
Non-Adiabatic PPDF
Adiabatic PPDF
Complex Chemistry
CO CO2 H H2
Molecular weights 28.01 kg/kmol 44.009 kg/kmol 1.008 kg/kmol 2.016 kg/kmol
Density Ideal gas
Dynamic viscosity 1.716 x 10–5 Pas
Specific heat Determined using thermodynamic polynomial data
N2 O O2 OH
Molecular weights 28.014 kg/kmol 16 kg/kmol 31.998 kg/kmol 17.007 kg/kmol
Density Ideal gas
Dynamic viscosity 1.716 x 10–5 Pas
Specific heat Determined using thermodynamic polynomial data
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[39] Masri, A. R., Dibble, R. W., and Barlow, R. S. 1996. “The structure of turbulent nonpremixed flames
revealed by Raman-Rayleigh-LIF measurements” Prog. Energy Combust. Sci. 22(4): pp. 307-362.
[40] Barlow, R., and Frank, J. 2007. “Piloted CH4/Air Flames C,D,E, and F—Release 2.1” Sandia National
Laboratories, Livermore, CA.
[41] van Oijen, J.A. and de Geoy, L.P.H. 2000. “Modelling of premixed laminar flames using flamelet-
generated manifolds” Combust. Sci. Technol., 161:113.
Purpose
This simulation of a lifted hydrogen flame uses turbulent combustion with segregated flow and partially-
premixed combustion.
Overview
This lifted-flame burner consists of a jet flame in a coaxial flow of hot combustion products from a lean
premixed flow (vitiated flow). The central fuel jet with a diameter of 4.57 mm, an averaged velocity of 107
m/s, a temperature of 305 K, and a Reynolds number of 23,600, is surrounded by a slow and hot coflow with a
diameter of 210 mm and a velocity of 3.5 m/s. The central fuel is a 25/75% (by volume) H2/N2 mixture. The
vitiated coflow with a temperature of 1045 K is the product from a H2/Air flame with a mixture fraction of
0.25.
Physics
For steady segregated turbulent flow and a partially premixed flame, the Coherent Flame model is used with
the Gulder laminar flame speed method and Adiabatic PPDF. Turbulence is modeled with the Realizable K-
Epsilon Two-Layer model and two-layer all-y+ wall treatment.
Physics Models:
• Adiabatic Partially-Premixed CFM
• Adiabatic PPDF
• Axisymmetric
• CFM Partially-Premixed Reaction Model
• Coherent Flame Model (CFM)
• Gradients
• K-Epsilon Turbulence
• Multi-Component Gas
• Partially-Premixed Combustion
• PCFM Ideal Gas
• PPDF Flamelet
• Reacting
• Realizable K-Epsilon Two-Layer
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Steady
• Turbulent
• Two-Layer All-y+ Wall Treatment
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[42] Cabra, R., Myhrvold, T., Chen, J.Y., Dibble, R.W., Karpetis, A.N., and Barlow, R.S. 2002. “Simultaneous
Laser Raman-Rayleigh-LIF Measurements and Numerical Modeling Results of a Lifted Turbulent
H2/N2 Jet Flame in a Vitiated Coflow”, 29th International Symposium on Combustion, the
Combustion Institute, Sapporo, Japan.
Purpose
This verification case simulates a turbulent, non-premixed flame with soot.
Overview
The experimental work was focused on soot production in turbulent hydrocarbon flames. Three diffusion
flames produced by turbulent jets of pure ethylene in still air were considered in the experiment. Soot
production was measured using the optical system described and illustrated in [43].
The given test case simulates soot production for the first flame type described in the experimental work:
where X is axial distance and Xm is the location of axial maximum of mean Soot Volume Fraction.
Physics
The Participating Media Radiation (DOM) model is used with the Weighted Sum of Gray Gases (WSGG) method
for the mixture absorption coefficient. The Optical Path Length (OPL) property of the Weighted Sum of Gray
Gases method node was set to 0.012, and was obtained using the following expression:
3
OPL = 0.1 Vol (27)
where Vol is the volume of the computational domain that has a temperature greater than a threshold value
(for example 1000K), signifying that it is the combustion zone.
Physics Models:
• Axisymmetric
• Flamelet
• Gradients
• Gray Thermal Radiation
• High-y+ Wall Treatment
• Ideal Gas
• K-Epsilon Turbulence
• Multi-Component Gas
• Non-Adiabatic
• Non-Premixed Flame
• Participating Media Radiation (DOM)
• Radiation
• Reacting
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Enthalpy
• Soot Emissions
• Soot Moments
• Standard K-Epsilon
• Steady
• Steady Laminar Flamelet (SLF)
• Turbulent
• Wall Distance
Initial Conditions:
Mixture Fraction 0 (indicates that the domain is filled with the oxidizer)
Pressure (Pa) 0.0
Static Temperature (K) 300
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[43] Coppalle, A. and D. Joyeux, D. 1994. "Temperature and Soot Volume Fraction in Turbulent Diffusion
Flames: Measurements of Mean and Fluctuating Values", Combustion and Flame 96(3), pp.
275-285.
Lagrangian Multiphase
Contents:
Evaporation of Multi-Component Droplets in Dry Air
Turbulent Dispersion of Material Particles in Grid-Generated Turbulence
Droplet Atomization
Purpose
This verification case demonstrates using the Multi-Component Droplet Evaporation model of
Simcenter STAR-CCM+.
Overview
This case simulates the experiments of Daïf et al. [44]. Various fuel droplets were suspended on a permanent
holder within a thermal wind-tunnel and allowed to evaporate in a moving heated air stream. The droplets had
various compositions of heptane [C7H16] and decane [C10H22].
Physics
In the experiment, droplets were held stationary and the heated air was moving. In the simulation, however,
the situation is reversed. The air within the domain is static, and at a constant temperature, with various size
droplets moving at the prescribed experimental velocity and temperature. These conditions are given in the
following table:
The default properties are used for air and heptane vapor. Otherwise:
• The decane vapor properties are mostly from Abramzon and Sirignano [45].
• The liquid heptane properties are mostly default. The Wagner coefficients are taken from Reid, Prausnitz,
and Poling [46]. The latent heat function is a Watson correlation [46] using boiling-point data from
Perry's [47].
• The liquid decane properties are mostly from Reid, Prausnitz, and Poling [46] and Perry's [47]. The latent
heat function is a Watson correlation [46] taken from Abramzon and Sirignano [45].
• The molecular diffusivity of the air/fuel mixture is taken to be 0.06265 cm2/s. This value is a 50/50 blend
of the heptane and decane diffusiveness in air, from Brenn et al. [48].
The problem is solved as a three-dimensional, unsteady, laminar flow case with Lagrangian multiphase, multi-
component gas, multi-component droplet evaporation, and ideal gas options.
The experiments of Daïf et al. [44] are used for comparison, providing the droplet size as a function of time
and also temperature as a function of time for some droplets. The STAR-CCM+ results are found to be
comparable to results from Torres et al. [49].
Physics Models:
• Gradients
• Ideal Gas
• Lagrangian Multiphase
• Laminar
• Multi-Component Gas (Air, C7H16, C10H22)
• Non-reacting
• Segregated Flow
• Segregated Fluid Temperature
• Segregated Species
• Three Dimensional
Material Properties:
Air
Heptane (C7H16)
Decane (C10H22)
Initial Conditions:
Boundary Conditions:
Injector Conditions:
To download the final simulation file to your own directory structure, right-click the link in this column and
choose “Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[44] Daïf, A., Bouaziz, M., Chesneau, X. and Ali Chérif, A., "Comparison of multicomponent fuel droplet
vaporization experiments in forced convection with the Sirignano model", Exp. Thermal Fluid Sci.,
Vol. 18(4), 1998, pp. 282--290.
[45] Abramzon, B. and Sirignano, W.A., "Droplet vaporization model for spray combustion calculations",
Int. J. Heat Mass Trans., Vol. 32(9), 1989, pp1605--1618.
[46] Reid, R.C., Prausnitz, J,M. and Poling, B.E., "The Properties of Gases and Liquids", 4th ed., McGraw-
Hill, 1987.
[47] Perry, R.H. and Green, D., "Perry's Chemical Engineer's Handbook", 6th ed., McGraw-Hill, 1984.
[48] Brenn, G., Deviprasath, L.J., Durst, F. and Fink, C., “Evaporation of acoustically levitated multi-
component liquid droplets”, Int. J. Heat Mass Trans., Vol. 50(25-26), 2007, pp5073--5086.
[49] Torres, D.J., O'Rourke, P.J. and Amsden, A.A., “Efficient multicomponent fuel algorithm”, Combust.
Theory Model., Vol. 7, 2003, pp67--86.
Purpose
This verification case demonstrates using the turbulent dispersion model in Simcenter STAR-CCM+.
Overview
The simulation is based on the experiments of Snyder and Lumley [50], in which particles of different materials
are injected into a flow with decaying grid-generated turbulence. The domain represents the flow in a wind
tunnel downstream of the grid, with the inlet boundary positioned at the location of the grid. The inlet velocity
is the tunnel speed of 6.55 m/s. The working section of the tunnel has a 200 mm square cross-section, and is
oriented vertically so that gravity acts in the streamwise direction.
Particles of four different materials are used, as shown in the table below. 800 parcels of each type are injected
at the first time-step and allowed to disperse for 0.5 s. The mean square displacement is calculated for each
particle type.
Physics
The turbulence profile in the tunnel uses the following curve fit equations [50]:
U2 x
= 42.4 − 16.0 (28)
u2 M
U2 x
= 39.4 − 16.2 (29)
v2 M
1 2
k= u + v2 (30)
2
U3 1 2
ε= + (31)
2M x 2 x 2
42.4 M
− 16.0 39.4 M
− 16.2
In the experiments, the particles were injected at x/M = 20 through a small pipe, with a velocity equal to the
flow speed. Their positions were measured at a number of locations downstream of x/M = 68.4, with the
displacement being given relative to their position at x/M = 68.4. In the simulation therefore, the particles are
injected into the stream from a location at x/M = 68.4 with a velocity equal to the flow speed. The
displacements are then calculated relative to the injection point.
The problem is solved as a three-dimensional, unsteady, turbulent flow case with Lagrangian multiphase and
turbulent dispersion options. The gas is assumed to be air with constant material properties.
The experiments of Snyder and Lumley [50] are used for comparison. The experiments give the mean square
displacement and the fluctuation velocity of each particle type as a function of time.
Physics Models:
• Boussinesq Model
• Constant Density
• Gas (Air)
• Gradients
• Gravity
• High y+ Wall Treatment
• Implicit Unsteady
• K-Epsilon Turbulence
• Lagrangian Multiphase
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature
• Standard K-Epsilon
• Three Dimensional
• Turbulent
Material Properties:
Initial Conditions:
In order to set an initial condition that satisfies the equations of the Standard K-Epsilon model, the analytical
profiles for decaying turbulence are taken from Eqn. 10.57 in Pope [51].
t −n
k t = k0 (32)
t0
t − n+1
ε t = ε0 (33)
t0
When fitted to the experimental data at x/M = 41, the coefficients are found to be k0 = 0.059227 and
ε0 = 0.5430. User field functions for the initial k and ε fields are defined in the simulation by assuming
n = 1/0.92, t0 = 0.11856, x0 = 0.2648 and t = x − x0 /U.
Using these initial conditions for turbulence, along with the velocity U = 6.55 m/s , the flow satisfies the
equations exactly at the first time-step. This means that the particle injection can start immediately without
having to first let the flow develop.
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[50] Snyder, W.H. and Lumley, J.L., “Some measurements of particle velocity autocorrelation functions in
a turbulent flow”, Journal of Fluid Mechanics, Vol. 48, Part 1, 1971, pp.41-71.
[51] Pope, S.B., “Turbulent Flows”, Cambridge University Press, 2000, pp.498-500.
Droplet Atomization
Purpose
This verification case compares results of a Simcenter STAR-CCM+ spray simulation with experimental data
from the studies of Liu, Mather, and Reitz ([52], [53]).
Overview
Atomization forms an important stage in several applications that involve fuel spray combustion. Within its
Lagrangian multiphase modeling, Simcenter STAR-CCM+ provides several models for droplet breakup and
atomization.
The setup for the case comes from the experiment in which liquid fuel droplets (Benz UCF fuel) are released in
an air cross flow using a point injector.
The droplet diameter is 1.7 x 104 m and the injection velocity is 16 m/s. In the experiment [52], nine different
air jet speeds (Cases 1-9) are studied to observe the different droplet breakup modes. This verification case
compares Case 4, which has an air velocity of 100 m/s, a Weber number of 102, and a Reynolds number of
1133.
Geometry Mesh
Physics
The Lagrangian droplet breakup models in this instance are KHRT, TAB, and Reitz-Diwakar. This case uses the
bag breakup regime, in which the non-uniform pressure field around the droplet causes it to expand in the
low-pressure wake region and eventually disintegrate when surface tension forces are overcome.
Reference data has been added only to test results for trajectory and diameter for the Reitz-Diwakar model, as
tuned according to reference sources.
Physics Models:
• All-y+ Wall Treatment
• Gas
• Gradients
• Gravity
• Ideal Gas
• K-Omega Turbulence
• Lagrangian Multiphase
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature
• SST (Menter) K-Omega
• Steady
• Three Dimensional
• Turbulent
• Wall Distance
Material Properties:
Initial Conditions:
Boundary Conditions:
Injector Conditions:
To download the final simulation file to your own directory structure, right-click the link in this column and
choose “Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[52] Liu, A.B., Mather, D., and Reitz, R.D. 1993. "Modeling the Effects of Drop Drag and Breakup on Fuel
Sprays", SAE Technical Paper 930072.
[53] Liu, A. B. and Reitz, R. D. 1993. "Mechanism of Air Assisted Liquid Atomization," Atomization and
Spray, 3, pp. 55–75.
Multiphase
Contents:
Aeration Tank with Impermeable Boundary for Continuous Water Phase
Polynomial Specific Heat for Pressurized Water
Wall Heat Flux Condition
Bulk Boiling
Bozzano-Dente and Tomiyama Drag Laws (CGD)
Particle Acceleration by Drag
Bulk Boiling Using S-Gamma Size Distribution
Forced Flow Boiling of Sub-Cooled Water in a Vertical Pipe at 45 Bar (CGD)
Forced Flow Boiling of Sub-Cooled Water in a Vertical Pipe at 149.9 bar and 2.21 MW/m2 (CGD)
Rising Terminal Velocity of Bubbles with Drag Laws
Bubble Acceleration under Virtual Mass Force
Tomiyama Transverse Lift Coefficient with Air Bubbles
Bubbly Flow in Water with Variable Density
Fluid Acceleration as a Solid Body
Three Phase Particle Terminal Velocity
Hydrotransport with Turbulence
Hibiki Bubble Column Using S-Gamma Size Distribution Modeling
Increase of Bubble Size Due to Pressure Decrease in an Ideal Gas
Purpose
This verification case demonstrates using an impermeable boundary condition on one phase of a two-phase
flow.
Overview
The problem geometry consists of a two-dimensional, 2 m by 1 m tank, with a 0.1 m wide opening at the
bottom. The tank is initially filled with water. Air bubbles (3 mm in diameter and moving with a velocity of
0.25 m/s) are injected into the bottom opening. The buoyancy force causes the air bubbles to rise through the
tank and leave from the top.
Although no specific reference data is supplied for this case, the results are verified as physically plausible.
Over most of the free surface, away from wall effects, bubbles should be leaving at the rate at which they
approach the boundary from below -- that is, at the terminal velocity for the given bubble size. Based on a
hand calculation using one of the Tomiyama drag laws, this terminal velocity is 0.251 m/s for 3 mm bubbles in
water at 20 deg C. The slip velocity profile test in the accompanying startest case shows that for the cell
centers under the top boundary, about half the points show a slip velocity of around 0.252 m/s.
