Acceleration of DDT by non-thermal plasma in a single-trial detonation tube

Dinfeng ZHENG,Bing WANG

aCollege of Engineering,Peking University,Beijing 100871,China

bSchool of Aerospace Engineering,Tsinghua University,Beijing 100084,China

1.Introduction

In practical applications,it is important to shorten the De flagration-to-Detonation Transition(DDT)to increase the operating frequency of Pulse Detonation Engines(PDEs).Therefore,different approaches are employed either to intensify the ignition success by high-energy injection or to increase the turbulence with spoilers inside the detonation tube,by either using shock wave interactions or a pre-detonation tube.1–5

An electric spark plug is commonly used to ignite the mixture in the detonation tube because it is easy and cheap.However,long-duration DDTs usually occur and are quite dependent on the ignition energy and position of the spark plug even for appropriate air–fuel ratios.One can realize independent DDT,but one must increase the spark plug ignition energy,for instance,to higher than 0.5 J,6,7though this is quite expensive.The dead weight of such a spark plug is also very heavy.It has been found that low-temperature plasma can enhance ignition and assist in combustion.8,9Thus,plasmaassisted flow control,ignition,and combustion have been further studied for application in PDEs.10–12

Nanosecond pulsed high-voltage discharge technology has been validated for its capability in igniting different detonation mixtures.Wang et al.13,14used the 80 kV voltage and 50 ns pulsed discharge to ignite a C2H4/air mixture and found that the ignition delay time was significantly reduced compared with that by electric spark plug ignition.Singleton et al.15,16used a 90 kV voltage and 85 ns pulsed discharge or 60 kV and 54 ns pulsed discharge to ignite a C2H4/air mixture with wide equivalent ratios and indicated that the DDT distance was much shortened.Busby et al.17employed high voltages ranging between 55 and 62 kV and 50–75 ns pulsed discharge for the equivalent premixed gasoline/air mixture in singletrial detonation experiments.Brophy and his coworkers18,19and Hackard20used nanosecond pulsed discharge ignition to achieve the coordinated operation of a PDE with an 80 Hz working frequency.Brophy et al.validated C2H4and air,while Hackard tested gasoline/air.Starikovskii21validated the enhancement of ignition and the combustion of propane using a 10–20 kV and 77 ns nanosecond pulsed discharger.The validated experiments showed that the ions released by the transient plasma could shorten the DDT.22

Both the miniaturization and low cost of high-voltage power supplies makes this goal in engineering applications more challenging in PDEs.Instead of the nanosecond pulsed high-voltage discharge approach,non-equilibrium,lowtemperature plasma can also be generated by Alternating Current(AC)-driven Dielectric Barrier Discharge(DBD)technology,such as that described in works by Rosocha et al.2,3Hu et al.,24Kim et al.25and others.9,26However,the above research was conducted to improve the combustion or startup characteristics of industrial burners.Compared with high cost of nanosecond high-voltage pulsed discharge,an AC-driven DBD lowers the requirements of the power supply and holds a better control of the gas heating.Nano-pulsed DBD technology is widely used to generate low-temperature plasma for flow control,27such as in separation control for an airfoil,compressors or even in adjusting shock waves.However,in the direct initiation of detonation in PDEs,most of the nanosecond pulsed discharge is performed without a dielectric barrier between the two electrodes.To the best knowledge of the present authors,no such research has been performed to examine the effectiveness of non-equilibrium, lowtemperature plasma generated by an AC-driven DBD on the shortening of DDT and triggering detonation combustion in PDEs.

In this paper,we pursue an experimental study on a labscale platform and validate the effectiveness of non-thermal plasma for accelerating DDT in a single-trial PDE tube,based on the novel plasma generator proposed by the present authors.Three types of detonation mixtures of hydrogen,acetylene,and propane are tested in the experiments.The flame development,propagation speed,and pressure distributions are compared in the present study.

