2002 Rolls-Royce plc The information in this document is the property of Rolls-Royce plc and may not...
-
Upload
lamont-hooke -
Category
Documents
-
view
216 -
download
3
Transcript of 2002 Rolls-Royce plc The information in this document is the property of Rolls-Royce plc and may not...
2002 Rolls-Royce plcThe information in this document is the property of Rolls-Royce plc and may not be copied or communicated to a third party, or used for any purpose other than that for which it is supplied without the express written consent of Rolls-Royce plc.
Whilst this information is given in good faith based upon the latest information available to Rolls-Royce plc, no warranty or representation is given concerning such information, which must not be taken as establishing any contractual or other commitment binding upon Rolls-Royce plc or any of its subsidiary or associated companies.
Ref.ppt
Computational Fluid Dynamics
Engineering Simulation
Professor Peter Stow
RR Engineering Fellow – Computational Fluid Dynamics
Head of Aerothermal Methods
Rolls Royce plc
Design Considerations
Component GeometryComponent Geometry
AerodynamicsAerodynamics Heat TransferHeat Transfer
ManufactureManufacture
MaterialMaterial
WeightWeight
Stress LevelStress Level
VibrationVibration
Computer Aided Engineering
AERODYNAMICTHERMODYNAMIC
DESIGN
AERODYNAMICTHERMODYNAMIC
DESIGNCADCAD TEST
TESTMECHANICAL
DESIGNANALYSIS
MECHANICALDESIGN
ANALYSISMANUFACTURE
MANUFACTUREPRODUCT
PRODUCT
CFDMODELS
CFDMODELS
STRESSVIBRATION
MODELS
STRESSVIBRATION
MODELS
MANUFACTURINGPROCESSMODELS
MANUFACTURINGPROCESSMODELS
High Performance Computing Potential Impact
Reduced engineering analysis time & cost– more automated analysis
– more integrated analysis Improved accuracy of engineering analysis
– more accurate numerical model
– more accurate mathematical model of physical processes– achieved at acceptable cost– achieved in acceptable elapsed time
Advances in computers mean that this is becoming possible
CFD Applications
Combustion Two-Phase Combusting flow
external aerodynamics
fuel injector and combustercooling flows
Exhaust afterbody flow nozzle, mixer, jet flow
Turbine multistage aerodynamics
aero design and optimisation
end wall and blade heat transfer
film cooling
rotor shroud leakage
rim seals
unsteady vane/rotor flow
forced response
Engine Systems rotating disc cavity flows brush and labyrinth seals secondary air system losses
Nacelle aerodynamics crosswind effects installed engine/pylon/wing interaction
Fan aerodynamics fan flutter fan/OGV/pylon interaction IGV forced response
Compressor Multistage aerodynamics unsteady rotor/stator flow annulus leakage flow
Swept Fan Aerodynamics
Swept FanConventional Fan
Shock moreoblique, withlower losses.
Leading edgeswept rearwardsFlow
Leading edge
Shock (more radial)
Mn rel
Reduced axialvelocity at tip giveslower Mn and highblade angles
Fan Flutter - Issues
Fan Flutter
5% Margin required (before threats)
3% Stability Margin (after stack-up)
No Flutter post Bird-strike
Intake:- Acoustics- Inlet Distortion- Non-axisymmetry
Fan blade shape: - Aerodynamics- Untwist- Modeshape
Tip leakage & Treatment
Downstream Non-axisymmetric issues(pylon, reverser etc.)