Physics
The top boundary of the tank is set to phase impermeable for the continuous water phase. This means the
boundary behaves like a wall for this phase and does not allow water to leave from the top. The problem is
solved using the Wang Curve Fit for calculating the phase interaction drag coefficient and the standard K-
Epsilon turbulence model with a high y+ wall treatment for both phases.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Multiphase (Water, Air)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Turbulent
• Two Dimensional
Material Properties:
Water Air
Density (kg/m^3) 1000.0 1.0
Dynamic Viscosity (Pa-s) 0.001 1.0E-5
Initial Conditions:
Water Air
Pressure (Pa) 0.0 0.0
Turbulence Intensity 0.01 0.01
Turbulent Velocity Scale 1.0 1.0
(m/s)
Turbulent Viscosity Ratio 10.0 10.0
Inlet Velocity (m/s) X = Y = Z = 0.0 X = 0.0, Y = 0.1, Z = 0.0
Inlet Volume Fraction 0.999 0.001
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates using a polynomial to specify the specific heat of a liquid phase at high
pressure within a Multiphase model.
Overview
The problem geometry is a channel, 3.66 m in length and a 1 m square section. The applied symmetry
conditions on the sides of the channel result in a one-dimensional flow. A uniform heat source is applied to the
liquid phase and takes the liquid from the inlet temperature to the saturation point. The air particles, that are
introduced at the inlet, are assumed to be spherical, 0.01 mm in diameter. All physical properties are constant
at the saturation value, except the specific heat of the liquid, which a six-degree polynomial represents. The
initial fields are set up using inlet values plus a small slip.
Physics
The polynomial for the specific heat of the liquid is the derivative of the enthalpy polynomial fit. The
polynomial for the specific heat is shown below:
The Schiller-Naumann drag law and an interphase energy transfer Nusselt number of 2 are specified in the
simulation. The problem is solved using a second order differencing scheme.
For the reference solution, if a realistic model for specific heat is used, the temperature rise reaches saturation
temperature. The temperature rise is:
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase (Liquid, Vapor)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Phase Coupled Fluid Energy
• Steady
• Three Dimensional
Material Properties:
Initial Conditions:
Liquid Air
Pressure (Pa) 0.0 0.0
Velocity (m/s) 6.03 6.04
Volume Fraction 0.9999 1.0E-4
Boundary Conditions:
Liquid Air
Inlet T = 566.55 K; U = 6.03 m/s T = 617.95 K; U = 6.03 m/s
Outlet T = 617.95 K T = 617.95 K
Maximum Steps 50
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates using the multiphase enthalpy equation and the wall heat flux condition in
high Reynolds number turbulent flows.
Overview
The problem geometry is a duct, 3.81 m in length and 4.953 mm square section, representing an equivalent
diameter (De) of 19.81 mm. Symmetry conditions are applied on three sides of the duct, resulting in a two-
dimensional flow. The fourth side is treated as a heated wall. A homogeneous liquid and vapor mixture,
initially at 500 K, is introduced into the duct with a uniform heat flux applied on the wall. The maximum
temperature occurs on the wall next to the outlet, representing the heat transfer between the wall and the
bulk mixture temperature under fully developed conditions. To simulate homogeneous flow and temperature
conditions, the vapor bubbles are assumed to be small, with a constant diameter of 0.1 mm.
Physics
The verification test is based on a full range of vapor volume fractions that are covered in four different
simulations (0.0, 0.3, 0.8, 1.0). The mass flux Gis set to 1497 kg/m^2 and heat flux qwall to 526300 W/m^2 for
all cases.
The flow is assumed to be turbulent, implementing the high-Re K-epsilon turbulence model with a high y+ wall
treatment. The problem is solved using the Schiller-Naumann drag law and a Ranz-Marshall heat transfer
coefficient. A phase coupled energy model is used, with the enthalpy form of energy equation and wall heat
flux boundary condition. An under-relaxation factor of 0.3 is used on the volume fraction system to ensure that
all cases converge.
The maximum temperature occurs on the wall next to the outlet and represents the overall effect of the wall
on the phase heat transfer and convection and turbulent diffusion in the bulk. It is compared to results from
single phase simulations and reference values calculated using heat balance.
qwall 4L
Tout = Tin + (35)
Gcp De
qwall
Twall,out = Tout + (36)
H
where cp is the specific heat of the mixture and L is the length of the duct. The convective heat transfer
coefficient, H, is defined as
k
H = Nu (37)
De
Nu = 0.023Re0.8Pr0.4 (38)
Material Properties:
Liquid Vapor
Density (kg/m^3) 736.9 37.5
Viscosity (Pa-s) 9.38E-5 1.92E-5
Specific Heat (J/kg-K) 5373 5336
Thermal Conductivity (W/m- 0.5621 0.0647
K)
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates using the bulk boiling model, phase reference enthalpies and latent heat
that are computed directly from the phase enthalpy.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned differences in vapor volume
fraction between the simulation and the analytical calculation. The threshold value is shown the following
table.
Overview
The solution domain is a duct measuring 2 m in length with a 0.1 m x 0.1 m cross section. The applied
symmetry conditions on the sides of the channel result in a one-dimensional flow. The flow regime consists of
superheated water, which is defined as the continuous phase, entering the bottom of the channel with a 2K
superheat (that is, 2K above saturation). Some of the water evaporates into steam bubbles, and a water/steam
mixture emerges from the pressure boundary outlet at the top. The steam bubbles are assumed to be spherical
particles, 1 mm in diameter.
Physics
The flow is assumed to be laminar with a gravitational acceleration of 9.81 m/s^2 acting down the duct. A
liquid-side Nusselt Number is set to 2 and the Tomiyama drag law is implemented. The problem is solved using
a Phase Coupled Energy Model with relaxation factors of 0.7 for momentum, 0.3 for pressure, 0.3 for volume
fraction and 0.5 for energy. The values are sufficient for convergence, provided the reference density is
958 kg/m^3 and the initial fields contain a non-zero value of slip, such as 0.5 m/s for the liquid and 0.501 m/s
for the vapor.
Since the duct cross section is uniform, the total mass flux G is also uniform. Neglecting the contribution of
kinetic energy, there is no change in the system enthalpy ℎ though this adiabatic process. At equilibrium, the
system enthalpy determines the steam quality x. Neglecting the slip velocity leads to an estimate of volume
fraction α, mixture density ρ and mixture velocity u:
GLINℎLIN + GGINℎGIN
ℎ= (40)
G
ℎ − ℎLSAT
x= (41)
ℎGSAT − ℎLSAT
x
ρG
α= 1−x x
(42)
ρL
+ ρG
ρ = 1 − α ρL + αρG (43)
G
u= (44)
ρ
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase (Liquid, Vapor)
• Multiphase Equation of State
• Multiphase Interaction
• Phase Coupled Fluid Energy
• Steady
• Three Dimensional
Material Properties:
Initial Conditions:
Liquid Vapor
Pressure (Pa) 0.0 0.0
Inlet Velocity (m/s) 0.5 0.501
Inlet Volume Fraction 0.99 0.01
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[54] Collier, J. & Thome, J. 1996. Convective Boiling and Condensation. 3 ed. s.l.:Oxford Science
Publications.
Purpose
This verification case demonstrates using the Bozzano-Dente [55] and Tomiyama [56] drag coefficients to
determine the rising terminal velocity of differently sized bubbles in water.
Acceptance Criteria
The acceptance criterion for the comparison is the average relative difference in terminal velocity between the
predictions obtained from the simulation and the reference values. The average values were calculated from
all the bubble sizes considered for each drag coefficient. The threshold value is shown in the following table.
Overview
The problem geometry is a 5 m tall column, having a 1 m x 1 m cross section. The applied symmetry conditions
on the sides of the column result in a one-dimensional flow. The domain is divided into 10 cells and is set up
with water flowing at 1 m/s. A set of 10 dilute dispersed phases, with diameters ranging from 0.046 mm to
46 mm, are introduced at the bottom of the column. The dispersed bubbles accelerate to a terminal velocity
relative to the continuous water.
Physics
The Bozzano and Dente drag coefficient is:
a 2
CD = f (45)
Ro
48 1 + 12Mo1/3 Eo3/2
f= + 0.9 (46)
Re 1 + 36Mo1/3 1.4 1 + 30Mo1/6 + Eo3/2
2
and a/Ro is a deformation factor
a 2 10 1 + 1.3Mo1/6 + 3.1Eo
≅ (47)
Ro 10 1 + 1.3Mo1/6 + Eo
The Tomiyama drag coefficient for a moderately contaminated water system is:
24 8 Eo
CD = max 1 + 0.15Re0.687 , (48)
Re 3 Eo + 4
with Re , Mo and Eo being the Reynolds number, the Morton number and the Eotvos number, respectively.
The terminal velocity is reached when the buoyancy force balances the drag force and is calculated as
4 ρL − ρG gDo
UT = (49)
3 ρL CD
The flow is assumed to be laminar with a gravitational acceleration of 9.81 m/s^2 acting down the column.
The computed terminal velocity for each bubble is compared to the analytical solution found for the Bozzano-
Dente drag law simulation Eqn. (45) and for the Tomiyama (contaminated) drag law simulation Eqn. (48) after
insertion into Eqn. (49).
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase (Liquid, Bubbles 1-10)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Three Dimensional
Material Properties:
Initial Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[55] Bozzano, G. and Dente, M., “Shape and terminal velocity of single bubble motion: a novel approach”,
Computers and Chemical Engineering, vol. 25, p. 571-576, 2001.
[56] Tomiyama, A., Tamai, H., Zun, I., and Hosokawa, S., “Transverse migration of single bubbles in
simple shear flows”, Chemical Engineering Science, vol. 57, p. 1849-1858, 2002.
[57] Clift, R., Grace, J.R., and Weber, M.E., “Bubbles, drops and particles”, Academic Press, 1978.
Purpose
This verification case demonstrates the response of slow dense particles to a free stream gas velocity in a
rectangular duct.
Overview
The solution domain is a 2 m by 0.1 m two-dimensional channel. A trace concentration of heavy particles, each
0.5 mm in diameter, is introduced at the inlet with a velocity of 1 m/s. The gas within the duct is flowing at 5
m/s. The drag force that results from the multiphase interaction gradually accelerates the particles.
Physics
The problem is solved using laminar flow with a constant drag coefficient of 0.44. The computed particle
velocity profile along the duct is compared to a reference solution, determined using the equation of motion
for a particle:
2
duP 3 CDρA uP+−uA
ρPuP = (50)
dx 4 dP
where ρ is density, u is the velocity, CD is the drag coefficient and d is the diameter. Subscripts P and A denote
quantities for particle and air, respectively.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Laminar
• Multiphase (Air, Particle)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps 75
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates the behavior of the S-Gamma model for size distribution when used with
interphase mass transfer, in particular bulk boiling.
Overview
The model geometry, material properties and initial conditions are the same as in the bulk boiling case. In
addition to the bulk boiling test, the S-Gamma model is enabled for the vapor phase and the interaction length
scale is set equal to the Sauter mean diameter (d32) of the vapor. The inlet diameter is 1 mm with zero inlet
size variance. Breakup and coalescence are not enabled, meaning that the variance remains zero along the
duct.
Physics
If a mass transfer rate (in kg/m3) for a phase is equal to ṁ ′′′ [58], the interfacial area (and hence S2) source is:
Here, ρ is the density and α is the volume fraction of the dispersed phase. The effect of increasing the diameter
on the overall mass transfer is automatically included in the calculations when using the S-Gamma Sauter
mean diameter.
The problem is solved using the two equation option for the S-Gamma model, calculating both the d32 (found
from volume fraction and S2) and d30 (found from volume fraction and S0). As mass transfer does not affect
the steam bubble number density, the value of d30 remains independent of the S-Gamma implementation of
mass transfer. In this setup, the d30 and d32 remains equal, and are compared as a check on the mass transfer
in S-Gamma implementation.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase (Liquid, Vapor)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Phase Coupled Fluid Energy
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Liquid Vapor
Pressure (Pa) 0.0 0.0
Velocity (m/s) 0.5 0.501
Volume Fraction 0.99 0.01
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[58] Yao, W. and Morel, C., “Volumetric interfacial area prediction in upward bubbly two-phase flow”,
International Journal of Heat and Mass Transfer, Vol 47, p. 307 - 328, 2004.
Purpose
This verification case demonstrates the wall boiling model in Simcenter STAR-CCM+, comparing results to
experimental void and temperature data for very high-pressure forced convective boiling. The flow is of forced
convection type and the operating pressure is 45 bar.
Acceptance Criteria
The acceptance criteria for the comparison are shown in the following table. The acceptance criterion is the
difference in axial vapor mean volume fraction, averaged over all the measurements available in the reference.
The acceptance threshold is consistent with the measurement uncertainty in Bartolomei et al. ([59], [60]).
Overview
The geometry used is a Bartolomei ([59]) boiling tube, where subcooled water flows upward through a 2-
meter-tall vertical pipe. The pressure at the outlet is 45 bar. Heat is applied uniformly at the pipe wall.
The basic geometry is a tube of diameter 15.4 mm and heated length of 2 m. To allow some flow development
between the inlet and the heated section, a short unheated leader of 0.2 m has been added. Similarly so that
the outlet conditions in this developing flow do not influence the end of the heated section, a short trailer has
also been added, with 0.2 m length.
The geometry’s discretization comprises 2300 quad cells as shown in the following illustration. The number of
radial cells is 20 and the number of axial cells 115. All cells have equal dimensions.
Physics
The two phases, liquid and vapor, which are flowing through the section of vertical pipe represent water and
steam. Subcooled liquid is introduced at the inlet, with a subcooling of 59.96 K and a mass flux of 900 kg/m2-s.
As already mentioned, the system pressure is 45 bar. Results are compared with data from the literature ([59]).
A K-Epsilon model with a two-layer wall treatment is selected for the continuous phase so that it can be used
under a wide range of flow conditions. The Turbulence Response model is an appropriate model for the
velocity fluctuations that are experienced in the dispersed phase for this bubbly flow.
The S-gamma model is activated for the dispersed phase so that it can track bubble size starting from bubble
creation at the boiling wall, and including the effects of breakup, coalescence, and (most importantly)
condensation.
A combination of Hibiki-Ishii Nucleation Site Number Density and Kocamustafaogullari Bubble Departure
Diameter is used here, which is applicable for water boiling under a wide range of pressures.
The Ranz-Marshall correlation, in turn, is used to calculate the heat transfer coefficient for the heat transfer
from the bulk water to the bubble interface. Regarding the heat transfer between the vapor phase and bubble
interface, a fast superficial response is assumed due the effect of internal mixing within the vapor, represented
by selecting a large Nusselt number such as 26.0. The Tomiyama model is used for modeling bubbly flows.
Drag reduction is not expected for small spherical bubbles.
The Turbulent Dispersion Force setting is important for wall boiling calculations in the presence of
condensation because it treats the differencing of rapidly changing volume fraction gradients accurately.
In this high-pressure calculation where wall dryout is not expected, the vapor remains at saturation
temperature. Therefore it is valid to assume constant vapor properties and constant interface properties.
