Unsteady supercritical/critical dual flowpath inlet flow and its control methods

Jun LIU,Huacheng YUAN,Yunfei WANG,Ning GE

College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China Jiangsu Province Key Laboratory of Aerospace Power Systems,Nanjing 210016,China

Unsteady supercritical/critical dual flowpath inlet flow and its control methods

Jun LIU,Huacheng YUAN*,Yunfei WANG,Ning GE

College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China Jiangsu Province Key Laboratory of Aerospace Power Systems,Nanjing 210016,China

Airbreathing hypersonic vehicle; Dual flowpath inlet; Terminal shock oscillation; Turbine based combined cycle; Unsteady flow

The characteristics of unsteady flow in a dual- flowpath inlet,which was designed for a Turbine Based Combined Cycle(TBCC)propulsion system,and the control methods of unsteady flow were investigated experimentally and numerically.It was characterized by large-amplitude pressure oscillations and traveling shock waves.As the inlet operated in supercritical condition,namely the terminal shock located in the throat,the shock oscillated,and the period of oscillation was about 50 ms,while the amplitude was 6 mm.The shock oscillation was caused by separation in the diffuser.This shock oscillation can be controlled by extending the length of diffuser which reduces pressure gradient along the flowpath.As the inlet operated in critical condition,namely the terminal shock located at the shoulder of the third compression ramp,the shock oscillated,and the period of oscillation was about 7.5 ms,while the amplitude was 12 mm.At this condition,the shock oscillation was caused by an incompatible backpressure in the bleed region.It can be controlled by increasing the backpressure of the bleed region.

1.Introduction

A Turbine Based Combined Cycle(TBCC)engine is a reusable,low-cost,and high-durability propulsion system.1–3It is one of the most promising propulsion systems for next generation vehicles.This combined cycle engine can operatewell in low and high Mach numbers through a turbine engine and a dual-mode scramjet engine,respectively.The inlet should provide proper air flow for the combined cycle engine during the whole flight envelope.Therefore,the performance of inlet is critical to the whole propulsion system.Especially,when large-amplitude pressure oscillations and traveling shock waves occurred in the inlet,it would cause structural damage,engine surge,combustion flameout,or non-recoverable thrust loss.4

On one hand,the TBCC inlet has been investigated through experimental tests and numerical simulation.A tandem co-axial con figuration TBCC engine with an inlet was tested at a free in flow Mach number of 5.0.A terminal shock oscillation phenomenon was detected,and the net thrust of the engine was sensitive to the terminal shock posi-tion.5The aerodynamic design of a dual- flow hypersonic inlet for a TBCC propulsion system and the performance,starting characteristics,and mode transition of an over/under con figuration TBCC inlet were investigated through numerical simulation and experimental tests.6–10An inward-turning inlet design for a TBCC propulsion system was tested at NASA’s Langley wind tunnel.11On the other hand,shock oscillation in supersonic and hypersonic inlets has been investigated widely.Trapier et al.12investigated the onset of supersonic inlet buzz through the analysis of pressure records.Lee et al.13investigated flow characteristics of small-sized rectangular and axisymmetric supersonic inlets.The buzz phenomenon of a small-size inlet is identical to that of a large inlet,but a small inlet can be easily affected by separation bubble.The buzz phenomena in hypersonic inlets were investigated by Tan et al.14,15,Chang et al.16–18,and Li et al.19However,to our knowledge,few of them have investigated the unsteady flow in a TBCC inlet as it operates in supercritical/critical conditions.There are two main differences between a TBCC inlet and a normal supersonic inlet,which may lead to the differences in unsteady flow characteristics.Firstly,a TBCC inlet is characterized by a dual flowpath which provides air flow for turbojet and ramjet engines.During the inlet mode transition,two engines work together to provide the thrust.The backpressure of TBCC inlet is determined by the operation conditions of these two engines,so when the inlet operates in supercritical condition,the terminal shock position is determined by the backpressure from the turbojet and ramjet flowpath.Secondly,the operation range of a TBCC inlet is wider than that of a normal supersonic inlet.To provide the required air flow for turbine/ramjet engines,mass- flow should be bled at an off-design condition.That makes it different from others.The TBCC inlet researched in this paper could still remain start due to a large amount of bleed when the shock was located upstream from the throat,which may cause a normal supersonic inlet into unstart.At this condition,the terminal shock position of the TBCC inlet was affected by the backpressure from the diffuser and the bleed region.

