AFATL-TR-82-32

Ambient Moisture Effects and

Closed Vessel Studies of Gun
Propellant Combustion
0 K Heiney
0 L Burkett
I F Whidden
T W Shiver

BALLISTICS BRANCH
C DIRECT FIRE WEAPONS DIVISION

MAY 1982

<r FINAL REPORT FOR PERIOD FEBRUARY 1981-MARCH 1982

jApproved for public release; distribution unlimited

LLU
C:)

Air Force Armament Laboratory

AM FORCE SYSTEMS COMMAND*UNITED STATES AIR FORCE*EGLIN AIR FORCE BASE, FLORIDA
82 08 l1 130

L-2-L m •mmn • -" •" • mmi

 NOTICE

Please do not request copies of this report from the Air Force Armament Laboratory.
Additional copies may be purchased from:
National Technical Information Service
5285 Port Royal Road
Springfield, Virgnia 22161
Federal Government agencies and their contractors registered with Defense Technical
Information Center should direct requests for copies of this report to:

-Defense Technical Information Center

Cameron Station
Alexandria, Virginia 22314

Ni
 IINCTLAS.STPTIf

I
SECURITY CLASSIFICATION OF THIS PAGE (fticn Dam,Entere-?)ý

REPORT DOCUMENTAION PAGE T B~EACRE. 40XPLETZM&F01ORM!

1. REPORT NUMBER GOVT ACCESSION NO 3 RECIPIENT'S CATALOG NUMBER
AFATL-TR-82-32
4. TITLE (and Subtitle)
S. TYPE OF REPORT & PERIOD COVERE D

AMBIENT MOISTURE EFFECTS AND CLOSED VESSEL Final-Report. -February

STUDIES OF GUN PROPELLANT COMBUSTION 91Mch98
6 PERFORM-4G 01G. REPORT NUMBER

7. AUTHOR(s) B. C43NTRACT OR ;5RANT NUMBER(&)

Otto K. Heiney Terrell W. Shiver
Douglas L. Burkett
James F. Whidden ___-______

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10PROGRAM ELEMENT. PROJECT, TAbK

Ballistics Branch, Direct Fire Wveapons Division PAE:&R UNI6N02ER
Air Force Armament Laboratory jF:662
Eglin Air Force Base, Florida 325427 JON 2560-0819
Ii. CONTROLLING OFFICE NAME AND ADDRESS 12 REPORT CATE
Air Force Armament Laboratory v12
Aiinaxnenz Division '3 NitA8BER OF PAGES
Eglin Air Force Base, Florida 32542 I40
14. MONITORING AGENCY NAME &ADDRESSflf different fro., Contirolling CGffic.ý) IS SECURITY CLASS. (i-f this report)

~Unclassified
IaDECLASSI FICATION.'DOWNGRADING

16. DISTRIBUTION STATEMEZNT (of thzis Report)

Approved for public release; distribution unlimited.

17 ISTRIBUTION STATEMENT (of ~he abstract enter*& in Block 20, If different fromn Report)

18. SUPPLEMENTARY NOTES

Availability of this report is specified on reverse of front cover.

11-I
`-19. KEY WORDS (Continue on reverse side if necessary and identIfy by block number)
Gun Propellant
puz Solid Propellant Combustion
J Interior Ballistics
Combustion Efficiency
Gas Specific Heat Ratios
IKE- 20. ABSTRACT (Continue cn reverse side It necessary and identify by block number)
YOM -This report covers several topics in the field of solid gun propellant
'ombustion. They include:, (1) mass spectrometry to experimentally define the
gas products of gun propellant combustion and correlate these experimental
results with theoretical predictions; i2) the use of closed combustion bomb
pressurization rates to extract solid propellant linear burning rate values is
discussed, and correlations with strand burner data are considered; (3) the-
effects of moisture adsorption on solid propellant combustion rates and ener
are experimentally defined, and (4) a reference section of solid propellant-j
DD 1 3 1473 EDITION OF I NO0V 15 IS OBSOLETE UCASFE
SECURITY CLASSIFICATION OF THIS PAGE ("o~In Petit Entered)
 UNCLASSIFIED
SeCU.i.y.CLASSI.ICATIONOF THIS PAGE....n.Dat. Uteed)

20. ABSTRACT (CONCLUDED)

-combustion product specific heat ratio values is provided.