The liquid on the other hand can experience a significant change in density and in specific heat while it is
heated from a subcooling of 60 K towards saturation temperature. Nevertheless a useful analysis can be made
using constant properties for the liquid by respecting the following two conditions:
1. In order to obtain the correct heat balance, while also using the experimental inlet temperature, the
following definition of liquid specific heat was chosen:
ℎLSAT − ℎLIN
Cp = (52)
TSAT − TIN
• Constant Density
• Exact Wall Distance
• Flow
• Reynolds-Averaged Navier-Stokes
• Segregated Fluid Temperature
• Turbulent
• K-Epsilon Turbulence
• Liquid
• Realizable K-Epsilon Two-Layer
• Two-Layer All y+ Wall Treatment
• Gas
• Issa Turbulence Response Model
• Particle Size Distribution
• S-Gamma (One Equation)
• Turbulence Response
Phase Interactions (of Multiphase Interaction) and Settings:
• Boiling Mass Transfer Rate:
◦ Continuous Phase Nusselt Number: Ranz-Marshall Correlation
◦ Disperse Phase Nusselt Number: 26.0 (constant)
• Continuous-Dispersed Phase Interaction:
◦ Continuous Phase: Liquid
◦ Dispersed Phase: Vapor
• Drag Force
◦ Drag Coefficient: Tomiyama
◦ Drag Correction: Volume Fraction Exponent (continuous) = 0.0
• Interaction Area Density (Symmetric)
• Interaction Length Scale (Field function: Sauter Mean Diameter resulting from the S-Gamma solution,
limited to the range from 1 μ to 12 mm)
• Interphase Mass Transfer
• Multiphase Material:
◦ Boiling Temperature (K): 530.59
Material Properties:
Initial Conditions:
Reference Values
Boundary Conditions:
0.0, x<0 m
570.0, 0 ≤ x ≤ 2 m
0.0, x>2 m
(53)
Outlet Pressure Outlet - Pressure Static Temperature (K) = 470.63 Sauter Mean Diameter (m) = 0.001
= 0.0 Pa Turbulence Intensity = 0.01 Static Temperature (K) = 530.59
Turbulence Viscosity Ratio = 10.0 Volume Fraction = 0.0
Volume Fraction = 1.0
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[59] Bartolomei, G. G. & Chanturiya, V. M. Experimental Study of True Void Fraction When Boiling
Subcooled Water in Vertical Tubes, Thermal Engineering, 14(2), 1967.
[60] Bartolomei, G., Brantov, V., Molochnikov, Y. An experimental investigation of true volumetric vapor
content with subcooled boiling in tubes, Thermal Eng., 29(3), p. 132–135, 1982.
Purpose
This verification case demonstrates the wall boiling model in Simcenter STAR-CCM+, comparing results to
experimental void and temperature data for very high-pressure forced convective boiling. The flow is of forced
convection type and the operating pressure is 149.9 bar.
Acceptance Criteria
The acceptance criterion for the comparison is the difference in axial vapor mean volume fraction, averaged
over all the measurements available in the reference. The threshold value is 0.05, as shown in the following
table.
Overview
The geometry used is a Bartolomei ([61]) boiling tube, where subcooled water flows upward through a 2-
meter-tall vertical pipe. The pressure at the outlet is 149.9 bar. Heat is applied uniformly at the pipe wall, with
a heat flux of 2.21 MW/m2.
The geometry's discretization comprises 2,000 stretched quad cells, as shown in the following illustration. The
number of radial cells is 20 and the number of axial cells is 100. All cells have equal dimensions.
Physics
The two phases, liquid and vapor, flowing through the section of vertical pipe represent water and steam.
Subcooled liquid is introduced at the inlet, with a subcooling of 154.23 K and a mass flux of 2012 kg/s-m2. As
previously mentioned, the system pressure is 149.9 bar. Results are compared with data from the literature
([61]).
The Standard K-Epsilon model with high-y + wall treatment was selected for both the continuous and
dispersed phases.
The Kurul-Podowski model is activated for the calculation of the bubble sizes. A combination of Hibiki-Ishii
Nucleation Site Number Density and Kocamustafaogullari Bubble Departure Diameter is used here, which is
applicable for water boiling under a wide range of pressures.
The Ranz-Marshall correlation is used to calculate the heat transfer coefficient for the heat transfer from the
bulk water to the bubble interface. Regarding the heat transfer between the vapor phase and bubble interface,
a fast superficial response is assumed due to the effect of internal mixing within the vapor, represented by
selecting a large Nusselt number such as 26.0. The Tomiyama model is used for modeling bubbly flows. Drag
reduction is not expected for small spherical bubbles.
The Turbulent Dispersion Force setting is important for wall boiling calculations in the presence of
condensation because it treats the differencing of rapidly changing volume fraction gradients accurately.
Physics Models:
• Axisymmetric
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Multiphase (Liquid, Vapor)
• Multiphase Equation of State
• Multiphase Interaction
• Phase Coupled Fluid Energy
• Steady
• Turbulent
Phase Physics Models:
Models Common to All Phases:
• Constant Density
• Exact Wall Distance
• Flow
• High y+ Wall Treatment
• K-Epsilon Turbulence
• Reynolds-Averaged Navier-Stokes
• Segregated Fluid Temperature
• Standard K-Epsilon
• Turbulent
Material Properties:
Initial Conditions:
Reference Values
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[61] Bartolomei, G. G. & Chanturiya, V. M. Experimental Study of True Void Fraction When Boiling
Subcooled Water in Vertical Tubes, Thermal Engineering, 14(2), 1967.
[62] Bartolomei, G., Brantov, V., Molochnikov, Y. An experimental investigation of true volumetric vapor
content with subcooled boiling in tubes, Thermal Eng., 29(3), p. 132–135, 1982.
Purpose
This case checks the rising terminal velocity of bubbles of different size in water using the following drag
coefficients:
Acceptance Criteria
The acceptance criterion for the comparison is the average relative difference in terminal velocity between the
predictions obtained from the simulation and the reference values. The average values were calculated from
all the bubble sizes considered for each drag coefficient. The threshold value is shown in the following table.
Overview
A flow of water and 10 dilute dispersed phases with diameters covering the range 46 micron - 46 mm is
introduced at the bottom of a 5m tall duct, with a 1 m x 1 m cross section, as shown in the following
illustration. The applied symmetry conditions on the sides of the column result in a one-dimensional flow. The
dispersed bubbles accelerate to a terminal velocity relative to the continuous water. The computed slip velocity
is used for comparison with the work by Tomiyama and others [64], and by Wang [65].
Physics
The problem is solved with a laminar flow using either the Wang Curve Fit method or the Tomiyama method
for calculating the phase interaction drag coefficient. The Tomiyama method is set to a contamination state of
either Pure or Moderate.
Terminal velocity is reached when the buoyancy force is balanced by the drag force and is calculated as:
4 ρL − ρG gDo
UT = (54)
3 ρL CD
24 8 Eo
CD = max 1 + 0.15Re0.687 , (55)
Re 3 Eo + 4
where Re and Eo are the Reynolds number and Eotvos number respectively.
The empirical correlation proposed by Wang for the drag coefficient is:
Cd = exp a + b ln Re d + c ln Re d 2 (56)
with:
a b c
Red ≤ 1 ln 24 -1 0
1 < Red ≤ 450 2.699467 -0.33581596 -0.07135617
450 < Red ≤ 4000 -51.77171 13.1670725 -0.8235592
Red > 4000 ln (8/3) 0 0
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Three Dimensional
Material Properties:
Initial Conditions:
Pressure (Pa) 0
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[63] Clift, R., Grace, J.R., and Weber, M.E. 1978. Bubbles, drops and particles, Academic Press.
[64] Tomiyama, A., Tamai, H., Zun, I., and Hosokawa, S. 2002. “Transverse migration of single bubbles in
simple shear flows”, Chem. Eng. Sci., 57 pp. 1849-1858.
[65] Wang, D.M. 1994. “Modelling of bubbly flow in a sudden pipe expansion”, Technical Report II-34,
BRITE/EuRam Project BE-4098, 1994.
Purpose
The purpose of this test case is to confirm a physically plausible response of bubbles to the water free stream
velocity; and to protect against unexpected changes in the models, numerics, and solvers of the hydrodynamic
equation system.
The computed bubble velocity profile along the duct is retained as a regression test.
This test is related to the corresponding verification test on particulate flow, Particle Acceleration by Drag.
Overview
A trace concentration of bubbles is introduced at 3 m/s at the inlet to a duct. These bubbles are accelerated by
the interphase drag force towards the velocity of the surrounding water, flowing uniformly at 5 m/s. Virtual
mass force is also relevant for the bubble acceleration rate.
Physics
The Laminar model is used with steady Eulerian Multiphase (EMP) in a two-dimensional mesh, using a constant
phase interaction drag coefficient.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Laminar
• Multiphase (Water, Bubble)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Pressure (Pa) 0
Velocity (m/s) Liquid: U = 5.0
Bubbles: U = 3.0
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This case compares the transverse force on air bubbles in a simple shear flow with published experimental
results. The lift force on the bubbles is modeled using the Tomiyama transverse lift coefficient.
Acceptance Criteria
The acceptance criterion for the comparison is the average relative difference in the lift coefficient. The
average was calculated for all the bubble sizes considered. The relative differences were normalized using the
largest unsigned value from the reference. The threshold value is shown in the following table and consistent
with the correlation accuracy (Tomiyama and others [66]).
Overview
A simple shear flow is set upon lines of the experiment conducted by Tomiyama and others [66]. The physical
parameters for the water phase are listed below. Five different sized bubbles ranging from 2.8 mm to 5.68 mm
in diameter are released at the bottom in the middle of a simple shear flow. The drag force on the bubbles is
modeled using the Tomiyama drag coefficient for pure water systems. For modeling lift force, Tomiyama lift
coefficient is used, given by the following correlation:
Where Re is the Reynolds number, Eod is the modified Eotvos number that is based on the maximum
horizontal dimension of a bubble, and f Eod is given by:
Physics
The Laminar model is used with steady Eulerian Multiphase (EMP) in a two-dimensional mesh, using a
Tomiyama phase interaction lift coefficient.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[66] Tomiyama, A., Tamai, H., Zun, I., and Hosokawa, S. 2002. “Transverse migration of single bubbles in
simple shear flows”, Chem. Eng. Sci., 57 pp. 1849-1858.
Purpose
This test case verifies that Simcenter STAR-CCM+ is capable of modeling bubbly-flow calculations in water at
low pressure (1 bar). In the chosen test case, gas density varies significantly with depth of water.
Overview
The flow of a constant density liquid and an ideal gas is simulated in a 1D vertical channel with a velocity inlet
at the bottom and a pressure boundary at the top. The pressure at the top of the column is set to 1 bar while
the geometry, the fluid properties and the flow parameters are chosen in order to provide a pressure
difference between the bottom and the top of the channel ~1 bar, with the gas density halving from inlet to
outlet.
A small particle diameter is adopted to ensure that the slip between the continuous and the dispersed phase
due to buoyancy is small and negligible when compared to the mixture velocity. Therefore, the expected
results are computed assuming equal velocity for the two phases (homogeneous assumption).
Physics
The Laminar model is used with steady Eulerian Multiphase (EMP) in a two-dimensional mesh, using a
Bozzano-Dente phase interaction drag coefficient. Choosing pressure P as the independent variable, an
analytical solution can be easily derived for the depth x P , the gas density and volume fraction and the
mixture density and velocity:
ρ1 = ρ10 (59)
P
ρ2 = ρ20 (60)
P0
• Mixture density:
dP
= ρg (63)
dx
• Volume conservation:
α1 + α2 = 1 (64)
• Homogeneous assumption:
u1 − u2 « u1 (65)
• Flow quality:
α0ρ20 α0ρ20
X= = (66)
1 − α0 ρ10 + α0ρ20 ρ0
P0
x10 = (67)
ρ10g
P0
x20 = (68)
ρ20g
• Mixture depth:
P P0
x P = 1−X 1− x + Xln x (69)
P0 10 P 20
X
α= ρ2
(70)
P0 1 − X ρ10
+X
• Mixture velocity:
u0ρ0
u= (71)
ρ
where:
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This test verifies the implementation of the transient term and the source term in the multiphase momentum
equation.
Overview
An infinite domain with periodic boundary conditions is initialized with a stagnant mixture of air and water
phases. Each phase then independently undergoes the same solid body acceleration due to the same body
force per unit mass:
dui
ρiαi = 0.1ρiαi (72)
dt
ui = 0.1t (73)
The regression test simply checks the volume average velocity at the end of the integration period.
The full integration period is ten seconds, at the end of which both phases should be moving at 1m/s.
Physics
The Laminar model is used with unsteady Eulerian Multiphase (EMP) in a two-dimensional mesh.
Physics Models:
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This case simulates the terminal velocity of particles falling with a constant drag coefficient. It checks that
Simcenter STAR-CCM+ can continue to run three phases and that the two identical dispersed phases give
identical results.
Overview
A flow of water containing 1% volume fraction of heavy particles is introduced at the top of a 2m tall duct. The
particles accelerate as they fall until they reach a terminal velocity relative to the water downflow. There is a
corresponding reduction in volume fraction as particles accelerate to terminal velocity. For test purposes, the
particle flow can be split in arbitrary shares between two identical disperse phases without affecting the overall
results.
Physics
The problem is solved with a steady laminar flow using a constant phase interaction drag coefficient of 0.44.
The expected terminal velocity for a constant drag coefficient is independent of volume fraction:
4 ρd − ρc gDparticle
ur = (74)
3ρcCD
The terminal velocity can also be tabulated against total dispersed phase volume fraction, allowing a graphical
solution or a bisection solution for the asymptotic volume fraction corresponding to the known terminal
velocity.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This case validates multiphase particle/turbulence interactions against experimental data for a fully developed
particulate horizontal flow.
Overview
Water carrying 165 micron particles with a density of 2700 kg/m3, and at a volume fraction of 0.0995, is
introduced in a horizontal duct. The mean velocity of the mixture is 3.5 m/s. Periodic boundary conditions are
applied at both ends of the pipe. Mass and momentum source terms are used to regulate the mean volume
fraction and mean flow velocity.
Physics
This case is about fluidization of solid particles in a horizontal pipe. The gravity force is balanced by the
turbulent dispersion force. Diffusivity of the particles is higher than the eddy diffusivity.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Multiphase (sand, water)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Three Dimensional
• Turbulent
Material Properties:
Initial Conditions:
Boundary Conditions:
Field Functions
The following are user-defined field functions in this case, with their definitions. They contain expressions that
are used in the source terms to set and regulate the flow rate and the volume fraction.
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[67] Eskin, D. 2012. “A simple model of particle diffusivity in horizontal hydrotransport pipelines”, Chem.
Eng. Sci., 82, pp. 84-94.
Purpose
This case demonstrates the behavior of the S-Gamma model for size modeling when breakup and coalescence
terms are activated in a turbulent flow. The predicted mean diameter using these (tuned) models can be
compared to experimental data. The predicted Sauter mean diameter d32 is used as the interaction length scale
for the phase interaction, and the performance of the coupled Multiphase/S-Gamma models in converging
from initial conditions is also tested.
Acceptance Criteria
The acceptance criteria for the comparison are the relative differences in Sauter Mean Diameter and Second
Size Distribution Moment, each of them averaged over all the measurements available in the reference. The
relative differences were normalized using the largest respective value from the reference. The threshold
values are shown in the following table and are consistent with the work of Lo and Zhang [70].
Overview
An axisymmetric 2D grid of 20 x 150 cells is used to simulate a 3.06-meter vertical bubble column with an
internal diameter of 50.8 mm. The dispersed phase is air and the continuous phase is water, both using
standard properties and assuming constant density.
The data compared in this test is the Time-Averaged Sauter Mean Diameter against radius, for height
z/D = 53.5 and volumetric flux jf = 0.986 m/s from Figure 8 of [68]. (This is equivalent to z = 2.7 m and
U = 1.23 m/s, volume fraction equal to 0.8 for water boundary conditions.)
This setup is expected to give a flow in which air bubbles move towards the center of the column, where their
greater proximity to one another is expected to increase coalescence. Shear close to the wall promotes
breakup at that location.
Physics
The inlet volume fractions are 0.8 water and 0.2 air with inlet velocities of 1.23 and 1.61 m/s respectively. An
inlet bubble diameter of 3 mm is specified. The K-Epsilon turbulence model is used for both phases.
