Taxes & Accounting

A newin situ method for the measurement of ignition delays and the propagation of the ignition wave in gun charges

Description
A newin situ method for the measurement of ignition delays and the propagation of the ignition wave in gun charges
Published
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  IEEE TRANSACTIONS ON MAGNETICS, VOL. 39, NO. 1, JANUARY 2003 275 A New  In Situ  Method for the Measurementof Ignition Delays and the Propagationof the Ignition Wave in Gun Charges David Zoler, Gabriel Appelbaum, Noam Shafir, Shoshi Roshu, Chay Goldenberg, Shlomo Wald, and Moshe Shapira  Abstract— Thephysicalandchemicalphenomenainvolvedintheballistic process, both in conventional guns and in electrothermal-chemical(ETC)guns,areverycomplex,sothatoptimizationofgunperformance requires a detailed understanding of the influence of different parameters. Theinitialstageofpropellant ignitionis crit-icaltotherestoftheprocess,intermsofbothperformance(muzzlevelocity,repeatability)andsafety.Abetterunderstandingoftheig-nition stage is thus very important in achieving improved perfor-mance. Inthis paper, we propose anew method and ignition sensoraimed at providing direct experimental information about the pro-pellant ignition process. We further report on measurements oftheignition delays, and the propagation of the “ignition wave,” per-formed using these sensors in experimental devices such as closedbombs and gun simulators.  Index Terms— Ignition delay, plasma, sensors. I. I NTRODUCTION T HE STUDY of plasma–propellant interaction, and its ef-fectonthevariousstagesoftheballisticprocessinguns,isan ongoing research program at Soreq NRC, Yavne, Israel. Thephysicalandchemicalphenomenaoccurringinthegunchamberduring the interior ballistic process are very complex. It is gen-erallyacceptedthat,atthisstageofguntechnology,onlyabetterunderstanding of these processes may lead to a substantial im-provement in performance. Thus, the ability to understand howvarious changes in parameters such as the propellant mass, typeand geometry, as well as the gun geometry and the type of igni-tion, affect the interior ballistic process is very important.The ignition of the propellant is one of the most importantstages in the ballistic process. A “proper” ignition can ensurea smooth gun operation, e.g., in minimizing the probabilityof pressure waves, which can lead to undesired or even cata-strophic results, and ensuring the reproducibility of the ballisticprocess, which affects accuracy. Improved understanding of the propellant ignition process becomes even more importantwhen trying to optimize the gun performance by increasing thecharge density or by using new, unconventional, igniting agentssuch as, for example, plasma.In this paper, we propose a new method for the experimentalmeasurement of propellant ignition delays indicating the prop-agation of the “ignition wave” inside a gun charge. ManuscriptreceivedJanuary14,2002.ThisworkwassupportedbytheIsraeliMinistry of Defense and Soreq NRC.The authors are with the Propulsion Physics Laboratory, Soreq NRC, Yavne81800, Israel (e-mail: zoler@soreq.gov.il).Digital Object Identifier 10.1109/TMAG.2002.805927 II. P REVIOUSLY  U SED  M ETHODS A number of experimental methods have already been devel-oped in order to study the propellant ignition process. Optical[1] or spectroscopic methods rely on the observation of the lightemittedbytheburningpropellantsurfaceattheonsetofignition.Electrical methods are based on the use of thin wires (“breakingwires”), which are assumed to break due to their melting whenthe propellant begins to burn, or on thermocouples, which mayindicate the time when the propellant surface reaches the “ig-nition temperature” (which is not always well-defined and maydepend on the method of ignition).These methods suffer from several shortcomings. Opticalmethods are problematic for the following reason. For conven-tional ignition, the combustion products of the igniter may notbe transparent enough to enable the observation of the onsetof ignition. For plasma ignition, where the radiation intensityemitted by the plasma is quite high in the visible range, it mayobscure the lower intensity radiation