I:

DTIC V,3

_t 0
Dist spca

UNCLASSIFIED
 PREFACE

This report has been generated under the requirements of in-house work
units 2560-08-19 and 2560-08-20, which sponsor on-going evaluation of gun pro-
pellant performance, and continuing investigations into the nature of the
physics and thermodynamics, which control gun propellant combustion, and thereby
r the efficiency of the aircraft cannon ammunition systems.

The presented analytic and experimental data were generated at the

Interior Ballistics Laboratory of the Direct Fire Weapons Division, Air Force
& Armament Laboratory during the period from February 1981 to March 1982,

"The Public Affairs Office has reviewed this report, and it is releasable
to the National Technical Information Service (NTIS), where it will be avail-
able to the general public, including foreign nationals.

This technical report has been reviewed and is approved for publication.

-FOR THE COMMANDER

MARVIN J. WOCRING, olonel, U AF

Chief, Direct Fire Weapons Division

(The reverse of this page is blank)

 TABLE OF CONTENTS

Section Title Page -

I INTRODUCTION ....... ...................... .1...

II COMBUSTION PRODUCT ANALYSIS BY MASS SPECTROMETRY . . .... 2

Hydrogen ................................... 5
Water .......... ........................... 5
Molecular Fracture ......... .................... 5
Data Reduction ................................ 6

III CLOSED BOMB BURNING RATE ANALYSIS ..... ............ 13

iV MOISTURE EFFECTS ON PROPELLANT CCMB11STION .............. 21

V GAMMA VALUES FOR COMBUSTION GASES ...... .............. 25

References ........... ......................... 33

Mi

L• iii
 LIST OF FIGURES

Figures Title Page

I Time-of-Flight ........... ....................... 4

2 Raw Da*a: Single Bass Propellant Combustion Specie Mass

Fractions ....... ....... ........................ 7

3 Computer Output: Single Base Propellant Combustion Specie

Mass Fractions ........... ..... .................. 8

4 Single Base Propellant Combustion Products ..... ......... 10

5 Double Base Propellant Combustion Products ..... ......... 11

6 Nitramine Propellant Combustion Products ........ ...... 12

7 Typical Closed Bomb Pressure-Time Trace ............. .... 15

L9 8

10
Polynominal Fit to Experimental Pressurization Data

Closed Bomb Burning Rate from Differential of Pressurization

Data. .......... ..... ..... ..... ..... ..... ... ...

Experimental Burning Rates from Various Laboratories .....

.....

..
17

18

19

11 Peak Pressure Versus Conditioning Time ............... ... 22

12 Muzzle Velocity Versus Conditioning Time ..... ........... 23

13 Specific Heat Ratio 02. *.26

14 Specific Heat Ratio N2. * . 27

15 Specific Heat Ratio CO ........... ..... ..... ..... ... 28

16 Specific Heat Ratio Air ........ ................... 29

17 Specific Heat Ratio H2 . . .. .. . . . . . . . . . . . . . . 30

18 Specific Heat Ratio CO2 . . . . . . . . . . . . . . 31

3

19 Expansion Coefficient Water Vapor. ............... .... 32

LIST OF TABLES

Table Title Page

1 Theoretical Mole Fractions ...... ..... ..... ......... 3

iv
 3•CT •q
INTRODUC 1010.