The coalescence factor is left at its default value of 0.01. This number is in the range found to give reasonable
agreement with the Hibiki experimental data of which one set is used in this test [68]. All coalescence and
breakup options are left as standard and the viscous coalescence probability method is set to be Quasi-Steady
State. No interphase mass transfer terms are enabled. The phase interaction is set up with the Wang curve fit
for drag, a lift coefficient of -0.05, and turbulent dissipation force activated. The interaction length scale is set
to be equal to the S-Gamma Sauter mean diameter for air.
In the phase interactions, the S-Gamma model is used for breakup and coalescence.
Physics Models:
• Axisymmetric
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Multiphase (Water, Air)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Turbulent
Material Properties:
Initial Conditions:
Water Air
Pressure (Pa) 0.0 0.0
Turbulence Intensity 0.01 0.01
Turbulent Length Scale (m) 0.005 0.005
Turbulent Velocity Scale 1.0 1.0
(m/s)
Velocity (m/s) 1.23 1.61
Volume Fraction 0.8 0.2
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[68] Hibiki, T., Ishii, M., Xiao, Z., 2001. “Axial interfacial area transport of vertical bubbly flows.” Int. J.
Heat and Mass Transfer, 44, pp 1869 - 1888.
[69] Hill, D.P. 1998. “The Computer Simulation of Dispersed Two Phase Flows”, Ph.D. dissertation,
Imperial College, London.
[70] Lo, S. and Zhang, D. 2009. “Modelling of Break-up and Coalescence in Bubbly Two-Phase Flows”, J.
Comp. Multiphase Flows, 1(1), pp. 23-38.
Purpose
This case verifies that the S-Gamma size distribution model correctly predicts the increase in diameter of
bubbles of an ideal gas (air) as they move up a long vertical column containing water as the continuous phase.
The pressure and, therefore, density decrease up the column and this change is expected to lead to bubbles
increasing in diameter. Sauter mean diameter d32, which is produced by the S-gamma model, is compared
with the value that is expected from ideal gas expansion of bubbles injected with a given initial size and
density.
Overview
This test uses a simple 2 x 240 cell mesh to represent a bubble column 12 meters high. This test plots bubble
size against height and bubble density against height.
The reference gauge pressure is set to 0 Pa at the top of the column, and the reference density is 1000 kg/m3.
The inlet air volume fraction is 0.01 and the inlet d32 is 1 mm. The water inlet velocity is 1.0 m/s and the air
velocity is 1.17 m/s.
Physics
The continuous phase is water, which has constant density, and the dispersed phase is air. The Ideal Gas and S-
Gamma models are used. The phase coupled fluid isothermal energy model is used to allow ideal gas
modeling, the temperature is set to a constant 300 K, and gravity is enabled. The phase interaction forces used
are Schiller-Naumann drag force and virtual mass force. To isolate the variable density response of S-Gamma
for this test, a constant interaction length scale of 1 mm is used rather than d32.
Physics Models:
• Eulerian Multiphase (EMP)
• Gradients
• Gravity
• Laminar
• Multiphase (Water, Air)
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Water Air
Pressure (Pa) 0.0 0.0
Velocity (m/s) X = 1.0, Y = 0 X = 1.2, Y = 0
Volume Fraction 0.99 0.01
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Multiphase VOF
Contents:
Free Surface Flow and Heat Transfer in a Straight Channel
Flow in a Capillary Pipe
Overset and Non-Overset: Water Impact of a Wedge
Purpose
This verification case demonstrates pressure gradient calculation and heat transfer in combination with using
the Volume of Fluid (VOF) model.
Overview
The problem geometry is a two-dimensional channel, measuring 6.0 m by 6.3 m, that consists of an inlet for
water and air, outlet and slip wall boundaries. The water inlet measures 0.7 m. A supercritical free-surface flow
(Fr > 1) is set up in the channel.
Physics
The problem is solved using the Volume of Fluid multiphase model. The analytical solution corresponds to
when steady-state conditions are reached. The pressure gradient is then constant at ∇p = 10 Pa/m for air and
∇p = 10000 Pa/m for water. The temperature remains constant and equal to 300 K during the simulation.
Physics Models:
• Gradients
• Gravity
• Implicit Unsteady
• Laminar
• Multiphase
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Segregated Flow
• Segregated Multiphase Temperature
• Two Dimensional
• Volume of Fluid (VOF)
Material Properties:
Liquid Vapor
Density (kg/m^3) 1000.0 1.0
Viscosity (Pa-s) 8.8871E-4 1.85508E-5
Specific Heat (J/kg-K) 4181.0 1003.0
Thermal Conductivity (W/m- 0.6203 0.02603
K)
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates surface tension, contact angle, and free surface tracking using the Volume
of Fluid (VOF) model.
Overview
Two parallel plates are in contact with water. The plates are connected at the top, so that flow is possible only
at the bottom. Initially the space between the plates is partially filled with water.
This case calculates the pressure jump, which is caused by the surface tension forces, across the interface
between the phases in a capillary pipe.
Physics
The problem is solved using the Volume of Fluid multiphase model. The analytical solution corresponds to
when steady-state conditions are reached. Then the velocity field is constant (U = 0, V = 0), and the pressure
jump across the free surface is 2σ cos(contactAngle) /d , where σ is the surface tension.
Physics Models:
• Gradients
• Implicit Unsteady
• Laminar
• Multiphase
• Multiphase Equation of State
• Multiphase Interaction
• Multiphase Mixture
• Segregated Flow
• Segregated Multiphase Temperature
• Two Dimensional
• Volume of Fluid (VOF)
Material Properties:
H2O Air
Density (kg/m^3) 997.561 N/A
Dynamic Viscosity (Pa-s) 0.01 0.01
Molecular Weight (kg/kmol) N/A 28.9664
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case simulates a drop test in 2.5 dimensions, which is of relevance for investigating slamming
problems. Numerical results for overset and non-overset methods are presented along with the comparison
against experimental results.
Overview
A test case for which experimental data exist was chosen to verify the overset method for the time-accurate
simulation of plane motions with 3-DOF.
The test case consisted of a high aspect ratio (0.61 m x 2.44 m) prismatic wedge with 20° deadrise, that is
dropped from different heights, with different weights, and with initial zero or non-zero heel angles. The
wedge was instrumented with accelerometers for recording the roll and vertical acceleration time history.
The drop test conditions that are chosen for the verification case are:
The drop height for all simulations was 0.61 m, measured from the water to the keel of the wedge. The
position of the center of gravity was at the symmetry plane and 0.216 m above the keel for the light-weight
condition. In the numerical model, the configuration is reduced to one of 2.5 dimensions. Thus, weight and
moments of inertia were reduced to the model slice of 0.003125 m.
The computational domain extends 2 m to the sides, 1 m from the keel upwards, and 1.5 m from the keel
downwards. This volume is filled with water to a depth of 0.9 m at the initial condition.
Since no side current is present in this configuration, no inlet or outlet boundary conditions are necessary.
Pressure-type boundary conditions are applied, where the known hydrostatic pressure (and the volume
fraction) is specified at all flow boundaries (side, top and bottom). Pressure disturbances are assumed minimal
on these boundaries.
At time t0 = 0, the vertical position is +0.61 m, and force and velocity are zero. The vertical acceleration, which
is normalized with the acceleration of gravity = 9.81 m/s2, is -1 (the vertical axis points upwards). The wedge
hits the water surface after 0.35 s, when its vertical velocity is at its maximum.
The verification case starts from a simulation file that is saved at 0.35 s and runs until the Maximum Physical
Time reaches 0.45 s. The simulation includes 3 DOFs (DFBI Free Motion: Translation Z, Translation X, Rotation
Y). The following quantities are used for comparing overset and non-overset results with experimental data
[71]:
Physics
The drop test motion is simulated using the DFBI Free Motion model (3 DOFs: Translation Z, Translation X,
Rotation Y) modeling water and air in the domain with the Volume of Fluid multiphase model:
Physics Models:
• Gradients
• Gravity
• Implicit Unsteady
• K-Epsilon Turbulence
• Multiphase (Air and Water)
• Multiphase Equation of State
• Multiphase Interaction
• Realizable K-Epsilon Two-Layer
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Three Dimensional
• Turbulent
• Two-Layer All y+ Wall Treatment
• VOF Waves
• Volume of Fluid (VOF)
• Wall Distance
Material Properties:
H2O Air
Density (kg/m^3) 997.561 1.18415
Dynamic Viscosity (Pa-s) 8.8871 x 10–4 1.85508 x 10–5
Initial Conditions:
Boundary Conditions:
Overset:
Non-Overset:
Top
Symm1 Symmetry N/A
Symm2
WedgeSides Wall No-Slip
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[71] Azcueta, R. 2001. “Computation of Turbulent Free-Surface Flows Around Ships and Floating Bodies”,
PhD thesis Technical University of Hamburg-Harburg.
Contents:
Stress Analysis of a Two-Dimensional Rectangular Plate with a Hole (CGD)
Finite-Element Bending of a Cantilever Beam
Deformation Analysis of an Elastic Solid Due to Exerted Fluid Pressure (CGD)
Finite-Element Scordelis Lo Roof
Fluid Structure Interaction Using Small Displacement Theory (CGD)
Fluid Structure Interaction Using Large Displacement Theory (CGD)
Purpose
This verification case demonstrates using the solid stress solver in predicting different stress components. The
obtained solution is compared with analytical data.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences in the
components of the shear stress tensor between the simulation and the analytical solution. The average is
performed over all the cells in the domain. The relative differences were normalized using the unidirectional
tensile load F defined below. The threshold value is 1%, as shown in the following table.
Overview
The geometry is a three-dimensional plate (4 m x 4 m x 0.1 m) with a hole in its center. The radius of the hole
is 0.5 m. Given the symmetric nature of the problem, only the top right quadrant of the plate is used in the
simulation. A vertical interface is created at the center of the quadrant. The material properties are uniform
across the interface.
Physics
The problem is solved using the solid stress, linear isotropic elastic material models with constant density.
The components of the stress tensor from the simulation are compared to analytical results computed from
[72]:
R2 R4
σxx = F 1 − (1.5cos 2θ + cos 4θ + 1.5cos 4θ (76)
r2 r4
R2 R4
σyy = F ‐ 2
0.5cos 2θ − cos 4θ − 1.5cos 4θ 4 (77)
r r
R2 R4
σxy = F ‐ 0.5 sin 2θ + sin 4θ + 1.5sin 4θ (78)
r2 r4
where F = 1 × 104 Pa is the unidirectional tensile load on the plate, r = x2 + y2 and θ = arctan y/x . In this
problem, x, y are measured from the center of the hole.
Physics Models:
• Material Law Models
◦ Isotropic Linear Elasticity
◦ Linear Elasticity
• Solid (Al)
• Solid Stress
• Steady
• Three Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Stopping Criteria:
Maximum Steps 3
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[72] Young, W. C. & Budynas, R. G., 2002. Roark’s Formulas for Stress and Strain. 7th ed. s.l.:McGraw-Hill.
Purpose
This case is a validation of the exact Timoshenko solution of bending of a cantilever under plane strain
conditions [73].
Acceptance Criteria
The acceptance criteria for the comparison are the unsigned relative differences in maximum displacement in
Y-direction and maximum bending stress, between the prediction of the simulation and the analytical
calculation, as shown in the following table.
Overview
This simulation calculates displacements and stresses in a cantilevered rectangular beam that is subjected to
pressure loads.
The solution is valid for all components of the displacement and the stress field. It is not simply a comparison
to a 1D Euler beam solution. The beam is loaded with a parabolic shear stress using the continuous option for
the traction load on the right end face. The constraints on the left end of the beam are based on the exact
solution which allows for some rotation of the beam.
To make this an even more rigorous test, the beam is not aligned to the laboratory coordinate system. The
loads and constraints are defined by user-defined field functions which are in a coordinate frame that is
aligned with the beam (the same coordinate system that Timoshenko uses). In this coordinate system, the
loaded end is at x = 0 m and the free end is at x = 30 m. The beam thickness is 1.0 m.
The display of the displacements and stresses are transformed in the coordinate system that is aligned with the
beam.
Simcenter STAR-CCM+ simulations make liberal use of the vector and tensor transformations to compute the
error between the exact solution and the FEA solution in the aligned coordinate system. The only difference
between the simulations is the Simcenter STAR-CCM+ mesh operation.
If one refines the mesh with either the directed mesher or the thin tetrahedral mesher, the solution does
approach the exact solution. Note that the second-order elements with the full integration rule allows the
nodal loads to be computed exactly from the parabolic shear load with just one element through the thickness.
This would not be the case with linear elements - that is, the nodal loads would not be correctly computed.
The sides of the cross section of the beam are not located along the laboratory coordinate axes, but are at a
45° angle as shown in the following diagram:
where:
• AB = 2.0
• BC = 1.0
• BE = ABcos 45 =2.0⋅0.707106781=1.414213
• BF = BC cos 45 =1.0⋅0.707106781=0.707106781
• BH = AB cos 45 =1.414213
• GH = AD cos 45 =0.707106781
Physics
This test is a cantilever beam in bending, modeled with Hex20 solid elements. All element types are expected
to approximate the exact continuum solution from Timoshenko as posted here. The Hex20 approximates the
exact solution in the displacements to single-precision accuracy. The result is expected with two elements
through the thickness and quadratic shape functions in each element.
Physics Models:
• Linear Isotropic Elastic
• Solid
• Solid Stress
• Steady
• Three Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
In the following table, faces are the part surfaces that are associated with boundaries. x, y, and z are in the
local coordinate system.
Analytical Solution
Traction at (x = 0) is:
y 2
τxy = 1.5P 1 − (79)
ℎ/2
∫
2
P= ℎ τxydy = 1.0 × 106 (80)
−
2
2P 1 − ν2 2−ν 2 1
Ux = y 3x2 − y − 3 l2 −
E 1−ν 2 1−ν
2P 1 − ν2 3ν (81)
Uy = − xy2 − x3 + 3xl2 − 2l3
E 1−ν
Uz = 0
Force and displacement stopping criteria have been considered. The solution converged in 2 iterations.
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[73] Timoshenko, S. P. and Goodier, J. N. 1982. Theory of Elasticity, 3rd Ed., McGraw-Hill, pp. 41-46.
Purpose
The verification case demonstrates the one-way interaction at a solid fluid interface in support of applications
of fluid structure interaction (FSI) simulations.
Acceptance Criteria
The acceptance criterion for the comparison is the unsigned relative difference in vertical displacement of the
solid-liquid interface, between the predictions of the simulation and the analytical calculations, as shown in
the following table.
Overview
This fundamental test case was designed to test the one-way coupling solution of fluid-solid interaction. A fluid
mass with a volume of 1 m3 exerts a constant pressure of 1 x 105 Pa on a block of elastic solid material with a
volume of 1 m3.
Physics
A steady flow interacts with a 3D solid using Realizable K-Epsilon Turbulence with two-layer all-y+ wall
treatment.
For analytical data, this model is a case of plane strain in the direction of pressure (the Y-direction). The
displacement profile is:
P/E 1 + ν 1 − 2ν
Y= H−y (82)
1−ν
where P is the pressure applied, E is the Young’s modulus, and H is the height of the domain [74].
Physics Models:
Fluid Continuum:
• Constant Density
• Gradients
• K-Epsilon Turbulence
• Liquid
• Realizable K-Epsilon Two-Layer
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Turbulent
• Two-Layer All-y+ Wall Treatment
• Wall Distance
Solid Continuum:
Material Properties:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[74] Timoshenko, S. P. & Goodier, J. N. 1982. Theory of Elasticity, 3rd ed., McGraw-Hill International Book
Company, pp. 41-46.
Purpose
The Scordelis Lo Roof case is a widely used benchmark test to check the accuracy of finite element types for
curved shell problems. This case comprises a moderately thin shell that is subject to gravity loading. It has
inextensional bending modes which are a very important feature of this problem.
