Reduce tip leakage flow using squealer tip in an axial turbine = Giảm xoáy đầu mút cánh turbine dọc trục sử dụng đầu mút lõm436

12 Figure 9: Stator blade geometric parameters and profile pressure distribution 12 Figure 10: Rotor blade geometric parameters and profile pressure distribution 12 Figure 11: Stator and

HANOI UNIVERSITY OF SCIENCE AND TECHNOLOGY

THESIS REDUCE TIP LEAKAGE FLOW USING

SQUEALER TIP IN AN AXIAL

TURBINE

DO DINH CHINH

ID: 20202603M CLASS: 20BKTHK

Department: Department of Aerospace Engineering

Faculty: School of Transportation Engineering

Hanoi, 05/2021

Advisor’s sign

C NG HÒA XÃ H I CH Ộ Ộ Ủ NGHĨA VIỆT NAM

Độ ậ – ự c l p T do H nh phúc – ạ

B N XÁC NH N CH NH S A LU Ả Ậ Ỉ Ử ẬN VĂN THẠC SĨ

H và tên tác gi ọ ả luận văn: Đỗ Đình Chinh

Đề tài lu ận văn: Giảm xoáy đầu mút cánh turbin d c tr c s dọ ụ ử ụng đầu mút lõm

Chuyên ngành: K thuỹ ật Hàng không

Mã số SV: 20202603M

Tác giả, Người hướng d n khoa h c và Hẫ ọ ội đồng ch m luấ ận văn xác nh n tác gi ậ ả đã sửa ch a, b sung luữ ổ ận văn theo biên bản h p Họ ội đồng ngày 06/09/2021 v i các n i dung sau: ớ ộ

1 Chỉnh s a hình th c, l i ch b n (lử ứ ỗ ế ả ỗi đánh máy, ạt i ph n m c lầ ụ ục

thêm ph n kầ ết luận và tài li u tham kh o) ệ ả

2 B sung danh m c t ổ ụ ừviết tắt, kí hiệu

3 Đánh số các phương trình và dẫn nguồn các phương trình

I assure that this thesis was my independent research under the instructions

of my advisor PhD Dinh Cong Truong This research is not a copy from any previous research paper

Hanoi, 15 July 2021

Do Dinh Chinh

this time I am grateful to my parents who taught me to cherish excellence Without their support, this work would have not been completed This is the first time I did a project in this field of study it is inevitable that there are some soshortcomings Finally, I would like to thank the commitee for take the time in reading this research work and I look forward to receiving the comments and corrections to complete this study

Student

Do Dinh Chinh

TABLE OF CONTENTS

CHAPTER 1 INTRODUCTION 1

1.1 Introduction 1

1.2 Previous research 2

1.3 Tip clearance of rotor blade 4

1.4 Tip leakage flow 5

1.5 Flat tip and squealer tip 8

CHAPTER 2 NUMERICAL ANALYSIS 10

2.1 Turbine model Error! Bookmark not defined. Turbine “LISA” 10

2.1.1 2.1.2 Stator and rotor blade geometry 11

2.2 Numerical method 16

2.2.1 Turbine performance curves 16

2.2.2 The fundamental equations of fluid dynamics 17

2.2.3 Simulation procedure 21

2.3 Meshing 22

2.4 Boundary conditions 25

2.5 Convergence criteria 27

CHAPTER 3 RESULTS AND DISCUSSIONS 28

3.1 Grid dependency test and validation 28

3.2 Pressure, velocity and temperature contours 32

3.3 Effect of tip clearance 36

3.4 Reduce tip leakage flow using squealer tip 38

3.5 Effects of the squealer tip on aerothermal performance ……… 38

CONCLUSION AND FUTURE WORK

REFERENCES

TABLE OF FIGURES

Figure 1 Location of turbine in aircraft engine 1

Figure 2 High pressure shrouded (left) and unsrhouded (right) turbine rotor blades 5

Figure 3 Illustration of tip leakage flow over a flat tip 6

Figure 4 Outline of the flow in the region of an unshrouded turbine rotor blade 6 Figure 5: Rotor tip with flat and squealer show clearly the cavity squealer tip with a small figure at the tip region 8

