Defence Science Journal, Vol. 71, No. 1, January 2021, pp. 18-24, DOI : 10.14429/dsj.71.

16007
© 2021, DESIDOC

Statistical Evaluation of Burning Rate Data of Composite Propellants Obtained from

Acoustic Emission Technique
Ehtasimul Hoque, Chandra Shekhar Pant*, and Sushanta Das
DRDO-Advanced Centre for Energetic Materials, Nasik - 422 207, India
*
E-mail: [email protected]

Abstract
The acoustic emission technique has been considered to be one of the most reliable and robust methods for
the measurement of the steady burning rate of composite propellants. In this work, attempts were made to quantify
the measurement variability of the burning rate of composite solid propellants by acoustic emission method using
statistical tools. A total of 1100 individual measurements were subjected to statistical treatment. The combination
of confidence interval and repeatability limit delineated the extent of natural dispersion in the burning rate
measurement data. The very high coefficient of variation values for the propellant compositions, having a burning
rate of more than 25 mm s–1 raised concerns about the suitability of the acoustic emission method for high burning
rate compositions. The Reliability interval approach was employed to determine the statistically significant sample
size for different composite propellants having a burning rate range of 5–31 mm s–1. The entire set of data was
screened for identification of outlying observation using the Dixon Q test, and the extent of contamination was
quantified. Moreover, the application of statistical techniques could have far-reaching implications for quality control
perspectives of burning rate measurement by acoustic emission and could be implemented as reference tolerance
limits and preventive measures for ensuring the good health of the instrument as well as propellant processing.
Keywords: Acoustic emission burning rate; Confidence interval; Repeatability limit; Coefficient of variation;
Outliers

1. Introduction friction sensitivity and burning rate of non-aluminised and

The burning rate of composite solid propellant is a aluminised composite propellants respectively.
convoluted function of several physical, mechanical, and A considerable amount of work has been carried out
chemical factors, such as temperature, pressure, heat and mass to develop non–intrusive methods that do not perturb the
flux, chemical reactions in the gas, interface, and condensed combustion process, using advanced technics such as ultrasonic
phases, etc.1-3. It is one of the most important measured quality waves, X-ray radiation, microwave, videography, plasma
characteristics due to its ability to change the ballistics of the capacitance, fuse wire, acoustic emission etc.11-14. An initial
rocket motor. Therefore, the measurement of the burning rate study by Caveny15, et al. reported 94% accuracy in the burn
more economically and efficiently way is a task of paramount rate measurement by acoustic emission technique for the single
importance for every propellant processing facility. Since to base, double base, triple base, and HMX composite propellants
date no theoretical model is capable of predicting burn rate with at high pressure. Koury16 & Christensen17 established a
accuracies within 1%, therefore burn rate can only be measured statistically significant correlation between the solid strand
by experimentally4. Over the years, the strand burner methods burning rates (7 – 9 mm s–1) measured by acoustic emission
have emerged as a standard technique for steady burning rate and full-scale motor, and proposed use of solid strand burning
measurement. However, ensuring the precision and accuracy rate technique in lieu of ballistic motors which would incur
was considered to be the main challenge of solid propellant considerable savings in expenditure and time. In addition,
burning rate measurement5,6. Jordan7 pointed out that the effect Rampichini18, et al. showed that the accuracy of acoustic
of the burning rate variability on the flight performance of emission was higher (0.7%) as compared to videography
the rocket motor was five times higher as compared to other (4%) and fuse wire (4.4%) technique. The other methods
parameters. This higher variability results in increases the on- like ultrasonic, X-ray, and microwave require specialised
target delivery cost. Liu8 demonstrated the uncertainties in instrumentations, and NATO Standardisation Agreement
the pressure exponent (n) value due to the variability in the 467414 recommends exclusion of the X-ray and microwave
burning rate measurement data. Furthermore, Kubota9, et al. measurement techniques as a routine ballistics characterisation
and Hoque10, et al. observed a direct correlation between tools. Therefore, acoustic emission is the most used, reliable, and
fast technique for solid strand burning rate measurement, and
Received : 24 June 2020, Revised : 18 September 2020 it had been widely implemented as an essential quality control
Accepted : 14 October 2020, Online published : 01 February 2021 tool in propellant processing facilities across the globe.