Overview
This simulation uses two derived parts, a vector warp surface and a point probe, to calculate displacement in
the y-direction at the mid-point of the free edge.
Physics
This 3D case is modeled with solid elements. The roof is subject to gravity loads.
Physics Models:
• Gravity
• Linear Isotropic Elastic
• Solid
• Solid Stress
• Steady
• Three Dimensional
The y-displacement value is compared to the value that is published in MacNeal [75]. All element types must
converge to this value as the mesh is refined.
Material Properties:
Initial Conditions:
Segment Conditions:
Maximum Steps 2
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[75] MacNeal, R. 1994. Finite Elements: Their Design and Performance, Marcel Dekker, Inc., pp. 444-446.
Purpose
This verification case demonstrates solving problems involving fluid-structure interaction (FSI) using small
displacement theory.
Acceptance Criteria
The acceptance criterion for the comparison is the unsigned relative difference in the period of the oscillation
between the predictions of the simulation and the analytical calculations, as shown in the following table. The
threshold is derived from the work of Paidoussis [77] and Jweeg & Ntayeesh [76].
Overview
A slender pipe receiving water is used to predict the pipe displacement due to a load of 1 kg. The geometry
consists of a simple half-pipe model, 0.02 m in diameter and 1.20 m in length, where flow enters through a
velocity inlet and exits through a pressure outlet. The inner wall of the pipe acts as a solid boundary to the fluid
flow, but is dynamically coupled to the fluid domain at the middle section of the overall geometry. In this
section, the pipe is represented as a solid domain 0.0005 m thick and 0.40 m long. Both ends of this section
are fixed and its outer surface is stress-free. The rest of the pipe is treated as a rigid boundary.
Physics
The problem is solved as a transient, three-dimensional, laminar, compressible flow case. Water with constant
velocity of 0.01 m/s is used. A user-defined density for water is prescribed using a field function.
The frequency of the vibrating pipe can be computed analytically as follows ([78], [77], [76]):
3
480 ⋅ E ⋅ I 1 − B / L / masssolid + massfluid
f= 2π
where B = 2ρAfv2L2 / 56 ⋅ E ⋅ I
and ρ is fluid density, Af is fluid cross sectional area, L is length, v is fluid velocity, E is Young’s modulus and I
is the moment of inertia given by
The analytical solution to this problem has a period of 2.34e-3 s, with maximum displacement of 5.91e-7 m.
• Implicit Unsteady
• Three Dimensional
Fluid Physics Models:
• Gradients
• Laminar
• Liquid (H2O)
• Segregated Flow
• Solution Interpolation
• User Defined EOS
Solid Physics Models:
• Constant Density
• Material Law Models
◦ Isotropic Linear Elasticity
◦ Linear Elasticity
• Solid (Al)
• Solid Stress
Water Properties:
Solid Properties:
Initial Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[76] Jweeg, M. J. & Ntayeesh, T. J. 2015. Dynamic Analysis of Pipes Conveying Fluid Using Analytical,
Numerical and Experimental Verification with the Aid of Smart Materials, International Journal of
Science and Research (IJSR), 4(12).
[77] Paidoussis, M. 2013. Fluid-structure Interactions: Slender Structures and Axial Flow, s.l.:Academic
Press.
[78] Paidoussis, M. & Issid, N. 1974. Dynamic Stability of Pipes Conveying Fluid, Journal of Sound and
Vibration, 33(3), pp. 267-294.
Purpose
This verification case demonstrates solving problems involving fluid-structure interaction (FSI) using large
displacement theory.
Acceptance Criteria
The acceptance criterion for the comparison is the unsigned relative difference in the period of the oscillation
between the predictions of the simulation and the analytical calculations, as shown in the following table. The
threshold is derived from the work of Paidoussis [80] and Jweeg & Ntayeesh [79].
Overview
A slender pipe receiving water is used to predict the pipe displacement due to a load of 1 kg. The geometry
consists of a simple half-pipe model, 0.02 m in diameter and 1.20 m in length, where flow enters through a
velocity inlet and exits through a pressure outlet. The inner wall of the pipe acts as a solid boundary to the fluid
flow, but is dynamically coupled to the fluid domain at the middle section of the overall geometry. In this
section, the pipe is represented as a solid domain 0.0005 m thick and 0.40 m long. Both ends of this section
are fixed and its outer surface is stress-free. The rest of the pipe is treated as a rigid boundary.
Physics
The problem is solved as a transient, three-dimensional, laminar, compressible flow case. Water with constant
velocity of 0.001 m/s is used. A user-defined density for water is prescribed using a field function.
The frequency of the vibrating pipe can be computed analytically as follows ([81], [80], [79]):
3
480 ⋅ E ⋅ I 1 − B / L / masssolid + massfluid
f= 2π
where B = 2ρAfv2L2 / 56 ⋅ E ⋅ I
and ρ is fluid density, Af is fluid cross sectional area, L is length, v is fluid velocity, E is Young’s modulus and I
is the moment of inertia given by
• Three Dimensional
Fluid Physics Models:
• Gradients
• Laminar
• Liquid (H2O)
• Segregated Flow
• User Defined EOS
Solid Physics Models:
• Fluid Structure Coupling
• Material Law Models:
◦ Isotropic Linear Elasticity
◦ Linear Elasticity
• Nonlinear Geometry
• Solid (Al)
• Solid Stress
Water Properties:
Solid Properties:
Initial Conditions:
Displacement (m) 0, 0, 0
Solid velocity (m/s) 0, 0, 0
Fluid Pressure (Pa) 0.0
Fluid Velocity (m/s) 0, 0, 0.001
Boundary Conditions:
where the Field Function is defined as 0.0, Fy, 0.0 , and Fy is defined as follows:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[79] Jweeg, M. J. & Ntayeesh, T. J. 2015. Dynamic Analysis of Pipes Conveying Fluid Using Analytical,
Numerical and Experimental Verification with the Aid of Smart Materials, International Journal of
Science and Research (IJSR), 4(12).
[80] Paidoussis, M. 2013. Fluid-structure Interactions: Slender Structures and Axial Flow, s.l.:Academic
Press.
[81] Paidoussis, M. & Issid, N. 1974. Dynamic Stability of Pipes Conveying Fluid, Journal of Sound and
Vibration, 33(3), pp. 267-294.
Electromagnetism
Contents:
Electric Potential in a Square Domain
TEAM Problem 13: 3D Non-Linear Magnetostatic Model
TEAM Problem 20: 3D Static Force Problem
TEAM Problem 23: 3D Static Force Problem in a 36-Degree Body
TEAM Problem 30a: 2D Induction Motor Analyses
Purpose
This verification case demonstrates electric potential models.
Overview
The domain is a 2D box 1 m x 1 m. At the left side and at the bottom of the domain, zero specific electric
current (Insulator) is set. At the top and right side, the electric potential is specified according to the given
potential.
The test testMaxAbsoluteError examines the difference between an analytical solution and
Simcenter STAR-CCM+ calculations in solving the Electric Potential equation. An analytical solution is set up
internally in Simcenter STAR-CCM+ with the field function AnalyticalSolution.
Physics
The Electrodynamic Potential model is used in a solid continuum with a 2D mesh. The analytical solution is
calculated as ϕ = x2 − y2.
Physics Models:
• Electrodynamic Potential
• Electromagnetism
• Gradients
• Solid
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
Maximum Steps 35
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[82] Ida, N. 2004. Engineering Electromagnetics, 2nd ed., Springer Publishing Company, Inc.
Purpose
This verification case compares numerical to experimental results for the 3D finite-element electromagnetic
capability of Simcenter STAR-CCM+.
Overview
The experiment involves measuring magnetic flux densities at selected locations. Between two steel channels
are a coil and a steel plate, with electric (DC) current of 3000 AT passing through the coil to generate the flux.
Search coils, which are a standard tool for measuring the saturation level of flux within a solid, are placed at
the selected locations.
The thin gap between the U profiles and the vertical plate, together with the partial overlap, poses an
additional challenge for the numerical method of the Simcenter STAR-CCM+ simulation, as the saturation level
and field orientation change on a short range.
In this case, symmetrical and periodic boundary conditions are used to simulate 1/4 of the model.
The first 25 out of 36 search coils are represented with plane section derived parts, and the remaining ones,
which are in a measured line in the experiment, are simulated with a line probe derived part.
Physics
For the steel, the material data is given in [91] in a tabular form with a functional description for flux densities
above 1.8T:
where:
To combine these descriptions, the tabular data is extended according to the appropriate values for higher flux
densities so that a single profile method can be used. The assumption of complete saturation above 2.2T, the
second case in Eqn. (84), is an integral part of the Table(B,H) method employed in this simulation. In the
simulation the table is used for flux densities of up to 2.2T.
Physics Models:
• Electromagnetism
• Finite Element Magnetic Vector Potential
• Gradients
• Solid
• Steady
• Three Dimensional
Coil:
• Excitation Coil
Material Properties:
Initial Conditions:
Maximum Steps 15
In a non-linear scenario like this one a Newton update will already provide a direction of descent for a changed
converged tolerance of 1.0E-4 and improve the performance.
• A perfect electric conductor (PEC), a theoretically ideal material that has infinite electrical conductivity, is
used on the boundaries that represent the outer surfaces: Top, Default, Face, and Outside.
• The use of the Magnetic Vector Potential Specification Method, with a setting of Constant, for the
Outside boundary has the same effect as the PEC setting.
• Perfect Magnetic Conductor (PMC) settings are used in modeling the symmetry at the bottom of the
numerical simulation domain.
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[83] Akima, H. 1970. "A New Method of Interpolation and Smooth Curve Fitting Based on Local
Procedures", Journal of ACM, 17, 4, 589.
[84] Bossavit, A. and Verite, J.C. 1983. "The "TRIFOU" Code : Solving the 3-D Eddy-Currents Problem by
Using H as State Variable", IEEE Trans. Magnetics, MAG-19, 6, 2465.
[85] Emson, C.R.I. 1988. "Methods for the Solution of Open-Boundary Electromagnetic-Field Problems",
lEE Proc., 135, Pt. A, 3, 151.
[86] Hwang, J.H. and Load, W. 1974. "Finite Element Analysis of the Magnetic Field Distribution inside a
Rotating Ferromagnetic Bar", IEEE Trans. Magnetics, MAG-lO, 4, 1113.
[87] Kettunen, L., Forsman, K., Levine, D., and Gropp, W. 1994. "Solutions of team problems 13 and 20
using a volume integral formulation", Proc. Aix-les-Bains TEAM Workshop, Aix-les-Bains, France.
[88] Nakata,T., Takahashi, N., Fujiwara, K. and Ahagon, A. 1988. "Periodic Boundary Condition for 3-D
Magnetic Field Analysis and its Applications to Electrical Machines", IEEE Trans. Magnetics, MAG-24,
6, 2694.
[89] Silvester, P.P., Cabayan, H.S., and Browne, B.T. 1973. "Efficient Techniques for Finite Element
Analysis of Electrical Machines", IEEE Trans: PA&S, PAS-92, 6, 1274.
[90] Sonneveld, P. 1984. "CGS, a Fast Lanczos-Type Solver for Nonsymmetric Linear Systems", Report
84-16, Department of Mathematics and Informatics, Delft University of Technology, The
Netherlands.
[91] "Team Problem 13, 3-D Non-Linear Magnetostatic Model", International Compumag Society, http://
www.compumag.org/jsite/images/stories/TEAM/problem13.pdf.
[92] Tong, P. and Rossetos, J.N. 1977. "Finite-Element Method (Basic Technique and Implementation)",
MIT Press.
[93] Zienkiewicz, O.C. 1977. The Finite Element Method (Third Edition), McGraw-Hill.
Purpose
This verification case compares numerical to experimental results for the 3D finite-element electromagnetic
capability of Simcenter STAR-CCM+ with a multi-part solid.
Overview
The experiment involves calculating the magnetic force applied to the pole at selected ampere-turn
excitations. An excitation coil has a steel center pole core and a steel yoke closing the magnetic circuit with the
pole through an air gap. DC current excites the coil. The ampere turns are 1000, 3000, 4500, and 5000 AT.
In this case, a symmetrical boundary condition is used to simulate 1/4 of the model.
Physics
Physics Models:
• Three Dimensional
Coil:
• Excitation Coil
• Solid
Multipart Solid
• Multi-Component Solid
• Multi-Part Solid
Material Properties:
Al Ash Fe
Magnetic Permeability 1.25666E-6 H/m 1.25663 Tabulated B-H curve, partially
7E-6 H/m calculated from the analytical
expression
Initial Conditions:
This unsteady simulation runs up to a Maximum Physical Time of 3 seconds, during which the current is
ramped up. This procedure gives a comparison of the force generated as a result of the currents applied to the
coils.
To download the final simulation to your own directory structure, right-click the link in this column and choose
"Save Link As" (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[94] "Team Problem 20, 3-D Static Force Problem", International Compumag Society,http://
www.compumag.org/jsite/images/stories/TEAM/problem20.pdf.
[95] Kettunen, L., Forsman, K., Levine, D., and Gropp, W. 1994. "Solutions of team problems 13 and 20
using a volume integral formulation," Proceedings of Aix-les-Bains TEAM Workshop, Aix-les-Bains,
France.
Purpose
This verification case compares numerical to experimental results for the 3D finite-element electromagnetic
capability of Simcenter STAR-CCM+ with permanent magnets in three dimensional geometries.
Overview
An excitation coil is placed next to a magnet. DC current excites the coil. The ampere turns are 1400 AT. The
objective is to calculate the magnetic force on the coil.
In the original experiment [96], the measurements consist of axial and restoring forces for two configurations,
both involving small magnets and coils. The configuration that is simulated by this verification case uses
Samarium-Cobalt magnet specifications.
Physics
Physics Models:
• Electromagnetism
• Finite Element Magnetic Vector Potential
• Gradients
• Solid
• Steady
• Three Dimensional
Magnet:
• Permanent Magnet
Material Properties:
Air Magnet
Magnetic Permeability (H/m) 1.25666E-6 1.41666E-6
Air Magnet
Remanence Flux Density (T) - 1.02
Initial Conditions:
Maximum Steps 5
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[96] "TEAM Problem 23", International Compumag Society, http://www.compumag.org/jsite/images/
stories/TEAM/problem23.pdf.
Purpose
This verification case compares the computed torque of an induction motor with the experimentally
determined torque.[97]
Overview
The experiment involves an induction motor case where time harmonic currents on the stator and the rotor
rotation induce eddy currents. The model is 2D, but it presents a particular challenge in the form of the
induced electric field and the high rotation speeds, expressed as . A volume-based finite element code
such as Simcenter STAR-CCM+ must include this expression. Simcenter STAR-CCM+ produces results very close
to the experimental values.
The geometry and mesh of this case appear as follows in Simcenter STAR-CCM+:
The verification case is for a three-phase exposed winding motor. Each stator winding phase spans 45 degrees.
The current density J is maintained constant at 310 A/cm2. The phase groups A, B, and C are located on 60-
degree centers and lag each other in phase by 120 degrees.
The object is to predict the torque, power dissipated, and stator terminal voltage induced for rotor speeds
ranging from 0 to 1200 rad/s, roughly three times faster than the stator field speed of 377 rad/s. The three
phase winding is excited at 60 Hz. Additional quantities, also computed analytically, are:
Speed Ωr (rad/s) Torque (N/m) Voltage/turn (V/m/ Rotor Loss (W/m) Steel Loss
turn) (W/m)
0 3.825857 0.637157 1455.644 17.40541
Physics
Physics Models:
Excitation Coil:
RotorLinear:
Stator Steel:
Void:
Material Properties:
Initial Conditions:
Boundary Conditions:
The outer boundary of the Void region is a Wall boundary. All other region boundaries form contact interfaces
with their adjacent regions.