2.Description of the experiment system

2.1.Experiment setup

A photograph of the experimental system is shown in Fig.1(a),which consists of dual detonation tubes with the spark plug and plasma igniter installed,respectively.In each tube,a gas charging system,ignition system,vacuum system,and data recording system are included.The outlet of each detonation tube is sealed with a Poly Tetra Fluoro Ethylene(PTFE) film.

The rectangular detonation tube,enlarged in the Fig.1(b),is 1500 mm in length and 60 mm×60 mm in the cross-section.There are10rectangular-ambulatoryplane-typespoilers installed inside each tube.The first one is located 150 mm from the tube head,with an interval of 60 mm between two spoilers.The total blockage ratio of the spoilers is about 43%.The front plane of the detonation tube is made of PTFE at 1450 mm×60 mm to visualize the flame development.Four inhouse built ion probes are located at equal intervals of 300 mm along the detonation tube,starting 200 mm from the tube head.Two dynamic pressure transducers are fixed on the detonation tube at 800 and 1250 mm.

The plasma generator or the electric spark plug is installed close to the head of each detonation tube.The spark plug is fixed 30 mm away from the tube head and the center of the plasma generator is also 30 mm away from the tube head.This means that the ignition kernel forms at the same position inside the two tubes.

Three types of detonation mixture are tested in the experiment,the species of which are shown in Table 1.To obtain detonation easily,additional oxygen is added to the mixture of propane and air.The mass fraction is used in the table.All the experiments are conducted under conditions of 298 K and 1 atm(1 atm=101,325 Pa).The theoretical C–J velocity of the detonation combustion is also shown in Table 1.

2.2.Low-temperature plasma generator

The con figuration of the non-thermal plasma generator is shown in Fig.2(a),consisting of a plasma igniter,AC power supply,and(Automatic Frequency Control)AFC system.

The novel plasma generator consists of(1)ceramic tube,(2)High Voltage(HV)electrode,(3)Low Voltage(LV)electrode with(4)gas holes,and(5)insulator in Fig.2(b).The HV electrode and the insulator are joined by screws,while the ceramic tube is glued to the HV electrode.Different con figurations of the plasma generator by DBD are proposed and compared.The scheme described below is speci fied and used in all the experiments.The HV electrode is 20 mm in diameter and 40 mm in length(LL),and the barrier medium is made of a corundum tube with an outer diameter of 25 mm and 50 mm in length.There are 24 gas holes in the LV electrode with a diameter of Φ=4 mm.The diameter of the LV electrode is 33 mm,which means that the discharging gap between the LV and HV electrodes is 4 mm.

The AC power can supply 0–40 kV output voltage and 30 kHz frequency sinusoidal waves.Controlled by the synchronous controller,the discharge of the plasma igniter can be continually adjusted up to a maximum frequency of 500 Hz,and the single discharging duration can be controlled from 0.1 to 1000 ms.Thus,the required discharging waveforms can be obtained by the AFC system.

Both the spark plug and plasma igniter are operated at 1 Hz in the present experiment.The work mode can be regarded as the single trigger mode,because the combustion process in the single-trial detonation tube is finished well before 1 s.

Table 1 Mixture parameters in experiments.

The electric power of the spark plug and plasma igniter can be compared to each other.The discharge duration of the spark plug is between 0.2 and 0.25 ms;but the discharge duration of the plasma igniter is controlled at 0.5 ms.The nominal power of the plasma igniter is 400 W,and thus,the energy of a single trigger is around 0.2 J.The effective energy of the single trigger spark plug is around 0.5 J.

The continuous discharge of the plasma igniter is observed in an atmospheric environment,shown in Fig.2(c).The single discharge of voltage-current is shown in Fig.3,t is time.The single pulse discharge within 0.5 ms includes 10 waves.