This report describes the procedurc; and results o, -me neer

effects analysis , d closed vessel in-house gun propeIIa;:t ,e.. - t: develop-
Sment condacted in the Interior Ballistic Laborotory of che Air Ptt .rvaiaent
Laboratory.
The initial two topics which will be discussed are somewhat disparcte and
related only through the fact that both are closed vessel approaches to anolyze
the combustion properties of gun propellants. The first is a combustion bomb
mass spectrometer assembly in which small (3-gram) samples of gun propellant
are burned and immediately fed into a time-of-flight mass spectrometer for com-
bustion product analysis. These approaches and results will be discussed in
Section II.
The second technique which will be covered is o
the use :i' c obestion
bomb to extract linear burning rates. The approach historically used at Eglin
and described in Reference 1 is that of the very high pressure prepressurized
linear strand burner, as opposed tc a closed combustion bomb. The closed vessel
work was undertaken as part of a JANNAF work shop, and the experimental results
acquired were surprisingly consistent. These results and the mathematical data
extraction used are presented in Section III.
The third area which will be addressed is that of ambient moisture effects
on solid propellant combustion performance. This topic arose from occasional
drastic experimental performance degradation in the GAU-8 gun from a given lot
of propellant of known baseline performance.
A final short reference section is included to provide specific heat ratio
values for the typical gun propellant combustion specie constituent gases as a
function of temperature and pressure. This data is frequently required for
various interiox ballistic computations and is not conveniently available in
any single reference.
 SECTION II

COMBUSTION PRODUCT ANALYSIS BY MASS SPECTROMETRY

Various types of mass spectrometers have been available to the combustion

community for some time and the limitations applying to their use as a tool to
quantitatively analyze gun Dropeilant equilibrium comb•!stion products are well
appreciated. The primary drawbacks are that (I) two of the main combustion pro-
ducts, CO and N2 fall at the same molecular weight of 28; (2) water will immedi-
ateiy condense on the relatively cool bomb walls and be virtually undetectable;
and (3) hydrogen tends to be absorbed or adsorbed by almost everything in the
sampling system and is difficult to handle quantitatively. Also, the relative
violence of the sampling process in the mass spectrometer results in the decompo-
sition of a portion of the formed species. In standard test practice, an elec-
tron beam energy of 70 electron volts is used to knock an electron off the
molecules to be sampled. This can split, for example, a molecule such as CH
'4
into CH3 , CH2 , and CH; or CO into C and 0. Alternatively, the electron beam
may knock two electrons rather than one off CO2 at mass unit 44, with the
result, that it appears to have a mass unit of 22. Due to these drawbacks,
researchers typically tend to rely on theoretical thermochemical combustion
specie computations, such as that described in Reference 2, rather than experi-
mental analysis.
The persistent problems with the A-10 aircraft ingesting GAU-8 muzzle gases
lent impetus, however, to a program oriented toward determining what could be
learned from an experimental examination of the combuztion of various propell-
ant types and their resultant combustion gas species. The types investigated,
along with their theoretical fr-P energy mole fraction combustion products,
are given on Table 1. The propellants consi:t of a single base, a double base,
and a nitramine formulation. The theoretical mole fraction values were deter-
mined by means of a NASA Lewis free-energy program modified for constant volume
applications (Reference 2). The mole fraction values labeled "hot" are effec-
tively the gases expected at combustion chamber conditions while the values
labeled "cold" are the resultant theoretical values subsequent to an expansion a
to ambient pressure and temperature conditions.
lExperimentally, the propellant was burned in a Paar bomb located directly
at the input of a CVC MA-2, time-of-flight mass spectrometer (Figure 1). The

2
 0
r-J 001 P
"4 00

-v-4

"4 0-

0) 01 "M4 0

cfl U
Ln- Ch t 0 000

F~ 00 Q4 "4

"44
U3

1- 4 -4 0

00 N1 00 00

I-
Mon4

1:1 ii)8
........

11-

464-,

awd)
M I"hot" data were repeatedly sampled within one minute of the firing. The cold
data were generated by bleeding the Paar bomb pressure to 25 psig and sampled

after a 12-minute cooling cycle.

These type formulations were experimentally evaluated with the basic
results as shown on Figares 4 through 6.
Prior to a discussion of the correlation of species between theory and
experiment, it is useful to discuss how the time-of-flight mass spectrometer
i•is quantitatively calibrated. This process is somewhat Edisonian and largely
accomplished through the use of calibration gases of known specie percentages.

