3D FLOW ANALYSIS OF AN ANNULAR DIFFUSER WITH AND WITHOUT STRUTS
-
Upload
iaeme-publication -
Category
Engineering
-
view
89 -
download
2
Transcript of 3D FLOW ANALYSIS OF AN ANNULAR DIFFUSER WITH AND WITHOUT STRUTS
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM -2014
17 – 19, July 2014, Mysore, Karnataka, India
222
3D FLOW ANALYSIS OF AN ANNULAR DIFFUSER WITH AND WITHOUT
STRUTS
Pramod B A1, Dr. B Sadashive Gowda
2
1(M.Tech Student, Department ofMechanical Engineering, Vidyavardaka college of Engineering, Mysore, Karnataka)
2(Principal, VVCE, Mysore 570002, Vidyavardaka college of Engineering, Mysore, Karnataka)
ABSTRACT
Numerical investigations have been carried out for an annular type gas turbine exhaust diffuser with inlet
guidevanes and with and without struts. Numerical simulations were carried out to determine the pressure recovery
coefficient, for a divergence angle 13o by keeping the diffusion length constant. The flow conditions at the inlet are
varied to evaluate how they affect the flow development in the passage. The velocity at inlet is varied from 80 m/s to 160
m/s in the steps of 40 m/s. In the present study a (1/6)th part of the model is considered for the analysis, due to symmetry.
The results for with and without struts indicates how the pressure recovery coefficient affects the efficiency of the
turbine.
Keywords: Annular diffuser, Exhaust diffuser, Gas turbine, Numerical simulations, Pressure recovery coefficient.
1. INTRODUCTION
A diffuser is a device that increases the pressure of a fluid at the expense of its kinetic energy [1]. The cross
section area of diffuser increases in the direction of flow, therefore fluid is decelerated as it flows through it, causing a
rise in static pressure along the stream. Annular diffusers are extensively used in axial flow compressors and turbines to
convert the kinetic energy of the exhaust flow into pressure. This makes the diffuser a critical element in the performance
of the turbine, which is often neglected.
The exhaust diffuser of an industrial gas turbine recovers the static pressure by decelerating the turbine
discharge flow. In the modern turbine the Mach number at the exhaust is around 0.4-0.45 and the total energy produced
by the turbine is approximately 350KJ/kg, in consequence, a 10% of the total energy of the turbine which is 35KJ/Kg is
being wasted or loss at the exhaust by entering into atmosphere[9]. Only a very few studies on experimental and
numerical investigations on simple diffusers are available [2], [3], [4] and the factors influencing their performance are
predominantly the area ratio and the length of the flow path over which diffusion occurs. At diffusers inlet, the intensity
of the turbulence is usually very high due to the swirl nature of the flow. It is a well known fact that, within diffusers the
flows are characterized by strong adverse pressure gradients which tends to flow to separate from the walls. In the
literature a very few researches are available on experimental analysis concerning on annular diffusers downstream to the
turbine [5], [6] or a compressor [7], [8]. In an annular diffuser, a number of different geometric variables can influence
the variation of pressure recovery and inlet condition of flow.
Struts are structural members which serve as both load bearings support and as passages for cooling air and
lubricating oil. The overall performance of the diffuser is highly influenced by the presence of struts. In the present study
the geometrical details of the struts are taken from NACA-0021 profile, which is one of the NACA four-digit wing
section series.
INTERNATIONAL JOURNAL OF MECHANICAL ENGINEERING
AND TECHNOLOGY (IJMET)
ISSN 0976 – 6340 (Print)
ISSN 0976 – 6359 (Online)
Volume 5, Issue 9, September (2014), pp. 222-231
© IAEME: www.iaeme.com/IJMET.asp
Journal Impact Factor (2014): 7.5377 (Calculated by GISI)
www.jifactor.com
IJMET
© I A E M E
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM
2. GEOMETRICAL DETAILS AND MATHEMATICAL FORMULATION
Diffuser Geometry Numerical investigations have been carried out in a gas turbine
Diffuser) to study the effect of the divergence angle and the Reynolds number. A series of 24 guide vanes and 6 struts
have been used, which provides a means of introducing swirl and aerodynamic blockage into the test section. The
Diffuser assembly of the scaled down gas turbine exhaust diffuser with and without strut is as shown in Fig. 1. The
geometrical details of 35% scaled down gas turbine exhaust diffuser is shown in Fig. 2, where the diffusing length is
450mm, inlet and outlet diameters are 190mm and 320mm, respectively and half cone angle is 13
Fig. 1: Diffuser assembly with and without struts
Fig. 2: Geometry of 35% Gas Turbine Exhaust Diffuser [9, 10]
3. GRID INDEPENDENCE STUDY
A commercially available meshing tool is used to generate tetrahedral grids. An unstructured mesh is
generated with the Gambit (Ver. 2.3.16) software meshing tool. The mesh density is increased near the surface of the
blades and struts to ensure accuracy, as velocity changes in th
highly concentrated at the surface of the blades. This is done by performing a number of simulations with different
mesh sizes, starting from a coarse mesh and refining it until physical results a
The tetrahedral cells are retained to have good discretization accuracy. Four mesh refinements of 1159917, 1645161,
2099319 and 2475980 cells are tested and compared with a velocity which is a physical parameter of th
cross section of the annular diffuser as shown in Fig. 3. For good accuracy of the results, the case with 2099319 cells
can be employed as this gives no significant difference compared to that with 2475980 cells. A fine mesh of 2099319
cells is chosen for further analysis.
International Conference on Current Trends in Engineering and Management ICCTEM
17 – 19, July 2014, Mysore, Karnataka, India
223
2. GEOMETRICAL DETAILS AND MATHEMATICAL FORMULATION
Diffuser Geometry Numerical investigations have been carried out in a gas turbine exhaust diffuser (Annular
Diffuser) to study the effect of the divergence angle and the Reynolds number. A series of 24 guide vanes and 6 struts
have been used, which provides a means of introducing swirl and aerodynamic blockage into the test section. The
Diffuser assembly of the scaled down gas turbine exhaust diffuser with and without strut is as shown in Fig. 1. The
geometrical details of 35% scaled down gas turbine exhaust diffuser is shown in Fig. 2, where the diffusing length is
et diameters are 190mm and 320mm, respectively and half cone angle is 13o.
Diffuser assembly with and without struts
Geometry of 35% Gas Turbine Exhaust Diffuser [9, 10]
meshing tool is used to generate tetrahedral grids. An unstructured mesh is
generated with the Gambit (Ver. 2.3.16) software meshing tool. The mesh density is increased near the surface of the
blades and struts to ensure accuracy, as velocity changes in the fluid are expected to occur in these regions. The mesh is
highly concentrated at the surface of the blades. This is done by performing a number of simulations with different
mesh sizes, starting from a coarse mesh and refining it until physical results are no more dependent on the mesh size.
The tetrahedral cells are retained to have good discretization accuracy. Four mesh refinements of 1159917, 1645161,
2099319 and 2475980 cells are tested and compared with a velocity which is a physical parameter of th
cross section of the annular diffuser as shown in Fig. 3. For good accuracy of the results, the case with 2099319 cells
can be employed as this gives no significant difference compared to that with 2475980 cells. A fine mesh of 2099319
International Conference on Current Trends in Engineering and Management ICCTEM -2014
19, July 2014, Mysore, Karnataka, India
exhaust diffuser (Annular
Diffuser) to study the effect of the divergence angle and the Reynolds number. A series of 24 guide vanes and 6 struts
have been used, which provides a means of introducing swirl and aerodynamic blockage into the test section. The
Diffuser assembly of the scaled down gas turbine exhaust diffuser with and without strut is as shown in Fig. 1. The
geometrical details of 35% scaled down gas turbine exhaust diffuser is shown in Fig. 2, where the diffusing length is
meshing tool is used to generate tetrahedral grids. An unstructured mesh is
generated with the Gambit (Ver. 2.3.16) software meshing tool. The mesh density is increased near the surface of the
e fluid are expected to occur in these regions. The mesh is
highly concentrated at the surface of the blades. This is done by performing a number of simulations with different
re no more dependent on the mesh size.
The tetrahedral cells are retained to have good discretization accuracy. Four mesh refinements of 1159917, 1645161,
2099319 and 2475980 cells are tested and compared with a velocity which is a physical parameter of the fluid for each
cross section of the annular diffuser as shown in Fig. 3. For good accuracy of the results, the case with 2099319 cells
can be employed as this gives no significant difference compared to that with 2475980 cells. A fine mesh of 2099319
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM
Fig. 3: Variation of velocity for annular diffuser with different mesh cells for 13
4. GOVERNING EQUATIONS
The calculation procedure is based on the solution of the equations governing the c
momentum in the time averaged form for a steady incompressible flow. These equations can be written in tensor notation
as:
The quantities ( ) represent the turbulent Reynolds stresses.
The k- model:
The Reynolds stresses are linearly related to the mean rate strain as,
The turbulent viscosity µ t is expressed as
Where k and ε are the turbulence kinetic energy and dissipation rate of turbulence, the values of which are
obtained from the solution of the following transport equations:
International Conference on Current Trends in Engineering and Management ICCTEM
17 – 19, July 2014, Mysore, Karnataka, India
224
Variation of velocity for annular diffuser with different mesh cells for 13o degree casing angle
The calculation procedure is based on the solution of the equations governing the conservation of mass and
momentum in the time averaged form for a steady incompressible flow. These equations can be written in tensor notation
) represent the turbulent Reynolds stresses.
The Reynolds stresses are linearly related to the mean rate strain as,
are the turbulence kinetic energy and dissipation rate of turbulence, the values of which are
obtained from the solution of the following transport equations:
International Conference on Current Trends in Engineering and Management ICCTEM -2014
19, July 2014, Mysore, Karnataka, India
degree casing angle
onservation of mass and
momentum in the time averaged form for a steady incompressible flow. These equations can be written in tensor notation
are the turbulence kinetic energy and dissipation rate of turbulence, the values of which are
(2)
(3)
(4)
(5)
(6)
(1)
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM
The turbulence generation term G is written as:
The effective viscosity µeff is calculated from
The five emperical model constant are assigned
5. BOUNDARY CONDITIONS
There are four types of boundary conditions to specify for the computation in annular exhaust diffuser, they
are Inlet, Outlet, Wall and Symmetry boundary conditions.
discretization. The governing steady-state equations for mass and momentum conservation are solved with a segregated
approach. In this approach, the equations are sequentially solved with implicit linearization
interpolation scheme is used in the analysis in order to get the accurate results. The pressure
implemented using iterative correction procedure (SIMPLEC algorithm).
Inlet: The present analysis involves the velocity as the inlet boundary condition and is varied from 80m/s to 160m/s in
the steps of 40m/s. Turbulence intensity is specified as Medium Turbulence i.e 5% with respect to the equivalent flow
diameter.
Outlet: Atmospheric pressure condition is applie
is set to Atmospheric.
Wall: The no slip condition and smooth surface conditions are used for all walls (except side walls).
Symmetry: Symmetry boundary conditions are applied to
considered (1/6)th part of geometry for the analysis.
In the present analysis numerical simulations are carried out by choosing k
target of 1×10-6 is set to ensure the convergence of the equations.
6. FLUID PROPERTIES
The material which is used for the analysis is Air at 25
tabulated in Table. 1.
Table.1:
Thermodynamic
Thermal Expansion coefficient
Dynamic Viscosity
Thermal Conductivity
Density
International Conference on Current Trends in Engineering and Management ICCTEM
17 – 19, July 2014, Mysore, Karnataka, India
225
The turbulence generation term G is written as:
is calculated from
The five emperical model constant are assigned the following values:
There are four types of boundary conditions to specify for the computation in annular exhaust diffuser, they
d Symmetry boundary conditions. Fluent code uses the finite volume method for
state equations for mass and momentum conservation are solved with a segregated
approach. In this approach, the equations are sequentially solved with implicit linearization. A second
interpolation scheme is used in the analysis in order to get the accurate results. The pressure-velocity coupling is
implemented using iterative correction procedure (SIMPLEC algorithm).
elocity as the inlet boundary condition and is varied from 80m/s to 160m/s in
the steps of 40m/s. Turbulence intensity is specified as Medium Turbulence i.e 5% with respect to the equivalent flow
Outlet: Atmospheric pressure condition is applied at the outlet boundary where in the pressure at the exit of the diffuser
Wall: The no slip condition and smooth surface conditions are used for all walls (except side walls).
Symmetry: Symmetry boundary conditions are applied to the model at the either walls of the diffuser, because we
considered (1/6)th part of geometry for the analysis.
In the present analysis numerical simulations are carried out by choosing k-ε turbulence model. The residual
nsure the convergence of the equations.
The material which is used for the analysis is Air at 25oC. The material properties of the Air at 25
Table.1: Properties of the Air at 25oC
Thermodynamic State Gas
Thermal Expansion coefficient 0.003356 [k^-1]
1.831×10-5 [Kg m^-1 s^-1]
Thermal Conductivity 2.61×10-2 [W m^-1 k^-1]
1.185 [Kg m^-3]
International Conference on Current Trends in Engineering and Management ICCTEM -2014
19, July 2014, Mysore, Karnataka, India
There are four types of boundary conditions to specify for the computation in annular exhaust diffuser, they
Fluent code uses the finite volume method for
state equations for mass and momentum conservation are solved with a segregated
. A second- order upwind
velocity coupling is
elocity as the inlet boundary condition and is varied from 80m/s to 160m/s in
the steps of 40m/s. Turbulence intensity is specified as Medium Turbulence i.e 5% with respect to the equivalent flow
d at the outlet boundary where in the pressure at the exit of the diffuser
the model at the either walls of the diffuser, because we
turbulence model. The residual
C. The material properties of the Air at 25oC are
(7)
(8)
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM
7. CODE VALIDATION
The problem is solved using Fluent CFD code, the existing code is validated with the results of Vaddin Chethan
[11] for accuracy and correctness. For the modeling of annular diffuser, the values of diffusion length, casing angle, hub
length and boundary conditions are retained same as Vaddin Chethan [11] and validation is carried for velocity of 80 m/s,
It is found to agree well with the results of published works.
8. RESULTS AND DISCUSSION
This paper presents the performance of gas tur
measured in the scaled down model. In the graphs and contour plots, it is referred that the position of measuring point in
terms of axial position, considering that:
� Axial position of 0 mm is the leading edge of inlet of diffuser.
� Results are produced in the diffuser model with and without struts for various section from inlet to outlet.
The diffusers performances have been also determined by the following parameters:
• Pressure Recovery coefficient(Cp)
An ideal pressure recovery can be defined if the flow is assumed to be isentropic. Then, by employing the
conservation of mass, this relation can be converted to an area ratio for incompressible flow.
Where,
Vav1 represents the average velocity at the inlet.
P is the pressure at the point at which pressure coefficient is being evaluated.
is the pressure in the free stream.
A is the cross sectional area, and
R is the universal gas constant.
8.1 Analysis of the annular diffuser for 13
The pressure recovery coefficient and velocity distribution for with and without struts along the length of the
diffuser are shown in Fig. 5 (a, c, e) and Fig. 5 (b, d, f) respe
diffuser, the velocity continuously decreases towards the outlet because of the diffusion. From the Fig. 5(a), we can
notice that the pressure recovery coefficient is slightly reduced in the c
struts. This is because, as the fluid comes in contact with the strut interface, the pressure recovery coefficient decreases
International Conference on Current Trends in Engineering and Management ICCTEM
17 – 19, July 2014, Mysore, Karnataka, India
226
solved using Fluent CFD code, the existing code is validated with the results of Vaddin Chethan
[11] for accuracy and correctness. For the modeling of annular diffuser, the values of diffusion length, casing angle, hub
tained same as Vaddin Chethan [11] and validation is carried for velocity of 80 m/s,
It is found to agree well with the results of published works.
Fig 4: Velocity v/s X/L
This paper presents the performance of gas turbine annular diffuser in terms of pressure recovery coefficient
measured in the scaled down model. In the graphs and contour plots, it is referred that the position of measuring point in
the leading edge of inlet of diffuser.
Results are produced in the diffuser model with and without struts for various section from inlet to outlet.
The diffusers performances have been also determined by the following parameters:
)
An ideal pressure recovery can be defined if the flow is assumed to be isentropic. Then, by employing the
conservation of mass, this relation can be converted to an area ratio for incompressible flow.
resents the average velocity at the inlet.
P is the pressure at the point at which pressure coefficient is being evaluated.
Analysis of the annular diffuser for 130 casing angle with and without struts: The pressure recovery coefficient and velocity distribution for with and without struts along the length of the
diffuser are shown in Fig. 5 (a, c, e) and Fig. 5 (b, d, f) respectively, at different values of velocity. From the inlet of the
diffuser, the velocity continuously decreases towards the outlet because of the diffusion. From the Fig. 5(a), we can
notice that the pressure recovery coefficient is slightly reduced in the case of struts compared to the diffuser without
struts. This is because, as the fluid comes in contact with the strut interface, the pressure recovery coefficient decreases
International Conference on Current Trends in Engineering and Management ICCTEM -2014
19, July 2014, Mysore, Karnataka, India
solved using Fluent CFD code, the existing code is validated with the results of Vaddin Chethan
[11] for accuracy and correctness. For the modeling of annular diffuser, the values of diffusion length, casing angle, hub
tained same as Vaddin Chethan [11] and validation is carried for velocity of 80 m/s,
bine annular diffuser in terms of pressure recovery coefficient
measured in the scaled down model. In the graphs and contour plots, it is referred that the position of measuring point in
Results are produced in the diffuser model with and without struts for various section from inlet to outlet.
An ideal pressure recovery can be defined if the flow is assumed to be isentropic. Then, by employing the
The pressure recovery coefficient and velocity distribution for with and without struts along the length of the
ctively, at different values of velocity. From the inlet of the
diffuser, the velocity continuously decreases towards the outlet because of the diffusion. From the Fig. 5(a), we can
ase of struts compared to the diffuser without
struts. This is because, as the fluid comes in contact with the strut interface, the pressure recovery coefficient decreases
(9)
(10)
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM -2014
17 – 19, July 2014, Mysore, Karnataka, India
227
till the length of the strut. However, in the case of velocity distribution, the velocity increases with decrease in pressure
along the strut portion. This scenario is shown in the zoomed portion of the Figures. 5(a) to 5(f).
The above explained scenario is applicable for the remaining figures with different velocities.
Fig. 5: (a & b) Variation of pressure recovery coefficient & velocityfor the diffuser with and without struts at V=80m/s
Fig. 5(c & d) Variation of pressure recovery coefficient & velocityfor the diffuser with and without struts at V=120m/s
Fig. 5: (e & f) Variation of pressure recovery coefficient & velocityfor the diffuser with and without struts at V=160m/s
Fig. 5(a) Fig. 5 (b)
Fig. 5(c) Fig. 5(d)
Fig. 5(e) Fig. 5(f)
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM -2014
17 – 19, July 2014, Mysore, Karnataka, India
228
Fig. 6, Fig. 7 and Fig. 8 shows the contour plots for velocity and pressure distribution for both with and without
struts for different velocity cases. In these figures fluid characteristics like velocity, pressure are shown by different
color. Due to the change of kinetic energy into pressure energy there is continuous increase in the magnitude of pressure
from inlet to outlet, and also due to the change in cross sectional area, the velocity increases along the length of
guidevanes and struts, and goes on decreases till the outlet of the diffuser.
Fig. 6(a): Velocity contour for 130 casing angle with strut and velocity at 80m/s
Fig. 6(b): Pressure contour for 13
0 casing angle with strut and velocity at 80m/sec
Fig. 6(c): Velocity contour for 13
0 casing angle without strut and velocity at 80m/s
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM -2014
17 – 19, July 2014, Mysore, Karnataka, India
229
Fig. 6(d): Pressure contour for 130 casing angle without strut and velocity at 80m/sec
Fig. 7(a): Velocity contour for 13
0 casing angle with strut and velocity at 120m/s
Fig. 7(b): Pressure contour for 13
0 casing angle with strut and velocity at 120m/sec
Fig. 7(c): Velocity contour for 13
0 casing angle without strut and velocity at 120m/s
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM -2014
17 – 19, July 2014, Mysore, Karnataka, India
230
Fig. 7(d): Pressure contour for 13
0 casing angle without strut and velocity at 120m/sec
Fig. 8(a): Velocity contour for 13
0 casing angle with strut and velocity at 160m/s.
Fig. 8(b): Pressure contour for 13
0 casing angle with strut and velocity at 160m/sec
Fig. 8(c): Velocity contour for 130 casing angle without strut and velocity at 160m/s
Proceedings of the 2nd
International Conference on Current Trends in Engineering and Management ICCTEM -2014
17 – 19, July 2014, Mysore, Karnataka, India
231
Fig. 8(d): Pressure contour for 130 casing angle without strut and velocity at 160m/sec
9. CONCLUSION
Numerical investigations have been carried out to investigate the static pressure development and pressure
recovery coefficient through an industrial gas turbine exhaust diffuser with and without struts.
From the above discussion the following conclusions can be made:
1. The pressure recovery within the diffuser increases as the flow proceeds, consequently the pressure also increases
with the decrease in velocity level.
2. With increase in area ratio pressure recovery increases due to higher rate of diffusion but pressure recovery loss
also increases.
3. With increases in inlet velocity, there is increase in pressure recovery, since the entrance losses increases marginal
with velocity.
4. Comparing the two cases, with and without the struts, it is clear that the overall diffuser loss is significantly
increased by the struts and this loss rise mainly occurs in the axial region of the struts and in the endwall regions,
where flow separates from the hub and the casing.
REFERENCES
[1]. Ackert J, "Aspect of Internal Flow", Fluid Mechanics of Internal Flow, Ed. Sovran G., Elsevier Amsterdam,
pp.1, 1967.
[2]. Sovran. G and Klomp E. D., "Experimentally determined optimum geometries for rectilinear diffusers with
rectangular, conical or annular cross-section", Fluid Mechanics of Internal Flow (Ed. G. Sovran) (Elsevier,
Amsterdam) pp. 270–312, 1967.
[3]. R.K. Singh, R.S. Azada,"Measurements of instantaneous flow reversals and velocity field in a conical diffuser",
Experimental Thermal and Fluid science, Vol. 17, pp. 100- 106, 1995.
[4]. P.W. Runstadler, F.X. Dolan, R.C. Dean, "Diffuser data book", Creare, TB-186, 1975.
[5]. U. Desideri, G. Manfrida, "Flow and turbulence survey for a model of gas turbine exhaust diffuser", ASME
paper 95-GT- 139, 1995.
[6]. B. Djebedjian, J. Renaudeaux, "Numerical and experimental investigation of the flow in annular diffuser",
Proceedings of FEDSM98, Washington, DC, FEDSM98-4967, 1998.
[7]. H. Pfeil, M. Going, "Measurements of the turbulent boundary layer in the diffuser behind an axial compressor",
ASME Journal of Turbo machinery, Vol. 109, 405-412, 1987.
[8]. H. Harris, I. Pineiro, T. Norris, "A performance evaluation of three splitter diffuser and vaneless diffuser
installed on the power turbine exhaust of a TF40B gas turbine", ASME paper 98-GT-284, 1998.
[9]. S. Ubertini, U. Desideri, "Flow development and turbulence length scales within an annular gas turbine exhaust
diffuser", Experimental Thermal and Fluid Science, Vol. 22, Issue 1-2, pp. 55-70, August 2000.
[10]. Stefano Ubertini, Umbert Desideri, "Experimental performance analysis of an annular diffuser with and without
struts", Experimental Thermal and Fluid Science, vol. 22, pp. 183-195, 2000.
[11]. Vaddin Chetan, D V Satish, Dr. Prakash S Kulkarni, “Numerical Investigations of PGT10 Gas Turbine
Exhaust Diffuser Using Hexahedral Dominant Grid”, International Journal of Engineering and Innovative
Technology (IJEIT) Volume 3, Issue 1, July 2013.