FuelInjectionDynamicsandCompositionEffectsonRDEPerformance
PI:MirkoGamba
Co-I:Venkat RamanDepartmentofAerospaceEngineering
UniversityofMichigan
KickoffmeetingOctober26,2017UTSR/NETL
Summary• Title:
– FuelInjectionDynamicsandCompositionEffectsonRDEPerformance• Fundingagency:
– UniversityTurbineSystemsResearch/NETL– FundingOpportunityNumber:DE-FOA-000117– TopicArea6:PressureGainCombustionR&D– Projectmanager:RobinAmes– Award:DE-FE0031228
• Personnel:– PI:MirkoGamba,UniversityofMichigan– Co-I:Venkat Raman,UniversityofMichigan– Studentscurrentlyinvolved:• FabianChacon• JamesDuvall• TakumaSato• Supraj Prakash
– Keyexternalcollaborators:• Dr.JohnHoke,InnovativeScientificSolution,Inc.(ISSI)• Dr.FredSchauer,AFRLWP• Dr.Venkat Tangirala ,GE• Dr.WilliamHargus,AFRL/Edwards• Drs.AdamHolleyandPeterCocks,UnitedTechnologyResearchCenter(UTRC)• KyleMcDevitt,WilliamsInternational 2
Outline
• Programmaticoverview
• Introductiontotheproblemandgeneralapproach
• Experimentalactivities
• Computationalactivities
• Interactionsandcollaborations
3
Outline
• Programmaticoverview
• Introductiontotheproblemandgeneralapproach
• Experimentalactivities
• Computationalactivities
• Interactionsandcollaborations
4
Overarchingobjectives
• Objective1:Developacomprehensiveunderstandingofinjectordynamics,couplingwithdiffuserback-reflections,andtheirimpactonRDEmixing,operationandperformance.
• Objective2:Developacomprehensiveunderstandingofmulti-componentfuels(syngasandhydrocarbonblends)onRDEdetonationstructureandpropagation,operationandperformance.
• Objective3:DevelopadvanceddiagnosticsandpredictivecomputationalmodelsforstudyingdetonationpropagationinRDEs,witharbitraryfuelcompositionandflowconfiguration.
5
Expectedoutcomes:RDEphysicsadvancements
• Outcome1:Fundamentalphysicalunderstandingofdetonation-induceddynamicsofpracticalinjectionsystems:– Couplingwithdiffusergeometry;– TheireffectonRDEmixing,detonationstructure,operabilityandperformance.
• Outcome2:Fundamentalphysicalunderstandingofmulticomponentfuel(MCF)chemistryeffects:–Detonationwavestructureandpropagation– RDEoperabilityandperformance;
6
Expectedoutcomes:RDEmethodsadvancements
• Outcome3:DetailedexperimentalmeasurementsundervariousinjectionschemesandMCFs– RDEperformanceanddetonationstructure• Creationofexperimentaldatabases
–MCFsrepresentativeofsyngasandnaturalgascharacterizedbyarangeofdetonability properties,ignitiondelaysandWobbe index• RelevancetoDOEandindustryprograms
• Outcome4:ComputationallyefficientdetonationmodelsforMCFsforusewithDNS/LESmodelingofdetonationsinrelevantfull-systemRDEsunderpracticalMCFs– Implementationintoopen-sourceplatforms;e.g.,openFoam– Transferofdetonationcomputationalmodelstoindustry
7
ADVANCINGPHYSICALUNDERSTANDINGOF:MIXING
OPERABILITYPERFORMANCE
Objectivesandtasks
8
FuelinjectiondynamicsandcompositioneffectsonRDEperformance
Objective1Developacomprehensiveunderstanding
ofinjectordynamics,couplingwithdiffuserback-reflections,andtheirimpact
onRDEmixing,operationandperformance
Task 2.2Investigatehowlowlossinjector
dynamicsaffectsdetonationmixingandstructureinopticalRDEusinglaser
diagnostics
Task 2.1Experimentalstudyoflowlossinjector
dynamics,theircouplingwithdetonationandexhaustdiffuserdynamics,andtheir
effects onRDEoperabilityandperformance
Objective2DevelopacomprehensiveunderstandingofMCF(syngas andhydrocarbonblends)
onRDEdetonationstructureandpropagation,operationandperformance
Task 3.2DNSstudiesofdetonationstructurein
multicomponentfuels
Task 3.1ExperimentalstudyofRDEdetonationpropertiesandperformancewithhydrocarbonMCFsinopticalRDE
Objective3Developadvanceddiagnosticsandpredictivecomputationalmodelsfor
studyingdetonationpropagationinRDEs,witharbitraryfuelcompositionandflow
configuration
Task 4.2Developmentofstructure-basedmodelofdetonationwavesinstratifiedmixtureswitharbitrarymulticomponentfuels
Task 4.1DevelopmentoflaserdiagnosticsforthestudyofdetonationwavesinopticalRDE
underrelevantconditions
Task 2.3Full-systemsimulationofcoupled
dynamicsofair/fuelinjector,detonationandexhaustdiffuserandtheireffecton
RDEoperation
Task 3.3ConductLESanalysisofRDE
configurationstounderstandtheeffectofmulticomponentfuelsonperformance
RDEINJECTORDYNAMICS
MULTICOMPONENTFUELS
METHODSDEVELOPMENT
Task Description Start Finish 2017
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
1.0 Project meeting and planning 10/17 9/20
1.1 Project meetings and progress reports 10/17 9/20
2.0 Study of the effects of low loss injector dynamics on detonation mixing and structure, and RDE operability and performance
10/17 9/20
2.1 Experimental study of low loss injector dynamics, their coupling with detonation and exhaust diffuser dynamics, and their effects on RDE operability and performance
10/17 3/19
2.2 Investigate how low loss injector dynamics affects detonation mixing and structure in optical RDE using laser diagnostics
4/18 3/20
2.3 Full-system simulation of coupled dynamics of air/fuel injector, detonation and exhaust diffuser and their effect on RDE operation
10/18 9/20
3.0 Study of the effects of relevant multicomponent hydrocarbon fuels on detonation structure, RDE operability and performance
4/18 9/20
3.1 Experimental study of RDE detonation properties and performance with hydrocarbon MCFs in optical RDE
10/18 9/20
3.2 DNS studies of detonation structure in multicomponent fuels 4/18 9/19
3.3 Conduct LES analysis of RDE configurations to understand the effect of multicomponent fuels on performance
10/18 9/20
4.0 Develop advanced diagnostics and structure-based detonation models for multicomponent fuels
10/17 9/20
4.1 Development of advanced laser diagnostics for the study of detonation waves in optical RDE under relevant conditions
10/17 9/19
4.3 Development of structure-based model of detonation waves in stratified mixtures with arbitrary multicomponent fuels
10/17 3/19
2018 2019 2020
Timelineoftheproject
9
Task1
Task2
Task3
Task4
Outline
• Programmaticoverview
• Introductiontotheproblemandgeneralapproach
• Experimentalactivities
• Computationalactivities
• Interactionsandcollaborations
10
OverviewofRDEoperationandPressureGain(PG)