18
 Hoque, et al.: Statistical Evaluation of Burning Rate Data of Composite Propellants Obtained from Acoustic

Despite the considerable use of acoustic emission burning burning rate value and the standard deviation were reported as
rate data, little attention has been paid to develop an integrated quality statistics.
idea about the variability and limitations associated with the Of particular note, the acoustic emission method has
measurement methodology. No comprehensive trend analysis a fundamental limitation of measuring a lower value of the
and precision statistics are reported in the literature covering a burning rate as compared to the main motor. The reason for
wide burning rate range and diverse propellant compositions. this aberration is thought to be the higher heat loss to the
Considering these factors, the goal of this study is to surrounding water medium for the usage of a thin strand of
analyse and quantify the measurement variability associated propellant during testing. Higher loss of heat produces a cooler
with the acoustic emission strand burning rate technique. The combustion zone at the propellant burning surface which
data set covered1100 burn rate data from diverse propellant in turn lowers the burning rate6,19-20. The acoustic emission
compositions with a burning rate range of 5–31 mm s–1. method is also inefficient to identify combustion instabilities
The approach broadly involves quantification of due to heterogeneity, small defects, and porosity. Besides, the
repeatability statistics by determination of confidence interval applicability of acoustic emission methodology for higher burn
(CI) and coefficient of variation (CV) of burning rate data of rate formulations (more than 20 mm s-1) has also not been well
different formulations. Reliability interval (L) was used to delineated in the literature.
determine the appropriate sample size to estimate the sample
statistics with a specified precision and accuracy. Furthermore, 2.2 Data Collection
detection statistics have been employed for ‘Outlier’ which is The measured acoustic emission strand burning rate data
an assignable cause of variability in the experimental data. were acquired from the in-house quality control data bank,
which consisted of a large quantity of coherence burning
2. Method rate data, accomplished from the last 10 years of disciplined
2.1 Burning Rate Measurement of Composite processing of composite solid propellant. The burning rate
Propellants by Acoustic Emission data of composite propellants were classified according to their
The burning rate of the composite propellant was differences in formulations, described in Table 1. The majority
determined in a modified Crawford’s bomb5 using acoustic of the burning rate measurement data were of HTPB/AP/Al-
emission technique. The schematic representation of the based composite propellants which were broadly grouped in
assembly is shown in Fig. 1. Solid propellant strands were two subcategories; P01: non–catalysed compositions and P02:
milled from propellant cartons. The dimensions of each strand catalysed by transition metal oxides. Furthermore, we studied
were 6 mm × 6 mm × 130 mm. A nichrome igniter wire was the burning rate of active binder/AP/Al/Nitramine based
threaded through the one end of the propellant strand keeping composite propellant. The reported burning rate data ranges
approximately 3 mm of propellant above the igniter wire, and from about 5–31 mm s–1.
this prepared sample was ignited while immersed in water The propellant formulations: P01 and P02 were further
inside the closed bomb. The water acted as a natural inhibitor augmented in five derivative compositions based on the coarse
to prevent side burning, and medium to carry acoustic signals to fine ammonium perchlorate (AP) ratio and the amount of
generated by propellant burning. The bomb was pressurised catalyst loading. This expanded set of propellant compositions
with nitrogen gas as per the required burning conditions, and increased the resolution of the study.
a firing pulse was given. These acoustic signals were captured
by a sensor mounted externally on the combustion bomb. The Table 1. Summary of burning rate data of composite propellant
from acoustic emission measurements
output electronic signal was recorded and displayed in voltage
vs. time domain by a data acquisition system. By monitoring Burning No. of
the time required for the flame to consume the known length Categories Description rate data
of propellant at a preset pressure, the burning rate of the (mm s–1) points
propellant was determined. To obtain a reliable, precise, and P01 HTPB/Al/AP (Non–catalysed) 5–7 500
unbiased result, repetitive burning rate measurements were P02 HTPB/Al/AP (Catalysed) 14–31 528
carried out from a single propellant batch, and the average P03 Active Binder/AP/Al/Nitramine 11–13 65

Figure 1. Burning rate measurement using acoustical emission technique.