emitted by the ignitedpropellant.As for thebreaking wiremethod, the stronggas flowor collisions among propellant grains may cause wire breaking,leading to false detection of ignition. In addition, a significanttime lag may exist between the time the propellant reaches theignition temperature ( 200 C) and the time when the metallicwire reaches its melting point and breaks (at 1089 C for acopper wire). This may lead to inaccuracies in the measuredignition times. Thermocouples may be similarly vulnerable tothe mechanical stresses mentioned for the breaking wires, andalso typically provide quite poor signal-to-noise ratios (SNRs),leading to low precision. For plasma ignition, an additionaldifficulty occurs due to electrical interference from the pulsedpower system.III. P ROPOSED  M ETHOD Anewmethodfor ignitiontimemeasurementmustovercomeat least some of the problems mentioned above. We propose amethod which uses an ignition sensor based on large changesexpected to occur, at the onset of ignition, in the electrical re-sistance of a metallic wire wound around a propellant grain, asshown in Fig. 1.Theoperationofthisignitionsensorisbasedonthefollowingprinciple: the propellant grain and the wire wound on it are sub- ject to the same heating flux. However, it is assumed that theirtemperature will not increase at the same rate due to their dif-ferent thermal properties. It is expected that the propellant sur-face temperature will increase much faster than that of the wire. 0018-9464/03$17.00 © 2003 IEEE  276 IEEE TRANSACTIONS ON MAGNETICS, VOL. 39, NO. 1, JANUARY 2003 Fig. 1. Schematic drawing of the proposed ignition sensor. At the ignition onset, the hot, dense gases emitted from the sur-face of the propellant grain on which the wire is wound willstrongly increase the heat flux incident on the wire. This willsignificantly change the rate of the wire’s resistance increaseand will indicate the onset of ignition.Using this method, it would be possible, by placing a numberof such sensors at different positions inside the combustionchamber, to measure the ignition times at these points, and fromthese times to obtain a description of the spatial progress of the “ignition wave” in the combustion chamber. An additionalpossible capability is the measurement of the gas temperatureinside the chamber after steady-state combustion is reached.In order to be used as a heat (ignition) sensor, the metallicwire has to possess the following properties: 1) high resistivitywithalargeenoughtemperaturecoefficient;2)lowspecificheattoallowarapidincreaseinwiretemperatureforagivenincidentheat flux; 3) high thermal diffusivity to enable a rapid equilibra-tion of the temperature in the whole wire volume; 4) sufficientelastic range to ensure a good contact between the wire and thepropellant grain, as well as in order to withstand the mechan-ical stresses which may appear due to gas flow or collisionsamong propellant grains; and 5) high enough melting temper-ature. Parameters such as the number of turns (the wire length),the wire diameter, and the electrical current flowing through it(which is required in order to measure its resistance) should bechosen based on the following requirements. The resistance hasto be large enough in order to allow a good SNR but, on theother hand, not so high that the Joule heating due to the currentflowing through it results in a negligible heat flux to the propel-lant compared to that due to the igniting agent.In order to evaluate the proposed method, the heating processof the wire and the propellant has been studied using a modelsimilar to the solid propellant ignition model presented in [2].The main assumption of the model is that, as mentioned above,the propellant and the wire are subject to the same heating flux(this is reasonable since the wire is tightly wound around thepropellant grain). Therefore, the temperature distribution insidebothmaterials,aswellastheirsurfacetemperature,willonlyde-pend on their thermal properties. The present calculations referto plasma ignition and use an empirical correlation for the heatflux based on experimental data obtained from propellant igni-tion experiments in a closed bomb [3]. The results of