11
2 M. Hishida et al.
Fig. 1 A rotating detonation engine and its two-dimensional CFDmodel
primary purpose of the present study is (1) to run a stablypropagating detonation, (2) to describe interesting physicalphenomena and (3) to discuss whether the rotating detona-tion engine is realistic in application to thrusters or powergenerators.
Assuming that the distance between the two coaxial cylin-ders is much smaller than their diameters and that the dia-meter is not too small to create a strong centrifugal force tothe flow, the flowfield can be approximated as a plane two-dimensional area 200 mm × 200 mm, as shown in Fig. 1b.Except that the upstream and downstream conditions areidentical to the cylindrical flow, the upper and lower sidesare connected by periodic boundary conditions.
2.1 Boundary conditions
Modeling of the boundary conditions is one of the mostimportant issues in the present analysis. In view of the existing
experiments, the boundary conditions are set to be as simpleas possible to make the analysis easy, without losing generalcharacters and holding consistency with Euler Equations anda part of experimental conditions.
(a) The inflow condition is set up as follows: At t = 0,the reservoir pressure and temperature are adjusted togive a specified Mach number in the fictitious uniformregion (the pressure 0.2 MPa, the temperature 300 K)in the quiescent toroidal area, after choking and sub-sequent isentropic expansion prevail. For this adjustedreservoir condition, at t>0 after a real rotating detona-tion has started running, the injection velocity is decidedat each azimuthal location by its local pressure, depen-ding upon where the place is along the azimuthal axis.(i) Immediately behind the detonation front, the pres-sure is usually higher than the reservoir pressure; insuch circumstances the injection velocity is assumedzero. (ii) Far from the detonation front, the pressurecould be lower than the choking value; then the injectionvelocity is given by the choking condition. (iii) In theintermediate region where the pressure is between thereservoir and choking ones, the injection velocity is cal-culated by the isentropic expansion from the reservoircondition.
(b) The reservoir condition: The reservoir condition is assu-med unaffected by the downstream condition, stayingconstant throughout the analysis. In reality a reverseflow into the reservoir can happen when the flowfieldpressure exceeds the reservoir one, which certainlynecessitates taking account of modeling a large size ofreservoir area. In the present modeling, however, suchphenomenon is not included to simplify the analysis.
(c) The headwall temperature: Due to constant exposure toa hot burnt gas behind the detonation front, the headwalltemperature can be high, supplying a heated unburntmixture flow even if it is cooled to some extent by coldfuel from the reservoir. This phenomenon is not consi-dered in order to be consistent with the use of EulerEquations.
(d) Thus, the entire flowfield is initially filled with anAr-diluted oxyhydrogen mixture (2H2 + O2 + 7Ar)at P1 = 0.2 MPa and T1 = 300 K, except near thehead wall region where an azimuthally propagatingC–J detonation wavelet is artificially placed assumingits structure identical to a 1-dimensional detonation. Theheadwall boundary is considered nonslip, adiabatic andnon-catalytic.
(e) At exit plane: The specific pressure boundary condi-tion is used where no other constraints are imposedon the flow properties because the outgoing flow isknown to be always supersonic for a steady rotatingdetonation.
123
the combustion chamber, and is exhausted out the far end of the combustion chamber. Similar to thePDE, the RDE has the advantage of being operated under a wide range of conditions and Mach numbers.Unlike the PDE, it does not have to refill and initiate a detonation 20 to 100 times every second, and italso provides a steady source of thrust, and so may be an attractive alternative to the PDE. The RDEdoes have technical challenges. For instance, at start-up, a detonation wave must be initiated in a singledirection, whereas most initiators will propagate a detonation wave in both directions from the initiationsite. Also, since the detonation wave continually runs near the head-end section of the combustionchamber, the inlet micro-nozzles are subjected to intense pressures and temperatures which may limittheir life, or cause back flow into the premixture plenum. Conditions within the combustion chamber,too, are less well understood than conditions within a PDE, so that designing a combustion chamber towithstand the forces and heat-fluxes typical in an RDE may be more problematic.
The feasibility of RDE has been experimentally shown at the Lavrentyev Institute of Hydrodynamics2.Additionally, there have recently been several numerical investigations into RDE’s3,4,5. These investiga-tions have typically been two-dimensional, although preliminary three-dimensional results have also beendemonstrated5. These numerical studies have focused on an overall description of the flow-field within anRDE combustion chamber. We have developed a similar code for simulating two-dimensional and three-dimensional RDE combustion chambers using the same algorithms that have been applied very success-fully to our PDE and of our general detonation research6−12. This paper focuses on some preliminaryresults using this model and numerical algorithms to give a better idea of the flow-field within an RDE,especially near the inlet and exit planes.
