in observed values that can occur depending upon the experi-Data obtained in this type of study can be plotted to yield inmental conditions employed to measure the ignition points. Ratio teresting information, as shown in Figure 5.7.
106
Chemistry of Pyrotechnics
Ignition and Propagation
107
100
200
300
400
500
REFERENCE TEMPERATURE, °C
FIG. 5.5 Thermogram of pure potassium chlorate, KCIO
FIG. 5.6 The potassium chlorate/sulfur system. Sulfur endo-3 . No
thermal events are observed prior to the melting point (356°C).
therms are seen near 105° and 119°C, as expected. A violent Exothermic decomposition occurs above the melting point as oxy-exothermic reaction is observed below 150 1C. The ignition tem-gen gas is liberated.
perature is approximately 200 degrees below the melting point of the oxidizer (KC1O 3 m.p. = 356°C). Ignition occurs near the temperature at which S 3 molecules fragment into smaller units.
Data from time versus temperature studies can also be plotted as log time vs. 1/T, yielding straight lines as predicted by the Arrhenius Equation (eq. 2.4). Figure 5.8 illustrates this con-Ignition temperatures can also be determined by differential cept, using the same data plotted in Figure 5.7. Activation en-thermal analysis (DTA), and these values usually correspond well ergies can be obtained from such plots. Deviations from linear to those obtained by a Henkin-McGill study. Differences in heat-behavior and abrupt changes in slope are sometimes observed in ing rate can cause some variation in values obtained with this Arrhenius plots due to changes in reaction mechanism or other technique. For any direct comparison of ignition temperatures, complex factors.
it is best to run all of the mixtures of interest under identical
"Henkin-McGill" plots can be quite useful in the study of ig-experimental conditions, thereby minimizing the number of vari-nition, providing us with important data on temperatures at which ables.
spontaneous ignition will occur. These data can be especially use-One must also keep in mind that these experiments are mea-ful in estimating maximum storage temperatures for high-energy suring the temperature sensitivity of a particular composition, compositions - the temperature should be one corresponding to in which the entire sample is heated to the experimental tempera-infinite time to ignition (below the "spontaneous ignition temperature. Ignition sensitivity can also be discussed in terms of the ture," minimum - S.I.T (min) - shown in Figure 5.7). At any relative ease of ignition due to other types of potential stimuli, temperature above this point, ignition during storage is possible.
including static spark, impact, friction, and flame.
108
Chemistry of Pyrotechnics
Ignition and Propagation
109
TABLE 5.3 Ignition Temperatures of Pyrotechnic Mixtures TABLE 5.4 Ignition Temperatures of Magnesium-Containing Mixturesa
Ignition
temperature,
Ignition
Component a
Melting point, oC
oC
temperature,
Oxidizer
o C b
I.
KC1O 3
356
150
S
119
NaNO 3
635
II.
KC10 3
356
195b
Ba(N0 3 ) 2
615
Lactose
202
Sr(N0 3 )2
610
III.
KC1O 3
356
540b
Mg
649
KNO 3
650
IV.
KNO
KC10 4
715
3
334
390b
Lactose
202
Note:
All mixtures contain 50% magnesium by weight.
V.
KNO 3
334
340
aReference 5.
S
119
bLoading pressure was 10,000 psi.
VI.
KNO 3
334
565b
Mg
649
VII.
BaCr0 4 (90)
Decomposition at
685c
high temperatures
B (10)
liberate sufficient energy, in a sulfur mix, to generate a self-2300
propagating process. A greater quantity of material must react at once to produce ignition.
aMixtures were in stoichiometric proportions unless other-Another important factor is the thermal stability and heat of wise indicated.
decomposition of the oxidizer. Potassium chlorate mixtures tend bReference 1.
to be much more sensitive to ignition than potassium nitrate com-CReference 4.
positions, due to the exothermic nature of the decomposition of KC1O 3 . Mixtures containing very stable oxidizers - such as ferric oxide (Fe 2O 3 ) and lead chromate (PbCr0 4) - can be quite difficult to ignite, and a more-sensitive composition frequently has to be used in conjunction with these materials to effect ig-SENSITIVITY
nition.
A mixture of a good fuel (e.g., Mg) with an easily-decomposed Sensitivity of a high-energy mixture to an ignition stimulus is in-oxidizer (e.g., KC1O 3 ) should be quite sensitive to a variety of fluenced by a number of factors. The heat output of the fuel is ignition stimuli.
A composition with a poor fuel and a stable ox-quite important, with sensitivity generally increasing as the fuel's idizer should be much less sensitive, if it can be ignited at all!
heat of combustion increases. Mixes containing magnesium or alu-Ignition temperature, as determined by DTA or a Henkin-McGill minum metal, or charcoal, can be quite sensitive to static spark study, is but one measure of sensitivity, and there is not any or a fire flash, while mixes containing sulfur as the lone fuel are simple correlation between ignition temperature and static spark usually less sensitive, due to the low heat output of sulfur. Ig-or friction sensitivity.
Some mixtures with reasonably high ig-
nition of a small quantity of material by static energy does not nition temperatures (KC1O 4 and Al is a good example) can be
110
Chemistry o f Pyrotechnics
Ignition and Propagation
111
1 50
200
250
300
TEMPERATURE, °C
FIG. 5.7
Time to explosion versus temperature for nitrocellulose. As the temperature of the heating bath is raised, the time to explosion decreases exponentially, approaching an instantaneous value. The extrapolated temperature value corres-FIG. 5.8
"Henkin-McGill Plot" for nitrocellulose. The natural ponding to infinite time to explosion is called the spontaneous logarithm of the time to ignition is plotted versus the reciprocal ignition temperature, minimum (S.I.T. min). Source of the of the absolute temperature (°K). A straight line is produced, data: reference 6.
and activation energies can be calculated from the slope of the line. The break in the plot near 2.1 may result from a change in the reaction mechanism at that temperature. Source: reference 6.
quite spark sensitive, because the reaction is highly exothermic and becomes self-propagating once a small portion is ignited. Sensitivity and output are not necessarily related and PROPAGATION OF BURNING
are determined by different sets of factors. A given mixture can have high sensitivity and low output, low sensitivity and Factors
high output, etc. Those mixtures that have both high sensi-The ignition process initiates a self-propagating, high-tempera-tivity and substantial output are the ones that must be treated ture chemical reaction at the surface of the mixture. The rate with the greatest care. Potassium chlorate/sulfur/aluminum at which the reaction then proceeds through the remainder of