HYDROGEN. In general, the detectors of the type used in the equipment

available for this program are less sensitive to hydrogen than the higher mass
unit gases. To determine the sensitivity factor, a known mixture of 98 percent
air and 2 percent hydrogen is introduced into the machine. As only 25.1 percent
of the hydrogen known to have been introduced is detected, a factor of 3.99 is
used to enhance the detected level for subsequent experimental testing.
WATER. Subsequent to propellant combustion in the bomb, the gas phase
moisture will immediately condense on the relatively cooler walls of the bomb.
Careful recovery of this water from the bomb walls and end plates typically
results in values of 11 percent by weight or 15 percent by mole fraction of
water from the combustion process. This is virtually 100 percent )f the theo-
retical water formed during combustion. Thus, on the computer data reduction,
a water correction factor is inserted to correct the mole fraction output
results for thi: water which does not enter the mass spectrometer sampling
ysstem.
MOL2CULAR FRACTURE. Due to the violent nature of the electron beam impact
used to charge the gas particle to be sampled, a fixed percentage of each type
of gas encountered will be broken into fragments. For a given electron beam
intensity, in this case 70 eV, the fractions remain constant and are given
below. These fractions are determined by repeatedly sampling calibration gases
of well known composition. For the fragment fractions given, nubers
'hc are
references to 100 percent for the peak of the sampled gas itself. t should be
remembered that these values are system and energy level specific.
CO 100% CO
4% C
1% 0

5
 CO22 100% CO2
12% 0

11% CO

6% C

N2 10 N2
2
6% N

CH 100% CH4
4
67% CH3

13% CH2

7% Ch

2% C
From the above, it is seen that molecules such as CO2 and CH4 will frag-
ment very badly during the sampling process. The fragmentation fraction of N2
N., peak at mass unit 28. The
kick allows the quantitative separation of the CO -
component due to N2 at ,WU28 will be 16.5 times the value" at AMU 14, The
"remainder at this mass unit will then be CO.
DATA REDUCTION. The raw data from the mass spectrometer, which provides
typical initial output as per Figure 2, is fed into a Tektronix 4054 Data
Processor. Processing oF data is conducted in accordance with the discussion
in the paragraphs above. The hydrogen detection level is enhanced, a manually
recovered water factor is iincluded, the N2 - CO ratio is determined from the N
fraction at AMU 14, and compeasation is made for the known fracturing of mole-
cules
c by the electron beam energy. Figure 2 zonýists nf raw .elative specie
data for a single base nitrocellulose propellant. It is seen that almost 85 per-
cent of the output is located at the peaks of 44 (C02 ) and 28 (CO, N2 ). This data,
along with those pertaining to the minor constituent species, is then machine
processed by the techniques discussed above to provide an output as per Figure
3. The logic flow is to first split the mass peak at 28 into components using
the measured intensity values. The hydrogen sensitivity factor and molecular
fracture percentages are accounted for in the corrected intensity column.
Finally, the overall percentage numbers are again modified to account for the
manually recovered water from the combustion bomb interior.

6
 -4!-

U)

:5 4-)

3 0)

CA

4-)0

RAE

'-U,- (4

w~~~ l
j4-)ZWH
 z x

WW c w. . . . .-
- La -

CL 0 r- =

0J
v 0
z
-)

3.