Figure 6: Schematic view of LEC’s LISA research axial turbine 10

Figure 7: Sketch of the turbine first stage with the relevant dimensions 11

Figure 8: Rotor blade (left) and stator blade (right) 12

Figure 9: Stator blade geometric parameters and profile pressure distribution 12 Figure 10: Rotor blade geometric parameters and profile pressure distribution 12 Figure 11: Stator and rotor blade design in Ansys Design Modeler 14

Figure 12: Conceptual view of rotor blade without squealer tip (WST) and with cavity squealer tip (CST) 15

Figure 13: Rotor blade with cavity squealer tip and fillet radius at the hub 15

Figure 14: 3D mesh of the stator blade 22

Figure 15: 3D mesh of rotor blade 23

Figure 16: Diffuser computational domain 23

Figure 17: The computational domain of turbine with WST 24

Figure 18: 3D mesh of the computational domain with CST 25

Figure 19: Complete computational domain when mirrored around the rotational axis 25

Figure 20: Mesh dependency test results 29

Figure 21: contours on stator and rotor blade for Mesh 1 to Mesh 4 29

Figure 22: Measured point and computed total pressure ratio compare two interface cases and Blanco’s computed results 30

Figure 23: Measured point and computed adiabatic efficiency compare two interface cases 32

Figure 24: Static pressure contour of the stator and rotor at the mid-span plane 33

Figure 25: Relative Mach number of the stator and rotor at the mid-span plane 33

Figure 26: Total pressure and Static entropy at the outlet plane of the stator 34

Figure 27: Total pressure and Static entropy at the outlet plane of the rotor 34

Figure 28: Temperature contour on the surface of stator and rotor blades 35

Figure 29: Static pressure contour at the end wall of the rotor 35

Figure 30: Flow visualization of the recirculation bubble over the rotor tip surface 36

Figure 31: Peak efficiency at different rotor blade tip clearance 37

Figure 32: Efficiency at mass flow rate of 11.7 kg/s 37

Figure 33: Squealer tip parameters 38

Figure 34: Pressure and Static entropy contour at the rotor outlet plane 39

Figure 35: Aerothermal performance of LISA turbine with cavity squealer tip 42

Figure 36 Distribution of temperature [K] on the stator blade 43

Figure 37: Pressure [Pa] contours on the shroud casing of rotor blade without squealer and with w/τ = 100% 43

Figure 38: Streamline through the tip clearance of rotor blade with w/τ =100% 45

Figure 39: Static entropy contours on the blade with w/τ = 100% 45

Figure 40: Nu contours on the blade with h/τ= 150% 47

Figure 41: Nu contours on the blade with w/τ = 200% 47

LIST OF TABLES

Table 1 LISA research turbine facility controlling parameters 11

Table 2 Design parameters of the first stage blades 13

Table 3: Design specifications of cavity squealer tip 16

Table 4 Measured operating condition at turbine design 16

Table 5 Thermodynamic properties of the gas used in the CFD analysis 26

Table 6 Boundary conditions in CFD analysis 26

Table 7: Mesh dependency test results 28

Table 8 Pressure ratio compared to Blanco’s results and measured point 31

Table 9 Maximum efficiency and stall point of turbine stage 32

Table 10 Flow angle at the inlet and outlet locations 36

Table 11 Values of tip clearance investigated and computed adiabatic efficiency 37

Table 12 Aerodynamic performance at different cases 39

Table 13: Effect of cavity squealer on aerodynamic and aerothermal performances for LISA turbine 41

1

CHAPTER 1 INTRODUCTION 1.1 Introduction

In the aviation industry, increasing the performance of aircraft is the most

important thing to improve aircraft operating cost and reduce emissions In addition to improvements in aerodynamics and materials of the structure, engine improvement is the top concern of many studies Turbines are always mentioned