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[97] "International TEAM Workshop Problem 30a", International Compumag Society, http://
www.compumag.org/jsite/images/stories/TEAM/problem30a.pdf.
Compressible Flow
Contents:
One-Dimensional Shock Tube
Supersonic Flow Over a Flat Plate
Turbulent Subsonic Flow through an Aero Intake
Turbulent Transonic Flow over an RAE-2822 Airfoil
Quasi-1D Converging-Diverging Nozzle Flow
Two-Dimensional Shock Reflection
Purpose
This verification case demonstrates the use of a gradient-based, second-order accurate spatial differencing
scheme in capturing the shock and expansion waves inside a shock tube.
Overview
This is a classical case, first considered by Sod [98]. A long thin tube that is filled with air is initially at rest but
separated into two sections (by a diaphragm) with different pressures and densities on either side. When the
diaphragm is burst, a shock wave and contact discontinuity propagates into the low-pressure region, while an
expansion fan propagates into the high-pressure section.
Physics
The flow is assumed to be laminar and inviscid. The calculations are performed in transient mode using the
coupled explicit solver. The case is run to a maximum time of 0.0061 seconds (as in the classical case from Sod
[98]).
Since shock tube problems are special cases of a Riemann problem, they can be solved analytically (see, for
example Hirsch [99]). In this case, therefore, the predicted pressure, density, temperature, velocity and Mach
number profiles along the tube are compared.
Physics Models:
• Coupled Energy
• Coupled Flow
• Explicit Unsteady
• Gas (Air)
• Gradients
• Ideal Gas
• Inviscid
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[98] Sod, G. A. (1978), “A survey of Several Finite Difference Methods for Systems of Nonlinear Hyperbolic
Conservation”, J. of Computational Physics, vol. 27, pp. 1-31.
[99] Hirsch (1995), “Numerical Computation of Internal and External Flows”, Wiley, vol. 2.
Purpose
This verification case demonstrates the development of a turbulent boundary layer for supersonic flow.
Overview
The problem geometry is a two-dimensional trapezoidal wedge. At zero pressure gradient, the heat transfer
process has a negligible effect on the boundary layer development. Therefore this problem is equivalent to
simulating flow over an isolated flat plate with an adiabatic wall.
Physics
During the experiment, the transition of the boundary layer from laminar to turbulent is artificially initiated by
tripping the flow at 23 mm downstream of the leading edge. Based on a free-stream Mach number for air of
1.724, the Reynolds number is 15.35E6. The two-layer variant of the realizable K-Epsilon turbulence model
with all-y+ wall treatment is used.
Experimental data is supplied by the Military Aircraft Division of British Aerospace Defence Ltd. It includes
descriptions of the boundary layer profiles, including flow velocity, Mach number, density, static temperature
and total temperature at a point 0.191 m downstream of the leading edge.
Physics Models:
• Coupled Energy
• Coupled Flow
• Gas (Air)
• Gradients
• Ideal Gas
• K-Epsilon
• Realizable K-Epsilon Two-Layer
• Reynolds-Averaged Navier-Stokes
• Steady
• Turbulent
• Two Dimensional
• Two-Layer All y+ Wall Treatment
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates subsonic flow in a three-dimensional configuration using a coarse mesh.
Overview
Free-stream air flowing at Mach 0.21 encounters a NACA-type axisysmmetric inlet. The flow captured by the
inlet is forced into an S-bend before reaching the engine face. The predicted static pressures at the top and
starboard wall are compared with the measured pressures [100].
Even though this test is run turbulent, the mesh is really more suited for inviscid flow. To keep the mesh
resolution requirements down, no attempt is made to resolve the boundary layers. The results are thus
essentially what one would obtain from an inviscid simulation.
Non-conformal interfaces are used in the free-stream and downstream sections to in an effort to make the
mesh cell count as low as possible.
Physics
Compressible subsonic flow is modeled with coupled and segregated flow models.
Physics Models:
• Coupled Energy (for Coupled Flow only)
• Coupled Flow
• Gas
• Gradients
• High-y+ Wall Treatment
• Ideal Gas
• K-Epsilon Turbulence
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature (for Segregated Flow only)
• Standard K-Epsilon
• Steady
• Three Dimensional
• Turbulent
Material Properties:
Initial Conditions:
Boundary Conditions:
Coupled Solver:
Segregated Solver:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[100] Fluid Dynamics Panel Working Group 13. 1991. Agard Advisory Report 270 “Air Intakes for High
Speed Vehicles”, pp. 139-162.
Purpose
This verification case demonstrates external aerodynamics on an airfoil. In addition to the second-order
scheme, a third-order MUSCL (Monotonic Upstream-Centered Scheme for Conservation Laws) scheme is also
tested. The MUSCL scheme is used for the resolution of flow features such as vortices and sound propagation.
Overview
This case simulates the compressible, turbulent flow of air over an isolated two-dimensional RAE-2822 airfoil
section at 2.92 degrees angle-of-attack and 0.729 free stream Mach number.
Physics
The flow field is modeled using the coupled and segregated flow solvers, with either second-order or third-
order (MUSCL) discretization schemes for the convection term. In four out of the five cases, the turbulence
Re = 6.5 × 106 is modeled using the Spalart-Allmaras turbulence model with the viscous layer resolved. The
fifth case uses Realizable K-Epsilon Two-Layer turbulence.
The data used for comparison are Cp coefficient values on the airfoil surface. The coupled cases also include
monitor plots of lift and drag coefficients vs. iteration.
The third-order MUSCL scheme is activated with the following model properties set to MUSCL 3rd-order/CD:
• Gas
• Gradients
• Ideal Gas
• Low-y+ Wall Treatment (for Spalart-Allmaras only)
• K-Epsilon Turbulence
• Realizable K-Epsilon Two-Layer
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature (for Segregated Flow only)
• Spalart-Allmaras Turbulence
• Standard Spalart-Allmaras
• Steady
• Two Dimensional
• Two-Layer All-y+ Wall Treatment (for K-Epsilon only)
• Turbulent
Material Properties:
Initial Conditions:
Boundary Conditions:
Coupled Solver:
Segregated Solver:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[101] Cook, P.H., McDonald, M.A., and Firmin, M.C.P. 1979. “Aerofoil RAE 2822 - Pressure Distributions,
and Boundary Layer and Wake Measurements Experimental Data Base for Computer Program
Assessment”, AGARD Report AR 138.
http://www.grc.nasa.gov/WWW/wind/valid/raetaf/raetaf.html.
Purpose
This verification case demonstrates compressible inviscid flow through a quasi-1D nozzle.
Overview
This case emulates the compressible flow through a quasi-1D nozzle. An area distribution is suggested by
Demirdzic et al. [103].
Physics
Data sets are compared for Mach number and pressure. The theoretical distributions are obtained from 1D
compressible flow theory. The area distribution, suggested by Demirdzic et al. [103], is given by the following
equation:
2
A = 1 + 1.035 ⋅ 1 − x/5 (85)
Two cases are presented, corresponding to inlet Mach numbers of 0.7 and 1.5, using coupled and segregated
flow.
Physics Models:
• Coupled Energy (for Coupled Flow only)
• Coupled Flow
• Gas
• Gradients
• Ideal Gas
• Inviscid
• Segregated Flow
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[103] Demirdzic, I., Lilek, Z., and Peric, M. 1993. “A collocated finite volume method for predicting flows at
all speeds”, Int. J. for Numerical Methods in Fluids, 16, pp. 1029-1050.
Purpose
This verification case demonstrates inviscid flow with oblique shock waves.
Overview
Inviscid supersonic flow passes a smooth surface. Inlet conditions are from oblique shock relations. The outlet
is a supersonic (pressure) outflow.
The static pressure variation along the line y = 0.5 is compared with the theoretical results.
Physics
This case uses inviscid segregated flow.
Stagnation inlet values for the velocity inlet were obtained using oblique shock wave relations.
Physics Models:
• Gas
• Gradients
• Ideal Gas
• Inviscid
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[104] Yee, H. C., Warming, R. F., and Harten, A. 1985. “Implicit Total Variation Diminishing (TVD) Schemes
for Steady-state Calculations”, J. Comp. Physics, 57, pp. 327-360.
Turbulence
Contents:
Two-Dimensional Single-Hill Flows
Turbulent Compressible Flow in a Two-Dimensional Converging-Diverging Nozzle
Fully Developed Turbulent Pipe Flow
Turbulent Natural Convection in a Heated Cavity of Aspect Ratio 1
Turbulent Flow Over a Surface-Mounted Rib
Turbulent Flow in a Channel Using a Wall-Function Mesh and a Low-Re Mesh
Two-Dimensional Transonic Diffuser
Turbulent Flow and Heat Transfer over a Backward-Facing Step
Heat Transfer In Turbulent Pipe Flow (CGD)
Friction Factors for Smooth Circular Pipe Flow (CGD)
Axisymmetric Impinging Turbulent Jet with Heat Transfer
T3A Transition
Turbulent Flow in a Swirling Diffuser
Curved Turbulent Channel Flow
Transitional Boundary Layer
Turbulent Natural Convection in a Heated Cavity of Aspect Ratio 5
Ascending and Descending Mixed Convection Flow in a Vertical Heated Pipe (CGD)
Purpose
This verification case demonstrates turbulent flow past a surface-mounted obstacle in a channel using different
turbulence models.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences between the
velocities measured in the simulation at different probes and the reference, with a threshold value of 10%, as
shown in the following table. The acceptance criteria threshold is defined on the bases of the CFD solution
provided by Davroux, et al. [106].
Overview
This case is based on one of the two test cases that were prepared for the Fourth ERCOFTAC/IAHR Workshop on
Refined Flow Modeling (ERCOFTAC ‘94 Case 2a). It simulates turbulent flow past a surface-mounted obstacle in
a channel. The geometry and mesh appear in the following figure.
The obstacle, representing a hill, is an inverse, fourth-order polynomial and extends across the whole channel
width. The maximum height and length of the hill model is 28 mm and 108 mm, respectively. Based on the
mean centerline velocity of water at the inlet and the obstacle height, the Reynolds number is 60,000.
Physics
Three turbulence models are compared:
Almeida et al. [105] performed the experimental measurements. Their data files contain the velocity profiles of
the first- (U and V) and second-order moments (u’2, v’2 and u’v’) at 14 specified measurement locations.
Physics Models:
• Constant Density
• Gradients
• Liquid (H2O)
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Steady
• Turbulent
• Two Dimensional
The following table summarizes the turbulence physics models employed in the different cases:
Material Properties:
Initial Conditions:
Boundary Conditions:
K-Epsilon/K-Omega 0.8
Turbulence URF
K-Epsilon/K-Omega 1.0
Turbulent Viscosity URF
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[105] Almeida, G.P., Durao, D.F.G., Heitor, M.V., and Simoes, J.P. 1990 “LDV measurements of fully-
developed turbulent channel flow”, Proc. 5th Int. Symp. on Applying Laser Techniques to Fluid
Mechanics, Lisbon, pp. 9-12.
[106] Davroux, A., Hoa, C. & Laurence, D. 1995. Flow Over a 2D Hill – Reference Solutions for k-e and
Second Moment Closure Turbulence Models, University of Karlsruhe, Germany, ERCOFTAC.
Purpose
This verification case demonstrates compressible, turbulent flow through a converging-diverging nozzle.
Overview
High pressure ideal gas from a plenum flows through a two-dimensional nozzle and exhausts into ambient air.
Physics
The problem is solved using a coupled implicit solver and a standard K-epsilon model. Nozzle pressure ratio is
compared to the reference experimental data found in [107].
Physics Models:
• Coupled Energy
• Coupled Flow
• Gas (Air)
• Gradients
• High y+ Wall Treatment
• Ideal gas
• K-Epsilon Turbulence
• Reynolds-Averaged Navier-Stokes
• Standard K-Epsilon
• Steady
• Turbulent
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[107] Mason, M. L., Putnam, L. E. and Re, R. J., “The Effect of Throat Contouring on Two-Dimensional
Converging-Diverging Nozzles at Static Conditions”, NASA Technical Paper 1704, 1980.
Purpose
This validation case demonstrates the use of turbulence models to simulate fully developed pipe flow. The
radial profile of the non-dimensional velocity u+ is compared against a selected reference.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences between the
non-dimensional velocities measured in the simulation at different probes and the reference, with a threshold
value of 5%, as shown in the following table.
Overview
A two-dimensional axisymmetric mesh represents the problem geometry, which appears in the following
figure.
Utilizing periodic boundary conditions, which allow a fully developed hydrodynamic boundary layer to be
generated, only three cells are used in the axial flow direction. As the following figure shows, the wall-function
type mesh has 20 uniform cells in the transverse direction and the low-Re mesh has 150 stretched cells.
Physics
The simulation is set up for turbulent flow with Re = 44,000. The reference data come from direct numerical
simulations [109].
• D. The Standard K-Epsilon Low-Re turbulence model with low-y+ wall treatment.
• E. The two-layer variant of the Realizable K-Epsilon turbulence model with all-y+ wall treatment.
• F. The SST (Menter) K-Omega turbulence model with low-y+ wall treatment.
• G. The Blended EVM Low-Reynolds Number K-Epsilon model with all-y+ wall treatment.
Physics Models:
• Axisymmetric
• Constant Density
• Gas (Air)
• Gradients
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Steady
• Turbulent
The following table summarizes the turbulence physics models employed in the different cases:
Material Properties:
Initial Conditions—Common:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[108] Schlichting, H. (1968) ‘Boundary Layer Theory’, 6th Edition, McGraw-Hill.
[109] Wu, X. & Moin, P. 2008. A direct numerical simulation study on the mean velocity characteristics in
turbulent pipe flow, Journal of Fluid Mechanics, Vol. 608, pp. 81-112.
Purpose
This verification case demonstrates solving a turbulent natural convection problem at a low turbulent Rayleigh
number.
Acceptance Criteria
The acceptance criteria for the comparison are the overall averages of the unsigned relative differences in the
velocities, temperatures, and Nusselt number between the simulation and the reference, with threshold values
as shown in the following table. The threshold values are derived from the study of Tian and Karayiannis [110],
which indicated a larger uncertainty on the Nusselt number measurement related to the imperfect boundary
conditions. These differences are converted into relative values by normalizing them using the maximum
unsigned value found in the reference for each compared profile. One average is then calculated for the
velocity, one for the temperature, and one for the Nusselt number, using all the available relative differences
in each case.
Overview
The problem geometry is a square 1 m x 1 m cavity, heated on one vertical wall and cooled on the other. The
two horizontal walls are not insulated and have a specified temperature distribution. An 80 x 60 stretched
square mesh is used on the cavity.
Physics
The model is set up to correspond to a Rayleigh number of 2.75E10. The problem is solved using the implicit
coupled solver with the Standard K-Epsilon Low-Re turbulence model. Density is assumed to be constant.
The velocity and temperature profiles at mid-height and the Nusselt numbers on the vertical walls are
compared to the experimental data found in [110].
Physics Models:
• Boussinesq Model
• Constant Density
• Coupled Energy
• Coupled Flow
• Exact Wall Distance
• Gas (Air)
• Gradients
• Gravity
• K-Epsilon Turbulence
• Low y+ Wall Treatment
• Reynolds-Averaged Navier-Stokes
• Standard K-Epsilon Low-Re
• Steady
• Turbulent
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
Note: The plots HotWallNu and ColdWallNu use a common Reference Temperature variable. When the
Reference Temperature is set to 283.0 K, the HotWallNu shows its correct values. When the Reference
Temperature is 323.0 K, the ColdWallNu shows its correct values. To set the Reference Temperature in the
simulation file, select Tools > Field Functions > Nusselt Number and specify the appropriate value in the
Reference Temperature property.
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[110] Tian, Y.S. and Karayiannis, T.G., “Low turbulence natural convection in an air-filled square cavity.