To sum up,unsteady subcritical flow in supersonic and hypersonic inlets have been investigated widely,but few of the research has been conducted on TBCC inlets.Therefore,this paper focuses on the investigation of unsteady flow in a TBCC inlet.Firstly,the characteristic of pressure oscillation near the terminal shock was investigated through wind tunnel tests.Then,the characteristic of terminal shock oscillation was explored by numerical simulation.Finally,based on the analyses of terminal shock oscillation,effective control methods were put forward.

2.Method

2.1.Descriptions of the TBCC inlet model

A dual- flowpath inlet is designed for a tandem co-axial con figuration TBCC propulsion system.The sketch of this inlet model is shown in Fig.1.It is devised to work from takeoff to Mach number 3.0.As shown in this figure,the inlet shares the same external compression ramps and rectangular-toround diffuser.The angles of the second and third ramps vary with free in flow Mach number.At Mach number 2.0,the inlet achieves external air compression using three ramps inclining at 6.0°,2.0°,and 4.0°,respectively.The throat of the inlet is a rectangular duct with a constant cross section of 80 mm wide by 30 mm high.The geometry of the diffuser is rectangle-toround shape transition based on the mathematic method mentioned in Ref.20.The area ratio of exit to entrance section of the diffuser is 3.7.Downstream from the diffuser’s exit plane,the duct is separated into two flowpaths.The inner round duct is the turbojet flowpath and the outer annular duct is the ramjet flowpath.The area ratio of turbojet flowpath to ramjet flowpath is 0.77.

2.2.Experimental conditions and measurements

Experimental tests were conducted in the NH-1 high-speed wind tunnel at Nanjing University of Aeronautics and Astronautics(Fig.2).The tunnel was operating in a blown-down mode with a usable run time longer than 40 s.The tunnel has a rectangular working section with a constant cross section of 600 mm wide by 600 mm high.The length of the working section is 1580 mm.The upstream working section is an interchangeable two-dimensional Laval nozzle,providing nominal free-stream Mach numbers from 0.5 to 2.0.For current tests,a Mach number 2.0 nozzle was used.The total pressure of free-stream was 208 kPa,and the total temperature was 300 K.The unit Reynolds number was 2.5×107m-1.During inlet mode transition tests,the angle of attack and yaw angle were 0°.

Time-accurate pressure measurements were performed by six dynamic pressure transducers to monitor unsteady flow patterns of the inlet.The probes of dynamics pressure transducers are shown in Fig.1.There are two dynamic pressure transducers in the third compression ramp(R3-R4),three in the lower surface of the diffuser(D1-D3),and one in the upper surface of the diffuser(U1).These transducers have an accuracy of±0.1%of the full range and a natural response frequency of 50 kHz.The dynamic pressure along the surface was measured by a TST5913 data acquisition system.The data sampling frequency was set to 10 kHz per channel during the test.

2.3.Numerical simulation

Numerical simulation of the unsteady flow field in the TBCC inlet was performed by FLUENT solver.The Reynolds averaged Navier-Stokes equations in three dimensions were solved by using a finite volume spatial discretization method.During the computation,the inviscid flux scheme was Roe’s method,and the Monotonic Upwind Scheme for the Conservation Laws(MUSCL)approach was used for variable extrapolation.The viscid flux scheme was discretized by a second-order central difference scheme.21,22The turbulent flow was modeled by the two-equation standard k-ε.The fluid was treated as compressible ideal gas.The computational domain includes the flow field surrounding the tandem TBCC inlet,and its boundary condition was set as pressure far- field.Turbojet/ramjet mass- flow plugs were set at the exit of each flowpath which was consistent with the experimental model.These two flowpaths were discharged into a common plenum.No-slip adiabatic wall conditions were imposed on the solid wall boundaries.The geometry was symmetrical,and the yaw angle of the model was equal to 0°.The flow field was symmetrical,too.Therefore,the computational domain was chosen just as a half of the actual one.A typical computational mesh is shown in Fig.3.

3.Results and discussion

3.1.Pressure oscillation at supercritical/critical conditions

The motion of the terminal shock can be inferred from the analyses of dynamic pressure signals.Fig.4 shows the dynamic pressure signals in a test.There were six steps in this test,namely six different throttle ratios.In this test,the throttle ratio of the ramjet flowpath was fixed as 80%,and the throttle ratio of the turbojet flowpath were increased from 55%to 73%.The specific values were 55%,60%,63%,66%,69%,and 73%.As Fig.4 shows the static pressure ratio(π)of probes D1-D2 and R3-R4 varied with flow time(t).In this figure pressure oscillated in the second step of probe D1 and the third and fourth steps of probe R4.In the second step of probe D1,pressure oscillated in an irregular pattern,but in the third step of probe R4,the law of pressure oscillation was similar to a sine curve with different amplitudes.We can infer that there was terminal shock oscillation around these probes.This section is going to reveal the characteristics of pressure oscillation as the TBCC inlet operates at supercritical and critical conditions.