19
 Def. SCI. J., Vol. 71, No. 1, January 2021

2.3 Statistical Methods sp is the pooled estimator of population standard deviation (σ).
2.3.1 Statistics of Variability The value of margin of error (E) was used to calculate the
The confidence interval (CI), repeatability limit (R), and reliability interval (L) as per the following equation:
coefficient of variation (CV) were evaluated as variability 100 × E
statistics21. For a sample size <30, the equation for confidence Reliability Interval, L = (5)
Z
interval on the mean ( X ) with sample standard deviation (s)
where Z is the historical mean of the population.
and sample size (n) is as follows:
The value of reliability interval (L) and, its improvement
tα ×s tα ×s with the increase in repetitive measurements decide the
, n −1 , n −1
X− 2
≤µ≤ X + 2
(1) number of replicates. It can be interpreted as if there is a great
n n improvement in reliability interval (L) by carrying out replicates
tα ×s then only the analyst should perform it otherwise repetitive
, n −1 α measurements should be restrained to avoid unnecessary time
where 2
is the 100 ×
percentage point in t distribution
n 2 and expenditure.
with n – 1 degrees of freedom.
The test conditions corroborated with the repeatability 2.3.3 Detection of Outliers: Assignable Cause of
criteria, the burning rate measurements for individual samples Variability
were carried out with the same standard method on identical In view of the problem definition i.e. variability in burning
propellant compositions in the same laboratory using the same rate measurement, the number of replicates was always less
instruments, although, there were minor deviations pertaining than 10, and Dixon24 criteria for single outlier was the most
to the operator and the time of testing which can be considered suitable detection technique as per ASTM–E178. The equation
negligible. of the criteria is as follows:
The fundamental precision statistics for repeatability X suspect − X closest
limit is repeatability standard deviation, and it was obtained Qcal = (6)
X max − X min
by combining the standard deviation values of each set of
measurements. Repeatability standard deviation is calculated where Xsuspect is the suspected data point and Xclosest is the most
using the following equation21: nearby data point of the suspected one. The denominator
represents the range of the experimental data.
( S12 + S 22 + S32 + ....... + S n2 The critical values calculated: Qcal in each of the cases were
Repeatability Standard Deviation, S r =
n equated to the standardised statistical criterion: Q0 described
(2) by ASTM–E178. If Qcal > Qo for a certain significance level,
where S1, S2……Sn are standard deviations of measured values it is an indication that the doubtful data is an outlier24-26. All
of respective batches of identical propellant. burning rate measurement data were scanned for outliers using
Using the repeatability standard deviation (Sr), the Minitab18 statistical software27.
repeatability limits at 95% of probability, was calculated
according to the following equation21: 3. Results and Discussions
Repeatability Limit, R = 2.8 × S r (3) 3.1 Analysis of Variability
Considering the wide burning rate range of 5-31 mm The primary quantification of variability associated with
s–1 and different magnitudes of variation, the coefficient of the acoustic emission burning rate measurement technique
variation (CV) was used for a meaningful comparative study. was done by calculating the confidence interval (α=0.05)
The coefficient of variation (CV) provided a relative measure for all the burning rate data as summarised in Table 2. The
of variability and weighted the standard deviation relative to determination of confidence interval (CI) avoided the situation
the mean. of reporting a single value and provided an interval around the
average burning rate value, in which the value of true burning
2.3.2 Determination of Replication Number rate can be found with a certain confidence level, usually
The desired level of measurement precision must be
Table 2. Summary of confidence interval and repeatability
prescribed for the important quality characteristic such as limit for composite propellant burning rate from the
the burning rate of solid rocket propellant. Statistically, it is acoustic emission method
expressed in terms of margin of error (E), which is defined
as the maximum acceptable difference between the true mean Burning Average Repeatability
(µ) and the sample mean ( X ). This maximum allowable error Categories rate confidence limit
(mm s–1) interval (α=0.05) (α=0.05)
can also be described as a percentage to the historic mean or
true mean (µ) of the population which is known as reliability P01 5–6 ± 0.05 0.13
interval (L)22, 23. 6–7 ± 0.06 0.15
For t-distribution, the margin of error is expressed as: P02 14–16 ± 0.44 1.08
tα × sp 17–20 ± 0.6 1.06
, n −1
Margin of Error, E = abs ( X − µ) = 2
(4) 25–31 ± 1.42 4.01
n P03 11–13 ± 0.2 0.6