Acoustic Liners
Fan Internal & Disk:- Mistuning- Damping
Downstream Acoustics
L/E Shape & Erosion
Engine-to-engine Variability
Fan Flutter
Aerodynamic Damping
unstable
Unsteady
Fan Flutter - Non-linear vs. Linear Unsteady
-0.1000
0.0000
0.1000
0.2000
0.3000
0.4000
0.5000
0.6000
0.7000
0.8000
-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180
Interblade phase angle
log
decr
emen
t
C113 AU3D
C113 SLiQ
unstable
stable
• Steady Viscous Flow
• Unsteady Linearised Flow
• Vibration Modeshape from SC03
Steady
SC03 Mode
stableSurface Work
Fan-OGV-Pylon Design
StaticPressure
Fan-OGV-Pylon Design
Pylon field transmitted through vanes
-0.1
-0.075
-0.05
-0.025
0
0.025
0.05
0.075
0.1
-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180
Circumferential angle
Pylo
n ge
nera
ted
pertu
rbat
ion
uniform (datum) vanes
nonuniform vanes (5 types)
top dead centrepart overcamber
full-overcamber
datum
part undercamber
full undercamber
full overcamber
datum
full undercamber
part overcamber
part undercamber
Circumferential variation in vane type
Fan-BOGV-Pylon Interaction
Fan-BOGV-ESS Interaction
Body force
Body force
Determ
inistic stress Determ
inistic stress
Inlet boundary conditions
Inlet boundary conditions
Exit pressure
Exit pressure
PV
M m
essages
PV
M m
essages
MULTISTAGE ANALYSIS
Unsteady Rotor-Stator Interaction
Contours of Vorticity
3D Rotor-Stator Interaction
Turbine Forced Response
Turbine Forced Response
Wake Shaping
Original Wake P0
Shaped Wake P0
Engine Intake Analysis
Installed Nacelle Ground Plane Effects
25 Knot Forward Speed 35 Knot CrosswindNon-Dimensional Mass Flow = 2307 (full scale)
Freestream P = 380kPa T = 298K Driving P = 292kPa
25 Knot Forward Speed 35 Knot CrosswindNon-Dimensional Mass Flow = 2312 (full scale)
Freestream P = 380kPa T = 298K Driving P = 294kPa
Horizontal Cross SectionWindward Lip
Horizontal Cross SectionWindward Lip
25 Knot Forward Speed 35 Knot Crosswind Windward Lip Separation
25 Knot Forward Speed 35 Knot CrosswindNon-Dimensional Mass Flow = 2307 (full scale)
Freestream P = 380kPa T = 298K Driving P = 292kPa
25 Knot Forward Speed 35 Knot CrosswindNon-Dimensional Mass Flow = 2312 (full scale)
Freestream P = 380kPa T = 298K Driving P = 294kPa
Fan Face Plane Fan Face Plane
25 Knot Forward Speed 35 Knot Crosswind Windward Lip Separation
Engine Intake Analysis
Installed Nacelle Analysis
1,930,378 Tetrahedral Cells316,303 Points91,316 Boundary Faces
Inviscid Flow Solution:
Pictures: Visual3
Solution on Chordwise Cut through Inboard Nacelle
Flow Solution on Aircraft Surface - Fuselage, Wing, Nacelles and Pylons - Half Model
Installed Nacelle Analysis
Traditional 3D Design by Analysis
1st Design
2nd Design
nth Design
Original Design
AnalysePrepare next 3D CFD runSubmit o/night runWait for results
Continue sequence
AnalysePrepare next CFD runSubmit o/night runWait for results
2-D design toolsCloningstacking
FAITH Design System
Seven Perturbations
Automatic creation of perturbations and CFD submission for overnight runs
No Communications No Communications No CommunicationsNo Communications
FAITH - 3D Forward Linear Design
2nd Re-design(= nth design incurrent process)
OriginalDesign
Check linear design withovernight CFD run
Linear Design
Linear Design
Linear Design
Interactive day-time analysis
45
55
65
75
85
95
105
0 25 50 75 100 125 150 175
Axial Distance (mm)
Sta
tic P
ress
ure
/kp
a
FAITH - 3D Inverse Linear Design
2nd Re-design(= nth design incurrent process)
OriginalDesign
Check linear design withovernight CFD run
Original
Linear Design
User defined target flow field
Interactive day-time analysisNo need to iterate
Target
Non-Axisymmetric Endwalls
Application Areas(Whole Engine Modelling)
Aerodynamic Requirements:Performance
Efficiency (SFC)Surge Margin
Noise
Mechanical Requirements:Maximum Stress
F.O.D.
Manufacturing Processesand Cost
Weight
Aeroelasticity:Force Response
Flutter
Shape Optimisation Life
Heat-transferCooling
Rolls-Royce SOPHY Design System
PADRAM-HYDRA-SOFT Aerodynamic Design System Advance Parametric design system – R-R business application focus
– Turbomachinery blading
– Fan – BOGV - pylon
– Intakes
– By-pass Exhaust nozzle & afterbody
– Fan tone noise, Fan-BOGV noise
– Water-jet pump Direct links to CAD system(s) - Parametric representation Rapid automatic meshing – structured, unstructured, mixed grids State of the art versatile Optimisation system for design capability
Integrated Automatic Design Optimisation System
SOPHY: SOFT-PADRAM-HYDRA
Base desig
n
SOFT
HYDRA
PADRAM
Costconstraint
s
jm56
jm52
Design Review
OK?