0 S-
CU)

44

~U~t
a II- U 0 ~ . w4.)
wr-t0@wwwN~CD

to x x0 'A. J

w Nw

Z H O
> EM
I LD MN RA

~ j~ ~
N~w0~boerSog~1 2
63i

8y Wm N
 Hot and cold theoretical and experimental combustion specie results for a
single base, double base, and nitramine gun propellant are contained on Figures
4, 5, and 6, respectively. Some general results for the three cases are immedi-
ately apparent. in all cases; experimental rdsults for hydrogen are lower than
theoretically predicted. This is no doubt due to adsorption of a portion of
the hydrogen onto bomb parts prior to introduction into the spectrometer
sampling system. The experimental water recovered is consistent at about 15 per-
cent irrespective of the sampling temperature. This would be expected as the
combustion water probably immediately condenses on the cooler metal bomb walls.
Virtually all the water formed during-combustion is recovered from the bomb
and does not enter the spectrometer sampling system. This is not the case for
moisture in the air as humidity, which is veýry readily detected during routine
system background checks.
For the N2 specie, experimental results tend to be consistently higher
than theoretical predictions, while for CO the experimental values are consis-
tently lower than the free energy predictions. This is due to a biased shred-
out of the mass peak at 28, which Drobably occurs due to the existence of other
fragments at mass 14 that were not there during the calibration process.
In conclusion, it can be stated that the mass spectrometer is a less than
ideal tool for specie verification. The physics and thermodynamics of the
inlet process and the violent dynamics of the sampling process result in a
nested array of cor--ction factors that somewhat tarnish the scientific ele-
gance of the experiment. Nonetheless, the data presented on Figures 4, 5, and
6 are suitable to determine the general quantity and type of combustion pro-
ducts emitted during cannon firing and that are libely to accumulate in an
onboard gun-firing bay.

9
>
w U,
4-

j---
1Ec1 t I T
u~

to

(9( 9 M 3)i

I 010
 4-

I 0

00

I SI
0
-~ Ca

tus

4 s-

4 Z C D

54 I

0 C-7

LL

1LV

U) N -zý
 - C-

LLI
LLI 4J

LIL
M

IE

4 4
I ji

[*J ILI

H 00

I- ýa E

H 12
 II
SECTION III

CLOSED BOMB BURNING RATE ANALYSIS

The technique which has traditionally been used at Eglin to determine

linear burning rates of advanced and conventional gun propellants is that of
the pre-pressurized strand burner. This is a high pressure (b0,000 psi) vari-
ant of the standard rocket industry Crawford bomb. However, due to the high
initial capital costs and intensive maintenance required with multiple stage
"intensifier high pressure strand burners, much of the gun propellant community
has, over the years, attempted to extracz burning rates from closed vessel
combustion bombs.
The closed bomb approach uses a much simpler experimental set-up but
requires complex computer processing of pressurization rate data to generate
the desired burning rates. A JANNAF committee has been established for some
time to coordinate and correlate the approaches from various government aud
industry laboratories for a solution of the burning rate determination prob-
lem. The purpose of this report is to describe the methodology used at Eglil
to extract linear burning rates from bomb pressurization data.
Closed impetus bombs have been available to the propellant industry for
many years. The type used at Eglin is a 90-cubic centimeter Technoproducts
model as described in Reference 3. The advantage of using this device is a
very rapid load, fire, and turnaround capability. The disadvantage is that
main seal leakage problems start at pressures around 30,000 psi and consequently

in routine usage, the operating pressure of the bomb is held to a relatively

low 25,000 psi. For impetus calculation, the bomb operation is quite simple
and uses only the final peak pressure and a Nobel-Abel equation of state
P(V-n) = nRT()
where RT = Fp; the impetus, and
n Cw charge mass, with
n Cw - n' is specific covolume
This covolume n' has a typical value of 29 cubic inches/pound for gun propellants
and accounts for the fact that combustion products are not mass points but rather
take up a finite volume of the available combustion chamber.
Thus, to express impetus in the customary units of foot/pounds, Equation (1)
P •becomes