20
 Hoque, et al.: Statistical Evaluation of Burning Rate Data of Composite Propellants Obtained from Acoustic

95%, and the length of the interval conveyed the precision technique for transition metal oxide catalysed HTPB/AP/Al
of the estimation. Furthermore, the precision statement for compositions with burning rate beyond 25 mm s–1. Liu8 also
the acoustic emission method was extended by reporting the reported a similar trend of the variability in the Crawford bomb
repeatability limit corresponding to each range of measured fuse–wire burning rate data for high burning rate compositions
burning rate as illustrated in Table 2. The repeatability limit of 25–35 mm s–1.
values represented the extreme allowed difference between
two replicates of a strand burning rate sample measured using 3.2 Calculation of Sample Size
the acoustic emission method with a probability of 95%. Herein, measurement of reliability interval (L) has been
The burn rate for the first composition of P01 was used to estimate the number of replicates to be included
measured under 4.9 MPa at 24 °C, and the burn rate for the in a random sample of composite propellant to determine
rest of the propellant compositions was reported at, 6.9 MPa the burning rate with a prescribed precision using acoustic
at 24 °C. The average burning rate of non-catalysed propellant emission technique. This approach made use of the historical
composition: P01 exhibited a narrow average confidence interval average burning rate (Z) and pooled standard deviation (sp) for
of ± 0.05 and ± 0.06, and the corresponding repeatability limits identical propellant formulations for calculation of reliability
were 0.13 and 0.15 respectively. It was found that for catalysed interval as described in equation 5. First, the reliability interval
compositions: P02, the length of the average confidence interval of the burning rate of different propellant compositions was
broadened up to ± 1.42 for the burning rate range of 25–31 determined at different replicate numbers (Fig. 3). Following
mm s–1, and the associated repeatability limit was 4, which the standard procedure, the percentage difference in the
is significantly higher as compared to that of others. Active reliability interval (Ln–1–Ln) was measured with each repetitive
binder/AP/Al/Nitramine based composite propellant: P03 measurement. The results are summarised in Table 3. It has
displayed a narrow confidence interval, and the repeatability
limit was determined as 0.6.
The scope of confidence interval and repeatability
limit is generally limited to the one-dimensional variability
of the method, and they are not statistically appropriate
for comparison of variability associated with the different
magnitude of measurement scale. For instance, the
confidence interval and repeatability limit of the burning
rate of range 5–6 mm s–1 should not be directly compared
with that of a high burning rate of 25–31 mm s–1. In this
context, the coefficient of variation (CV) was used to carry
out a statistically eloquent comparative study of variability
associated with the measurement of solid strand burning
rate by acoustic emission method. Figure 2 outlined the
two-dimensional tightness of the burning rate data across
all measurement scale in the form of the coefficient of
variation vs. burning rate plot. Figure 2. Plot of the coefficient of variation against the burning rate
On comparison of the coefficient of variation, it was of different propellant compositions ranging 5–31 mm s–1.
found that the non–catalysed propellant composition:
P01 showed the most precise burning rate result, and the
determined average coefficient of variation was 1% for the
data sets comprising 500 data points. For the P02: catalysed
propellant composition, the acoustic emission method
generated the coefficient variation of average 2% for the
burning rate range of 14–20 mm s–1. For active binder/
AP/Al/Nitramine based P03 compositions the average
coefficient of variation was also around 2%. Moreover,
the coefficient of variation was mostly less than 5% for
the propellant compositions having a burning rate range
of 5–20 mm s–1, these values indicated the reliability of
the acoustic emission solid strand burning rate technique
for the propellant compositions with burning rate range of
5–20 mm s–1.
However, the variability drastically increased for
the high burning rate compositions of 25–31 mm s–1. The
average coefficient of variation was found to be more Figure 3. Plot of reliability interval against replicate numbers for the
than 5% which is statistically undesirable, and it raised burning rate of different propellant compositions ranging
a question about the applicability of acoustic emission 5–31 mm s–1.