2.3.Data recording system

High-frequency pressure sensors(PCB®,Type 113A22 with 500 kHz response frequency)are used to measure the detonation wave pressure.Through the synchronous controller,the data from the pressure transducers and ion probes are recorded by the 16-channel NI PXI-1042Q Data Acquisition System(DAS),with a 2.5 MHz data acquisition frequency.A Redlake®(HG100K)high-speed camera,produced by Kodak,is employed to record the flame development in the dual tubes.The 30 K frames-per-second mode was selected in the experiments.

3.Results and discussion

3.1.Experimental measurements

At time t=0 ms,both the spark plug and plasma igniter were triggered synchronously.The time sequences for the flame development were recorded by the high-speed camera.The up row corresponds to the case of the spark plug tube and the down one to the plasma-igniter tube.

Taking the propane mixture as an example,as shown in Fig.4,a bright spot was captured by the camera in the spark plug tube initially at t=0.043 ms,but disappeared gradually.At t=1.390 ms,the flame had already propagated to around 0.34 m in the down tube,and the flame was very luminous;however,in the up tube,the flame kernel was still growing.Until t=2.990 ms,after the flame initialized by the plasma igniter had gone out of the tube,the flame ignited by the spark plug became brighter,similar to that of a plasma igniter at t=1.257 ms.The temporal flame sequences show that the development was much faster for the flame ignited by the plasma generator.

In observing the flame exhaust at the tube exit,a mushroom-shaped flame plume was formed,which implies a kind of rapid expansion of the supersonic flow.This occurred in both cases when the flame went out of the tube.It can be concluded that somehow,the detonating combustion had already been formed in the detonation tube,but we will validate this later using wave pressure measurements.

Very similar flame developments were observed for the other two cases,in which the flame ignited by the nonthermal plasma generator was faster than that of the spark plug.As presented,comparable ignition energy is infused in the dual-detonation tube,but the one in the plasma-igniter tube is a slightly smaller.

The plasma ignition behaves as bulk ignition,which means multiple flame kernels form and then spread in the discharge region between the HV and LV electrodes,and thus,the bulk of the high-temperature burned gas will ignite a fresh mixture.The positive effects of flame propagation will accelerate the flame in the tube.However,for the spark plug,the flame kernel is concentrated very locally.The flame kernel is usually developed slowly from the initial kernel point due to the thermal effect.

Through the DBD,multiple streamer discharges form in the gap between the LV and HV electrodes.The gas holes in the LV electrode can release the discharges and form the airflow that can adjust the ion distribution.Thus,bulk lowtemperature discharge air flows will be generated around the plasma igniter.The radicals of OH,H,and O and various other ions,and electrons,contained in the bulk discharged air will induce a rapid chemical reaction.Once the chain reaction is triggered,the combustion will accelerate.

Fig.5 compares the flame speed for three detonation mixtures.The flame speed is calculated by the temporal signal triggering the ion probes and the known location information.

The flame is rapidly accelerated because of the turbulence from the spoilers inside the dual detonation tubes.It can be seen that an/a(overdriven)detonation combustion is formed close to the exit of the tube,because the measured flame speed is comparable with the C–J theoretical speed for the hydrogen and air mixture.However,the early stage of the flame development is quite different.With the plasma igniter,the flame speed at x=0.1 m is higher than 100 m/s,but that ignited by the spark plug is far below 100 m/s.At x=0.65 m,the flame of the hydrogen mixture ignited by the plasma igniter has already reached a speed of 1898.7 m/s,which can be regarded as a detonation wave.Therefore,the ignition method can significantly influence the early-stage flame development,which significantly determines the DDT distance.

The flame-propagating trend inside the tubes for different detonation mixtures has a very similar process.However,the final flame speed measured in this study is dependent on the mixture properties.The measurements show that the DDT in the non-thermal plasma tube is shorter than that in the spark plug tube.The flame propagation is accelerated by the assistance of the non-thermal plasma ignition of more than 60%of the hydrogen mixture,90%of the acetylene mixture,and 55%of the propane mixture.It should be emphasized that the maximum acceleration of the flame propagation occurs during the DDT process.