Inlet operation conditions can be divided into supercritical,critical,and subcritical,which was determined by the position of the terminal shock.In this paper,when the terminal shock locates near the throat,namely probe D1 in this combined cycle inlet,the inlet operates in supercritical condition;when the terminal shock locates near the shoulder of the third ramp,namely probe R4,the inlet operates in critical condition;when the terminal shock locates in front of the cowl leading edge,namely probe U1,the inlet operates in subcritical condition.These operation conditions can be obtained through the analyses of static pressures along the upper and lower surfaces.Fig.5 shows the steady static pressure distribution along the flowpath as the terminal shock located at the throat and shoulder of the third compression ramp.The numerical simulation results fit well with the experimental data,except some deviation in the second bend of the diffuser.The difference could be attributed to over-prediction of the separation bubble.The flowpath was narrowed by the bubble,whereas the flow was accelerated.This caused the pressure in CFD to be smaller than the experimental data.The results of a normal grid(600000 cells)was consistent with those of a fine grid(1230000 cells).The difference of parameters in the diffuser exit plane was smaller than 0.1%in these two cases.As Fig.5(a)shows,the upper and lower surface static pressures increased dramatically at the throat,so the terminal shock located at the throat.In Fig.5(b),the static pressure increased dramatically at the shoulder of the third compression ramp,so the terminal shock located at this place.

The positions of the terminal shock can be inferred from the steady pressure distribution,and the motion of the terminal shock can be inferred from dynamic pressure signals.After the analyses of Figs.4 and 5,we inferred that the terminal shock oscillated at supercritical and critical conditions with different patterns.The characteristics of terminal shock oscillation will be revealed through numerical simulation in next section.

3.2.Characteristics of the terminal shock oscillation

3.2.1.Supercritical operation condition

A three-dimensional unsteady numerical simulation method was adopted to investigate terminal shock oscillation in supercritical condition.Fig.6 shows the Mach number contour and velocity vector along the flowpath as the inlet operated at supercritical condition.As shown in Fig.6(a),the terminal shock oscillated from the upstream of the throat to its downstream.A separation bubble occurred on the sidewall and developed gradually along the flowpath.To illustrate the development of the separation bubble clearly,seven sections(S1-S7)along the flowpath and in each section three lines(L1-L3)of velocity vector are shown in Fig.6(b).Section S1 is the exit plane of the throat,section S6 is the exit plane of the diffuser,section S2-S5 are the planes between throat and diffuser exit plane,section S7 is the entrance of turbine/ramjet flowpath.In these sections,L1 is the line in the symmetrical plane,L2 is the line 20 mm away from the symmetrical plane,and L3 is the line 40 mm away from the symmetrical plane.A reverse flow appeared on line L3 of section S1.This reverse flow was a separation bubble caused by the terminal shock and boundary layer interaction.The reverse flow appeared on line L2 of section S2 which was downstream from section S1.That means the separation bubble grew along the flowpath.In the exit plane of the diffuser,namely section S6,the separation disappeared,but the velocity in the lower zone of this section was low.

As Fig.7(a)shows,the shock positions varied with time.It oscillated around x=330 mm,and the period of oscillation was about 50 ms.The fitting curve of these numerical results in a cycle is shown in Fig.7(b),x-coordinate is the nondimensional time(T),y-coordinate is the positions of terminal shock.In this figure,the triangle points are the numerical results,and the solid line is the fitting of these values.The law of terminal shock was similar to a sine curve,and the amplitude of shock oscillation was about 6 mm.

3.2.2.Critical operation condition

As the backpressure of the diffuser increased,the terminal shock moved upstream to the third compression ramp,and then the operation condition of the inlet changed from supercritical to critical.At this condition,the terminal shock was affected by the backpressure from the diffuser and the bleed region(Pb).Fig.8 shows the Mach number contour and velocity vector along the flowpath at this condition.As Fig.8(a)shows,when the shock moved to its most upstream position,the flow separated from the shoulder of the third compression ramp and a weak shock appeared at the entrance of the throat;when it moved to the most downstream position,the shape of the terminal shock turned into λ.The separation onset on the sidewall and developed gradually along the flowpath,as illustrated in Fig.8(b).

As Fig.9(a)shows,the shock positions varied with time.It oscillated around x=300 mm,and the period of oscillation was about 7.5 ms.The fitting curve of these numerical results in a cycle is shown in Fig.9(b),x-coordinate is the nondimensional time(T),y-coordinate is the positions of terminal shock.The law of terminal shock was similar to a sine curve,and the amplitude of shock oscillation was about 12 mm.