2

turbine were evaluated based on efficiency and Nusselt number The turbine studied in this investigation is an axial annular turbine named “LISA”, which was experimentally tested at the Laboratory for Energy Conversion (LEC), Institute

of the ETH Zürich, Switzerland Numerical calculations have been performed using 3-D Reynolds Averaged Navier-Stokes (RANS) equations with the shear stress transport (SST) turbulence model and “total energy” option with “mixing-

plane” option between rotor and stator interfaces The impact on aerothermal

Currently, there are many studies on aerodynamic enhancing methods to limit

the influence of the tip leakage flow One of the most popular methods is to design the squealer tip for the blade In one study, Heyes et al [19] showed that blade tip geometry had a positive effect on the aerodynamic performance of axial turbine cascades by limiting the undesirable effects of the tip leakage flow In terms of thermodynamics, Ameri et al [20] showed that a squealer tip directly slowed down the leakage flow, while also increasing the total heat transfer

numerical simulation was performed by Kavurmacioglu et al [22] and the authors pointed out a reduction in aerodynamic loss on a suction side squealer when compared to the conventional flat tip Key and Arts [23] compared the flat tip with the cavity squealer tip, which was based on the flow characteristics at

both low and high-speed conditions in a linear cascade They discovered that

squealer tips would lower aerodynamic loss in the case of the flat tip under specified conditions An experiment by Newton et al [24] measured the heat transfer coefficients and the pressure coefficients in a linear cascade with flat tip, suction side squealer and cavity squealer Their results indicated a fall in heat transfer when using the squealers Krishnababu et al [25] investigated the effects

of the blade tip’s geometry and concluded that cavity tip increased the aerodynamic performance and heat transfer Lee and Kim [26] studied the

influences of the tip gap’s height on aerodynamic performance when using a

cavity squealer tip in a linear cascade turbine Schabowski and Hodson [27] investigated the aerodynamic effects of various tip designs in a low-speed linear turbine cascade and found that the cavity squealer tip led to a lower aerodynamic loss

At present, the number of studies on the simultaneous effects of aerodynamics and heat of the tip leakage flow is very limited When Lee and Chae [26] studied the effects of squealer rim height on aerodynamic losses downstream of a high-

turning turbine rotor, they came to the conclusion that by increasing the rim’s height, the aerodynamic loss height reduced until the squealer rim’s height- -tochord ratio reached 2.75% Zhou and Hodson [27] conducted the experimental and numerical works of the squealer geometry’s effect on the aerothermal performance of the tip leakage flow of cavity tips They reported that squealer height affected aerodynamic loss complexly, and the heat transfer coefficient reduced with increasing the height and reducing the width while reducing the width mitigated aerodynamic loss Kang and Lee [28] investigated the effects of

In this study, squealer tip configurations with varied squealer’s width and

height were studied to find the effect of squealer tip on the aerodynamic efficiency, thermodynamic performance, and leakage mass flow rate of the axial turbine in comparison with the case without the squealer tip

1.3 Tip clearance of rotor blade

In normal, the tip clearance is not too big to prevent losses, but it could not

be too small, because the expansion of rotor blade because thermal expansion and inertial force during operation could damage the casing

One practiced method of mitigating the over the tip leakage flow is achieved

by introducing a shroud to the rotor blade In Fig 2 [11], two high pressure turbine rotor blades are depicted: one shrouded and the other free-tip Both rotor blades present orifices from which fluid (air bled-off of the HP compressor) is ejected to create a boundary layer of cooled air (~700 K) that protects the metal (through film cooling) from burning and from the combustion products

5

Figure 2 High pressure shrouded (left) and unsrhouded (right) turbine rotor

bladesFig 2 shows the structure of a shrouded turbine blade (left) and an unshroud turbine blade (right) Even though the shroud over the rotor increases the aerodynamic efficiency of the turbine stage, the added weight at the tip of the blade creates considerable mechanical (mainly centrifugal) stresses at the root of the blade and to the disc itself Therefore, the rotational speed of a shrouded blade will have a lower limit compared to an unshrouded one Since the work output is proportional to square of the blade rotational speed (via Euler turbine equation), an advantage of using unshrouded blades becomes apparent Nonetheless, the shroud damps out the blade vibrations which is an advantage as compared to unshrouded rotor

1.4 Tip leakage flow

6

Figure 3 Illustration of tip leakage flow over a flat tipThe over the tip leakage (OTL) flow has its origins on the static pressure difference that occurs at each side of the rotor airfoil at the tip In this gap the fluid is not deflected by the blade and hence does not contribute to the work output of the stage The fluid enters the gap on the pressure side of the rotor blade and continues to the other side where it mixes with the core flow and rolls

up into a vortex An additional vortex due to the endwall boundary layer of the casing (outer passage vortex) interacts with the over the tip leakage vortex [15] These sequences can be observed from the sketch of Fig 4