The plasma igniter can generate a large number of active particles,high-energy electrons,and free radicals in the discharge region,which can thus induce the acceleration of a chemical element reaction.Using the AC-driven power supply,the plasma igniter can realize a high-frequency discharge.The spark plug works at low frequency but the discharge area is much smaller.However,both the spark plug and the plasma igniter can yield to the frequency requirement that in practice,is usually less than 100 Hz in PDE.

The frequency of a PDE is related to not only the working frequency of the igniter but also to the discharge energy and discharge pattern of the igniter.Considering the geometry and charge of the mixture in the PDE tubes,the operating frequency of the PDE must also be limited to the frequency of the charge and discharge of the fresh mixture.Only a single trial detonation tube is studied in the present experiment,and the flame speed can be significantly enhanced depending on different mixtures.In general,from the experimental results,the ignition delay time can be shortened by more than 50%,and even up to 90%,and correspondingly the working frequency can be double or five times that of the spark plug case.However,the operation of a PDE also depends on the charge and discharge of the fresh mixture,the frequency of which was not tested in the present experiments.Therefore,it cannot be concluded how much the frequency could be improved in a single-trial detonation tube,which is considerably different from multi-trial PDEs.

Fig.6 shows the wave pressures measured by pressure transducers at x=0.8 m and 1.25 m inside the dual tubes.It is shown that detonation waves are observed at those positions,where the pressure peak is higher in the plasma igniter tube than that in the spark plug tube.The pressures are measured by the transient pressure transducers fixed at different tube wall locations.With the same mixture,the spatiotemporal pressures along the tube are hardly changed.The experiments have good repeatability.None of the averaged treatments in the pressure measurements were collected in the present study.

However,the pressure peak is dependent on the mixture property.The oxygen-rich mixture of the propane has the highest peak pressure of the detonation waves.All the cases show that the detonation wave triggered by the plasma igniter is easier to generate inside the tube under the quiescent condition.

In the present experiment,the discharge energy of the plasma igniter is kept constant,and the geometry and the spoilers inside the tube are fixed for different mixtures.With the same initial pressure and temperature,the different pressure peaks and different locations of the pressure peaks are found.

The discharge in the plasma generator can generate large amounts of radicals such as OH,H,and O and electrons,which induce chemical reactions differently for different fuel types.For the hydrogen and air mixture,the free radicals are mainly generated as H,OH,and O by the impact of electrons on molecules of oxygen and hydrogen.For C2H2and the air mixture,the CHO,OH,and CH radicals are mainly generated in the discharge region.For the oxy-mixture of C3H8and air,a more excited atomic oxygen O(1D)is easily generated due to higher concentrations of oxygen,which induce the large chemical reaction rates.Thus,the different chemical reactions induced by the active particles will occur downstream in the detonation tube.Thus,the reactivity and heat release will be different,and the flame accelerations are diversi fied.The different pressure peaks can then be measured and their locations will vary.

3.2.Discussion of the mechanisms

The low-temperature plasma igniter has a larger discharge volume compared with the spark plug,and the mixture is ionized as active particles with a lower activation energy.For the present hydrogen-carbon fuel,the initial chain reaction can be represented as below with the normal ignition:

The reaction rate coefficients for reaction(1)are roughly related to the mixture temperature for the above induction reaction.However,with the plasma ignition,the higher energy electrons generated in the discharge process induce the dissociation reactions below with the CHsand O2in the mixture:

As we know,the ionization energy is lower for oxygen.Large amounts of O(1D),the excited atomic oxygen,will significantly contribute to the induction of a chain reaction with the other molecules or particles.Therefore,the reaction rate coefficients for reactions(2)and(3)are at least two orders higher in magnitude than that of the initial chain reaction.This is the inherent characteristic of the plasma ignition to enhance the chemical reaction rate and shorten the ignition delay time.