21
 Def. SCI. J., Vol. 71, No. 1, January 2021

Table 3. Values of reliability interval (RI) and percentage difference with an increase in replicate numbers of burning rate samples
for all the propellant compositions
Categories P01 P02 P03
Burning rate (mm s ) –1
5–6 6–7 14–16 17–20 25–31 11–13
Replicate size (n) L % Diff L % Diff L % Diff L % Diff L % Diff L % Diff
2 8.3 6.9 22.1 21.2 46.3 16.2
3 2.3 6.0 1.9 5.0 6.1 16.0 5.9 15.4 12.8 33.5 4.5 11.7
4 1.5 0.8 1.2 0.7 3.9 2.2 3.8 2.1 8.2 4.6 2.9 1.6
5 1.1 0.3 0.9 0.3 3.1 0.9 2.9 0.8 6.4 1.8 2.2 0.6
6 1.0 0.2 0.8 0.1 2.6 0.5 2.5 0.5 5.4 1.0 1.9 0.3
7 0.9 0.1 0.7 0.1 2.3 0.3 2.2 0.3 4.8 0.6 1.7 0.2
8 0.8 0.1 0.6 0.1 2.1 0.2 2.0 0.2 4.3 0.5 1.5 0.2
9 0.7 0.1 0.6 0.1 1.9 0.2 1.8 0.2 4.0 0.3 1.4 0.1
10 0.7 0.0 0.5 0.0 1.8 0.1 1.7 0.1 3.7 0.3 1.3 0.1

been observed that after a certain number of repetitive Table 4. Summary of identification of ‘Outliers’ in the composite propellant
measurements, there was no significant improvement burning rate data from the acoustic emission method
in reliability interval. Burning Rate No. of No. of Batches Probability
The reliability interval data suggested that for Categories
(mm s–1) batches with ‘Outliers’ (%)
P01: non–catalysed propellant formulations, the
P01 5–7 104 8 7.6
reliability interval was substantially improved by
carrying out the third repetitive measurement i.e. 6% P02 14–31 100 7 7
and 5% respectively, and no substantial improvement P03 11–13 13 1 7.6
was achieved by performing more than three replicates.
Hence, three replicates would be sufficient for P01 Despite huge variability corresponding to high burning
compositions to determine the burning rate with considerable rate compositions: 25–31 mm s–1, moderate contamination
accuracy. in the full set of burning rate data was found. Thus, the
The comprehensive results of Table 3 advocated carrying apparently large coefficient of variation (CV) observed in
out four replicate measurements for the first two propellant the experimental data of high burning rate composition was
compositions of P02 (catalysed by burning rate modifier) not due to the presence of outlying observation, rather it
and active binder/AP/Al/Nitramine based P03 formulation might be an extreme manifestation of the random variability
to obtain a strand burning rate with an acceptable level of inherent to that particular propellant composition.
accuracy. However, it was found that the acoustic emission In addition, the probability of the appearance of
technique of burning rate measurement would require at least outlying observation for a single batch was around 7% for
seven repetitive measurements for the propellant composition all the propellant compositions. Table 5 depicts the extent
having a burning rate range of 25–31 mm s–1 to produce a of contamination in the average burning rate and standard
more explicit result. The percentage difference rendered deviation due to the presence of outlying observations in the
improvement around 2% and 1% with carrying out fourth and sample.
fifth repetitive measurements respectively. The choice of seven There was no significant contamination in the average
replicates was attributed to the value of reliability interval burning rate for all the compositions, but the precision of the
which should be less than 5% as better statistical practice. data set drastically increased when the outlying observations
were removed. For instance, when the ‘Outlier’ was omitted,
3.3 Identification of ‘Outliers’ in the Experimental
Data Table 5. The extent of contamination in average burning
rate and standard deviation due to the presence of
The origin of observed measurement variability in the
‘Outliers’ in the composite propellant burning rate
burning rate data was further investigated. The common data
assignable cause of measurement variability is one-time
systematic errors in the experimental procedure of burning Contamination due to ‘Outliers’
Burning Min to Max
rate determination, and it may lead to aberrant observations; in Categories rate
statistics, they are known as ‘Outliers’. These anomalous data Average burning Standard deviation
(mm s–1)
points result in contamination in the accuracy and precession rate (mm s–1) (mm s–1)
statistics of burning rate measurement. P01 5–7 0.01 to 0.04 0.02 to 0.07
The statistical criterion of the Dixon Q test was applied to P02 14–31 0.11 to 0.63 0.17 to 1.00
the full set of data to identify the outliers, and the observations
are summarised in Table 4. P03 11–13 0.13 0.24