Figure 4 Outline of the flow in the region of an unshrouded turbine rotor bladeResearches have been published recently on the tip leakage flows in turbines from theoretical study Study of Rain [1] has found the models of flow through the tip gap of an axial compressor Rain has gigured out the structure of flow on

7

the tip gap surface of a compressor Moore et al has presented the effect of Reynolds number to flow on a tip gap Research showed that there is a large separated flow ay the sharp edge of blade with the high Reynolds number over

10000 Moore et al [2] also calculated turbulence model with a high Reynolds number from 100 to 10,000 [2] Bindon [3] showed the development of vortex along the leading edge to trailing edge of the blade The detailed development of tip clearance loss from the leading to trailing edge of a linear turbine cascade was measured and the contributions made by mixing, internal gap shear flow, and endwall secondary flow were identified, separated, and quantified for the first time Only 13 percent of the overall loss arises from endwall secondary flow and

of the remaining 87 percent, 48 percent is due to mixing and 39 percent is due to internal gap shear All loss formation appears to be dominated by phenomena connected with the gap separation bubble [3] Yamamoto [4] has found that the

clearance gaps size and the cascade incidences were chosen as the most important variables affecting the mechanisms Flows close to the endwall and inside the clearance were surveyed in great detail using a micro five-hole pitot

(Prasad and Wagner [7], Stephan et al [8], Xiao et al [9], McCarter et al [10], and Blanco [11]) Sjolander [12] presented an overview of the tip leakage flow,

of vortex tubes in a flow with an initial normal vorticity distribution results in a streamwise component of vorticity at the exit of the passage [15, 16] One of the

8

most important sources of losses is due to over the tip leakage (OTL) flow in rotor blades since it accounts for over one third (>1/3) of the overall turbine stage losses [17] At the trailing edge of each turbine blade there is a momentum deficit

1.5 Flat tip and squealer tip

Usually in the manufacture of rotor blades, the tip is often made with the

shape flat (Flat tip) Therefore, a number of studies have been conducted to evaluate the effect of the tip with other profiles on turbine performance One of the design improvements that have proven to be effective is the use of a squealer tip By piercing down the tip in a specific profile, the losses are reduced In this article we will use the same profile as the flat tip Fig 5 illustrates the flat p ti(left) and squealer tip (right):

Figure 5: Rotor tip with flat and squealer show clearly the cavity squealer tip

with a small figure at the tip region

In fact the squealer tip method has been studied on compressor blades to increase aerodynamic performance And for turbines, researchers have applied it

to improve the cooling capacity of the blades With that idea, applied to a specific object, in this project, we conduct a study on changing the end ted ip configuration from a flat shape to a squealer shape and evaluate its influence on the vortex in the turbine passage, and also to improve turbine performance in this

9

case We will also evaluate the impact of this method on thermal performance in the later part of the study

10

CHAPTER 2 NUMERICAL ANALYSIS 2.1 Turbine model

2.1.1 Turbine “LISA”

The turbine studied in this investigation is an axial annular turbine named

“LISA” This turbine was tested at the Laboratory for Energy Conversion (LEC)

Figure 6: Schematic view of LEC’s LISA research axial turbine

The air circulates in a quasi-closed loop; an opening to the atmosphere exists

at the exit of the turbine The mass flow rate through the compressor is altered by adjustable inlet guide vanes and is measured by a calibrated venturi nozzle To control the turbine inlet temperature the air passes through a water-cooled heat

exchanger The control of the turbine rotational speed is done with a DC generator to an accuracy of ±0.1 rpm Some characteristics of the LISA turbine are described below [9]:

11

Table 1 LISA research turbine facility controlling parameters

Compressor power 750 kW Turbine speed (max.) 3000 rpm

Compressor mass

flow rate 6 to 13 kg/s Turbine inlet temperature 33 to 55 ⁰C Generator power 400 kW Turbine exit pressure Atmospheric Working fluid Air Turbine tip diameter 800 mm

Therefore the measurement planes positions are defined as follows:

; ;

Where is the axial chord distance of the respective blade rows as

demonstrated in Fig 7:

Figure 7 Sketch of the turbine first stage with the relevant dimensions :

At the position, there are measurement probes to mersure some characteristics needed

2.1.2 Stator and rotor blade geometry

Every blade has a different parameters, based on these parameters we design the turbine using design module ANSYS Design Modeler 19.1 Stator and rotor blade row is shown in Fig 8