II. RDE Description
A basic RDE is shown in Figure 1. The com-bustion chamber is an annular ring, where themean direction of flow is from the head end(bottom in figure) to the exit plane (top). For apropulsive engine, a nozzle would be fixed to theexit plane, however, this is not incorporated intoour model. The micro-nozzles flow in a premixtureof fuel and air or oxygen, and a detonation propa-gates circumferentially around the combustionchamber consuming the freshly injected mixture.The gas then expands azimuthally and axially,and is either subsonic or supersonic (or both),depending on the back pressure at the outletplane. The flow has a very strong circumferentialaspect due to the detonation wave propagation.Because the radial dimension is typically smallcompared to the azimuthal and axial dimension,there is generally little variation radially withinthe flow. This allows the RDE to be “unrolled”into two dimensions, and we will do this todescribe the main features of an RDE.
Micro-nozzlesHead-End
Nozzle-End
Detonation Wave
Figure 1. Schematic of three-dimensional RDE showingdetonation wave.
46th AIAA/ASME/SAE/ASEE Joint Propulsion Conference Topic: High Speed Air-Breathing Propulsion
2
From: (left) Hishida M., Fujiwara T. and Wolanski P., Shock Waves, 19:10-10, 2009(right) Schwer D. A. and Kailasanath K., AIAA 2010-6880
OverviewofRDEoperationandPressureGain(PG)
12
2 M. Hishida et al.
Fig. 1 A rotating detonation engine and its two-dimensional CFDmodel
primary purpose of the present study is (1) to run a stablypropagating detonation, (2) to describe interesting physicalphenomena and (3) to discuss whether the rotating detona-tion engine is realistic in application to thrusters or powergenerators.
Assuming that the distance between the two coaxial cylin-ders is much smaller than their diameters and that the dia-meter is not too small to create a strong centrifugal force tothe flow, the flowfield can be approximated as a plane two-dimensional area 200 mm × 200 mm, as shown in Fig. 1b.Except that the upstream and downstream conditions areidentical to the cylindrical flow, the upper and lower sidesare connected by periodic boundary conditions.
2.1 Boundary conditions
Modeling of the boundary conditions is one of the mostimportant issues in the present analysis. In view of the existing
experiments, the boundary conditions are set to be as simpleas possible to make the analysis easy, without losing generalcharacters and holding consistency with Euler Equations anda part of experimental conditions.
(a) The inflow condition is set up as follows: At t = 0,the reservoir pressure and temperature are adjusted togive a specified Mach number in the fictitious uniformregion (the pressure 0.2 MPa, the temperature 300 K)in the quiescent toroidal area, after choking and sub-sequent isentropic expansion prevail. For this adjustedreservoir condition, at t>0 after a real rotating detona-tion has started running, the injection velocity is decidedat each azimuthal location by its local pressure, depen-ding upon where the place is along the azimuthal axis.(i) Immediately behind the detonation front, the pres-sure is usually higher than the reservoir pressure; insuch circumstances the injection velocity is assumedzero. (ii) Far from the detonation front, the pressurecould be lower than the choking value; then the injectionvelocity is given by the choking condition. (iii) In theintermediate region where the pressure is between thereservoir and choking ones, the injection velocity is cal-culated by the isentropic expansion from the reservoircondition.
(b) The reservoir condition: The reservoir condition is assu-med unaffected by the downstream condition, stayingconstant throughout the analysis. In reality a reverseflow into the reservoir can happen when the flowfieldpressure exceeds the reservoir one, which certainlynecessitates taking account of modeling a large size ofreservoir area. In the present modeling, however, suchphenomenon is not included to simplify the analysis.
(c) The headwall temperature: Due to constant exposure toa hot burnt gas behind the detonation front, the headwalltemperature can be high, supplying a heated unburntmixture flow even if it is cooled to some extent by coldfuel from the reservoir. This phenomenon is not consi-dered in order to be consistent with the use of EulerEquations.
(d) Thus, the entire flowfield is initially filled with anAr-diluted oxyhydrogen mixture (2H2 + O2 + 7Ar)at P1 = 0.2 MPa and T1 = 300 K, except near thehead wall region where an azimuthally propagatingC–J detonation wavelet is artificially placed assumingits structure identical to a 1-dimensional detonation. Theheadwall boundary is considered nonslip, adiabatic andnon-catalytic.
(e) At exit plane: The specific pressure boundary condi-tion is used where no other constraints are imposedon the flow properties because the outgoing flow isknown to be always supersonic for a steady rotatingdetonation.
123
the combustion chamber, and is exhausted out the far end of the combustion chamber. Similar to thePDE, the RDE has the advantage of being operated under a wide range of conditions and Mach numbers.Unlike the PDE, it does not have to refill and initiate a detonation 20 to 100 times every second, and italso provides a steady source of thrust, and so may be an attractive alternative to the PDE. The RDEdoes have technical challenges. For instance, at start-up, a detonation wave must be initiated in a singledirection, whereas most initiators will propagate a detonation wave in both directions from the initiationsite. Also, since the detonation wave continually runs near the head-end section of the combustionchamber, the inlet micro-nozzles are subjected to intense pressures and temperatures which may limittheir life, or cause back flow into the premixture plenum. Conditions within the combustion chamber,too, are less well understood than conditions within a PDE, so that designing a combustion chamber towithstand the forces and heat-fluxes typical in an RDE may be more problematic.
The feasibility of RDE has been experimentally shown at the Lavrentyev Institute of Hydrodynamics2.Additionally, there have recently been several numerical investigations into RDE’s3,4,5. These investiga-tions have typically been two-dimensional, although preliminary three-dimensional results have also beendemonstrated5. These numerical studies have focused on an overall description of the flow-field within anRDE combustion chamber. We have developed a similar code for simulating two-dimensional and three-dimensional RDE combustion chambers using the same algorithms that have been applied very success-fully to our PDE and of our general detonation research6−12. This paper focuses on some preliminaryresults using this model and numerical algorithms to give a better idea of the flow-field within an RDE,especially near the inlet and exit planes.