Part 1: the thermal and fluid flow fields”, International Journal of Heat and Mass Transfer, 43, p.
849-866, 1999.
Purpose
This verification case demonstrates flow separation that occurs over a surface mounted rib.
Overview
In 1994, Computational Dynamics Ltd. (CD) collaborated with the Danish Maritime Institute (DMI) in an EU
sponsored project that is aimed at evaluating the performance of various commercial CFD codes. Under this
collaboration, DMI provided the test data and CD performed the CFD simulations. This case simulates two-
dimensional turbulent flow past a surface-mounted obstacle (such as a fence or rib) that extends across the
width of a channel. Based on the mean inlet velocity of air and the obstacle height, the Reynolds numbers
were ranged between 1,500 - 3,000.
The DMI test case of the crossflow over a surface-mounted rib is shown below: the computational mesh (a)
and an enlarged view of locally refined mesh (b).
Physics
Four turbulence models are compared:
• The standard K-Epsilon (SKE) turbulence model with high-y+ wall treatment
• The realizable K-Epsilon (RKE) turbulence model with high-y+ wall treatment
• The two-layer variant of the realizable K-Epsilon (TLRKE) turbulence model with all-y+ wall treatment
• The SST K-Omega (SSTKw) turbulence model with all-y+ wall treatment.
The experimental data files were supplied by DMI and contain the velocity profiles of the first- (U and V),
second- and higher-order moments, plus information which is related to the experimental arrangements. Only
the data that were obtained at Re = 3,000 are used for this case.
Physics Models:
• Constant Density
• Gas (Air)
• Gradients
• High y+ Wall Treatment
• K-Epsilon Turbulence
• Realizable K-Epsilon
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Steady
• Turbulent
• Two Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Purpose
This verification case demonstrates fully developed, wall bounded turbulent flow on a wall-function type mesh
and on a low-Re mesh. The radial profiles of axial velocity, turbulent kinetic energy and turbulent dissipation
rate (when available) are compared with reference data.
Acceptance Criteria
The acceptance criteria for this case, shown in the following table, are the average values of the unsigned
relative differences in the radial profiles of the axial velocity, turbulent kinetic energy and turbulent dissipation
rate.
Overview
Turbulent flow is modeled in a planar channel. A pressure gradient drives the flow. Two different mesh types
are used to test various turbulence models. The following figure provides an overview of the geometry.
The wall-function type mesh has 19 uniform cells in the transverse direction and the low-Re mesh has 39
stretched cells. Both meshes have 4 cells in the longitudinal direction, as can be seen in the following figure.
Physics
The simulation is set up for turbulent flow with Re = 589.19. The reference data comes from direct numerical
simulations [111].
Three turbulence models are compared on the wall-function type mesh:
• D. The Standard K-Epsilon Low-Re turbulence model with low-y+ wall treatment.
• E. The two-layer variant of the Realizable K-Epsilon turbulence model with all-y+ wall treatment.
• F. The SST (Menter) K-Omega turbulence model with low-y+ wall treatment.
• G. The Blended EVM Low-Reynolds Number K-Epsilon model with all-y+ wall treatment.
Physics Models:
• Constant Density
• Coupled Flow
• Gradients
• Liquid (H2O)
• Reynolds-Averaged Navier-Stokes
• Steady
• Turbulent
• Two Dimensional
The following table summarizes the turbulence physics models employed in the different cases:
Material Properties:
Initial Conditions—Common:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[111] Moser, R. D., Kim, J. and Mansour, N. N., “Direct Numerical Simulation of Turbulent Channel Flow up
to Re_tau = 590”, Physics of Fluids, Vol. 11, No 4, 1999.
Purpose
The purpose of this test case is to demonstrate the ability of specific turbulence models (here Spalart-Allmaras)
to predict shocks and flow separation for compressible flows. The present case corresponds to the weak-shock
case described by Georgiadis et al. [112]. (More details can be found on the NASA Web site: Sajben Transonic
Converging-Diverging Diffuser.)
Overview
Air flows through a converging-diverging diffuser. The inlet Mach number is about 0.46. A weak shock occurs
at the throat. Although the case as set up here has been scaled to give a throat dimension of 1, the physical
setup has a throat dimension of 0.044, resulting in a Reynolds number of about 590,000.
Physics
The Spalart-Allmaras Turbulence model is used with steady coupled flow in a two-dimensional mesh.
The reference data consists of pressure values on the top and bottom walls, and velocity profile data at four
stations downstream of the throat.
Physics Models:
• Coupled Energy
• Coupled Flow
• Gas
• Gradients
• Ideal Gas
• Low-y+ Wall Treatment
• Reynolds-Averaged Navier-Stokes
• Spalart-Allmaras Turbulence
• Standard Spalart-Allmaras
• Steady
• Turbulent
• Two-Dimensional
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[112] Georgiadis, N.J., Drummond, J.E., and Leonard, B.P., 1994. “Evaluation of Turbulence Models in the
PARC Code for Transonic Diffuser Flows”, NASA TM-106391.
Purpose
This case demonstrates complex turbulent flow and heat transfer with a wall-function type mesh.
Acceptance Criteria
The acceptance criteria for the comparison are the respective averages of the unsigned relative differences in
the wall shear stress and Nusselt numbers between the simulations and the reference, measured at the probe
locations from the reference. The criteria are shown in the following table and are obtained from the 6th
ERCOFTAC Workshop results [113]. The relative differences were normalized by using the largest unsigned
values of the wall shear stress and Nusselt number found in the reference, respectively.
Overview
Developing flow in a channel encounters a backward-facing step. The ratio of channel height to step height is
5. Reattachment of the flow occurs 6.67 step heights (with height, H = 1 m) downstream. Developing
boundary layer profiles are specified on the top and bottom walls. The inlet is specified 3.8 step-heights
upstream of the step, and the outlet is 30 step-heights downstream of the step.
The mesh is sufficiently coarse to permit a wall function solution at this relatively low Reynolds number. Values
of y+ typically range from 30 to 50 on the walls. However, immediately downstream of the step, y+ values
drop to as low as 9.
Inlet (m) 4
Step (m) 1
Outlet (m) 5
Physics
The constant-density turbulent flow has a Reynolds number of 28,000 (based on the step height, H = 1 m, and
the free stream velocity). The bottom wall is heated with a heat flux of 500 W/m2.
Two turbulence models are tested on a two-dimensional mesh, with steady coupled flow:
Material Properties:
Density (kg/m2) 2
Dynamic Viscosity (Pa-s) 7.142 x 10–5
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
• Clear the data using Solution > Clear Solution from the top menu bar. Click OK in the Clear Solution
dialog that appears.
• Select the Stopping Criteria > Maximum Steps node. Increase the Maximum Steps property to 1500.
• To ensure convergence, select the Solvers > Coupled Implicit node. Lower the Courant Number property
to 10. To get a quicker convergence, gradually increase the Courant number during the run:
For the first ten iterations, set it to 10.
For the next 100 iterations, set it to 50.
For the remaining iterations, set it to 1000.
• Choose Solution > Run from the top menu bar.
Bibliography
[113] Rodi, W. & Mulas, M., 2005. Engineering Turbulence Modelling and Experiments 6. s.l., ERCOFTAC
International Symposium on Engineering Turbulence and Measurements-ETMM6. Elsevier.
[114] Vogel, J. C. and Eaton, J. K. 1985. “Combined Heat Transfer and Fluid Dynamic Measurements
Downstream of a Backward-Facing Step”, AIAA J. 23(2) pp. 163-171.
Purpose
This verification case demonstrates heat transfer in fully developed, wall bounded turbulent flow inside of a
circular pipe for fluids with different Prandtl numbers. The predicted Nusselt numbers are compared with
correlations. Two meshes were employed, a wall-function type mesh and a low-Re mesh.
Acceptance Criteria
The acceptance criteria for the comparison are the average value of the unsigned relative differences in the
Nusselt number between simulations and correlations. The Petukhov ([115]) correlation is used for Prandtl
numbers 1, 10, 100, where it has shown an accuracy of 10%. Further, the Sleicher correlation ([116]) is used
to compare a larger range of Prandtl numbers (0.01 < Pr < 100), where the wide applicability of the correlation
comes at the cost of a larger 30% error. For each case, the average was calculated between all the Prandtl
numbers simulated.
Overview
Heat transfer in a turbulent flow is modeled in a circular pipe. An imposed mass flow rate which drives the flow
with a Reynolds number of 50,000 was analyzed.
Two different mesh types are used to test various turbulence models. In the wall-function type mesh, the near
wall cell thickness is calculated to ensure a y+ ≅ 50 and the number of cells across the section is set to 13, to
maintain a uniform cell size. In the low-Re type mesh, the near wall cell thickness is calculated to ensure a
y+ ≅ 1 whilst the number of transverse cells is 150. In both cases 3 cells are used in the longitudinal direction.
Physics
The simulation is set up with a turbulent flow at a Reynolds numbers of 50,000. Fluids with different Prandtl
numbers were employed, as shown in the following table. The reference data is based on the correlations from
Petukhov ([115]) and Sleicher ([116]).
• A. The two-layer variant of the Standard K-Epsilon (TLSKE) turbulence model with all-y+ wall treatment.
• B. The two-layer variant of the Realizable K-Epsilon (TLRKE) turbulence model with all-y+ wall treatment.
• C. The Elliptic Blending K-Epsilon (EBKE) model with all-y + wall treatment.
• D. The Modified Quadratic K-Epsilon (MQKE) high Reynolds model with high-y+ wall treatment.
• E. The two-layer variant of the Standard K-Epsilon (TLSKE) turbulence model with all-y+ wall treatment.
• F. The two-layer variant of the Realizable K-Epsilon (TLRKE) turbulence model with all-y+ wall treatment.
• G. The Elliptic Blending K-Epsilon (EBKE) model with all-y + wall treatment.
• H. The Modified Quadratic K-Epsilon (MQKE) low Reynolds model with low-y+ wall treatment.
Physics Models:
• Axisymmetric
• Constant Density
• Liquid
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Turbulent
The following table summarizes the physics models employed in the different cases:
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[115] Petukhov, B., 1970. Heat transfer and friction in turbulent pipe flow with variable physical
properties. Adv. Heat Transfer, Volume 6, pp. 503-565.
[116] Sleicher, C. & Rouse, M., 1975. A convenient correlation for heat transfer to constant and variable
property fluids in turbulent pipe flow, International Journal of Heat and Mass Transfer. 18(5), pp.
677-683.
Purpose
This verification case demonstrates fully developed, wall bounded turbulent flow in a circular pipe at different
Reynolds numbers. The predicted friction factors are compared with correlations and experimental data. Two
meshes were employed, a wall-function type mesh and a low-Re mesh.
Acceptance Criteria
The acceptance criterion for the comparison is the average value of the unsigned relative differences in friction
factors between simulations and the Colebrook correlation [118]. For each case, the average was calculated
from all the Reynolds numbers simulated. The acceptance criterion represents a desired accuracy level for CFD
simulations regardless of the specific turbulence model adopted. The performance of a specific model allows
much improved accuracy from the acceptance level.
Overview
Turbulent flow is modeled in a circular pipe. An imposed mass flow rate drives the flow and several Reynolds
numbers were analyzed. Two different mesh types are used to test various turbulence models. A number of
turbulence models are employed for each mesh, and several cases with different Reynolds numbers are run for
each case in the range of 20,000 – 500,000.
The meshes are changed depending on the Reynolds number to ensure a constant y+ amongst the cases. In
the wall-function type mesh (a), the near wall cell thickness is calculated in each case to ensure y+ ≅ 50 and
the number of cells across the section is recalculated to maintain them at uniform size. In the low-Re type
mesh (b), the near wall cell thickness is calculated to ensure y+ ≅ 1 whilst the number of transverse cells is
kept at a constant value of 150. In all cases 3 cells are used in the longitudinal direction.
Physics
The simulation is set up for turbulent flow at a series of different Reynolds numbers, as shown in the following
table. The reference data come from the Colebrook correlation [118] and the Princeton Superpipe experiment
[119].
Reynolds Number
20,000
50,000
100,000
250,000
Reynolds Number
500,000
• A. The two-layer variant of the Standard K-Epsilon turbulence model with all-y+ wall treatment.
• B. The two-layer variant of the Realizable K-Epsilon turbulence model with all-y+ wall treatment.
• C. The Elliptic Blending K-Epsilon model with all-y+ wall treatment.
• D. The Modified Quadratic K-Epsilon high Reynolds model with high-y+ wall treatment [117].
Four turbulence models are compared on the low-Re mesh:
• E. The two-layer variant of the Standard K-Epsilon turbulence model with all-y+ wall treatment.
• F. The two-layer variant of the Realizable K-Epsilon turbulence model with all-y+ wall treatment.
• G. The Elliptic Blending K-Epsilon model with all-y+ wall treatment.
• H. The Modified Quadratic K-Epsilon high Reynolds model with high-y+ wall treatment [117].
Physics Models:
• Axisymmetric
• Constant Density
• Liquid
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Turbulent
The following table summarizes the turbulence physics models employed in the different cases:
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
Save Link As... (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[117] Baglietto, E. 2006. Anisotropic Turbulence Modeling for Accurate Rod Bundle Simulations, Miami
USA., Proceedings of ICONE14 Conference.
[118] Colebrook, C. F. & White, C. M. 1937. Experiments with Fluid Friction in Roughened Pipes,
Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences. 161
(906): 367–381.
[119] McKeon, B. J. et al. 2004. Friction factors for smooth pipe flow, J. Fluid Mech., Vol. 511, pp. 41- 44.
Purpose
This case demonstrates stagnation heat transfer using a wall-function type mesh.
Overview
A jet emanating from fully-developed pipe flow impinges on a heated wall that is perpendicular to the jet. The
jet is isothermal and the wall is heated with a constant heat flux. The flow is axisymmetric in a two-
dimensional mesh. The heated wall is placed 10 jet diameters away from the jet. The pipe inlet is 5 diameters
long. A pressure boundary is placed 10 diameters from the jet centerline.
Two different mesh resolutions are used to test various turbulence models.
The mesh is oriented so that the jet flow is on the x-axis and the heated wall is aligned with the x-y plane.
Physics
The constant-density turbulent flow has a Reynolds number of 23,750. Four variants of the K-Epsilon
turbulence model are tested on an axisymmetric two-dimensional mesh, with steady coupled flow:
• A. The Realizable K-Epsilon Two-Layer model with two-layer all-y+ wall treatment and wall-function type
mesh.
• B. The Realizable K-Epsilon Two-Layer model with two-layer all-y+ wall treatment and low-Re type mesh.
• C. The V2F Low-Reynolds Number K-Epsilon model with V2F all-y+ wall treatment
• D. The Blended EVM Low-Reynolds Number K-Epsilon model with all-y+ wall treatment
The reference data come from experimental data for the Nusselt number on the bottom wall [120].
Physics Models:
• Axisymmetric
• Constant Density
• Coupled Energy
• Coupled Flow
• EB K-Epsilon
• EB All-y+ Wall Treatment [EB K-Epsilon]
• Gas
• Gradients
• K-Epsilon Turbulence
• Realizable K-Epsilon Two-Layer
• Reynolds-Averaged Navier-Stokes
• Steady
• Turbulent
• Two-Layer All-y+ Wall Treatment [Realizable K-Epsilon]
• V2F All-y+ Wall Treatment [V2F K-Epsilon]
• V2F K-Epsilon
Material Properties:
Density (kg/m3) 1
Dynamic Viscosity (Pa-s) 8.421 x 10–5
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[120] Baughn J. W. and Shimuzu, S. 1989. “Heat Transfer Measurements from a Surface With Uniform
Heat Flux and an Impinging Jet”, J. Heat Transfer 111 pp. 1096-1098.