22
 Hoque, et al.: Statistical Evaluation of Burning Rate Data of Composite Propellants Obtained from Acoustic

the maximum change in average burning rate for P02 Propellants, Explos. Pyrotech., 1991, 16(1), 43-47.
composition was 0.63 mm s–1, but the improvement in standard doi: 10.1002/prep.19900150603.
deviation was as high as 1 mm s–1. Hence, from a quality 10. Hoque, E., Pant, C. S. & Das, S. Study on Friction
control perspective, it is imperative to identify the outlying Sensitivity of Passive and Active Binder based Composite
observations in the experimental data to rectify, adjust, and Solid Propellants and Correlation with Burning Rate. Def.
improve the measurement procedure. Sci. J., 2020, 70(2), 159-165.
doi: org/10.14429/dsj.70.14802.
4. Conclusion 11. Bozic, V. S. Measurement System for Determining
In summary, the variability in 1100 repeated measured Solid Rocket Propellant Burning Rate Using Reflection
acoustic emission solid strand burning rate data was assessed Microwave Interferometry. J. Propul. Power, 1997, 13(4),
and quantified using confidence interval (CI) and repeatability 457–462.
limit (R). The use of the coefficient of variation (CV) indicated doi: 10.2514/2.5197.
that the acoustic emission method might not be the appropriate 12. Deepak, D.; Jeenu, R.; P. Sridharan, P. & Padmaabhan,
methodology for measuring the burning rate of more than M. S. Application of Ultrasonic Burn Technique Rate
25 mm s–1 for HTPB based composite propellants, catalysed by for Measurement Propellants of Instantaneous of Solid
transition metal oxides. Furthermore, statistically reasonable Propellants. Def. Sci. J., 1998, 48(2), 197–204.
sample numbers were obtained for steady burning rate doi: 10.14429/dsj.48.3899
measurement of composite propellants having different ranges 13. Gupta, G.; Jawale, L. & Mehilal, Bhattacharya, B. Various
of burning rates. Methods for the Determination of the Burning Rates of
Additionally, the whole data set was investigated for Solid Propellants - An Overview. Cent. Eur. J. Energ.
the identification of ‘Outliers’ and moderate contamination Mater., 2015, 12(3), 593–620.
was noted. However, it has been envisioned that the ‘Outlier’ 14. STANAG-4674: Non-Intrusive Methods for Measuring
detection statistics should be implemented as a regular quality the Burning Rate of Solid Rocket Propellants. NATO
control activity to prevent unintended adulteration in the Standardization Office, Brussels, Belgium 2018.
burning rate measurement data. 15. Caveny, L. H.; Pokrocos, L. M. & Felshein, C. R. High
Pressure Burning Rate of Multi-base Propellants. Report
References AMS-1377, Department of Aerospace and Mechanical
1. Kubota, N. Propellants and Explosives. Wiley-VCH, Sciences, Princeton University, New Jersey, USA 1977.
Germany, 2007. 16. Koury, J. L. Solid Strand Burn Rate Technique for
2. Ishita, K. & Ramakrishna, P. A. Studies on the role of iron Predicting Full-Scale Motor Performance. Report AF-
oxide and copper chromite in solid propellant combustion. RPL-TR-73-49, Air Force Rocket Propulsion Laboratory
Combust. Flame, 2014, 161(10), 2717–2728. Edwards Air Force Base, California, USA 1973.
doi: 10.1016/j.combustflame.2014.03.015. 17. Christensen, W. N. Development of an Acoustic
3. Krishnan, S. & Jeenu, R. Combustion characteristics of Emission Strand Burning technique for Motor Burning
AP/HTPB propellants with burning rate modifiers. J. Rate Prediction. In AIAA/SAE 14th Joint Propulsion
Propul. Power, 1992, 8(4), 748–755. Conference, Las Vegas, NV, USA, July 25- 27, 1978.
doi: 10.2514/3.23545. 18. Rampichini, S.; Ruspa, D. & DeLuca, L. T. Acoustic
4. Fry, R. S.; DeLuca, L.; Frederick, R.; Gadiot, G.; Emission of Underwater Burning Solid Rocket Propellants.
Strecker, R.; Besser, H. L.; Whitehouse, A.; Traineau, J. In Combustion of Energetic Materials, edited by K. K.
C.; Ribereau, D. & Reynaud, J. P. Evaluation of Methods Kuo & L. T. DeLuca. Begell-House Inc, New York, 2002,
for Solid Propellant Burning Rate Measurement. Report pp. 274–283.
NATO RTO-MP-091, NATO RTO AVT Working Group 19. Blair, D. W. The Influence of Diameter of the Burning
016, August 2001. Velocity of Strands of Solid Propellants. Combust. Flame,
5. Crawford, B. L. & Huggett, C. Direct Determination of 1973, 20(1), 105-109.
Burning Rates of Propellant Powders. Anal. Chem., 1947, doi: 10.1016/S0010-2180(73)81261-1.
19(9), 630–633. 20. Herrington, L. E. Correlation of Motor and Strand
6. Sutton, G. P. & Biblarz, O. Rocket Propulsion Elements. Composite Propellant Burning Rate. AIAA J., 1964, 2(9),
Wiley-Interscience Publication, New York, 2001. 1671-1673.
7. Jordan, F. W. Burn Rate Factors to Consider. In Proceedings doi: 10.2514/3.2650.
of 34th JANNAF Combustion Meeting, West Palm Beach, 21. ASTM Standard E691-18, Standard Practice for
Florida, 1997. Conducting an Interlaboratory Study to Determine the
8. Liu, T. Correlations of Uncertainties of Composite Precision of a Test Method, ASTM International, West
Propellant Strand Burner Burning Rate Measurement for Conshohocken, PA, USA, 2018.
Quality Control. Propellants, Explos. Pyrotech., 2011, 22. Jeffery, G. H.; Bassett, J.; Mendham, J. & Denney, R.
36(2), 131–139. C. Vogel’s Textbook of Quantitative Chemical Analysis.
doi: 10.1002/prep.200900098. Longman Scientific & Technical, New York, 1989, pp.
9. Bazaki, H. & Kubota, N. Friction sensitivity mechanism 142–143.
of ammonium perchlorate composite propellants. 23. ASTM Standard E122-17, Standard Practice for