New design
Yes
No
Optimum design
Convergence history
Design parameters
Optimizer
Additional design
parameters/cost
Automation Flowchart Geometry created & meshed parametrically CFD boundary conditions and mesh pre-processed in batch Costs and constraints extracted in batchLibrary of optimisers available
Current Design System
PADRAMHYDRASOFT
LPC System OGV pre-diffuser
Casingtreatment
Nacelle/Intake
Exhaustnozzle
Water pump
Nozzle design
Multi-stage design
Non axi- endwalls
Forced response
Nacelle Automation and Optimisation
SOFT GEMO
HYDRAJL09-PAX
JM52
Design Parameters
cost and constraint functions
mesh(RAMIN)RAMIN(Mesh)
Nacelle/Intake Design Space
The Design parameters – 77 parameters:• Intake Scarf Angle• Intake Centreline Angle (relative to engine centreline)• ……..
Secondary geometric parameters :• Internal Lip Tilt (angle between intake C/L and local lip
axis topline, sideline and bottom line)• Non-planer Highlight
scarf angle: +10 deg
+10°
2200200
RAMIN: Rapid Meshing for Intakes/Nozzles
RAMIN Mesh on Symmetry Plane
HYDRA solution
New Design Capability
PADRAMHYDRASOFT
Low noise fan and OGV design
Automatic design of
Casingtreatment
Parametric Exhaust Nozzle
System
General Endwall design capability
Parametric Collector box
Internal cooling
passages
Rotorperformance
StatorPerformance Pylon
Inletdistortion
Noise:
ToneBroadbandBuzzsaw
Under platform Cavity Well
design
High Performance Computing Potential - New Simulations
Cheaper/more affordable computing offers vast prospects for new
simulations Multistage Analysis - Steady/Unsteady CFD
– Component optimisation over whole operating range
– Whole engine optimisation Aeroelasticity - CFD <-> Stress/Vibration analysis
– Component optimisation for performance & structural integrity Noise Analysis - Unsteady CFD
– Design optimisation for Noise as well Performance & Stability Turbine Heat Transfer -CFD<->Heat Conduction<->Stress Analysis
– Performance & life optimisation Multi-disciplinary Optimisation
– CFD<->Structural Analysis<->Manufacturing Turbulence & Transition modelling
– Direct/Large Eddy Simulation
Turbulence & TransitionDNS & LES
Tone Noise Sources and Propagation
Distortion NoiseFull 3D Non-linearUnsteady
Buzz-SawFull AnnulusNon-Linear
Rotor AloneNon-LinearSteady
Fan/OGVOGV Geometry Fan/OGV Ratio
Bypass3D Bypass Liner/Geometry Optimisation
Radiation and Transmission Thru’ Shear Layer
RadiationIntake LinersIntake Geometry
LP TurbineMulti-stageUnsteady
Broadband Noise Sources
Rotor BBSelf-Noise Interaction with Inlet B.L.
Fan Wake/OGV Interaction OGV Self Noise
Jet Noise
Fan-OGV BB Sources due to turbulence interacting with (blade) surfaces
Jet Noise due to Turbulence (and shocks)
Low Tone Noise Fan Blade Design
HYDRA, PADRAM and SOFT used to demonstrate low tone noise fan blade design optimisation
~ 9dB reduction
Cost Function (Pa)
Design space covered axial and circumferential movement (lean & sweep) of blade sections over outer 20% of blade span (4 design parameters)
Design Iterations
datum cost
Each iteration around 2 hours on PC cluster initial optimum achieved in ~ 2 days
~ 75% span ~ 75% span
~ 95% span ~ 95% span
Low Tone Noise Fan Blade Design
Datum fan blade Low tone noise blade
Design optimisation introduces forward sweep of blade sections over outer 20% of span - leads to swallowed shock at tip compared to expelled shock of datum blade
contours of static pressurecontours of static pressure
Hydra QTD Buzz-Saw Analysis
26 blade full annulus Measured Static Stagger Angle Variation
Hydra CFD mesh ~55M nodesRun time per aerodynamic point (96 cluster CPUs) ~ 10 days
Viscous end wallsRotor tip gap includedAcoustic Liner included
Acoustic Liner
Broadband Jet Noise
Plane Jet DNS Far-field Noise
by Acoustic Analogy
Exhaust Nozzle LES
2002 Rolls-Royce plcThe information in this document is the property of Rolls-Royce plc and may not be copied or communicated to a third party, or used for any purpose other than that for which it is supplied without the express written consent of Rolls-Royce plc.
Whilst this information is given in good faith based upon the latest information available to Rolls-Royce plc, no warranty or representation is given concerning such information, which must not be taken as establishing any contractual or other commitment binding upon Rolls-Royce plc or any of its subsidiary or associated companies.
Ref.ppt
Computational Fluid Dynamics
Engineering Simulation
Professor Peter Stow
RR Engineering Fellow – Computational Fluid Dynamics
Head of Aerothermal Methods
Rolls Royce plc