Figure 9 Stator blade geometric parameters and profile pressure distribution :

Figure 10 Rotor blade geometric parameters and profile pressure distribution :

Tab 2 presents some design parameters of the rotor and stator blade:

Axial chord

length [mm] 49.61 49.71 49.82 43.41 46.83 50.08 Chord length [mm] 85.37 80.88 76.40 59.68 59.72 60.46 Pitch [mm] 69.81 63.70 57.60 46.54 42.47 38.40

Leading edge

radius [mm] 6.20 7.00 8.00 3.24 4.00 4.30 Trailing edge

thickness [mm] 1.21 1.30 1.38 0.98 1.10 1.28 Throat

diameter [mm] 20.01 18.53 17.07 16.08 14.42 13.25

Fillet radius [mm] 2.00 X 2.00 X X 3.00 Aspect ratio [-] 0.82 0.87 0.92 1.17 1.17 1.16

Chord / pitch

ratio [-] 1.22 1.27 1.33 1.28 1.41 1.57 Zweifel

14

Based on these parameters, the geometry was designed by using ANSYS Design Modeler 19.1 as illustrated in Fig 11

Figure 11 Stator and rotor blade design in Ansys Design Modeler :

Because of the large thickness design at the leading edge of the blades, an acceleration and deceleration occurs in the front suction side as depicted in the profile pressure distribution of Fig 6 and Fig 7 The stator and rotor profiles are both stacked along radial lines that intersect the axis of rotation The stator along its leading edge and the rotor along its center of gravity to avoid bending stresses caused by the centrifugal forces originated during rotation

The difference between this design and the one in the validated paper is the rotor hub fillet and the stator fillet The rotor hub fillet radius is 3 mm and the stator fillet radius is 2 mm These fillets strengthen the structure prevent damage

in operation In addition, these fillets make the simulation much more accurate due to the accurated flow phenomenon This is an improvement that Blanco d idnot have

rotor blades were set at 2 mm and 3 mm, which were used in this work However, these stator and rotor fillet radius values were not considered in the work of Blanco [20] The blades were designed with fillet to reduce structural failure and damage, which was an improvement compared to Blanco’s design as shown in Fig 3 The design specifications of axial turbine and cavity squealer tip are listed in Tab 2 and Tab 3, respectively

Figure 13: Rotor blade with cavity squealer tip and fillet radius at the hub

16

Operating conditions at the turbine design point are given in Tab 3 [9] These values will be used as boundary conditions in the CFD analysis and the results will be compared to the experiment measurement

Table 3: Design specifications of cavity squealer tip

Minimum Maximum Steps

Table 4 Measured operating condition at turbine design

Pressure ratio (1.5 stages, total to static) [-] 1.60

Total inlet pressure [bar- absolute Norm] 1.405

Rotor tip clearance/ blade span ratio [%] 1.0

Stage

Blade row relative exit Mach number (average) [-] Stator 0.54

Rotor 0.50

2.2 Numerical method

2.2.1 Turbine performance curves

In general, the operating characteristics of a turbine, or turbine performance, can be deduced from two important parameters: total pressure ratio, adiabatic efficiency

Total pressure ratio is commonly calculated as follow:

(1)

It is common that when the mass flow rate increases, the pressure ratio decrease Generally, a point is reached at which the pressure raise is maximized and further increase in mass flow may lead to instability

Thermal performance of squealer tips is evaluated calculating average Nusselt number, , is given by:

2.2.2 The fundamental equations of fluid dynamics

The continuity equation:

Based on the conservation of mass, we obtain the continuity equation for the fluid flow [30] :

The energy equation:

Since energy is conserved, we obtain the energy equation [30]:

(8)

The Navier-Stokes equations:

Navier-Stokes equations are the governing equations of Computational Fluid Dynamic It is based on the conservation law of physical properties of fluid The principle of conservational law is the change of properties, for example mass, energy, and momentum, in an object is decided by the input and output [7]

We obtain the Navier-Stockes equations in Cartesian coordinates ]: [30

In fluid dynamics, turbulence or turbulent flow is fluid motion characterized

by chaotic changes in pressure and flow velocity Nowadays turbulent flows may

be computed using several different approaches:

Large eddy simulation (LES)