II. RDE Description
A basic RDE is shown in Figure 1. The com-bustion chamber is an annular ring, where themean direction of flow is from the head end(bottom in figure) to the exit plane (top). For apropulsive engine, a nozzle would be fixed to theexit plane, however, this is not incorporated intoour model. The micro-nozzles flow in a premixtureof fuel and air or oxygen, and a detonation propa-gates circumferentially around the combustionchamber consuming the freshly injected mixture.The gas then expands azimuthally and axially,and is either subsonic or supersonic (or both),depending on the back pressure at the outletplane. The flow has a very strong circumferentialaspect due to the detonation wave propagation.Because the radial dimension is typically smallcompared to the azimuthal and axial dimension,there is generally little variation radially withinthe flow. This allows the RDE to be “unrolled”into two dimensions, and we will do this todescribe the main features of an RDE.
Micro-nozzlesHead-End
Nozzle-End
Detonation Wave
Figure 1. Schematic of three-dimensional RDE showingdetonation wave.
46th AIAA/ASME/SAE/ASEE Joint Propulsion Conference Topic: High Speed Air-Breathing Propulsion
2
From air/fuel manifold
To diffuser and turbine
Pressure waves feed back into manifold
Waves can reflect back
From: (left) Hishida M., Fujiwara T. and Wolanski P., Shock Waves, 19:10-10, 2009(right) Schwer D. A. and Kailasanath K., AIAA 2010-6880
Coupling,dynamicsandlossofpressuregain
13
• Unsteadyoperationofinjectionsystem– Injectoreffectivelytransitionfromastifftoanon-stiffinjector
– Post-detonationproductsbackflowintoplenums
– Exciteplenumdynamics
From:(top)Nordeen etal.,AIAA2011-0803
American Institute of Aeronautics and Astronautics
2
Understanding the operation of a RDE requires a basic thermodynamic model. The requirements for this model are driven by its suitability as an initial analysis tool of a RDE in much the same way that a Brayton cycle model is used for preliminary analysis of gas turbines. The model must be one-dimensional and independent of flow geometry. There must be means to account for the first order effects of thermodynamic states and an accounting of loss mechanisms. An assessment of efficiency and performance must be made with a reasonable degree of fidelity. Common thermodynamic equations of state should be used and the chemistry of combustion should be manifest only as heat added and appropriate gas constants. Above all, the model must be understandable at a fundamental level.
A thermodynamic assessment is made of a rotating detonation wave engine for the purpose of creating a parametric model. This model is based on a ZND (Zeldovitch-von Neumann-Doring)6 analysis modified by the use of the Rankine-Hugoniot equations and the application of a vector analysis of the upstream conditions. This model is compared to the thermodynamic cycle based on data from a computational simulation of an RDE.
With some adjustments, the modified ZND model approximates many features of the computational model. Further refinements should improve the predictability of the model. This model provides a reasoned thermodynamic basis for theoretical understanding, design and testing of RDE’s.
II. Numerical Simulation The simulation method is documented in a separate paper by Schwer and Kailasanath7 and will not be discussed
in detail. In summary, a premixture of hydrogen-air is injected through micro-nozzles along the inlet wall. The model is a two-dimensional Euler computation without heat or viscous diffusion. The chemistry of combustion is an induction parameter model.
The modeled chamber is 14 cm in diameter by 17.7 cm long and is modeled on a 0.2 mm x 0.2 mm grid. The heat added is 3.5500e10 erg/gm. The molecular weight of the reactants is 20.9167. Specific heats were extracted from the simulation are 1.4256 for the reactants and 1.2412 for the products. The gas constants are 3.975e6 erg/gm/K for reactants and 3.477e6 erg/gm/K for products.
III. RDE General Features A proper model of the thermodynamic cycle requires an understanding of the transfer of energy in an RDE.
There are many processes involved, and only the most significant will be discussed. The wave will be conceptually treated as a shock wave with heat addition, as in the traditional ZND analysis. The transfer of energy through the wave can be followed through a series of vector diagrams along streamlines of relative flow in the rotating frame of reference, and the corresponding path lines in the fixed frame of reference. These same streamlines form the basis for an enthalpy-entropy cycle analysis. For a number of reasons, the streamlines exhibit distinct thermodynamic cycles. However, the streamline cycles are not so different as to exclude a generalized RDE cycle that will be the basis of the one-dimensional model. Before the streamlines are discussed, a description of the basic features of the RDE will create a useful vocabulary. Investigators including Hishida8 have explored many of these features.
Figure 2. Unrolled RDE contour of stagnation enthalpy and major features.
Dow
nloa
ded
by U
NIV
ERSI
TY O
F M
ICH
IGA
N o
n M
arch
15,
201
5 | h
ttp://
arc.
aiaa
.org
| D
OI:
10.2
514/
6.20
11-8
03
From air/fuel manifold
To diffuser and turbine
Stiff injectorEffectively non-stiff injector (intermittent)
• Stronglycoupledsystem– Responseofinjectionsystemtovarying– Back-reflectionsfromdiffuser(impedancemismatchandwavereflections)
– Mixingdynamicsandeffectiveness• Incompletefuel/airmixing• Fuel/airchargestratification
– Detonationwavedynamicsandstructure• Mixtureleakage(incompleteheatrelease)• Parasiticcombustion
Exampleofupstreampropagationofblastwave
14
11
• Initial schlieren results obtained at 25k frames/s
• Allows full FOV of test section
• Significant upstream shock propogation
• Clear product backflow in air plenum
• Faint plume seen in fuel plenum
• Frame rate limits shock tracking and determination of backflow characteristics due to large frame-to-frame movements
25k frames/s Schlieren Results
Reflectedshocks
Combustion product backflow
t=0.01216 secP=2116 Pa
Note: Flow is not choked here! Shock propogation into inlet plenums may occur differently in an actual RDE.
