low pressure combustion of composite propellant

8
Propellants, Explosives, Pyrotechnics9,149-156 (1984) 149 Low Pressure Combustion of Composite Propellant H. F. R. Schiiyer Department of Aerospace Engineering, Delft University of Technology, NL-2629 HS Delft (The Netherlands) P. A. 0. G. Korting Prins Maurits Laboratory, Netherlands Organization for Applied Scientific Research, NL-2280 AA Rijswijk ( e Netherlands) Niederdruckverbrennung von Cornposit-Treibstoffen Der VerbrennungsprozeD von zwei Satzen eines neuen Composit- Treibstofftypswurde untersucht im Druckbereich zwischen 5 kPa und 915 kPa absolut. Erne neuentwickelte Technik zur gleichzeitigen Be- sthung der Brenngeschwindigkeit, des Verbrennungswir- kungsgrades und der Empfindlichkeit des Treibstoffs fir Verbren- nungsinstabilitat wird zusammen mit einem speziell entwickelten datenreduzierenden Kode beschrieben. Trotz des niedrigen Druckes wiihrend der Versuche konnte kein niedrigeres Deflagrationslimit nachgewiesen werden, aber es zeigte sich, daS ein Gesetz fiir die Brenngeschwindigkeitnach Vieille anwendbar ist fiir den ganzen frag- lichen Druckbereich. Der groDe Unterschied im Brennverhalten konnte darauf zuriickgefiihrt werden, daS die beiden Satze hergestellt wurden mit Ammoniumperchlorat von verschiedenen Lieferfirmen. Der Verbrennungswirkungsgrad war schlecht und die Verbrennungs- instabilitat trat im ganzen Druckbereich auf. Combustion ir basse pression de propergols composites On a examine le comportement B la combustion de deux lots d’un nouveau type de propergol composite dans une gamme de pression comprise entre 5 kPa et 915 kPa. On dkcrit une nouvelle technique mise au point pour determiner B la foi la vitesse de combustion, le rendement de la reaction de combustion et la sensibfit6 du propergol aux effets d’instabilitt lors de la combustion, ainsi qu’un code de calcul dkveloppk specialement pour la reduction des donn6es. Malgrt la basse pression utiliske pour les experiences, il n’a pas Btt possible de deceler un seuil de dkflagration plus bas mais on a pu montrer qu’une loi de combustion de Vieille s’applique B l’ensemble de la gamme de pression qui intbresse. La grande difference constatee du point de vue comportement 21 la combustion des deux lots de propergol peut &re attribute au fait que le perchlorate &ammonium provenait de deux fournisseurs diffkrents. Le rendement de la reaction de combustion etait faible alors que des instabilith de combustion se manifestaient sur l’ensemble de la gamme de pression. Summary The combustion behaviour of two batches of a new type of compos- ite propellant was investigated in the pressure range between 5 kPa and 915 kPa absolute. A newly developed technique to determine simultaneously the burning rate, combustion efficiency, and the pro- pellant’s sensitivity for combustion instability is described together with a specially developed data reduction code. Notwithstanding the low pressure during the experiments no lower deflagration limit could be traced but a Vieille type burning rate law appeared to be applicable for the whole pressure range of interest. A large difference found in the combustion behaviour could be attributed to the fact that the two batches had been manufactured with Ap from different suppliers. Combustion efficiencywas poor while combustion instability occurred in the whole pressure range. Nomenclature A a C* CP F i k L M Ma mP n P RJ Rb r T t i Cross-sectional V velocity of sound X characteristic velocity Y specific heat at constant pressure A e w ( )’ wave number eigenfrequency (Y length (Y molar mass ( 11 Mach number ( 12 propellant mass ( If area V r burning rate exponent ( )tt pressure universal gasconstant burning rate temperature time response function ( )xx volume velocity longitudinal coordinate ratio of specific heats small increment density function, see Eq. (2) small perturbation imaginary part real part tube system surge tank flame second partial derivative wrt time second partial derivative wrt distance 1. Introduction The combustion behaviour of a propellant is of prime importance for its applicability. Therefore it is mandatory to establish the combustion behaviour experimentally in an effi- cient manner. A method is described where the burning rate characteristics, the combustion efficiency and the sensitivity for combustion instability of a propellant may be established during one series of testruns. This technique requires a sophisticated data reduction and analysis. To this end a special code has been developed. Therefore, the main features of this code are also discussed in this paper. Subsequently the results obtained by this testing technique and code are presented. 2. Experimental Methods and Apparatus For the simultaneous determination of the combustion behaviour of a propellant, i.e. burning rate, combustion effi- ciency and sensitivity for combustion instability, a Crawford bomb or a strandburner is less suited. Therefore another approach has been followed, which also allowed for measure- ments at subatmospheric pressures. This technique consists of burning a disk of propellant in a combustion chamber connected to a large surge tank that can be brought to any desired pressure level. During combustion the pressure in the system hardly changes, as the amount of propellant is small as compared to the volume of the system, while the large heat capacity of the system acts as a heat sink keeping the temperature of the gases in the surge tank at about ambient level. The combustion chamber may be operated in two different modes: - with a nozzle to achieve chamber pressures different from - without a nozzle so that the chamber pressure essentially 0721-3115/84/0510-0149$02.50/0 the surge tank pressure equals the pressure in the surge tank. 0 Verlag Chemie GmbH, D-6940 Weinheim, 1984

