Inlet Isolator and Combustion Physics at Take- Over Region of Scramjet Engines

Transcription

1 Inlet Isolator and Combustion Physics at Take- Over Region of Scramjet Engines Brendan McGann, Damiano Baccarella, Tonghun Lee* Department of Mechanical Science & Engineering, University of Illinois Venkat Narayanaswamy* Department of Mechanical & Aerospace Engineering, NC State North Carolina State University Grant: FA University of Illinois, Urbana-Champaign AFOSR Review Meeting (July, 2017)

2 Main Issue: Unstart in Supersonic Engines Main Issue: Turbulent High Re# Flames: and Boundary Layer Separation can Result in Choking of the Flow Path Flame Stabilization Mach 4.5 OH PLIF Unstart due to Flow Choking 2 What happens at the operational stability limits of Unstart? What transition takes place in the inlet isolator boundary layer? How does stability mapping correlate with Ka# space? Flame structure How does the inlet isolator physics impact combustion and vice versa? What are the rate limiting or dominant physics at the stability limits?

3 Inlet Unstart (Mass Injection) Background: Previous Work in Unstart Cold Unstart with Mass Injection H. Do et al., 2011 (Stanford, USA) Previous Studies of Unstart with Visualization Onset of Choking with Heat Release Laurence et al., 2014 (DLR, Germany) Upper BL still laminar t=2.5ms t=3.0ms t=3.5ms unstart shock Unstart shock appears in field of view Separated flow reaches leading edge generating stronger incident shock Unstart shock propagates upstream t=4.0ms t=4.5ms t=5.0ms t=5.5ms t=6.0ms t=6.5ms t=7.0ms 3 Mach 5 airflow, P 0 = 350 kpa, T 0 = 300K 25% CO 2 by volume, Rayleigh Scattering Mach 8, 27 km altitude conditions airflow, P=2.1kPa, T=260K

4 Diagnostics Approach: Inlet Physics & Unstart Inlet Isolator Physics: high speed Schlieren, Rayleigh scattering pp Single Shot dd 2 = MM pp uu for MM uu 4.5 uu Normalized Average General Instability Monitoring: absorption spectroscopy (SCLAS) Flow Field Visualization, Turbulence Characterization (PIV) 2 2 t f l f / v l f Ka = = = 2 2 t η / v η η 3 v η = ε 1/ 4 Interaction Flame Structure & Ignition Mapping: nano-libs Combustion Dynamics: High Speed PLIF 4 N. Peters

5 Some Potential Inlet Isolator & Combustion Interaction 5 Understanding the Synergistic and Mutual Impact on Unstart Normal shock SBLI without separation should have steep change in P Pressure smearing/steepening along unseparated isolator shock train will indicate a subsonic channel propagation Enhanced momentum deficit within parts of boundary layer can cause mass flux in circumferential fluid migration Changes in the lnlet isolator flow and shock wave propagation can have dramatic impact on the combustion process no shock oblique shock

6 Planned Results & Benefits to the Air Force Planned Results Novel Optical and quantitative unstart data for both inlet isolator and combustion (T, P, ɸ, Ka#, Ma#) Data in a generic axis-symmetric geometry Unstart phenomena over full development to disgorge X-51 scramjet engine, US Air Force Benefits to the Air Force Understanding of scramjet unstart as a complete interaction between inlet isolator and combustor Development of comprehensive models and numerical simulation to capture the full unstart process Novel design guidelines for inlet and combustor design Development of new strategies for unstart mitigation 6

7 Where we left off laser year Arc Heated Combustion Tunnel ACT-I (Professor Hyungrok Do, Notre Dame) New Project Start December Mach 4.5 Flow 3D PLIF Venkat presented a vision talk last year focused on ACT-I diagnostics New unstart project officially kicked off December of 2016 (Joint funding from Dr. Ivett Leyva and Dr. Chiping Li, AFOSR) Preparation and shakedown of the ACT-II at the University of Illinois

8 New Development: ACT-II at University of Illinois ACT-II (Arc-heated Combustion Test-rig) Pulsed, arc-heated hypersonic facility for supersonic combustion research. Optimized for laser diagnostics on high-speed reactive flows. High flexibility of flow conditions and geometric configurations. ACT-II test section Free flow model scramjets (2D and axisymmetric) Overview of ACT-II facility ACT-II arc-heater 8 Direct-connect circular combustor ID 28 mm Direct-connect circular combustor ID 35 mm