In addition,for the H2/air mixture,under normal temperature,the reaction below

has a reaction rate coefficient more than 40000 m3/s,which is much higher than the reaction involving O,the ground-state atomic oxygen,being about 10000 cm3/s.This is commonly regarded as the kinetic acceleration mechanism of nonthermal plasma.

The realization of the DDT must be due to the coupling of the leading shock wave with the chemical reaction heat release.As a result of the discharge of the igniter,whenever the spark plug or plasma igniter is used,the leading shock wave will be generated(or possibly developed from the compression waves)and propagate upstream and downstream.Once the plasma igniter is utilized,the large numbers of radicals,electrons,and activated particles trigger faster chemical reaction and heat is then released downstream from the discharge region.The fresh unburned mixture is compressed by the leading shock wave and can further accelerate the combustion.This is regarded as the effects of impact from the generation of plasma.However,if the spark plug is used,the chemical reaction process will be slower,and it takes time for the flame to couple with and supply heat energy to the leading shock wave.Thus,it is faster for the combustion and leading shock wave to be coupled with each other using the plasma generator,compared with the case of ignition using a conventional spark plug.

4.Conclusions

This investigation successfully conducted an experimental validation of DDT shortening and detonation combustion in the speci fied single trial dual-detonation tubes based on the novel con figuration scheme of a plasma generator proposed by the present authors using AC-driven DBD technology.

Three different types of detonation mixtures of hydrogen,acetylene,and propane were ignited in the experiments.The flame kernel formation and flame development were captured by a high-speed camera.The flame speed and pressure waves were tracked by ion probes and high-frequency pressure transducers.

The measurements show that the bulk of the flame kernels were observed in the plasma igniter tube,and the flame propagation was rapidly accelerated downstream.The detonation wave could be observed in both tubes due to the effects of spoilers,judged from the exhaust flame plume and the pressure distribution.The flame speeds were calculated by the temporal signals from the ion probes,a significant flame acceleration was found during the DDT process,and the flame propagation was enhanced by more than 50%for the three mixtures.

The mechanisms of low-temperature plasma on initiation of detonation were analyzed,and the observations of different pressure peaks and their locations for different mixtures were also explained.

Acknowledgements

The authors are grateful for the support of the National Natural Science Foundation of China(Nos.51176001 and 51676111),the Tsinghua University Initiative Scientific Research Program(No.2014Z05091)and the Importation and Development of High-Caliber Talents Project of Beijing Municipal Institutions.

References

1.Ting JM.,Bussing TRA.,Hinkey JB.Experimental characterization of the detonation properties of hydrocarbon fuels for the development of a pulse detonation engine.Reston:AIAA;1995.Report 415 No.:AIAA-1995-3154.

2.Li J,Lai WH,Chung K.Amplification factor on de flagration-todetonation transition of propane-oxygen mixtures.Reston:AIAA;2006.Report No.:AIAA-2006-0554.

3.New TH,Panicker PK,Lu FK,Tsai HM.Experimental investigations on DDT enhancements by Shchelkin spirals in a PDE.Reston:AIAA;2006.Report No.:AIAA-2006-0552.

4.Farinaccio R,Stowe RA,Harris PG.Multi-pulse detonation experiments with propane-oxygen.Reston:AIAA;2002.Report No.:AIAA-2002-4070.

5.Meyers JM,Lu FK,Wilsons DR.Performance enhancements on a pulsed detonation engine.Reston:AIAA;2003.Report No.:AIAA-2003-1173.

6.Sinibaldi JO,Brophy CM,Robinson JP.Ignition effects on de flagration-to-detonation transition in gaseous mixtures.Reston:AIAA;2000.Report No.:AIAA-2000-3590.

7.Kaneshige M,Shepherd JE.Detonation database.California:GALCIT Technical Report;1997.Report No.:FM97-8(1997).