3.3.Control method for terminal shock oscillation

3.3.1.Supercritical operation condition

According to the analyses above,the separation induced by the terminal shock and boundary layer interaction was the main factor to cause terminal shock oscillation.According to Ref.20,area distributions along the diffuser would control the streamwise pressure gradient imposed upon the flow(and thereby the flow separation).Therefore,the length of the diffuser was extended to reduce the gradient of pressure along the flowpath.

In the new model,the length of the diffuser was extended from 400 mm to 650 mm.The Mach number contour and velocity vector along the flowpath as the inlet operated at supercritical condition are shown in Fig.10.The terminal shock located in the throat which was the same as before,but it was steady in the throat.According to Fig.10(b),the reverse flow only appeared at the bottom of line L3 in sections S1,S2,and S3.The separation bubble in the diffuser was smaller than that in the baseline model.Thus,the development of the separation bubble was well controlled.

To illustrate the effectiveness of the extending diffuser length method,the total pressure recovery(σ)at the diffuser exit planes of both models are shown in Fig.11.Fig.11(a)shows the total pressure recovery of the baseline model as the shock located at the downstream position of a cycle.We de fine the total pressure recovery less than 0.74 as a low total pressure.The boundary of the low-total pressure zone is the dash line shown in Fig.11.The minimum value of the baseline model was about 0.65.While in the new model,the low total pressure zone was smaller than that of the baseline model,the minimum value was 0.72 which was greater than that of the baseline model.

3.3.2.Critical operation condition

The characteristics of terminal shock oscillation in critical operation condition has been revealed in Section 3.2.At this condition,the terminal shock was affected by the backpressure from the diffuser and the bleed region,so the motion of the terminal shock can be controlled by the backpressure of the bleed region.In this section,the backpressure of the bleed region was set as 60 kPa,80 kPa,and 85 kPa,respectively.Fig.12 shows the Mach number contour of the inlet and the bleed exit plane as the backpressure of the bleed region was set as 60 kPa and 85 kPa.As shown in Fig.12(a),the Mach number contour of the bleed exit plane can be divided into two zones(i.e.,subsonic zone and supersonic zone),of which the boundary is shown as a dash line.This Mach number contour was similar to that in Fig.8(a)where the backpressure of the bleed exit plane was 40 kPa.The amplitude of shock oscillation was consistent with each other about12 mm,butthefrequency of shock oscillation decreased from 100 Hzto 80 Hzasthebackpressure increased.When the backpressure increased up to 80 kPa,the amplitude of shock oscillation was only 6.9 mm.Finally,when the backpressure increased up to 85 kPa,the terminal shock was stable in the shoulder of the third compression ramp.Downstream from the terminal shock,a small weak shock appeared in front of the bleed region.The Mach number contour of the exit plane in the bleed region is shown in Fig.12(b),in which the subsonic zone in this condition was greater than the former one.Additionally,to avoid unstart of the inlet,the backpressure of the bleed region should be lower than 94 kPa.

4.Conclusions

The characteristics of terminal shock oscillation in a turbine based combined cycle inlet and the control methods of the shock oscillation were investigated experimentally and numerically.The main conclusions are as follows:

(1)According to the experimental data acquired from highresponse and steady-state pressure transducers,the terminal shock oscillated in the inlet throat and the shoulder of the third compression ramp.

(2)When the inlet operated in supercritical condition,namely the terminal shock located in the throat,the shock oscillated,and the period was about 50 ms,while the amplitude was 6 mm.In this condition,shock oscillation can be suppressed by extending the length of the diffuser from 400 mm to 650 mm.

(3)When the inlet operated in critical condition,namely the terminal shock located at the shoulder of the third compression ramp,the shock oscillated,and the period was about 7.5 ms,while the amplitude was 12 mm.In this condition,shock oscillation can be suppressed by increasing the backpressure of the bleed region up to 85 kPa.

Acknowledgements

This study was co-supported by the Funding for Outstanding Doctoral Dissertation in NUAA of China(No.BCXJ16-01),Funding of Jiangsu Innovation Program for Graduate Education(No.KYLX16_0393),and Foundation of Graduate Innovation Center in NUAA of China(No.KFJJ20160204)which is supported by the Fundamental Research Funds for the Central Universities and the Aerospace Science and Technology Innovation Fund of China Aerospace Science and Technology Corporation.