the calcu-lations up to the ignition onset are shown in Fig. 2(a)–(c).The results presented suggest some conclusions. First, for thesame time interval, the heat wave penetrates much deeper intothe metallic wire than into the propellant grain (by almost anorder of magnitude) and the temperature distribution in the wireis much more homogeneous. Second, the surface temperatures Fig. 2. Calculated temperature distribution inside (a) the propellant and (b)metallic wire for five different times, and (c) the surface temperature as afunction of time, up to the onset of ignition (which occurs at            ms). reached at the same time by the two materials are verydifferent:the wire’s surface temperature strongly lags behind that of thepropellant. This lag excludes the possibility that heat flux fromthe wire itself may ignite the propellant. Thus, the theoreticalpredictions support the feasibility of the proposed method.IV. C ALIBRATION AND  P RELIMINARY  T ESTS In order to evaluate the suitability of the wire as an ignitionsensor, we performed a calibration of its resistance temperaturecoefficient, as well as preliminary ignition tests.For the calibration (which is important for the evaluation of both gas temperature and the expected resistance change at ig-nition onset), the sensor, made of a metallic wire wound on acylindrical grain made of wood, was placed in the center of anelectrical furnace whose length is substantially larger than thatof the grain, ensuring a homogeneous temperature in its envi-ronment. A source of dc electrical current was used to provide  ZOLER  et al. : MEASUREMENT OF IGNITION DELAYS AND THE PROPAGATION OF THE IGNITION WAVE IN GUN CHARGES 277 Fig. 3. The wire’s relative resistance versus temperature.Fig. 4. The wire’s relative resistance versus time during the propellant grainheating and ignition. a constant electrical current flowing through the sensor. The de-pendence of the wire’s resistance on the temperature was mea-sured during the slow cooling of the propellant from 300 C to40 C. The results are shown in Fig. 3. The dependence is linearto a high accuracy.As a preliminary test of the sensor, we replaced the woodgrain by a propellant grain. The voltage difference on the wireas a function of time was measured by an oscilloscope whilethe furnace temperature was slowly increased until the propel-lant was ignited. The time dependence of the wire’s relative re-sistance is given in Fig. 4. The two regions defining the inertheating (up to about 2.3 s; the heating is very slow so that theslope is not apparent on the scale used here), and the post-ig-nition heating are clearly delimited. During the propellant grainburning,thesensorresistance reachesamaximumvaluemorethan3.5timesitsinitialvalue, .Thepostignitionheating rateis relatively low (1.3 s until the maximum value of the resis-tance is obtained), which can be explained by the fact that theexperimentisperformedunderatmosphericpressure,sothattheburning rate is relatively low.Theresultsofthisexperimentsupporttheassumptionthatthesensor has thecapabilityof indicatingthetime of ignitiononset.V. C LOSED -B OMB  E XPERIMENTS Further tests were performed in a closed bomb, which pro-vides an environment closer to that of a gun chamber duringfiring, using both conventional and plasma ignition. Fig. 5. Schematic drawing of the closed bomb configuration used in the firstexperiment with conventional ignition.Fig. 6. Electrical measurement circuit used for the sensor.  A. Conventional Ignition We present the results of two closed-bomb experiments withconventional ignition. A schematic view of the closed-bombconfiguration used in the first experiment is given in Fig. 5. Theclosed bomb is filled with 0.61 kg of inert (polyacetal) grainsand 0.05 kg of propellant grains. The combustion gases fromthe igniter (made from a small amount of propellant which isheated and ignited by a conducting wire) ignite the propellantgrains, leading to a strong increase in the pressure. The time de-pendenceofthepressureduringtheprocessofignitionandcom-bustion is measured using two piezoelectric pressure gaugesand . The ignition sensor was positioned on the side of theclosed bomb opposite to that of the