8.Stange Sy,Kim Y,Ferreri V,Rosocha LA,Coates DM.Flame images indicating combustion enhancement by dielectric barrier discharges.IEEE Trans Plasma Sci 2005;33(2):316–7.

9.Rosocha LA,Coates DM,Platts D,Stange S.Plasma-enhanced combustion of propane using a silent discharge.Phys Plasmas 2004;11(5):2950–6.

10.Wu Y,Li YH,Min J,Pu YK.Experimental investigation into characteristics of plasma aerodynamic actuation generated by dielectric barrier discharge.Chin J Aeronaut 2010;23(1):39–45.

11.Wu Y,Li YH,Min J,Song HM,Liang H.Effect of pressure on the emission characteristics of surface dielectric barrier discharge plasma.Sens Actuators A:Phys 2013;203(1):1–5.

12.Zhao GY,Li YH,Liang H,Han MH,Wu Y.Flow separation control on swept wing with nanosecond pulse driven DBD plasma actuators.Chin J Aeronaut 2015;28(2):368–76.

13.Wang F,Liu JB,Sinibaldi J,Brophy C,Kuthi A,Jang C,et al.Transient plasma ignition of quiescent and flowing air/fuel mixtures.IEEE Trans Plasma Sci 2005;33(2):844–9.

14.Wang F,Kuthi A,Gundersen MA.Technology for transient plasma ignition for pulse detonation engines.Reston:AIAA;2005.Report No.:AIAA-2005-0951.

15.Singleton DR,Sinibaldi JO,Brophy CM,Kuthi A,Gundersen MA.Compact pulsed-power system for transient plasma ignition.IEEE Trans Plasma Sci 2009;37(12):2275–9.

16.Singleton DR,Pendleton SJ,Gundersen MA.The role of nonthermal transient plasma for enhanced flame ignition in C2H4-air.J Phys D Appl Phys 2011;44:022001.

17.Busby K,Corrigan J,Yu ST,Wiliams S,Carter C,Schauer F.Effects of corona,spark and surface discharges on ignition delay and de flagration-to-detonation times in pulsed detonation engines.Reston:AIAA;2007.Report No.:AIAA-2007-1028

18.Brophy C.Initiation improvements for hydrocarbon/air mixtures in pulse detonation applications.Reston:AIAA;2009.Report No.:AIAA-2009-1611.

19.Brophy C,Dvorak T,Dausen D.Detonation initiation improvements using swept-ramp obstacles.Reston:AIAA;2010.Report No.:AIAA-2010-1336

20.Hackard CN.Ignition characteristics for transient plasma ignition of ethylene/air and JP-10/air mixtures in a pulse detonation engine[dissertation].Monterey:Naval Postgraduate School;2007.

21.Starikovskii AY.Plasma supported combustion.Pro Comb Inst 2005;30:2405–17.

22.Sinibaldi J,Rodriguez J,Channel B,Brophy C,Wang F,Cathey C,et al.Investigation of transient plasma ignition for pulse detonation engines.Reston:AIAA;2005.Report No.:AIAA-477 2005-3774.

23.Rosocha LA,Kim Y,Anderson GK,Abbate S.Combustion enhancement using silent electrical discharges.Int J Plasma Environ Sci Technol 2007;1(1):8–13.

24.Hu H,Song Q,Xu Y,Li G,Nie C.Non-equilibrium plasma assisted combustion of low heating value fuels.J Therm Sci 2013;22(3):275–81.

25.Kim W,Mungal MG,Cappelli MA.Flame stabilization using a plasma discharge in a lifted jet flame.Reston:AIAA;2005.Report No.:AIAA-2005-0931.

26.Stange S,Kim Y,Ferreri V,Rosocha LA,Coates DM.Flame images indicating combustion enhancement by dielectric barrier discharges.IEEE Trans Plasma Sci 2005;33(2):316–7.

27.Wu Y,Li YH.Progress and outlook of plasma flow control.Acta Aeronaut et Astronaut Sin 2015;36(2):381–405[Chinese].