1.Bilardo VJ,Curran FM,Hunt JL,Lovell NT,Maggio G,Wilhite AW,et al.The bene fits of hypersonic airbreathing launch systems for access to space.Reston:AIAA;2003.Report No.:AIAA-2003-5265.

2.Chen M,Tang HL,Zhu ZL.Goal programming for stable mode transition in tandem turbo-ramjet engines.Chin J Aeronaut 2009;22(5):486–92.

3.Chen M,Zhu ZL,Zhu DM,Zhang J,Tang HL.Performance analysis tool for turbine based combined cycle engine concept.J Astronaut 2006;27(5):854–9.

4.Trapier S,Duveau P,Deck S.Experimental study of supersonic inlet buzz.AIAA J 2006;44(10):2354–65.

5.Ohshima T,Enomoto Y,Naskanishi H,Futamura H,Yanagi R,Mitani T.Experimental approach to the HYPR Mach 5 ramjet propulsion system.Reston:AIAA;1998.Report No.:AIAA-1998-3277.

6.Cindy W,Albertson SE,Carl AT.Mach 4 test results of a duallfowpath turbine based combined cycle inlet.Reston:AIAA;2006.Report No.:AIAA-2006-8138.

7.Saunders JD,Slater JW,Dippold V,Lee J,Sanders BW,Weir LJ.Inlet mode transition screening test for a turbine-based combinedcycle propulsion system.Jannaf 2008.

8.Sanders BW,Weir LJ.Aerodynamic design of a dual- flow Mach 7 hypersonic inlet system for a turbine-based combined-cycle hypersonic propulsion system.Washington,D.C.:NASA;2008.Report No.:NASA/CR-2008-215214.

9.Dippold VF.Computational analyses of the LIMX TBCC inlet high-speed flowpath.Washington,D.C.:NASA;2012.Report No.:NASA/TM-2012-217219.

10.Foster LE,Saunders JD,Sanders BW,Weir LJ.Highlights from a Mach 4 experimental demonstration of inlet mode transition for turbine-based combined cycle hypersonic propulsion.Washington,D.C.:NASA;2012.Report No.:NASA/TM-2012-217724.

11.O’Brien TF,Davis DO,Colville JR.The advanced combined-cycle integrated inlet test program–test results.Reston:AIAA;2008.Report No.:AIAA-2008-2637.

12.Trapier S,Deck S,Duveau P,Sagaut P.Time–frequency analysis and detection ofsupersonicinletbuzz.AIAA J 2007;45(9):2273–84.

13.Lee HJ,Lee BJ,Kim SD,Jeung IS.Flow characteristics of smallsized supersonic inlets.J Propul Power 2011;27(2):306–18.

14.Tan HJ,Sun S,Yin ZL.Oscillatory flows of rectangular hypersonic inlet unstart caused by downstream mass- flow choking.J Propul Power 2009;25(1):138–47.

15.Tan HJ,Li LG,Wen YF,Zhang QF.Experimental investigation of the unstart process of a generic hypersonic inlet.AIAA J 2011;49(2):279–88.

16.Chang JT,Wang L,Bao W,Qin J.Novel oscillatory patterns of hypersonic inlet buzz.J Propul Power 2012;28(6):1214–21.

17.Jiao XL,Chang JT,Wang ZQ,Yu DR.Mechanism study on local unstart of hypersonic inlet at high Mach number.AIAA J 2015;53(10):3102–12.

18.Chang JT,Li N,Xu KJ,Bao W,Yu DR.Recent research progress on unstart mechanism,detection and control of hypersonic inlet.Prog Aerosp Sci 2017;89:1–22.

19.Li ZF,Gao WZ,Jiang HL,Yang JM.Unsteady behaviors of a hypersonic inlet caused by throttling in shock tunnel.AIAA J 2013;51(10):2485–92.

20.Lee CC,Boedicker C.Subsonic diffuser design and performance for advanced fighter aircraft.Reston:AIAA;1985.Report No.:AIAA-1985-3073.

21.Cheng DS,Tan HJ,Sun S,Tong Y.Computational study of a high-performance submerged inlet with bleeding vortex.J Aircraft 2012;49(3):853–60.

22.Wang WX,Guo RW.Numerical study of unsteady starting characteristics of a hypersonic inlet.Chin J Aeronaut 2013;26(3):563–71.

14 August 2016;revised 26 June 2017;accepted 2 August 2017

Available online 16 October 2017

?2017 Chinese Society of Aeronautics and Astronautics.Production and hosting by Elsevier Ltd.Thisis an open access article under the CCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

*Corresponding author at:College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China.

E-mail address:yuan_hch7@sina.com(H.YUAN).

Peer review under responsibility of Editorial Committee of CJA.