9 Powering of ACT-II ACT-II power feeding circuit Prevent overcurrent during arc ignition. Stabilize arc I-V characteristic curve. Reduced after ignition to maximize arc power. ACT-2 Electric Circuit DC Power supply. Input power: 350 kw Max power output 250 kw (660 V, 400 A). Fully programmable and remotely controlled. Coaxial electric configuration. Wall stabilization by a constrictor. Separated injection of N2 and O2. Operation Sequence 1. Apply Voltage (close C1). 2. Open Argon valve. 3. Open Nitrogen valve. 4. Close Argon valve. 5. Reduce Resistance (close C2) 6. Cut-off Voltage (open C1 & C2). 7. Close Nitrogen valve. 9 Discharge parameters remarkably stable during the useful test time. Test time can be extended up to a second (limited by electrodes overheating). After a test, vacuum is quickly restored and a new test can be performed in a few minutes.

10 Shot-to-shot variability Repeatability of the flow conditions in ACT-2 AVG STDEV %STDEV Mass Flow N2 [g/s] Mass Flow O2 [g/s] Mass Flow Total [g/s] Supply Voltage [V] Arc Voltage [V] Arc Current [A] Total pressure [kpa] Total temperature 1 [K] Estimated by mass conservation 24 tests performed consecutively in the same nominal conditions. Relative standard deviation less then 2% for the total temperature, 1% or less for the other parameters. Maximum shot-to-shot variation less then +/-3% on all the parameters Maximum run time: 1 second 10

11 Diagnostics: Inlet unstart experiments Inlet unstart in an axisymmetric model scramjet Inlet/Diffuser Isolator Fuel injection ports Combustor Outlet/Nozzle Picture of the assembled model 11 Experimental targets: Isolator boundary layer profiles to study crossflow migration. PLS measurement using tracers to study the extent of subsonic boundary layer region in response to combustor fluctuations. High frequency heat flux measurements in nonseparated isolator flow to delineate the upstream migration of temperature fluctuations. High frequency pressure measurements in nonseparated isolator flow to delineate how much upstream the pressure fluctuations originating in combusting region propagates. Model dimensions in mm

12 Inlet unstart in an axisymmetric model scramjet CO 2 planar laser Rayleigh scattering in the Scramjet model at 10 khz Pure CO 2 freestream at Mach 4. Nitrogen injected downstream to induce blockage. PLRS visualization in the isolator. Pressure measurements in the inlet and combustor. Mach number scale Injectors Pressure sensors Kulite XTE-190M 100% CO 2 MACH 4 FREE STREAM 35 mm Inlet: area ratio 1.3 Pressure sensor Kulite XCS-062 Field of view (slowed 500 times) Isolator entrance pressure Combustor entrance pressure 12

13 Inlet unstart in an axisymmetric model scramjet t 0 Broadband chemiluminescence (Casio EX-F1 at 1200 fps) T 0 = 2500 K; P 0 = 120 kpa; M 1 = 4.5; φ = 3.0 Fuel injection begins in the supersonic stream. t 0 +4ms t 0 +13ms t 0 +17ms After a few ms a combustion reaches regime conditions. Choking occurs at the end of the combustor. Choked flow propagates upstream. Separation in the isolator. t 0 +27ms t 0 +31ms t 0 +45ms t 0 +55ms isolator fuel combustor Combustor entirely choked. Separated flow in the isolator. Flames propagate into the isolator through the separated region. Inlet unstart. Flow entirely subsonic throughout the model. Low speed and high temperature flow. Flames localized at the injectors. OH* chemiluminescence (intensified Photron SA-5 at 15 khz) UNCHOKED CHOKED 13 isolator injector injector combustor Strongest flame at the end of the combustor. No flame in the isolator. Strongest flame downstream of the injectors. Flames propagating into the isolator.

14 Thermal choking in a direct connect combustor OH* chemiluminescence and PLIF Time sequence of the pseudoshock propagation 280 mm t = 0 The flame is short and intense high burning rate. The pseudo-shock propagates upstream very fast. Pseudo-shock structure visualization 0.6 ms 0.9 ms 1.3 ms The flame is stretched and looses intensity low burning rate. The propagation velocity is reduced below the flow speed and the flame recedes. The flame shrinks and its burning rate increases again. The shock-train gains speed and return to propagate upstream. OH PLIF OH* chemiluminescence 14 FLOW DIRECTION

15 Thermal choking in a direct connect combustor Time resolved pressure measurements Direct connect, circular (ID 28 mm) instrumented combustor, 400 mm long. 18 pressure ports 20 mm apart. Flow conditions: P0 = 1 bar, T0 = 2200 K, M = 3 Pressure measured by 12 Kulite XTE-190 sensors. 4 radial jets fuel (ethylene) injector. Pressure signals at incipient choking Instrumented direct-connect combustor 15 Pressure distribution in the combustor at different equivalence ratios Choking observed for φ At φ = 0.58 pseudo-shock (PS) localized between 130 and 260 mm. PS moves upstream increasing φ and reaches the injectors at φ Before choking the pressure follows the supersonic branch of 1D solution. After choking the pressure follows the subsonic branch of 1D solution.