Direct numerical simulation (DNS)

19

and widely used in calculating fluid flow The RANS equations are described below [30]:

The continuity equation:

geometry and grid generation, setting-up a physical model, solving it and processing the computed data The way geometry and grid are generated, the set problem is computed and the way acquired data is presented is very well known

[8] K-epsilon (k ɛ- ) turbulence model is the most common model used in Computational Fluid Dynamics (CFD) to simulate mean flow characteristics for turbulent flow conditions K-ε model focuses on the mechanisms that affect the

dissipation, It is the variable that determines the scale of the turbulence, ɛ

whereas the first variable, k, determines the energy in the turbulence K-epsilon turbulence model has been shown to be useful for free-shear layer flows with relatively small pressure gradients Similarly, for wall-bounded and internal flows, the model gives good results only in cases where mean pressure gradients are small

20

variable that determines the scale of the turbulence, whereas the first variable, k, determines the energy in the turbulence

The SST k-ω turbulence model [Menter 1993] is a two-equation eddy-viscosity model which has become very popular The shear stress transport (SST) formulation combines the best one of two worlds The use of a k-ω formulation

in the inner parts of the boundary layer makes the model directly usable all the way down to the wall through the viscous sub- y hence the SST k-la er ω model can be used as a Low-Re turbulence model without any extra damping functions Therefore an ideal model should introduce the minimum amount of

complexity into the modeling equations, while capturing the essence of the relevant physics

In order to validate the correctness of the model, the y-plus value was used Y-plus is a non-dimensional distance It is often used to describe how coarse

approximately 300 and 100 A faster flow near the wall will produce higher values of Y-plus, so the grid size near the wall must be reduced The y-plus value

is defined by:

(13)

In which, is friction velocity, is the shear stress in an

distance and is the dynamic viscosity (kg/m s)

In this study, the k-ω SST model was used to calculate the performance of turbine The best value of y-plus to use the SST model is to mesh such that y-plus the smaller the better (near 0) However, to make y-plus smaller, the first grid size close to the wall is very small, resulting a huge increase in the number of

21

2.2.3 Simulation procedure

For the numerical calculation, three-dimensional RANS equations with the SST k-ω model and mixing plane were solved numerically using the commercial CFD code, ANSYS-CFX 19.1 [21] Design-Modeler® was used to design the stator, rotor blades, diffuser, and cavity squealer tip Turbo-Grid® and ICEM were employed to generate the meshes for rotor, stator passages, diffuser domain and cavity squealer tip domain, respectively ANSYS CFX-Pre, CFX-Solver, and

CFX-Post were used to define boundary conditions, solve the governing

equations, and post-process the results, respectively As shown in Fig 4, the entire computational domain is divided into four parts; a stator domain, a rotor domain, a cavity squealer domain and finally an exhaust diffuser domain The stator and annular diffuser domains were stationery parts, rotor and cavity domains were set to rotational with the same speed of 2700 rpm Hexahedral elements were used to mesh the computational domain An O-type grid was used near the blades, H/J/C/L-type grids were used in the other regions of the rotor and stator blocks using Turbo Grid, respectively The grid mesh of diffuser and cavity squealer domains are structured mesh with hexahedral cells by ICEM The working fluid was considered to be air ideal gas, which is the same gas that was used at the LEC LISA test facility The boundary conditions used in this

The number of nodes and elements for stator domain are 519456 nodes and

495897 elements The turbine stator stage row is composed of 36 blades but just one blade with a pitch of 10˚ in circumference is modeled in order to speed up

the convergence to a steady state solution The stator is connected to the downstream rotor domain via an interface of the type “Mixing plane”or “Frozen rotor” depends different cases

Figure 14 3D mesh of the stator blade : The mesh of rotor computational domain was generated using Turbo Grid, illustrated in Fig 15 As with the stator domain, the hexahedral cells were used to mesh the rotor block

diffuser domain via an interface of the type “Mixing plane”or “Frozen rotor” corressponding the interface between the stator, rotor and diffuser

For the diffuser domain, the grid used is of structured type with hexahedral

cells, it was done with the meshing program ANSYS ICEM CFD v19.1 that comes with the ANSYS CFX package It was meshed using a 2D profile and extruded around the axis of rotation to obtain the 3D domain, as shown in Fig 16

Figure 16: Diffuser computational domain