23
 Def. SCI. J., Vol. 71, No. 1, January 2021

Calculating Sample Size to Estimate, With Specified Contributors

Precision, the Average for a Characteristic of a Lot or
Process, ASTM International, West Conshohocken, PA, Mr Ehtasimul Hoque has completed MSc (Chemistry) from
USA, 2017. IIT Bombay in 2016. His current areas of work include quality
control and characterisation of high energetic materials.
24. Dixon, W. J. Processing Data for Outliers, Biometrics.
In the current study, he has carried out all the statistical
1953, 9(1), 77–89. analysis, and prepared the manuscript.
25. ASTM Standard E178-16a, Dealing With Outlying
Observations, ASTM International, West Conshohocken, Dr Chandra Shekhar Pant did his doctoral research in propellant
PA, USA, 2016. chemistry from Defence Institute of Advanced Technology,
26. Grubbs, F. E. Procedure for Detecting Outlying Pune. He is associated with the area of processing, testing and
Observations in Sample. Technometrics, 1969, 11(1), evaluation of high energy materials for 15 years.
1–21. In the present work he contributed for review of test data and
27. Minitab 18 Statistical Software, version 18.0, Minitab manuscript correction.
Inc, State College, PA, USA.
Mr Sushanta Das is MTech in Chemical Engineering from
IIT Bombay. He is associated with research and production of
ACKNOWLEDGEMENT high energy materials for last 18 years.
The authors are thankful to Mr Srinivasan Seshadri, GM, Contribution in the current study, he has planned and guided
ACEM for the encouragement to carry out the study, and the work.
permission to publish the work.

24