Upload: h-f-r-schoeyer

Post on 06-Jul-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Low Pressure Combustion of Composite Propellant

Propellants, Explosives, Pyrotechnics 9,149-156 (1984) 149

Low Pressure Combustion of Composite Propellant H. F. R. Schiiyer

Department of Aerospace Engineering, Delft University of Technology, NL-2629 HS Delft (The Netherlands)

P. A. 0. G. Korting

Prins Maurits Laboratory, Netherlands Organization for Applied Scientific Research, NL-2280 AA Rijswijk ( e Netherlands)

Niederdruckverbrennung von Cornposit-Treibstoffen Der VerbrennungsprozeD von zwei Satzen eines neuen Composit-

Treibstofftyps wurde untersucht im Druckbereich zwischen 5 kPa und 915 kPa absolut. Erne neuentwickelte Technik zur gleichzeitigen Be- s t h u n g der Brenngeschwindigkeit, des Verbrennungswir- kungsgrades und der Empfindlichkeit des Treibstoffs f i r Verbren- nungsinstabilitat wird zusammen mit einem speziell entwickelten datenreduzierenden Kode beschrieben. Trotz des niedrigen Druckes wiihrend der Versuche konnte kein niedrigeres Deflagrationslimit nachgewiesen werden, aber es zeigte sich, daS ein Gesetz fiir die Brenngeschwindigkeit nach Vieille anwendbar ist fiir den ganzen frag- lichen Druckbereich. Der groDe Unterschied im Brennverhalten konnte darauf zuriickgefiihrt werden, daS die beiden Satze hergestellt wurden mit Ammoniumperchlorat von verschiedenen Lieferfirmen. Der Verbrennungswirkungsgrad war schlecht und die Verbrennungs- instabilitat trat im ganzen Druckbereich auf.

Combustion ir basse pression de propergols composites On a examine le comportement B la combustion de deux lots d’un nouveau type de propergol composite dans une gamme de pression comprise entre 5 kPa et 915 kPa. On dkcrit une nouvelle technique mise au point pour determiner B la foi la vitesse de combustion, le rendement de la reaction de combustion et la sensibfit6 du propergol aux effets d’instabilitt lors de la combustion, ainsi qu’un code de calcul dkveloppk specialement pour la reduction des donn6es. Malgrt la basse pression utiliske pour les experiences, il n’a pas Btt possible de deceler un seuil de dkflagration plus bas mais on a pu montrer qu’une loi de combustion de Vieille s’applique B l’ensemble de la gamme de pression qui intbresse. La grande difference constatee du point de vue comportement 21 la combustion des deux lots de propergol peut &re attribute au fait que le perchlorate &ammonium provenait de deux fournisseurs diffkrents. Le rendement de la reaction de combustion etait faible alors que des instabilith de combustion se manifestaient sur l’ensemble de la gamme de pression.