16 New Developments Current Diagnostics Development Collaborative research with Aerospace Systems Directorate, WPAFB 16

17 Nano Second Laser Induced Breakdown Spectroscopy (LIBS) Mach 3, Research Cell, Cavity Flameholder 3 AFRL RC-19 Computer Burner Plasma Breakdown Beam Dump VPCC Lamp Q-Switch Signal Gen. Camera Spectrometer Mirror w/ Hole 532 nm mirror Thin Film Polarizer f=175 mm lens Pockels Cell Laser Nd:YAG 532 nm mirror Adv. Q- Switch Sync Adj. Half Wave Plate 17 Calibration setup for LIBS in variable pressure combustion chamber (VPCC)

18 No Distortion Wedge 2 Wedge 1 Effect of Distortion on Ignition and Fuel Mixing Fuel Mole Fraction Gas Density Wedge position 1 results in a shear layer deflected down and smaller recirculation zone Wedge position 2 results in a lifter shear layer in front of the shock impingement location Similar maximum fuel mole fraction in all cases despite significant difference in fuel injection rate 18

19 CH C-X PLIF Imaging: On Resonant Transition CH C-X Excitation piloted turbulent combustor 10 khz CH PLIF Herzberg & Johns, Astrophys. Journal, vol. 158, 1969 Problems On resonant transition: cannot be used near scattering surfaces (enclosed chamber) or simultaneously with PIV 19

20 Simultaneous 10kHz PIV-PLIF Results Camera View Instantaneous Velocity Average Velocity CH OH 20

21 Summary Laser and Optical Diagnostics High speed CH C-X PLIF, OH PLIF and PIV (10 khz) Application of nano-libs in WPAFB RC-19 ACT-II Development Hypersonic Combustion ACT-II: Integration of Stable Arc Heating System ACT-II: Flame Stability in Dual Jet Axis-Symmetric Configuration and modeling ACT-II: Direct Connect Unstart Experiments Future objectives Unstart Shakedown in ACT-II Freeflow Facility PLRS measurements for flowfield characterization Dual diagnostics for inlet isolator and combustor Pressure measurements at unstart conditions Unstart model development 21

22 Next Steps Anticipated new results Impact of Combustor on the Isolator Shock/Boundary Layer Interaction Impact of Inlet Isolator Physics on the Combustion Propagation during Unstart (what promotes or mitigates heat release) Any Breakthroughs in the Horizon First Simultaneously Imaging of Inlet Isolator and Combustor During Entire Unstart Process (Initiation to Inlet Disgorge) Integration of novel high speed laser diagnostics for inlet and combustor unstart Publications (Awards, Honors, Media Coverage) D. Baccarella, Q. Liu, T. Lee, The Supersonic Combustion Facility ACT-2, AIAA SciTech, AIAA , Questions

23 Technical Challenges Remaining Experimental Challenges Application of quantitative diagnostics for laser and optical imaging (quantification of PLIF signal) High contrast imaging of the inlet isolator section coupled with hot flow combustion cases (possible use of NO PLIF) Schlieren or shadowgraph imaging in an axis-symmetric geometry Analysis Challenges Methodology for quantifying the interaction between inlet isolator and combustor (modeling strategy) Qualitative characterization of the combustion (Ka#, Re#, etc.) flame dynamics for various unstart stages 23 Questions

24 Acknowledgements AFOSR (Program Manager: Dr. Ivette Leyva) AFOSR (Program Manager: Dr. Chiping Li) University of Illinois Foundation Dr. Campbell Carter (WPAFB AFRL) Dr. Timothy Ombrello (WPAFB AFRL) Dr. Stephen Hammack (WPAFB AFRL) Dr. Aaron Auslender for valuable feedback Dr. Hyungrok Do (Seoul National University) Grad Students: Brendan McGann, Damiano Baccarella, Qili Liu, Gyusub Lee 24