igniter. The electrical mea-surement circuit used for the sensor is shown in Fig. 6. The roleofthe15- resistoristoindicatewhenthesensorwireisbroken.The voltage difference on it, , will be zero when this occurs.Clearly, in this experiment the sensor will be heated and ig-nitedmostlybythecombustiongasesresultingfromtheburningofthe propellantgrainsplaced ontheoppositeside of theclosedbomb(Fig.5).TheresultsoftheexperimentareshowninFig.8.The heating process begins around s from the initi-ation of the igniter. As predicted by the theoretical calculations[Fig. 2(c)], during the inert heating stage the wire temperatureincreasesslowly,whichexplainsthesmallchangeofthewirere-sistance. This continues up to about s with a some-what larger rate after this time. The reason for this change isa faster increase in gas pressure (Fig. 7), which increases theheat flux incident on the ignition sensor. At s, asharp increase occurs in the wire resistance, indicating the igni-tion and burning of the sensor’s propellant grain. In the presentexperiment, the change in the wire resistance continues up tos, when the wire was broken (the voltage on the15- resistor becomes zero), probably due to melting or me-chanical breaking. Clearly, however, the lifetime of the wire islong enough to indicate the onset of ignition.  278 IEEE TRANSACTIONS ON MAGNETICS, VOL. 39, NO. 1, JANUARY 2003 Fig.7. Resultsofthefirstclosed-bombexperimentwithconventionalignition:the time dependence of the measured parameters.    is the voltage on the 15-   resistor (see Fig. 6). The arrow indicates the time of ignition onset.Fig. 8. Schematic drawing of the closed-bomb configuration used in thesecond experiment with conventional ignition. The second experiment provides even closer conditions tothose of actual firing: the ignition sensor is surrounded by pro-pellant grains (see Fig. 8). Another change in comparison to thefirstexperimentistheuseofanadditionalsensor.Onthissensor,the wire is wound on an inert grain made of polyacetal (PA).The two sensors are positioned close to each other. The purposeof the second sensor is to show that heating of the wire by thecombustion gases produced by the grain on which the sensor iswound clearly differs from the indirect heating of the sensor bythe gases produced by theburning of the surrounding propellantgrains.The results of the second experiment are presented in Fig. 9.During inert heating, the low rate of increase of the wire re-sistance for both sensors is quite similar. At a specific time,indicated by an arrow in Fig. 9, this rate for the sensor withthe real propellant grain increases sharply in comparison to thatof the inert sensor. This indicates the ignition of the propellantgrain.Notethatasignificantchangeintherateofincreaseofthewire’s resistance for the inert grain begins only after the pres-sure reaches large values, leading to a larger heat flux.  B. Plasma Ignition A closed bomb experiment was also performed with plasmaignition, where a high power plasma jet produced by an ablativecapillary electrical discharge [4]–[7] is injected into the closedbomb, igniting the propellant. The experimental setup is similartothatofthefirstexperimentwithconventionalignition(Fig.5),with the conventional igniter replaced by a plasma injector. Themeasurementsaremoredifficulttointerpretthanthose obtained Fig. 9. Results of the second closed-bomb experiment with conventionalignition: the time dependence of the measured parameters.Fig. 10. Results of a closed-bomb experiment with plasma ignition: thetime dependence of the measured parameters. The arrow indicates the time of ignition. with conventional ignition, mostly due to the electrical noiseinduced by the presence of the plasma and the pulsed powersystem, similar to that mentioned in [8].The time dependence of the pressure, as well as the(smoothed) sensor resistance is shown in Fig. 10. The igni-tion delay is significantly smaller than those obtained for aconventional ignition process (about 2–3 ms versus 15 ms).However, it is more difficult to identify the transition from inertheating to burning. These results indicate that the proposedmethod can also be used for plasma ignition, although the levelof electrical