300 slpm air, no He flow
CFD computation at U-MFrom NETL experiments (Ferguson) on AFRL injector
Exampleofdetonationchamber/plenumcoupling
15
U-M RDE
0 0.1 0.2 0.3 0.4 0.5Air mass flow rate, kg/s
0
1
2
3
4
5
Inle
t pre
ssur
e, a
tm
Air, cold flowAir, hot flowFuel, hot flow
0 0.1 0.2 0.3 0.4 0.5Air mass flow rate, kg/s
0.5
1
1.5
J
Detonation
Unsteady deflagration
Tran
sitio
n
J =
γpM 2!
!
f
γpM 2|o
DetonationDeflagration
Hot f
low
Cold
flo
w
Fuelrelevance:TowardoperationwithNGandsyngas• Mostofworkconductedsofar:
– Hydrogen/airoperation– Stabledetonationinlab-scaleRDEenabledbyfavorabledetonationproperties• Detonability• Cellsize
• Application(energyconversionsystems)require:– Naturalgasand/orsyngasoperation– Fuelflexibilityandbroadoperation(e.g.,fromparttofullload)
• Anticipatedchallenges– Stabilizationofdetonationwave• Limitsimposedbydetonationcellsize
– Fuelblendofrelevantspecies(H2/CH4/CO)impactsdetonationproperties• E.g.,inductionlength(cellsizeandstability)stronglyreducedwithH2 addition(contrarytopropane/ethylene)• E.g.,presenceofCO2 shiftsCOtoCO2conversionequilibrium,impactingheatrelease• E.g.,differenceinoxidationtimescaleofdifferentcomponentscanaffectoverallstructure
160.5 1 1.5Equivalence ratio
10-2
10-1
100
101
102
103
Indu
ctio
n le
ngth
, mm
1.5
2
2.5
CJ s
peed
, km
/s
H2
CH4
1 atm10 atm18 atm
0 0.2 0.4 0.6 0.8 1CO mole fraction
10-2
10-1
100
101
102
103
Indu
ctio
n le
ngth
, mm
1.5
2
2.5
CJ sp
eed,
km
/s
1 atm10 atm18 atm
0% CH4
5% CH4
10% CH4
ER = 1
H2 balance
10-2 10-1 100 101
Relative position, mm
1500
2000
2500
3000
Tem
pera
ture
, K
1 atm
18 atm10 atm
0% CH4
5% CH4
10% CH4
40% COH2 balanceER = 1
10-2 10-1 100 101
Relative position, mm
0
0.2
0.4
0.6
0.8
1
Mac
h nu
mbe
r
0% CH4
5% CH4
10% CH4
40% COH2 balanceER = 1
1 atm
18 atm10 atm
Overarchingobjectives
• Objective1:Developacomprehensiveunderstandingofinjectordynamics,couplingwithdiffuserback-reflections,andtheirimpactonRDEmixing,operationandperformance.
• Objective2:Developacomprehensiveunderstandingofmulti-componentfuels(syngasandhydrocarbonblends)onRDEdetonationstructureandpropagation,operationandperformance.
• Objective3:DevelopadvanceddiagnosticsandpredictivecomputationalmodelsforstudyingdetonationpropagationinRDEs,witharbitraryfuelcompositionandflowconfiguration.
17
Outline
• Programmaticoverview
• Introductiontotheproblemandgeneralapproach
• Experimentalactivities
• Computationalactivities
• Interactionsandcollaborations
18
RDEExperimentalInfrastructureatU-M• Injectorsectorsubassembly
– SectorofRDEinjectorformixingeffectivenessmeasurements
• Reduced-scaleRDE(6”RDEplatform)– Developedunderpreviousprojects– OperationalwithH2/Air,variousflowratesandequivalenceratios– Willbeexpandedtoinclude:
• MCFscapability• AdditionalinstrumentationtoinvestigateRDEdynamics
• OpticalRDE(Race-TrackRDE)– Fabricationbeingcompletedunderpreviousprojects– Equivalentto12”roundRDE– Usedforflowfield measurementsunderRDErelevantconditions
19
Injectorsectorexample(photograph)• Sectorof6”roundRDEgeometry
– Pintle geometryisidenticaltoRDE’s,justunwrapped
– AirplenumgeometryisdifferentthanRDE’s– Equivalentlengthabout1/8ofcircumferenceofcircularRDE
– Opticalaccessforlaserdiagnostics• Objective
– Conductflowvisualizationandnon-reactingmixingmeasurementsunderdifferent(jet-to-air)velocityanddensityratios(momentumfluxratio)
• ThissystemcanbeusedinsupportofRDEmeasurements(flowvisualizationandmixing)– Usedtoperformflowstructureimagingandmixturefractionmeasurementusingtracer-basedlaserinducedfluorescencemeasurements
– Variationofinjectiondesign– Parametricvariationofvelocityanddensityratios
• Canbeavailabletoprojectifneeded
20
Air port
Air plenum
Discrete injectorportholes
Side window for lateral laser sheet entrance
Front window for imaging
Laser Mixing region
Region of interest
Air
Schlieren imagingtoidentifyflowstructure(non-reactingmixing)
21
Air
He
He jet Air stream separation
Recirculation
Time
Radial stratification
3-Drenderingof6”roundRDEsystem
22
3-Drenderingof6”roundRDEsystem• Fuelinjectionsystem
– Red/blueplatepair–Modularandreadilyexchangeable
– Threedesignscurrentlyavailable
23
“Afterburner”
To exhaust
Air plenum
Fuel injector(semi-impinging jet shown)
Fuel Fuel
Air Air
Example injector plate
Threeinjectionschemesavailablefor6”RDEsystem
24
different configurations are important so that it can beassumed that in real conditions, the global conclusions ofthe present work will remain valid. In [14], Wolanski indi-cates that the calculation of the mixture formation withoutchemical reaction can be a useful first step to evaluate theefficiency of an injection device for combustors working inthe RD mode. In a possible experimental study of a CDWREcombustor, it should be important to investigate the pro-pellant mixing in order to qualify the injector before afire test.
2. Design of the CDWRE injector
Assuming that the area occupied by an injection elementis small with respect to the entire injector, the chambercurvature is not considered so that the injector wall elementhas a rectangular form and the corresponding domain for themixing flow simulation is a parallelepiped with addition ofthe H2 and O2 feeding pipes, as presented in Fig. 2.
It is now crucial to define the key parameters of themixing domain: sizes, locations of the injection holes onthe injector wall, lengths of the feeding pipes. For themixing domain we define:
! a the length along the x-axis (see Fig. 2 for the axisdefinition);
! b the length along the z-axis;! L the length along the y-axis;! A%inj the relative injection area.
As the feeding pipes can be tilted, the pipe outlet isgenerally an ellipse, whose major and minor axes are D and drespectively as shown in Fig. 3. For a feeding pipe we have:
! d the diameter or minor axis of the outlet;! D the major axis of the outlet;! ℓ the length between the inlet and outlet sections;! α the angle of the pipe axis with respect to the y-axis.
For each injected gas, a particular parameter set can bechosen.
The net injection area Ainj and the relative injectionarea A%inj are expressed by the following formulas:
Ainj ¼ πd2O2
þd2H2
4ð1Þ
A%inj ¼Ainj
abð2Þ
To define the injection hole diameters, it is assumed thatthe mixture is stoichiometric so the Equivalence Ratio (ER)is 1. The propellants are injected at the same total condi-tions to obtain a subsonic flow at the outlets with aprescribed Mach number. The injected momentum flux
ρV2! "
injis kept identical for both jets. The present study
has been done for three different configurations whoselayouts are presented in Fig. 4. The geometrical parametersof the feeding pipes are the same, only the relative positionsof the outlets vary. These configurations are identified asfollows:
(a) the “sheared injection” is designed to create a shearflow between the jets of different propellants. The axesof the two pipes lie in parallel planes;
(b) the “impinging jets” configuration has the axes of thetwo pipes in the same plane;
(c) the “semi-impinging jets” configuration uses both mixingprinciples so that the jets are partly impinging and partlysheared.
The geometrical concepts of two injection elements inthe injector wall plane are presented in Fig. 5. For the shea-red injection configuration (Fig. 5a), the centres of injectionholes are aligned on the z-axis; the holes are separated by adistance δ required for drilling. For the semi-impinging jetsFig. 3. Zoom on the modelled injection element.
Fig. 4. Layouts of the studied injection elements: (a) sheared injection; (b) impinging jets; (c) semi-impinging jets.
T. Gaillard et al. / Acta Astronautica 111 (2015) 334–344336
Semi-impinging (ONERA)
Pintle injector (NRL)Radial injection (AFRL)
FromGaillardetal.,ActaAstronautica,111:334-3442015
Fuel
Ox
6”roundRDEsystem:pintle injectordetails
25
BB 6.060
0.7
50
0.300
0.7
50
SECTION B-B
SCALE 1 : 1.5
SHEET 1 OF 1
REV
ASIZE
TITLE:
5 4 3 1
DRAWN BY:
11 2016
GAS DYNAMICS IMAGING LABUNIVERSITY OF MICHIGAN
SCALE 1:4
PART NO: RB-0
MATERIAL:
QTY: EDIT
2
UNLESS SPECIFIED:ALL DIMENSIONS ARE IN INCHES
EDIT
0.010.005
TOLERANCES:TWO PLACE DECIMAL THREE PLACE DECIMAL
Pintle injector plate
BB 6.060
0.7
50
0.300
0.7
50
SECTION B-B
SCALE 1 : 1.5
SHEET 1 OF 1
REV
ASIZE
TITLE:
5 4 3 1
DRAWN BY:
11 2016
GAS DYNAMICS IMAGING LABUNIVERSITY OF MICHIGAN
SCALE 1:4
PART NO: RB-0
MATERIAL:
QTY: EDIT
2
UNLESS SPECIFIED:ALL DIMENSIONS ARE IN INCHES
EDIT
0.010.005
TOLERANCES:TWO PLACE DECIMAL THREE PLACE DECIMAL
FuelAir Fuel Air
Outer body (wall)
Inner body (wall)
Injector plate (fuel)
Air plenum
Injector plate (air)
ProductsProducts
PicturesofRDEhardware(assembledwithexhaust)
26
Gas sampling (exhaust emission measurements)
Optical accessDump exhaust
RDE
To exhaustSupply and
control
6”RDEsystem:someinstrumentation
27
Air
Fuel
“Afterburner”
To exhaust
Air/fuelplenums
Fuel Fuel
Air Air
Detonationchannel
Smallformatopticalaccess
Air/fuelinjector
Smallformatopticalaccess&sampling
probeaccess
Toexhaust
CTAPanddynamictransducers,ionprobes
Diffuser(suddenexpansion)
3-DrenderingofRace-TrackRDEsystem(12”)• Designedwithopticalaccessinmind
– Allowsforopticalaccessofinjectionsystemanddetonationchamber
• Fuelinjectionsystem– FollowsmodulardesignapproachofroundRDE– Red/bluepair,withsimilarmodularity– Injectorsunderdesignandstudy
28
Laser sheet
Imaging region
Centerbody
Afterburner
Outerbody
RT-RDEBeingCompleted
29
ExampleofRDEoperation
30
Mixture: H2/airAir flow rate: 450 g/sEquivalence ratio: 1
Testsequenceandignitionprocess(acousticsignature)
31
0.8
I
II
III
IV
A
B
I. IgnitionII. Transition to detonationIII. Detonation termination and transition to deflagrationIV. Fuel off
Testsequenceandignitionprocess(acousticsignature)
32
!̇ = 175'() = 1
!̇ = 150'() = 1
!̇ = 200'() = 0.9
B
B
A ⟶ B ⟶ A
!̇ = 200'() = 1.2
!̇ = 450'() = 1.0
A
A
II
I
(a)
(b)
(c)
(d)
(e)
Couplingbetweenoperationmodeandplenum• Inletconditionsdependonoperationmode
– Indetonationmode,plenumpressurelower thanindeflagrationmode
33
0 0.1 0.2 0.3 0.4 0.5Air mass flow rate, kg/s
0
1
2
3
4
5
Inle
t pre
ssur
e, a
tm
Air, cold flowAir, hot flowFuel, hot flow
0 0.1 0.2 0.3 0.4 0.5Air mass flow rate, kg/s
0.5
1
1.5
J
Detonation
Unsteady deflagration
Tran
sitio
n
J =
γpM 2!
!
f
γpM 2|o
DetonationDeflagration
Hot f
low
Cold
flo
w
PR = plenum-to-channel pressure ratio
300 g/sf = 1
Pressure variation along detonation channel
PR � 2
PR � 1.7
Air H2 Detonation channel
Cold flow
Hot flow
Couplingbetweenoperationmodeandplenum
34
300 g/sf = 1
Pressure variation along detonation channel
Cold flow
Hot flow
0 0.1 0.2 0.3 0.4 0.5Air mass flow rate, kg/s
0
0.2
0.4
0.6
0.8
1
Thro
at M
ach
num
ber
Air, cold flowAir, hot flowFuel, hot flow
Estimation of Mach number at air and fuel injection throat
Air stream partiallychocked
PR � 2
PR � 1.7
Air H2 Detonation channel
PR = plenum-to-channel pressure ratio
• Inletconditionsdependonoperationmode– Indetonationmode,plenumpressurelower thanindeflagrationmode
• Airinjectionpartially(space/time)chokesathighflowrates– Detonationmodeisobserved(correlation?)– Flowpossiblyseparatesatinjectors(reducedcross-sectionalareaforflow)– Injectorstiffnessvaryoverdetonationcycle
Waterfallspectraofdetonationchamberdynamicpressure
• 3mainmodestypicallyobserved
– A:wavepropagationspeedat0.8fD– B:Toneat1fD– C:Toneat0.25fD
• Possiblycouplingofvariousdynamics– Plenumdynamics&detonationwave
35
150 g/s
175 g/s
200 g/s
250 g/s
300 g/s
ER = 0.8
A BC
Detonation chamber pressure variation
Complex dynamic pressure signature
Focusonproject• Systemsupgrade
– Instrumentationupgrades(dynamicpressure)toquantifydynamics– Additionofvariableareadiffuser– Airinjectorwithvariationinarearatio(stiffness)– Extendairplenumtoevaluateplenum/detonationwaveextentofcoupling
• Twomajoractivities/focus– Dynamics:injector/detonation/diffuserdynamics– Multicomponentfuelsoperation
• Dynamics– IdentificationandevaluationofRDEdynamics– Questions:
• whatarethey?– IdentificationfrommacroscopicobservablesonroundRDE
• Whatdotheydependon?– GeometricandfuelvariationonroundRDE
• Howdotheyeffectdetonationwavestructureandoveralloperation– Detailedlaserdiagnosticsforflowfield measurements(mixing,flowvelocityanddetonationstructure)– CombinedPIVandtracerPLIF(flamemarkerormixturefraction)studiesinRT-RDE
• Multicomponentfuels– Evaluationofuseofhydrocarbonson
• RDEoperabilityandperformancefrommacroscopicobservables• Detonationstructuredependence• Effectondetonationdynamics
– ImpactofCH4,andCOadditions– ImpactofC2H4 orC3H8 additions(contaminants,fuelflexibility)– ImpactofCO2 additiononchangingheatreleaseprofile 36
Exampleofspectralandcross-spectralanalysisforsystem’sdynamicsidentification
37
Oil flow visualization on bottom- andside-walls3-C PIV
Time-average(250images)
Instantaneous
Notenoughimagestoaverageintheseregions
u, m/s0-100 500
8
No other separation bubbles downstream in the mixing region
D- must be from the diffuser
8
No other separation bubbles downstream in the mixing region
D- must be from the diffuser
Speed iso-contours from 2-C PIV
M = 2
Figure 9. pi � pj cross-spectral results calculated using simultaneously measured pressure fluctuation timetraces located on the bottom-wall centerline: (a) coherence; (b) narrowband time lag.
27 of 33
American Institute of Aeronautics and Astronautics
Figure 9. pi � pj cross-spectral results calculated using simultaneously measured pressure fluctuation timetraces located on the bottom-wall centerline: (a) coherence; (b) narrowband time lag.
27 of 33
American Institute of Aeronautics and AstronauticsWall static pressure measurements
Shock structure visualized by schlieren imaging
(a)
(b)
(c)
Figure 12. Perturbation pathways: (a) 600 < f < 3500 Hz; (b) 300 < f < 600 Hz; (c) f < 300 Hz.
19 of 31
American Institute of Aeronautics and Astronautics
(a)
(b)
(c)
Figure 12. Perturbation pathways: (a) 600 < f < 3500 Hz; (b) 300 < f < 600 Hz; (c) f < 300 Hz.
19 of 31
American Institute of Aeronautics and Astronautics
(a)
(b)
(c)
Figure 12. Perturbation pathways: (a) 600 < f < 3500 Hz; (b) 300 < f < 600 Hz; (c) f < 300 Hz.
19 of 31
American Institute of Aeronautics and Astronautics
Sources (S, N, M) and propagation paths of frequency-dependent disturbances
Application to shock train dynamics (supersonic isolator)
CollaborationwithAFRL/Edwardsondiagnostics:augmentlaserdiagnosticsfordetonatingflows
38
V” = 0
V’ = 0 V’ = 1
A2Σ
X2Π
N”
N’
Absorption (rovibronicelectronic transition)
OH
Absorption spectrumLaser line
Transition/lasercoupling
A2Σ
PhotonAbsorp,on
VETandRET
PhotonEmission
Predissocia,onandPhoto-ioniza,on
CollisionalQuenching
State-resolvedenergy transfer
Collectionwindow
CollaborationwithAFRL/Edwardsondiagnostics:augmentlaserdiagnosticsfordetonatingflows
39
V” = 0
V’ = 0 V’ = 1
A2Σ
X2Π
N”
N’
Absorption (rovibronicelectronic transition)
OH
Absorption spectrumLaser line
Transition/lasercoupling
A2Σ
PhotonAbsorp,on
VETandRET
PhotonEmission
Predissocia,onandPhoto-ioniza,on
CollisionalQuenching
State-resolvedenergy transfer
• Thesemethodsandcollaborationgivesusaframeworkto:– EvaluateandoptimizeLIF-basedimagingtechniquefordetonatingflows–DemonstratemethodsinRDErelevantflowfields (RT-RDE)– PerformmeasurementsonRT-RDEatAFRL/Edwardsleveragingtheirinstrumentationandcapabilities• Anticipated3measurementscampaigns
Outline
• Programmaticoverview
• Introductiontotheproblemandgeneralapproach
• Experimentalactivities
• Computationalactivities
• Interactionsandcollaborations
40
GoalsofCFDProgram
• Developfully-resolvedadaptivemeshcompressiblesolversforcapturingdetonationprocesses– Studystructureofdetonationsinnon-premixedsystems–Developreduced-ordermodels– Studyfuelcompositioneffectsonstability
• AssistinthedevelopmentoftheexperimentalRDEconfigurations– Providedetailedsimulationdatatocomplementexperimentalmeasurements– Conductsimulationsoutsideofexperimentalparameterstoextenddatasets
• Developmentsandstudiesleverage–OpenFOAM suitesofcodes–U-MdetonationsolversUMDetFOAM
41
OpenFOAM CodeDevelopment
• AllcodesandmodelsdevelopedusingtheopenFOAM opensourcecodebase– 10+yearsexperienceinusingthistool• SeveralNETLprojectssuccessfullycompleted
– Easytransferofcodetoindustry/researchcommunity• PriorsolverstransferredtoSiemensInc.
• Highlyscalableandrunson10K+processors– Extensivecoderewritestoensurelinearscalability
42
CompressibleDetonationSolvedUMDetFOAM
• Fullyexplicitsolver– EulerandN-Sequations
• Severalfluxschemes– Locallyadaptabletoensureminimaldissipation
• CANTERA-basedchemistrymodule– Allowsanydetailedchemistrymechanismtobeused– Canhandlearbitrarynumberofspecies
• Adaptivemeshrefinement– Locallyadaptivegridstocapturedetonationstructures
43
CaseStudies- 1DdetonationwithAMR
• ConvergencetestwithH2/Airmechanism
• ThebasegridforAMRstudyisdx =0.4mm– Showsconvergenceasincreasingthelevelofrefinement
44
Product27.1 atm3000 K1975 m/s
H2/ Air, 1 atm, 300K, 0 m/s
2-Dethylenecase
• Cellularstructurevalidation– Longitudinaltracksfromtheintersectionpoints– 2cellstructureacrossthechannelwidth
45
Hayashi et al.
U of M
C2H4 / O2, 0.1 atm, 300 KD = 3 µm, h = 2 mm
AFRLandU-MFullGeometryModeling
46
AFRL geometry Pintle geometry
Legend:Iso-contour of density gradient (black)Iso-contour of H2 mass fraction (0.1), colored by temperatureIso-contour of OH mass fraction (0.0075), colored by temperature
AFRLInjectorResponse
• Flashbackoccurswhenadetonationwavemovesacrossinjector– Chockingisterminated– Post-combustiongasespropagatesbackintoplenum–Blastwavesmoveintoplenum
• Quickinjectorrecovery– Reversedflowpushedbackintochannelduetohighplenumpressure
47
Focusontheproject
• Simulatefueleffectsoncelldetonationsize– 2Dgeometry– 2Dunrolledgeometry– Fullscalegeometry
• Effectofwavestructureondetonationprocess– Couplingbetweeninflowanddetonationchamber– Interactionofdownstreamwavestructures
• Modelingdetonations– Atabulatedmodelingapproach
48
Outline
• Programmaticoverview
• Introductiontotheproblemandgeneralapproach
• Experimentalactivities
• Computationalactivities
• Interactionsandcollaborations
49
Interactions,collaborationsandsynergies• Strongcouplingbetweenexperimentsandcomputations
– Modeldevelopmentandvalidation– Experimentdesignandunderstanding– Strongcollaborationswithexternalpartners– CombinedinvestigationofthephysicsofdetonationsunderMCFs(relevancetoapplication)andimpactofinjector/detonation/diffuserdynamicsondetonationpropertiesandRDEperformance
• Keyexternalcollaborations– ISSI/AFRLWP(Drs.JohnHoke&FredSchauer)onRDEoperation,performanceandmodeling.– UTRC (Drs.AdamHolleyandPeterCocks)ondetonationandRDEmodelingforarbitraryfuelsandgeometries.
– GE (Venkat Tangirala)onRDEoperation,performanceandmodeling.– AFRL/Edwards (Dr.WilliamHargus)onthedevelopmentandapplicationofdiagnosticsappliedtorelevantRDEgeometries.
– WilliamsInternational(KyleMcDevitt)ondetonationandRDEmodelingforarbitraryfuelsandgeometries.
• Othercollaborations/interactions– NETL (Dr.Ferguson)onmodelingandRDEperformance&operation– UniversityofMaryland(Prof.Yu)onuseofexperimentaldataforvalidationinsimplegeometries– NRL (Dr.Kailasnath)oncodeandcombustionmodeldevelopment
50
51
Questions?
Top Related