- 12 C- (2)

i3
 From Equation (1), it is also seen
dn/dt v dP/dt

Thus, from the time rate of bomb pressurization, one can extract a gassifi-
cation rate dn/dt. Also if
dn/dt = rS C (3)
the gassification rate, as a function of pressure, will lead to a linear burn-
ing rate r as a function of pressure, given that S is at all times accurately
known. The basic problem associated with determinatior of linear burning rates
from closed vessel pressurization data is that S• may not be fixed or adequately
expressed analytically and severe bias errors may arise from only the gassifi-
cation rate being experimentally measured.
Figure 7 is a typical closed bomb pressure versus time profile acquired
from the Technoproducts bomb using a Kistler 607C gauge fed into a Nicolet
Explorer III Digital storage oscilloscope. These results are not typical, only
to the extent that they contain 60 microseconds of bad data due to either
momentary guage or scope analog-to-digital converter failure.
As the rate of pressurization is the parameter of primary interest, the
data must be differentiated either directly numerically or a polynominal fitted
FM and then the polynominal analytically differentiated. Only the center part of
the pressure rise curve can be utilized. The points to the left of the start
line on Figure 7 cannot be used as this area is part of the ignition cycle
rather than linear combustion, and all faces of the bulk grain burning surface
have not yet been ignited. Attempted data reduction in this regime will lead
to certain error, as would attempts to process data lying past that portion of
the curve to the right of the stop line, at which point propellant grains have
started to fracture and/or burn out.
As data is acquired every 20 microseconds, the approximately 2.5 milli-
seconds of data hetween the start and stop fiducial values on Figure 7 are
represented by 126 discrete values. As direct numerical differentiation has
proven in practice to be erratic and highly noisy, the AFATL approach has been
to least squares fit all points within the selected range to a 5-degree polyno-
mial per techniques in Reference 4. The detailed computer output includes a
series of constants A1 to A6 giving a best fit to the pressure time data. In
this approach,
2 3 4 5
1 2 +3 A4t +A 6 t; (4)

14

A
 40

cci)

1 15
 then,
• +3A~t2 4t3 5t4. 5
dP/dt A2 + 2A3t +3A4 + t SA6t
+4A (5)

The nature of this type fit to the data is shown on Figure 8 with the
- equation provided by Equation (4) compared to the 126 points of discrete data.
It is seen that the fit is outstandingly nice; in addition, a major advantage
of this approach is that if a few spurious points such as those centered at
13.6 milliseconds exist, they will be all overwhelmed by the mass of "good"
data contained in the other 123 po.nts. The analytic differential of Equation
(4) is Equation (5) which with Equation (3) and assuming the burning surface
is either fixed or known, provides a buining rate per Figure 9. This data is
for M-10 which was the subject of a JANNAF Round Robin evaluation of closed
vessel and strand burner techniques in various laboratories. AFATL strand
burner and closed vessel burning rate values as well as closed vessel results
from other laboratories are presented on Figure 10.
Figure 10 serves as a good illustrative example of some of the limitations
of using a closed bomb fur burning rate data extraction. It is seen that the
results of three laboratories show very low burning rates for M-10 at 1,000 psi
which subsequently rapidly rise to the 2,500 psi range and then level out.
This is a clear indication 'hat, in this range, the ignition cycle has not yet
been completed. Clearly what has been measured in this region by the time
differential of the pressure data is not a burning rate but rather an ignition
flame spread phenomenon.
The data on Figure 10 represent single perforate propellant which is also
the simplest case to e,:al with, as the burning surface can be considered to be
relatively constant (per Reference f). For a seven-perforate propellant,
the burning surface is not constant but rather progressive, and resort must be
made to either geometrical arguments or closed form "form functions" which are
PR described in detail in Reference 7. Whichever approach is utilized, the burn-
Ml ing surface data becomes invalid at the point of web burnout to slivers.
OL Irrespective of any sort of painstaking analysis at this point, the inherent
0 natures of manufacturing and fracturing irregularity render the process funda-
mentally non-analytic.
An additional problem that arises when dealing with seven-perforate pro-
pellant is that of what initial burning surface with which to start computations.

V 16
 CYY
-j 0)

LO) co
mmwwwm
w XMW

17,
 CD

LID

c~a
I-- -

I1MCC MqMLHZUW

* *18
 00

Uz
~
~ 04
0 so

-t~L
0
04-

ii Lc
This is particularly true if the mathematical differential is not initiated at
the start of burning, which, in general, it is not, but rather subsequent to
the ignition transient. The manner in which this is handled at the Air Force
Armament Laboratory (AFATL) is to assume linear mass consumption as a function
of start pressure to peak pressure and enter the form function relationship at
the appropriate mass consumed pressure ratio point. This is a reasonable
approximation but again can provide bias errors.
An additional consideration for which compensation is generally made is
that of heat loss to the walls of the bomb. Using the AFATL bomb, this is a
surprisingly small correction, as the impetus efficle.zy tends to run in excess
of 97 percent. The procedure used to make this minor heat loss correction is
to increment all burning rates by a factor determined by a ratio of the inverse
of the experimental efficiency. In this case
r' = r/.97 = 1.031r
This can be thought of as a distributed or PDOT heat loss correction approach.
The plot of the burning rates from the various techniques and differing
laboratories given on Figure 10 shows a high degree of scatter below 10,000
psi but a reasonable convergence above that point. As the main working pres-
sures in a typical gun are above 10,000 psi and the lower regime is largely
in the ignition transient of the ballistic cycle, closed bomb burning rate
data is of some utility for interior ballistic calculations. The primary
approach that will continue to be used at Eglin for burning rate determination
is that of the linear strand burner. For those organizations without access
to this type equipment and who, by necessity, must resort to closed bomb burning
rate data extraction, the AFATL Drograms are available programmed in either
Basic or Fortran upon application to AFATL/DLDL.

S20
 3-
1 10 T - - - - - - - - ----

SECTION IV

MOISTURE EFFECTS ON PROPELLANT COMBUSTION

The normal climate at Eglin Air Force Base is exceptionally humid, with
the result that stored gun propellant can often be exnosed to simultaneously
high levels of high temperature and humidity for extended periods. Often,
locally stored Class B propellant cannot be protected in hermetically sealed
bulk munition cans without reclassification as Class A, resulting in storage
compatibility problems. This normally results in bulk propellant being locally
stored in vented containers.
It has previously been reported that the GAU-8 gun, with its plastic
rotating bands, showed a high degree of performance variation with very slight
changes in barrel chamber design (Reference 8). Performance of a given lot of
locally loaded propellant, in the same chamber, has also experimentally been
seen to slowly degrade with time. It was intuitively postulated that the per-
formance decrease was due to ambient humidity effects on the nitrocellulose
propellant.
Recently an experimental firing program has been conducted at AFATL to
quantitatively define the effect of adsorbed moisture on this double base gun
propellant in the GAU-8 cartridge. All shots were fired with a 428-gram pro-
jectile with a 154-gram propellant charge from the same Mann barrel. Each
listed data point consists of an average of 5 fiTings at the specified time
of subsequent dryirg.
The double base propellant used for this test was taken from a single lot
at ambient storage. It was spread on open trays and exposed in an environmental
chamber to 95'F and 95 percent relative humidity for a period of 8 days. From
tray to tray the adsorbed weight of water, due to this treatment, was a very
uniform 1.13 percent.
The propellant was then dried at 110 F in a forced air oven at virtually
I+ 0 percent relative humidity (RH). Gun firings were conducted after 0, 1, 2, 4,
8, 24, and 67 hours of drying. The results are plotted on Figures 11 for
chamber pressure versus conditioning time and Figure 12 for muzzle velocity
versus conditioning time. It is immediately seen that the effect of adsorbed
moisture on performance is phenomenally strong.

21
÷Co

LO

Co

CY)

IA~4 to-.i

•o o

NN

na-4
224

too
 I--

00

> 0 >

UQ

N0

j2
 As the data has the natural appearance of a logarithmic relationship, a
Hewlett-Packard model 97 least squares routine was used to fit the points to a
relationship of the form.
y = a + b in t
For velocity, selected points give
v = 3177 + 53.67 in t
with a correlation coefficient of .RR
while for pressure the fitted cons,:ants are
P = 50,360 + 5096 In t
with a correlation coefficient of 0.997.
The results of this test program clearly demonstrate the very significant
degradation of gun propellant performance as a function of adsorbed moisture
level and prove the accuracy of the old adage "keep your powder dry".
In the future, this experimental program will be rerun to quantitatively

[ determine the percent of moisture remaining on the grains as a function of

the drying cycle.
time in

I
L2-

24

Rt__
 SECTION V

GAMMA VALUES FOR COMBUSTION GASES

This section will provide reference values for the specific heat ratio of
I various gases typically found as gun propellant combustion products and also
)• for those gases such as helium and hydrogen which are used as working fluids
in light gas guns. The specific heat ratio data for oxygen, nitrogen, carbon
monoxide, air, hydrogen, and carbon dioxide have largely been extracted from
References 9 and 10, and the data is presented on Figures 13 to 18. The gamma
values on the plots are given at 1, 10, and 100 atmospheres where Y C /C
P v
Water vapor has reasonably curious thermodynamic behavior, and the litera-
ture is somewhat reticent in providing tabulated specific heat ratio information.
Reference 11 provides data for a related parameter, the isentropic expansion
exponent K which is defined as

K =p

where, along an isentrope

pvk = constant
The data provided for water vapor on Figure 19 is this isentropic expansion
exponent.
For helium, a plot as a function of temperature and pressure is not
required, in the temperature ranges encountered in light gas gun operation,
and a constant value of Y = 1.66 is suitable for general interior ballistic
usage.
The gamma values are important for several interior ballistic applications,
particularly energy computations, in that while impetus or F is the value for
p
specific energy for a propellant usually quoted, actual available total elergy
is:
Q = CFp/(Y-l)

25
 CD~

CIm

x 26
 InI
%
CD L

z4
hiL

zC
(' a

S'

LL.

27)
 I-I U)

z
z

C-)Y

CD 4-

(D w 0
co Ixl

4 ci
I-- -

U) C
I 09

I &Q
ý-I-

mli

1±
 in

49C

CDI~ .

Ii 4-)

iii
C-

ME- J. VIC
-V

29
 S -

L~
0 CDY

06

z N

CD.

LL.

4-30

•a • 30
 4 U
I.
N CY 4-

iI--

k D
Li,

Z,, I4cl U,

3]1

31,=
 a--

44

N:- -

_ coWC- LLA

•-!

32
 REFERENCES

1. Moy, B.K., "Ballistic Characterization of Ingredients for Advanced Gun

Propellants," AFATL-TR-79-50, April 1979.

2. Heiney, O.K., "Theoretical Gun Propellant Thermochemical Evaluation,"

AFATL-TR-71-11, January 1971.

3. "A Simplified Method for Constant Volume Bomb Measurement of Specific

Impulse," Technoproducts Data Bulletin #626G3, San Cur)os, California
(undated).

4. Bevington, P.R., "Data Reduction and Error Analysis f-1 the Physical
Sciences," McGraw Hill, 1969.
S.. Robbins, F.W., "1JANNAF Round Robin Burning Rate Results for Closed Bomb
and Strand Burner Firings of M-10 Propellant," JANNAF Report in Print,
1982.

6. Heiney, O.K. and West, R.J., "Interior Ballistics, Muzzle Flash, and Gas
Gradients of Aircraft Cannon," AFATL-TR-76-34, March 1976.

7. Hunt, F.W., "Internal Ballistics," U.S. Ministry of Supply, London 1951.

Sj 8. Heiney, O.K. etal, "Experimental Interior Ballistic Variability Phenomena

in the 30mm GAU-8 Gun," AFATL-TR-79-64, August 1979.

9. "Tables of Thermal Properties of Gases," National Bureau of Standards

Circular 564 GPO, November 1955.
10. King, C.R., "Compilation of Thermodynamic Properties of Gaseous Hydrogen,"
NASA Technical Note D-275, April 1960.

11. Keenan, J.H. and Keyes, F.G., "Steam Tables," John Wiley, 1978.

33
(The reverse of this page is blank)