Summary

The combustion behaviour of two batches of a new type of compos- ite propellant was investigated in the pressure range between 5 kPa and 915 kPa absolute. A newly developed technique to determine simultaneously the burning rate, combustion efficiency, and the pro- pellant’s sensitivity for combustion instability is described together with a specially developed data reduction code. Notwithstanding the low pressure during the experiments no lower deflagration limit could be traced but a Vieille type burning rate law appeared to be applicable for the whole pressure range of interest. A large difference found in the combustion behaviour could be attributed to the fact that the two batches had been manufactured with Ap from different suppliers. Combustion efficiency was poor while combustion instability occurred in the whole pressure range.

Nomenclature

A

a C*

CP

F i

k L M Ma mP n P RJ R b r T t

i

Cross-sectional V

velocity of sound X characteristic velocity Y specific heat at constant pressure A

e w ( )’ wave number

eigenfrequency ( Y length ( Y molar mass ( 11

Mach number ( 12

propellant mass ( If

area V

r

burning rate exponent ( )tt pressure universal gasconstant

burning rate temperature time

response function ( )xx

volume velocity longitudinal coordinate ratio of specific heats small increment density function, see Eq. (2) small perturbation imaginary part real part tube system surge tank flame second partial derivative wrt time second partial derivative wrt distance

1. Introduction

The combustion behaviour of a propellant is of prime importance for its applicability. Therefore it is mandatory to establish the combustion behaviour experimentally in an effi- cient manner. A method is described where the burning rate characteristics, the combustion efficiency and the sensitivity for combustion instability of a propellant may be established during one series of testruns.

This technique requires a sophisticated data reduction and analysis. To this end a special code has been developed. Therefore, the main features of this code are also discussed in this paper. Subsequently the results obtained by this testing technique and code are presented.

2. Experimental Methods and Apparatus

For the simultaneous determination of the combustion behaviour of a propellant, i.e. burning rate, combustion effi- ciency and sensitivity for combustion instability, a Crawford bomb or a strandburner is less suited. Therefore another approach has been followed, which also allowed for measure- ments at subatmospheric pressures.

This technique consists of burning a disk of propellant in a combustion chamber connected to a large surge tank that can be brought to any desired pressure level. During combustion the pressure in the system hardly changes, as the amount of propellant is small as compared to the volume of the system, while the large heat capacity of the system acts as a heat sink keeping the temperature of the gases in the surge tank at about ambient level. The combustion chamber may be operated in two different modes: - with a nozzle to achieve chamber pressures different from

- without a nozzle so that the chamber pressure essentially

0721-3115/84/0510-0149$02.50/0

the surge tank pressure

equals the pressure in the surge tank.

0 Verlag Chemie GmbH, D-6940 Weinheim, 1984

T

ftp
Page 2: Low Pressure Combustion of Composite Propellant

150 H. F. R. Schoyer and P. A. 0. G. Korting

time construction of successive extremes

__c

+ + + - - - + + + - - - . + - - - + * + - - - 4

. . . . . . . . . . . . . . . . . . . . . . .

Propellants, Explosives, Pyrotechnics 9,149-156 (1984)

combustion chamber to mercury manometer

I I to surge tank, N 1 m3

propellant ' i interface section T to vacuum pump or compressor

@ pressure transducer

valve

Figure 1. Schematic of the experimental set-up for determining pro- pellant combustion characteristics at subatmospheric pressures.

laraer I 10 % pmax

t ime - Figure 2. Computerized determination of the instant of ignition or extinguishment.

During most of the experiments reported here the combustion chamber has been operated in the second mode. A schematic of the system is shown in Fig. 1.

Except for the pressure in the surge tank pressures have been measured continuously during an experiment. The loca- tions where pressures have been measured are indicated in Fig. 1. The pressure in the combustion chamber has been measured by piezoelectric transducers (Kistler and Sund- strand). Pressure measurements in the tube connecting the combustion chamber with the surge tank indicated that there were no pressure gradients due to boundary layer effects.

Finally the pressure in the surge tank has been obtained from a mercury manometer before and after an experiment.

The electric pressure signals have been recorded on an analogue tape recorder and a W oscillograph recorder. The manometer readings yielded the absolute values of the initial and final pressure in the surge tank.

Ignition of the propellant was achieved by a hot wire tech- nique('l'). To this end the surface of the propellant disk was coated with a compatible pyrotechnique lacquer in which the hot wire was embedded.

3. Data Reduction

The analogue recordings of the pressure signals have been digitized and subsequently computer processed by a specially developed computer code(', '). This code determines the

instants of ignition and extinguishment, the mean pressure level during a testrun, the pressure amplitudes and frequencies of the oscillations if these occur and finally it calculates the amount of propellant that has been consumed according to a Vieille burning rate law.

Before and after a testrun, the pressure hardly changes. In the region where the pressure initially (finally) is about con- stant to the location where the pressure is about 10% of the maximum pressure, the computer code determines the loca- tion where four times in succession the pressure rise (decrease) during a fixed time interval exceeds the previous pressure rise (decrease). This location is defined as the instant of ignition (extinction (see Fig.2). The mean pressure during a testrun is obtained by integrating the pressure history between ignition and extinction and dividing by the burning time.

If oscillations occur the computer code determines amplitudes and frequencies in a digital way i.e. the subsequent pressures are compared. If a pressure is higher than or equal to the previous pressure, this pressure is recorded as a maximum (+), otherwise as a minimum (-); after processing the whole pressure signal one is hence left with a sequence of successive extremes. Now, like extremes are removed except for the last one. So one has obtained the various maxima and minima in the pressure history (see Fig. 3). Due to noise, however, it is possible that the difference between a maximum and a minimum is smaller than the noise level. In that case such extremes are also removed. The code uses two noise levels; an absolute noise level which usually is dominant at low pressures and a relative noise level (signal to noise ratio) which domi- nates at the higher pressures (see Fig.4). The difference between a successive maximum and minimum is defined as the double oscillatory amplitude while the inverse of the time interval is double the frequency of the oscillation. The instant that is related to a frequency or amplitude is defined as the time of the corresponding maximum (see Fig. 3).

The code has shown to be extremely versatile and very fast in determining frequencies and amplitudes as only numbers have to be compared and no calculations have to be performed in contrast to the classical Fourier analysis.

Arbitrary Osci I lotor y Signal

Figure 3. Computerized determination of extremes, amplitudes, and frequencies.

Page 3: Low Pressure Combustion of Composite Propellant

Propellants, Explosives, Pyrotechnics 9,149-156 (1984) Low Pressure Combustion of Composite Propellant 151

By specifying a “Vieille” burning rate law which then is integrated one obtains the amount of propellant consumed at every instant. By comparing the calculated amount of propel- lant consumed at burnout with the actual consumed propellant one immediately gets an impression of the reliability of the burning rate law applied.

4. Propellant Description

The experiments concern a composite propellant that has been manufactured(’) by the Fraunhofer Institut fiir Treib- und Explosivstoffe, ICT. Its special feature is that it is able to burn in a controlled way at subatmospheric pressures. Its composi- tion is given in Table 1 whereas its theoretical Derformance is

- 0

ul m .-

Absolute Noise Level t

- time

Figure 4. Noise levels, indicated schematically by a sinusoidal signal and the absolute and relative noise levels for filtering a signal.

Table 1. Propellant Composition

Percentage (by weight) according to Constituant

Manufacturer Chemical Analysis

Ammonium NHaClOd 66 65.2

presented in Table 2. From the latter Table it’is evident that the pressure in the region of interest hardly affects the theoret- ical performance of the propellant. The adiabatic flame tem- perature is low due to the presence of nitroguanidine. The physical data of the propellant are given in Table 3. Except for the relatively low heat capacity the propellant does not distin- guish itself from other propellants; the thermal diffusivity has a normal value.

The propellant had been manufactured in two different batches at different times. The first batch was received in cups with an outer diameter of 5 cm; the second batch was pressed in cups with an outer diameter of 10 cm. The thickness of the propellant disks was approximately 1 cm. The cup served as an inhibitor during combustion so that a cigarette burning grain resulted. Although some disks showed contamination with traces of boron, no effect of this contamination could be observed during the experiments.

5. Steady State Combustion Perchlorate Nitroguanidine CH4N402 13 11.2 A total number of 62 5-cm-diameter propellant disks have Binder* (C4H6)” 21 23.6 been investigated at mean pressures between 5 kPa and 915

kPa absolute, while 13 10-cm-diameter propellant disks have * Styrenemutadiene Copolymer. been tested at mean pressures between 24 kPa and 130 kPa.

The initial temperature for all propellants was 17°C.

Table 2. Theoretical Performance according to Chemical Equilibrium Calculations

Thermodynamic Data at 101.33 kPa at 25.33 kPa

Tf [KI 1779 1777 e [kg/m31 0.1363 0.0341 M [kgkmol] 19.89 19.88 cp [JflrgKl 1940 1961 Y [-I 1.275 1.271 c* [ d s ] 1299 1299

Mole Fractions Combustion Products

5.1 Lower deflagration limit

The first remarkable result was that no lower deflagration limit could be detected. In the literature it is often argued that any propellant should exhibit a lower deflagration limit. For the first batch of propellant this implies that if the propellant has a lower deflagration limit, it must be lower than 6 kPa. Below this pressure we were not able to ignite the propellant, but it could not be established whether this is due to the igni- tion method or whether this is an inherent propellant prop- erty. For the second batch of propellant it has not been tried to ignite the propellant below 20 kPa.

0.3006 0.0331 0.1117 0.3191 0.1298 0.1057

0.3005 0.0331 0.1116 0.3190 0.1298 0.1056

5.2 Burning rate

It has been found that the first batch of propellant displays a Vieille burning rate law with a break in the slope at about 100 kPa:

Table 3. Physical Properties of Propellant

Density 1614 k 11 w m 3 1 Thermal conductivity at 297 K 0.21 f 0.01 [WdKI Heat Capacity at 293 K 743 f 50 [ J h K l Heat Capacity at 333 K 821 f 50 IJ/kgKl Thermal Diffusivity at 295 K (1.75 f 0.075) X lo-’ [m%]

0.005 0.09825 MPa < p:

where r is in mmls and p in MPa. The results have been plotted in Fig. 5.

The second batch, although manufactured according to the same specifications, displayed quite a different character. The only difference in its composition was that the AP had been obtained from a different supplier. It was found that above

MPa < p 5 0.09825 MPa: r = 4.6863 po.6976 r = 2.7650 p0.4702

Page 4: Low Pressure Combustion of Composite Propellant

152 H. F. R. Schoyer and P. A. 0. G. Korting

2 - 2 E E - 1 -

k a6- L

.w-

= 0.L-

E ; 0.2-

m C -

t 0.1

Propellants, Explosives, Pyrotechnics 9,149-156 (1984)

Second Batch of Propellant p 3 0.03 MPa : r = 21465 ~ 0 . ~ 5 ) '

,/ I

I I

I I

I I

I

4

~

- Y) . E 2.0 E -

Figure 7. The combustion efficiency based on the ratio of the measured

L

; 1.0 - 0

0,

a 0.E

E 0.4 m

.-

First Batch of Propellant

p 0.09825 Mlb : r = 4.6863 p0'6976 p w 0.09825 MPa : r: 2.7650 po.L702

0.001 0.004 0.01 Q.0L 0.1 0.4 1

Pressure. p lMPal

Figure 5. The propellant burning rate for the first batch of propellant (5-cm-dim. propellant disks).

0.01 0.04 0.1 0.4 1 - Pressure p IMPal

Figure 6. The propellant burning rate for the second batch of propel- lant (10-cm-dim. propellant disks).

1

-0 e! M v n a" 0.75 \ U

3 2

i t 2 0.5

OZ!

0

30 kPa the propellant burning rate again followed a Vieille burning rate law:

p 2 0.03 MPa: r = 2.1465 p0.4531

where r is in m d s and p in MPa. Below 30 kPa a burning rate exponent n = 5 was obtained

which is an extremely high value. However this information is only based on three data points and therefore may be less reliable. The results have been plotted in Fig. 6.

It is seen that the burning rate exponent n = 0.47 for the first batch and n = 0.45 for the second batch agree fairly well, but the pressure region for which they hold is quite different.

5.3. Combustion efjkiency

During nearly all testruns unburned propellant particles (oxidizer as well as binder) have been found in the system after a testrun. This was a significant indication that the combustion efficiency may not have been ideal. To obtain a more quantita- tive measure of the combustion efficiency one may compare the actual pressure rise in the system with the theoretically expected pressure rise. This ratio serves as a measure for the combustion efficiency. The theoretically expected pressure rise is obtained from

where T is taken to be ambient temperature. The reasoning for this is that the system has a very large heat capacity so that after a short time the combustion gases have cooled to ambient temperature. The results, in relation to the mean pressure, are shown in the Figs. 7 and 8 for the two batches.

For the first batch it is seen that the maximum combustion efficiency tends to increase with mean pressure. Also the scat-

5 cm propellant disks; first batch

. . . . . . * . . . .... * . . . .. ..

1

I and expected rise in vessel pressure for the first batch of propellant.

0.001 0.01 0.1 w Mean pressure IMPal

Page 5: Low Pressure Combustion of Composite Propellant

Propellants, Explosives, Pyrotechnics 9, 149-156 (1984) Low Pressure Combustion of Composite Propellant 153

1

-8 .+ W

0,

a

3 0.75 -. P E"

a" 0.5

vf

t 0.25

10cmpropellont disks ; second botch

-

- . .

.

0 1 0M)l a01 0.1 1 expected rise in vessel pressure for the

4 Mean mssure IMP01 second batch of nmnellant.

a

0 3

ul

L

a'

0

1.035C

1.0325

0300

Figure 8. The combustion efficiency based on the ratio of the measured and

Figure 9. A typical example of oscillatory burning (com- bustion instability) during

0 10 15 20 25 30 35 subatmosDheric burning of --D Time [sl

---+ Time[sl

propellant (first batch). -

Figure 10. A typical example of the combination of dpldt- extinguishment and chuffing (combustion instability) dur- ing the burning of propellant (first batch).

)

Page 6: Low Pressure Combustion of Composite Propellant

154 H. F. R. Schoyer and P. A. 0. G. Korting

Combustion chamber, $0.05

A P

Propellants, Explosives, Pyrotechnics 9,149-156 (1984)

Surge Tank (d 0 513

ter seems to increase with mean pressure. On the average a combustion efficiency between 50% and 75% is achieved.

For the second batch of propellant there is a strong tendency of increasing combustion efficiency with mean pressure while the scatter seems to remain the same. Again, a combustion efficiency between 50% and 75% is achieved.

T (I) \ A 1

Connecting Tube d 0 053

t

6. Combustion Instability

@ A 2

w

6.1. General observations

L. 2

During the experimental program, the following forms of combustion instability have been observed: - oscillatory combustion i.e. large amplitude oscillations with

very specific frequencies are superimposed on the mean pressure signal. Fig. 9 shows an example of such oscillatory combustion. Due to the compressed time scale one cannot discern the individual oscillations.

- combination of chuffing and dpldt extinguishment. Fig. 10 is an example of this phenomenon. During oscillatory combustion it is often observed that the

amplitudes grow exponentially, then reach a limiting, about constant, amplitude and at or near burnout decay exponen- tially. This exponential growth and decay allow for the deter- mination of the propellant response function.

If oscillatory combustion takes place the frequencies of the oscillation have to agree with the eigenfrequencies of the pro- pellant combustion mechanism and with the eigenfrequencies of the acoustic environment. For example, if a propellant has a strong tendency to generate oscillatory combustion at fre- quency F1 while the acoustic environment only allows for oscil- lations at frequencies far away from F1 one will hardly observe oscillatory combustion in such a system. On the other hand, if one of the acoustic eigenfrequencies and the frequency, F1, match, severe oscillations may be observed. As the theory of combustion instability has not been developed far enough to predict reliably the propellant eigenfrequency , F1, one usually investigates the sensitivity for oscillatory combustion in an acoustic environment with known acoustic eigenfrequencies. This method forms the basis for T-burned6) and Helmholtz resonator-burner(’) work. The analysis of combustion instabil- ity by means of an L*-burner(*) is of a quite different nature(g) and very much related to chuffing and dp/dt extinguishment.

L.9

6.2. Acoustic analysis

From the above it is obvious that the acoustic environment has to be well defined. As we have only observed low fre- quency (up to 70 Hz) oscillations a one dimensional analysis suffices.

A schematic of the system is given in Fig. 11. The acoustic wave equation may be written as:

a2%x - vtt = 0 (2)

where $ represents either the velocity or pressure perturba- tion, v’ and p’ respectively. The boundary conditions are:

x = o x = L

VI = 0 V’ = 0

(3) (4)

In addition we have from the continuity equation:

vi A1 = V; A2 (5 )

Table 4. Eigenfrequencies of the System for Velocities of Sound 350 m/s and 400 m/s

Type I I1 111 IV

Frequency [Hz] first harmonic 19.2-22.0 9.6-11.0 17.9-20.4 20.8-23.8 F1 second harmonic 38.5-44.0 28.8-33.0 53.6-61.2 62.5-71.4 F2

The solutiod2) of Eq (2) together with the boundary condi- tions, Eqs. (3) and (4), and the matching condition of the tubes with a different cross-section, Eq. (5 ) , yield the following eigenvalues:

2 j - 1 a 4 L

FIIj = - * - 2 j - 1 a Fir~j = - * - 4 L - L 1 2 j - 1 a 4 L1

F, =-.-

j = 1 , 2 , 3 , ...

The indices I through IV are associated with the various wave shapes. Solutions type I and type I1 are associated with a longitudinal wave where the wave length is related to the total system length L. Solutions type I11 are associated with a lon- gitudinal wave where the wave length is related to the length of the surge tank, while solutions type IV concern longitudinal waves where the wave length is related to the length of the tubing system.

During combustion the gases will cool rapidly as the system has a large heat capacity. On the other hand the analysis has been made assuming a uniform temperature distribution in the system. The authors are well aware that this assumption is highly questionable. Moreover during combustion a gradual rise in the mean temperature may be expected. Therefore we estimate the velocity of sound somewhere between 350 m/s and 400 m/s. The corresponding frequencies of the first two harmonics of the various wave forms are given in Table 4.

Page 7: Low Pressure Combustion of Composite Propellant

Propellants, Explosives, Pyrotechnics 9,149-156 (1984)

1001 I

80. T =O.m M R 5 cm 0 propellant disks

::k 0

0 10 20 30 40 50 €0

Low Pressure Combustion of Composite Propellant

100- 1 80-

p = 0,018 MPa 5 c m l propellant disks

A -

0 10 20 30 10 50 60 t

t t t t t t t t - Frequency [Hzl

I"".

0 10 20 30 10 50 60 '

1 1 1 t + t * - Frequency lHzl

= 0,016 MPa 5 cm$ propellant disks

u)

%

t 0 10 20 30 40 50 60 70

4- .-

I t t t t t t t + - Frequency [Hzl - Frequency [Hzl

4 l t t t t

155

Figure 12. The frequency spectra of the first batch of propellant at four different mean pressures. The intensities are related to the maximum intensity occuring at the particular mean pressure.

6.3. Correlation with experimental results

Frequency spectra have been obtained for various cases where oscillatory combustion has been observed. It was gener- ally found that the first batch of propellant would generate oscillations up to 70 Hz. The second batch of propellant, although the only difference in its composition being AP from a different supplier, only generated oscilla$ions below 30 Hz. Another difference with the second batch of propellant was that those all concerned 10-cm-diameter disks versus 5-cm- diameter disks for the first batch. It is not clear whether the different nature of the frequency spectrum is due to the com- position or geometry changes. Another interesting feature that has been noticed with the 5-cm-diameter propellant disks is that the character of the spectra is strongly affected by the mean pressure during combustion. This is illustrated in Fig. 12 which shows the frequency spectra for the first batch of propel- lant at four different mean pressures. It is also seen that the frequencies of the oscillations agree fairly well with the calcu- lated acoustic eigenfrequencies as given in Table 4.

6.4 Response function

If oscillatory combustion takes place it follows from linear theory(",') that the pressure fluctuation during growth of the oscillations varies as:

p' = exp{(ik' - ki)t} (7)

where k is given by

k = kT + ik' = {j n + y RL Ma - i(y RE, - l)Ma}/(L/a) (8)

It therefore follows that the response function strongly deter- mines the ability of a propellant to generate oscillatory com- bustion. By plotting, on a logarithmic scale, the pressure amplitude versus time as it may be determined from a testrun one obtains the imaginary part of the frequency k':

(9)

and vice versa the real part of the response function is found experimentally from

RL= -[-+1] 1 -k' y 2FMa

It may be shown that at zero frequency RE, = n, while for F + CQ , Rf, + 0. So in fact any value of Rf, > n expresses that the propellant has a tendency for oscillatory combustion. The imaginary part of the response function causes a minute shift

Page 8: Low Pressure Combustion of Composite Propellant

156 H. F. R. Schoyer and P. A. 0. G. Korting Propellants, Explosives, Pyrotechnics 9, 149-156 (1984)

in the frequency and therefore cannot be determined from this experimental program.

The real part of the response function has been determined for all cases where oscillatory combustion with exponentially growing amplitudes has been observed. To obtain the imagi- nary part of the eigenfrequency k’ the exponential growth and decay rates were determined. The decay is a result of losses due to heat transfer and viscosity. Assuming the dampening to be the same all over a testrun, k’ follows from adding the observed decay constant to the observed growth constant. It was found that the real part of the response function varied with frequency and mean pressure. There is a tendency for Ri to decrease with increasing pressures and increasing frequen- cies. The results are shown in Fig. 13.

7. Conclusions

During testing of two batches of composite propellant at subatmospheric pressures no lower deflagration limit could be established. If lower deflagration limits exist for the propel- lants tested they must lie below 6 kPa for the first batch and 20 kPa for the second batch of propellant.

It has been established that the Vieille burning rate law for these propellants may be extended to extremely low pressures.

A minute change in composition between the two batches, i.e. AP was obtained from two different suppliers, caused a substantial difference in combustion behaviour .

The combustion efficiency of both propellants was poor i.e. between 50% and 75%.

Oscillatory combustion has been observed in the whole pressure region investigated. The preferred oscillatory fre- quencies were low i.e. below 70 Hz.

2 I) 1 0

I

8. References

(1) R. S. de Boer and H. F. R. Schoyer, “Results of L*-Instability Experiments with Double Base Rocket Propellants”, Delft Uni- versity of Technology Dept. of Aerospace Eng./Technological Laboratory TNO, Report LR-224/TL-R 3050-1, DelftRijswijk, 1976.

(2) H. F. R. Schoyer and P. A. 0. G. Korting, “Low Pressure Com- bustion Characteristics of Composite Propellant”, Delft Univer- sity of Technology Dept. of Aerospace Eng./Prins Maurits Laboratory TNO, Report LR 367PML 1982-165, DelftRijswijk, 1982.

(3) F. Bebelaar, B. A. C. Ambrosius, and H. F. R. Schoyer, “Digi- tal Data Reduction Algorithm for Oscillatory Signals with an Application to L*-Instability”, Delft University of Technology Dept. of Aerospace Eng., Report LR-298, Delft, 1980.

(4) F. Bebelaar and H. F. R. Schoyer, AZAA J. 19 (11), 1498-1500 (1981).

(5) W. Klohn and A. Rassinfosse, Proceedings Infernationale Jahres- tagung ZCT, 1981, Karlsruhe, pp. 289-296.

(6) F. E. C. Culick et al., “T-Burner Manual”, CPIA Publication No. 191, Chemical Propulsion Information Agency, Silver Spring, 1969.

(7) H. F. R. Schoyer, J . Spacecr. Rockets 19 (2), 188-192 (1982). (8) H. F. R. Schoyer, J. Spacecr. Rockets 17 (3), 200-207 (1980). (9) H. F. R. Schoyer, Collection of Papers of the 25th IACAA, Tel-

(10) E. W. Price and F. E. C. Culick, “Combustion of Solid Rocket Aviv, 1983, pp. 55-72.

Propellants”, AIAA, New York, 1968.

(Received October 6, 1983, Ms. 23/83)

5 crn propellant disks

0