[121] Behnia, M., Parneix, S. & Durbin, P., 1998. "Prediction of heat transfer in an axisymmetric turbulent
jet impinging on a flat plate", Int. J. Heat and Mass Transfer, pp. 1845-1855.
[122] Craft, T., Graham, L. & Launder, B., 1993. "Impinging jet studies for turbulence model assessment—
II. An examination of the performance of four turbulence models", Int. J. Heat and Mass Transfer,
36(10), pp. 2685-2697.
T3A Transition
Purpose
This case verifies that the Gamma ReTheta model predicts correctly the onset of the transition and the
transition length for a high free-stream turbulence intensity. (Free-stream turbulence is considered high when
it is above 0.5%.)
Overview
This case uses a flat plate developing boundary layer subjected to decaying freestream turbulence.
Physics
The SST (Menter) K-Omega Turbulence model is used with steady coupled flow in a two-dimensional mesh.
User field functions are created specifically for the Gamma ReTheta Transition model to indicate the state of
the boundary layer. The correlations of these field functions provide the best results for a large range of
Reynolds numbers and free-stream turbulence intensities. The syntax of these field functions is as follows:
• Flength:
• ReCrit:
The Correlation Method expert property of the Gamma ReTheta Transition model is set to User Field Function,
and then the sub-nodes of the model are set as follows:
• Gas
• Gradients
• K-Omega Turbulence
• Low-y+ Wall Treatment
• Reynolds-Averaged Navier-Stokes
• SST (Menter) K-Omega
• Steady
• Turbulent
• Two-Dimensional
Material Properties:
Initial Conditions:
Intermittency 0.0
Pressure (Pa) 0.0
Specific Dissipation Rate (/s) 270.0
Turbulent Kinetic Energy (J/Kg) 0.05
Velocity (m/s) U = 5.4
V=0
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
References
ERCOFTAC Classic Database Case 20
Purpose
This verification case demonstrates turbulent predictions in a strongly anisotropic flow condition. Several
results are compared with reference experimental measurements (Clausen, et al. [126]).
• axial velocity
• swirling velocity
• turbulent kinetic energy
• pressure coefficient
All these relative differences are normalized with the largest unsigned respective value found in the reference
data. The threshold values employed are shown in the following table. The acceptance criteria reflect the
combination of experimental error and limitations in the reproduction of flow anisotropy typical of the
turbulence closures. The values have been selected on the basis of previous numerical studies by Armfield, et
al. [123] and Cho and Fletcher [reflink].
Overview
Swirling turbulent flow enters an axisymmetric diffuser with an included angle of 10 degrees. Inlet axial and
swirl velocity profiles are provided, together with Reynolds stress profiles. The constant inlet dissipation rate is
computed from preliminary calculation such that the streamwise variation of the skin-friction coefficient Cf
matches the theoretical evolution given by:
0.455
Cf = (86)
ln2 0.06 Re x − 60, 000
Actual dimensions are used for the mesh, but a unit inlet reference velocity has been used, necessitating a
modified viscosity to match the Reynolds number (202,000 based on inlet diameter).
Physics
The constant-density turbulent flow has a Reynolds number of 202,000. Three turbulence models with High-
y+ wall treatment, and one (Case B) with Two-Layer All-y+ wall treatment, are tested on a three-dimensional
mesh:
The experimental data is for the pressure coefficient on the top wall and Reynolds stress profiles at various
axial stations.
Physics Models:
• Constant Density
• Gas
• Gradients
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Steady
• Three-Dimensional
• Turbulent
Material Properties:
Initial Conditions:
Boundary Conditions:
Cases A, B, and C
Case D
Case D
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[123] Armfield, S., Cho, N., and Fletcher, C., 1990. "Prediction of turbulence quantities for swirling flow in
conical diffusers", AIAA, 28, No. 3 p. 453.
[124] Baglietto, E. 2006. Anisotropic Turbulence Modeling for Accurate Rod Bundle Simulations, Miami
USA., Proceedings of ICONE14 Conference.
[125] Cho, N. and Fletcher, C., 1991. "Computation of turbulent conical diffuser flows using a non-
orthogonal grid system", Computers and Fluids, 19 p. 347.
[126] Clausen, P. D., Koh, S. G., and Wood, D. H. 1993. “Measurements of a Swirling Turbulent Boundary
Layer Developing in a Conical Diffuser”, Experimental Fluid and Thermal Fluid Sci., 6 pp. 39-48.
Purpose
This case demonstrates the applicability of turbulence models to flow in presence of streamline curvature. The
effect of curvature correction modification for two-equation models is also assessed. The axial velocity profile
on a transverse plane component is compared against a selected reference.
Acceptance Criteria
The acceptance criterion for the comparison is the average of the unsigned relative differences between the
axial velocity measured in the simulation at different probes and the reference, with a threshold value of 5%,
as shown in the following table. The relative differences were normalized using the largest velocity value
measured in the reference at the selected transverse plane. The acceptance criteria represents a desired
accuracy level for CFD simulations regardless of the specific turbulence model adopted.
Overview
This case models periodic curved channel flow using a grid of 50 x 120 quadrilateral cells, as shown in the
following figure.
Physics
Four turbulence models are compared:
• A. The realizable K-Epsilon Two-Layer turbulence model with all-y+ wall treatment and curvature
correction (the Curvature Correction Option property activated)
• B. The SST K-Omega turbulence model with high-y+ wall treatment and curvature correction
• C. The realizable K-Epsilon Two-Layer turbulence model with all-y+ wall treatment (without any curvature
correction)
• D. The modified quadratic variant of the K-Epsilon turbulence model with high-y+ wall treatment
(without any curvature correction)
Physics Models:
• Constant Density
• Coupled Flow (for K-Omega turbulence only)
• Gas
• Gradients
• Reynolds-Averaged Navier-Stokes
• Segregated Flow (for K-Epsilon turbulence only)
• Steady
• Turbulent
• Two Dimensional
The following table summarizes the turbulence physics models employed in the different cases:
Material Properties:
Initial Conditions—Common:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[127] Baglietto, E. 2006. Anisotropic Turbulence Modeling for Accurate Rod Bundle Simulations, Miami
USA., Proceedings of ICONE14 Conference.
[128] Shur, M. L., Strelets, M. K., Travin, A. K., and Spalart, P. R. 2000. “Turbulence Modeling in Rotating
and Curved Channels: Assessing the Spalart-Shur Correction”, AIAA Journal, 38(5), pp. 784-792.
[129] Wattendorf, F. L. 1935. “A study of the effect of curvature on fully developed turbulent flow”, Proc.
Royal Soc., 148(865), pp. 565–598.
Purpose
This case demonstrates a transitional boundary layer on a low-Reynolds-number type mesh.
Acceptance Criteria
The acceptance criteria for the comparison is the average of the unsigned relative differences in the skin
friction coefficients measured at different probes, when comparing simulated and experimental values, as
shown in the following table. The relative differences were normalized using the largest unsigned values of the
laminar and turbulent skin friction coefficients, respectively, from the reference. While turbulence models are
not specifically designed to capture transitional boundary layers, this test allows quantifying the error that
would be introduced when applying the model to a transitional condition. The acceptance criteria represents a
desired accuracy level regardless of the specific turbulence model adopted.
Overview
The problem geometry is shown in the figure below and it consists of a two-dimensional rectangular duct, 1 m
in length. Flow enters a long duct parallel to a wall. Transition is suppressed until Rex = 60, 000, after which
the flow is allowed to become turbulent.
Physics
Two turbulence models are compared, both with all-y+ wall treatment, in a transitional boundary layer flow
0 < Rex < 300, 000 :
A comparison is made to formulas for skin friction as a function of Reynolds number based on streamwise
distance. In the turbulent suppression region, the formula is:
0.664
Cf = (87)
Reα1/2
0.455
Cf = 2
(88)
ln 0.06 Rex − 60, 000
Physics Models:
• Constant Density
• Coupled Flow
• Gradients
• Liquid
• Reynolds-Averaged Navier-Stokes
• Steady
• Transition Boundary Distance
• Turbulence Suppression
• Turbulent
• Two Dimensional
The following table summarizes the turbulence physics models employed in the different cases:
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[130] White, F. M. 1991. Viscous Fluid Flow, 2nd ed. McGraw-Hill.
Purpose
This verification case demonstrates a simulation of turbulent natural convection at a moderately high Rayleigh
number on a high-quality stretched mesh.
Acceptance Criteria
The acceptance criteria for the comparison are the overall averages of the unsigned relative differences in the
velocities and temperatures between the simulation and the reference. Threshold values for velocity and
temperature coefficients are selected on the basis of the assessment of Xu [132], as shown in the following
table. These differences are converted into relative values by normalizing them using the maximum unsigned
value found in the reference for each compared profile. One average is then calculated for the velocity and one
for the temperature, using all the available relative differences in each case.
Overview
A tall (aspect ratio 5) cavity is heated on one vertical wall, and cooled on the other vertical wall. The bottom
wall is insulated; a linear temperature profile is imposed on the top wall. The stretched mesh that is used for
the case consists of 49 cells in the X and Y directions.
Geometry Grid
Physics
To simulate natural convection at high laminar Rayleigh numbers Ra = 4.75 × 1010 , this case uses variants of
K-Epsilon turbulence with coupled and segregated solvers. Constant density is assumed.
The velocity profile at mid-height and the core temperature profile are compared to values provided by
Cheesewright et al. [131]. The case uses values of the actual mean temperature chosen by Xu et al. [132],
Tℎ = 338K and Tc = 293K.
Physics Models:
• Boussinesq Model
• Constant Density
• Coupled Energy (for Coupled Flow only)
• Coupled Flow
• Exact Wall Distance
• Gas
• Gradients
• Gravity
• K-Epsilon Turbulence
• Low-y+ Wall Treatment (for Segregated Flow only)
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature (for Segregated Flow only)
• Standard K-Epsilon Low-Re (for Segregated Flow only)
• Standard K-Epsilon Two-Layer (for Coupled Flow only)
• Steady
• Temperature Flux Model (for Segregated Flow only)
• Turbulent
• Two Dimensional
• Two-Layer All-y+ Wall Treatment (for Coupled Flow only)
Material Properties:
Initial Conditions:
Boundary Conditions:
Coupled Solver:
Segregated Solver:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[131] Cheesewright, R., King, K. J., and Ziai, S. 1986. “Experimental Data for the Validation of Computer
Codes for the Prediction of Two-Dimensional Buoyant Cavity Flows”, Significant Questions in
Buoyancy Affected Enclosure or Cavity Flows, J. A. C. Humphrey et. al. eds. 75-81, ASME Winter
Annual Meeting.
[132] Xu, W., Chen, Q., and Nieuwstadt, F.T.M. 1998. “A new turbulence model for near-wall natural
convection”, Int. J. Heat Mass Transfer, 41, pp. 3161-3176.
Purpose
This verification case demonstrates fully developed, wall bounded flow in a vertical heated pipe with a circular
cross section, at different mixed convection conditions; that is, when a forced convection flow is largely
affected by buoyancy. The friction factors and Nusselt numbers predicted with different turbulence models are
compared with reference data obtained with DNS simulations and experimental data.
Acceptance Criteria
The acceptance criteria for the comparison are the average values of the unsigned relative differences in local
friction factors and Nusselt numbers between simulations and the reference data. The average was calculated
from all the cases simulated with each turbulence model, including ascending and descending flows. The
relative differences were normalized using values for forced convection (no buoyancy effects) provided by the
reference.
The acceptance criteria represent desired accuracy levels for CFD simulations regardless of the specific
turbulence model adopted. The performance of a specific model allows much improved accuracy from the
acceptance level.
Overview
Turbulent flow in mixed convection condition is modeled in a circular pipe. A value of 5,300 is imposed for the
Reynolds number and the Prandtl number is equal to 0.71. The pipe is heated by imposing a constant heat flux
on the pipe walls. As the flow is assumed to be fully developed, a short portion of the pipe only need be
modeled in combination with fully developed interfaces. The buoyancy of the gas is accounted for using the
Boussinesq model. Different buoyancy conditions were tested for both ascending and descending flows. Four
low Reynolds turbulence models were included in the analysis.
The following image illustrates the geometry and the computational mesh employed. The mesh is of low
Reynolds type, with a near wall cell thickness calculated to provide y+ ≅ 0.5. The number of transverse cells is
80 and they are stretched smoothly to provide the desired near wall cell thickness. Three cells are used in the
longitudinal direction.
Physics
The simulation is set up for turbulent flow at different buoyancy levels. The buoyancy influence was
characterized with the Buoyancy parameter, as expressed by Jackson et al. ([135]):
Gr
Bo = 8 ⋅ 104 3.425
(89)
Re Pr 0.8
βgD4q̇
Gr = (90)
λν2
where β is the coefficient of volumetric expansion, g is the gravity, D is the pipe diameter, q̇ is the heat flux
imposed on the walls, λ is the thermal conductivity of the fluid, and ν is its kinematic viscosity.
A series of values of the Buoyancy parameter were tested, as shown in the following tables. Also as shown in
the tables, the number of simulations is larger than the cases for which reference data are available. The
reason is that in the ascending flow case there can be an impairment or an enhancement in the heat transfer
depending on the Buoyancy parameter value. The impairment has a very sharp transition and thus a large
number of points were included in that range to demonstrate the capability of Simcenter STAR-CCM+ to
capture the effect.
The reference data comes from the DNS simulations carried out by You et al. ([136]) and additionally the
experimental data from Carr, et al. ([134]).
The values for which reference data is available are marked with an asterisk.
A. The Standard K-Epsilon Low-Re (SKELR) turbulence model with low-y+ wall treatment
B. The Elliptic Blending K-Epsilon (EBKE) model with all-y+ wall treatment
C. The Modified Quadratic K-Epsilon (MQKE) low Reynolds model with low-y+ wall treatment
D. The V2F Low-Reynolds Number K-Epsilon (V2FKE) model with V2F all-y+ wall treatment
Physics Models:
• Axisymmetric
• Boussinesq
• Constant Density
• Gas
• Gradients
• Gravity
• Reynolds-Averaged Navier-Stokes
• Segregated Flow
• Segregated Fluid Temperature
• Steady
• Turbulent
Material Properties:
Initial Conditions:
Boundary Conditions:
To download the final simulation to your own directory structure, right-click the link in this column and choose
“Save Link As...” (or the equivalent choice for your browser).
To rerun the simulation, clear the data using Solution > Clear Solution from the top menu bar. Click OK in the
Clear Solution dialog that appears, then choose Solution > Run from the top menu bar.
Bibliography
[133] Baglietto, E., 2006. "Anisotropic Turbulence Modeling for Accurate Rod Bundle Simulations", Miami
USA., Proceedings of ICONE14 Conference.
[134] Carr, A. D., Connor, M. A. & Buhr, H. O., 1973. "Velocity, temperature, and turbulence measurements
in air flow for pipe flow with combined free and forced convection", ASME J. Heat Transfer 95, pp.
445-452.
[135] Jackson, J. D., Cotton, M. A., and Axcell, B. P., 1989. "Studies of mixed convection in vertical tubes",
Int. J. Heat Fluid Flow, Volume 10, pp. 2-15.
[136] You, J., Yoo, Y. & Choi, H., 2003. "Direct numerical simulation of heated vertical air flows in fully
developed turbulent mixed convection", Int. J. of Heat and Mass Transfer, Volume 46, pp.
1613-1627.
Europe
Stephenson House
Sir William Siemens Square
Frimley, Camberley
Surrey, GU16 8QD
+44 (0) 1276 413200
Asia-Pacific
Suites 4301-4302, 43/F
AIA Kowloon Tower,
Landmark East
100 How Ming Street
Kwun Tong, Kowloon
Hong Kong
+852 2230 3308
© 2022 Siemens
This software and related documentation are proprietary and confidential
to Siemens. A list of relevant Siemens trademarks can be found here.
Other trademarks belong to their respective owners.