noise should be reduced in order to enable a moreaccurate evaluation of the ignition time.VI. G UN  S IMULATOR  E XPERIMENTS Furthertestingofthemethodwasperformedusingagunsim-ulator (a schematic drawing of which is given in Fig. 11). Amore detailed description of this experimental system is givenin [7], which presents a series of experiments in which the gunsimulator was used as part of a study aimed at the optimiza-tion of the plasma ignition process, with both conventional andplasma ignition. In some of these experiments, several ignitionsensors as described in the present paper were used. Here, wereport only results of one plasma ignition experiment as an ad-ditional validation test of the proposed method.The results obtained in this experiment, using plasma igni-tion,arepresentedinFig.12.Thesmallarrowsindicate thetimeof ignition onset, which was determined from the time at whicha significant change in the slope of the resistance versus time  ZOLER  et al. : MEASUREMENT OF IGNITION DELAYS AND THE PROPAGATION OF THE IGNITION WAVE IN GUN CHARGES 279 Fig. 11. Schematic view of the gun simulator.Fig. 12. Ignition sensor signals (      versus time) for a gun simulatorexperiment with plasma ignition. curve is observed. Further examples are given in [7]. As shown,the sensors can indicate the onset of ignition at different pointsin the cartridge, and the time delays between these times at dif-ferent points (notice that one of the sensors in the present ex-periment was not ignited). We note that the ignition times inthese experiments are still much larger than those obtained inrealfiring,sincemostofthesimulatorisloadedwithinertgrainsandtherateofpressureincreaseisrelativelysmall.However,theresults arestill useful for the qualitativeevaluationof the effectsof different parameters on the ignition process [7]. Further ex-periments are required in order to examine the applicability of the method (in particular, the time resolution) under firing con-ditions.VII. C ONCLUSION A new method for experimental evaluation of ignition timesand the spatial progress of the ignition wave through a propel-lant bed is presented. The method was validated in several ex-periments, gradually approaching the conditions in actual firingconditions. It was found that the ignition sensor is capable of indicating the time of the onset of propellant ignition.Futureworkincludesthereductionofthenoise,whichaffectstheprecisionespeciallyforplasmaignition,aswellasthetestingof the method under conditions which are even closer to thosemet in actual firings.R EFERENCES[1] K. K. Kuo  et al. , “Flame-spreading phenomena in the fin-slot regionof a solid rocket motor,” in  Proc. 29th AIAA Joint Propulsion Conf. ,Monterey, CA, 1993.[2] R. Alimi, C. Goldenberg, L. Perelmutter, D. Melnik, and D. Zoler,“A solid phase model for plasma ignition of a solid propellant,”in  Fourth International Symposium on Special Topics in ChemicalPropulsion . New York: Begell House, 1997, pp. 668–678.[3] M. Sudai, “The interaction between a plasma beam and a solid propel-lant in constant and variable volume systems,” M.Sc. thesis, Tel-AvivUniversity, Tel-Aviv, Israel, 1999.[4] Z. Kaplan, D. Saphier, D. Melnik, Z. Gorelik, J. Ashkenazy, M. Sudai,D. Kimhe, M. Melnik, M. Smith, and A. Juhasz, “Electrothermal aug-mentation of a solid propellant launcher,”  IEEE Trans. Magn. , vol. 29,pp. 573–578, Jan. 1993.[5] D. Zoler, S. Cuperman, J. Ashkenazi, M. Caner, and Z. Kaplan, “A timedependent model for high pressure discharges in narrrow ablative capil-laries,”  J. Plasma Phys. , vol. 50, pp. 51–70, 1993.[6] W. Morreli and W. F. Oberle, “Electrothermal-chemical gun propulsionin the United States,” in  Proc. 5th Int. Gun Propellant and PropulsionSymp. , Nov. 1991.[7] C. Goldenberg, D. Zoler, N. Shafir, S. Roshu, S. Wald, and M. Shapira,“Plasma–propellant interaction at low plasma energies in ETC guns,”  IEEE Trans. Magn. , vol. 39, pp. TK–TK, Jan. 2003.[8] J.D.PowellandA.E.Zielinski,“Theoryandexperimentforanablating-capillary discharge and application to electrothermal-chemical guns,”,Aberdeen Proving Ground, MD, Tech. Rep. BRL-TR-3355, June 1992.
Search
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks