By R. Bunsen and L. Schischkoff.
Translated by Chas. E. Munroe, Prof. U. S. N. A.
In spite of the apparent simplicity which the phenomena of the combustion of powder present, a combustion upon which the mechanical effect produced depends, we are far from knowing exactly all the circumstances which surround it. The researches of Gay Lussac, completed more than thirty years ago, are the first and the most complete which were undertaken upon this subject; more recent investigations have led to results so contradictory that it has been impossible, up to the present day, to devise a chemical theory which would agree with the experimental results.
Let us admit that, practically the normal composition of powder corresponds to two molecules of saltpetre, one atom of sulphur and three atoms of carbon. If all the carbon is converted into carbon dioxide and all the nitrogen is set free we should have the following equation.
2KNO3+S+3C=3CO2+2N+K2S
Then one gram of powder containing
Saltpetre .7484 grms.
Sulphur .1184
Carbon .1332
1.0000
Will give after its explosion
Potassic sulphide 0.4078 grms
Nitrogen 0.1037 grms = 82.52cm3
Carbon dioxide 0.4885 grms = 248.40cm3
1.0000 330.92cm3
The volume of the gases will remain constant even though carbon protoxide should be formed in the place of carbon dioxide and nitrous oxide in the place of free nitrogen. Now if we except some slight traces of hydrogen and hydric sulphide and if there are no other gases produced but carbon dioxide, carbon protoxide, nitrogen and nitrous oxide we can regard this volume of gas 330.92 cm3 as being the maximum volume which can be formed by the combustion of one gram of powder having the normal composition. Gay Lussac and other more recent observers have found volumes much greater than this. These contradictions show the imperfections of the methods used in their researches and have led us to return to the question while employing a more exact method of research.
We set out to determine
1. The composition of the solid residue from the combustion of the powder.
2. The pulverulent matters which form the smoke.
3. The substances composing the gas.
4. The relative quantities of solid residue and of smoke for a given weight of powder.
5. The quantity of heat developed and the temperature of the flame.
6. The pressure exerted by the gas, the explosion having taken place in the space occupied by the powder in the state of grains, and supposing that no heat is lost by radiation or conduction.
7. The theoretical work which the powder can produce.
On account of the limited amount of time which we could command for our common work we have not been able to determine these experimental data but for a single species of powder, and under the ordinary atmospheric pressure. Therefore we offer our work only as an application of the method which we have followed, a method which could, with some slight modifications, be applied to the study of the combustion of powder under other circumstances.
We have determined the composition of the powder which we have employed by treating it first with water, which gave us the weight of saltpetre; the insoluble residue was then treated with carbon disulphide which dissolved out the sulphur; and the new residue burned with cupric oxide yielded the quantities of carbon, hydrogen and oxygen which it contained. We have found in this way for this powder:
Saltpetre 78.99 per cent.
Sulphur 9.84
Carbon 7.69
Hydrogen 0.41
Oxygen 3.07
Ash traces
100.00
We have used our different apparatuses according as we have wished to determine qualitatively or quantitatively the nature of the products of the combustion.
Qualitative Analysis.
For the qualitative analysis we have employed the following apparatus (Fig. 1, see plate.) We took a glass tube sufficiently large, closed at one end by a stopper (b) and carrying at the other end a tube for conducting off the gases. Through the stopper a small brass tube (a), 250 mm. long and 2 mm. in diameter, passes, in which the pulverized powder is put. The powder is ignited and when it begins to issue in a regular jet from the tube the stopper (b) is filtered into the large glass tube. The flame from the powder is liable to break the tube (d) on account of the high temperature which is developed in the part of the tube which surrounds the powder. In order to avoid this the tube (a) is enclosed quite to its extremity in a tube of perforated sheet iron or, what is quite as good, another tube of thin glass. The solid residue and the substance held in the pulverulent condition by the gas, under the form of smoke, remain in the tubes (a) and (d), the gases are received over a vessel of pure mercury after they have been allowed to pass freely until they have driven all of the air from the tube (d). The analysis showed that the solid residue was composed of
- Potassic sulphate
- Potassic carbonate
- Potassic hyposulphite
- Potassic sulphide
- Potassic hydrate
- Potassic sulphocyanate
- Potassic nitrate
- Carbon or charcoal
- Sulphur
- Ammonic carbonate
The gaseous products contain the following gases:
- Nitrogen
- Carbon dioxide
- Carbon protoxide
- Hydrogen
- Hydric sulphide
- Notable quantities (according to circumstances) of nitric oxide and even of nitrous oxide.
Quantitative Analysis.
In order to obtain easily and without danger a considerable quantity of these bodies for quantitative examination we have employed the apparatus shown in Fig. 2. It is composed of a glass bulb (d), which is heated by the aid of a gas lamp, in which the combustion is effected and where the solid residue remains. To one of the bent extremities of this bulb is adapted a glass tube 2.5 mm. in diameter and 1 m. long. To the upper part of this tube a brass collar is cemented to which is attached a vulcanized rubber tube (a) capable of holding from 15 to 20 grms. of powder. The collar (b) contains a diaphram pierced with a very narrow circular hole which will not allow the powder to run into the tube (c) except in the form of a very fine stream like the stream of sand which runs through an hour glass. The bulb (d) opens into a large glass tube, 1.5 to 2 metres long and 25 mm. in diameter, which serves to increase its length, where the pulverulent matters which are held in the smoke are deposited. In order to make the powder run in a continuous stream the caoutchouc tube is agitated by the aid of a small stick fixed to one of its ends as represented in Fig. 2. The combustion of the powder goes on quietly in the bulb when the flow takes place continuously, but interruptions in the flow by no means prevent the success of the experiment. The gaseous products are carried off into the atmosphere. They cannot be collected over either water or mercury, for the pressure which is exerted when the end of the conducting tube is placed under the liquid causes the flame produced by the combustion of the powder in the bulb (d) to rise through the glass tube (c) and explode all the powder contained in the tube (a). The sudden explosion of 15 to 20 grms. of powder is certainly very violent but it is not at all dangerous. The caoutchouc tube in fact offers only a feeble resistance and goes to pieces immediately while the glass tubes, on the contrary, resist very perfectly. Explosions will also occur if the opening in the bulb becomes obstructed during the evolution of the gas. In order to collect easily the gaseous products we have introduced into the extension (e,e) a glass tube which is joined to the bulb tubes (s,s) by means of rubber connecting tubes and the bulb tubes communicate with an aspirator. The bulb tubes (s,s) closed at first by pinch cocks are finally hermetically sealed by fusing their extremities in a lamp flame. We have analyzed separately 1st, the solid residue which is found in the bulb; 2d, the pulverulent deposit which is found in the extension (e,e); 3d, the gas collected in the tubes (s,s).
A. Residue from the Combustion.
It forms a grayish yellow, semi-fused mass which is soluble in water and leaves a slight residue of carbon. The following is the method of analysis which we pursued:
1. Carbon. This was collected and weighed on a tared filter, 7 grms. of the original substance dissolved in warm water being used. We thus obtained 0.0682 grms. which corresponds to 0.974 per cent.
2. Potassic sulphide. The solution separated by filtration from the carbon was placed in contact with calcined cupric oxide for two days, and agitated from time to time. The liquid, brown at first, finally becomes completely clear. The cupric sulphide, together with the excess of cupric oxide was collected upon a filter and dissolved in fuming nitric acid. Finally the sulphur, dissolved in the condition of sulphuric acid, was precipitated by the aid of baric nitrate. We obtained 0.1567 grms, of bark sulphate which corresponds to 2.239 per cent, of baric sulphate or to 1.058 per cent, of potassic sulphide, or to 1.077 of potassic hydrate or to 0.9043 per cent, of potassic oxide and to 0.308 per cent, of sulphur.
3. Potassic hyposulphite. For the following researches we have separated the filtrate obtained from the previous analysis into seven portions each of which, consequently, contains 1 grm. of the original substance deprived of its sulphur and carbon.
We treated the potassic hyposulphite with a titrated solution of iodine after having added to the original solution sufficient acetic acid to give a decided acid reaction. Let a be the quantity of iodine contained in one degree of the burette (a= 0.001242 grms.) and t the number of degrees necessary for imparting a blue color to the original solution after the addition of a little starch. We have then the following formula for determining the quantity of hyposulphite.
s=at(K2S2O3/I)
The reaction which takes place being as follows
2I+2K2S2O3=2KI+K2S4O6
We found t=40.7, a=0.001242 grms. Then the quantity of hyposulphite=40.001859 x 40.7=0.07566 grms. or 7.566 per cent., corresponding to 3.747 per cent of potassic oxide or to 2.5452 per cent, of sulphur and 8.0434 per cent, of potassic nitrate.
4. Potassic sulphate. The same quantity of the solution containing 1 grm, of the substance for analysis gave us 0.7582 grins, of bane sulphate which will give, for 100 grms. of the original substance, 56.623 of potassic sulphate corresponding to 10.397 of sulphur, or 30.624 of potassic oxide, or 65.721 of saltpetre.
5. Potassic sulphocyanate. Small quantities of this substance can be readily treated by calorimetric processes. For this purpose we prepared a solution containing in each degree of the burette 0.0004894 grrns, of potassic sulphocyanate. As the coloring liquid we used ferric chloride, conveniently diluted and acidified by hydrochloric acid. The quantity of iron which it contained was not ascertained. We took two equal volumes of it which we poured into two similar cylindrical vessels placed in the same position. To one of these we added the liquid containing the sulphocyanate to be tested, in the other we poured the normal solution of the sulphocyanate until the two liquids presented exactly the same color. At the same time we took the precaution to add water to the second vessel in order that the two liquids should be equally diluted. Let a be the quantity, by weight, of potassic sulphocyanate contained in one degree of the burette, and t the number of degrees which are required to render the tint of the liquid identical in the two glasses, we should have for the quantity of sulphocyanate
s=at,
a being equal to 0.0004894 one would have by this method of treating the sulphocyanate at least 0.0005 grms. We found t=17.5 which gives for 100 of the original substance 0.8564 of sulphocyanate, corresponding to 0.4145 of potassic oxide, or 0.2815 of sulphur, or 0.8896 of potassic nitrate.
6. Treatment of the ammonia. The ammonia is determined by the aid of a titrated solution of hydrochloric acid. The quantity of hydrochloric acid contained in the normal solution of this acid is first determined by means of a solution of silver. The liquid employed contained in one cubic centimetre 0.002357 grms, of anhydrous acid. Two volumes of this liquid, equal to 23.81 cm), are taken and one of the two is poured into a bottle which is put into a test glass filled with cold water. The solution containing the ammonia is introduced into a retort (Fig. 3) containing potash free from nitrates and submitted to distillation. The conducting tube leads the steam generated into the bottle e where it condenses, thus carrying off all the ammonia from the original liquid. Iii order to prevent the absorption of the ammonia in the conducting tube a small vulcanized rubber tube is adapted to the extremity of the latter. This is closed by a small rod of glass and then slit up a slight distance with a pen knife. This opening allows the vapor to pass out but it becomes hermetically sealed each time that a vacuum is formed inside the tube by cooling. When all the ammonia has passed over, which is determined by the volume of water which has distilled, the bottle (e) is allowed to cool and its contents are poured into a test glass. In a similar glass the second measured portion of hydrochloric acid is poured, and it is diluted with water until the volume of the two liquids is the same in both vessels. Equal quantities of litmus solution are added to the two solutions and the acid is treated with a titrated solution of ammonia of such a strength that twenty degrees of the burette are sufficient to neutralize the volume of acid used. Let a be the quantity of acid contained in the volume t which is used, and t be the number of degrees used for neutralizing the acid in the two cases, then the quantity of ammonia which has distilled over corresponds to a quantity of hydrochloric acid equal to
a(t1-t)/t1
and the quantity of ammonic carbonate contained in the liquid will be given by the formula
X=(NH4)2CO3/2HCl a t1-t/t1
We obtained equal values for t and t1, which showed that there was no ammonic carbonate in the original solution.
7. Potassic nitrate. Having proved that the original solution contained no ammonia, we took, as in the preceding case, a volume containing one grm. of the original substance, acidulated it slightly with sulphuric acid, and placed a strip of zinc in it. The liquid was kept cool and sulphuric acid was added to it from time to time to such a degree that the evolution of gas went on slowly. The potassic nitrate was thus converted into ammonia. Potassic hydrate was then added in sufficient quantity to redissolve all the oxide of zinc precipitated 4 and the ammonia was distilled as in the preceding case. Preserving the same notation the potassic nitrate is determined by the aid of the formula
X=KNO3/HCl a t1-t/t1
We found a=0.05612 grms., t1=27, t=18 from which we deduce x = 0.05185 grms.
8. Potassic carbonate. A. volume of the original liquid which contains 1 grm. of the residue is taken and the potassic carbonate is precipitated by the aid of a solution of manganous chloride, which has been previously fused. The precipitate is filtered off, washed, and then, together with the filter paper, introduced into a proper apparatus for treating the carbon dioxide. Dilute sulphuric acid is added and it is boiled for sometime. The carbon dioxide, determined by the loss of weight of the apparatus, weighed 0.0860 grms. which corresponds to 27.016 per cent, of potassic carbonate or to 2.3452 of carbon, to 18.417 of potassic oxide and 39.525 of potassic nitrate.
9. Potassic hydrate. By precipitating again the manganous oxide dissolved in the displacing apparatus (8), by the aid of sodic carbonate, we obtained 0.1654 grms. of Mn3O4. As a molecule of Mn3O4 corresponds to three molecules of manganous carbonate, for 0.0860 grms. of carbon dioxide we should have 0.1495 grms. of Mn3O4. On deducting this number from 0 1654 grms. which represent the total weight of the manganoso-inanganic oxide precipitated, there remain 0.0159 grins. This amount was not precipitated as manganous carbonate but in the condition of mangauons hydrate on account of the presence of the potassic hydrate. But, as Mn3O4 corresponds to three MnO and as for this quantity of manganous oxide six molecules of potassic hydrate are necessary, 0.0159 grms. of Mn3O4 correspond to 2.339 grms. of potassic hydrate. Again in the determination of the potassic sulphide we saw that this was transformed into potassic hydrate which is evidently so much to be deducted from the quantity which we have found.
Total quantity of potassic hydrate in 100 2.339
Quantity corresponding to potassic sulphide in 100 parts 1.077
Remainder 1.262
which is equivalent to 2.274 of potassic nitrate or to 1.0596 of potassic oxide.
10. Determination of the total potassic oxide in the residue. A volume of the liquid, containing 1 grm. of the residue, after haying been treated with sulphuric acid, dried and ignited, gave 1.0447 grins, of potassic sulphate corresponding to 56.497 per cent, of potassic oxide.
Summing up the results of these different analyses we have for the composition of the residue in per cents
Potassic sulphate 56.62
Potassic carbonate 27.02
Potassic hyposulphite 7.57
Potassic sulphide 1.06
Potassic hydrate 1.26
Potassic sulphocyanate 0.86
Potassic nitrate 5.19
Carbon 0.97
Ammonic carbonate 0.00
Sulphur 0.00
The proportion of potassic oxide contained in the different salts which compose the residue is 56.88; the proportion determined directly in analysis 10 is 56.50; the difference is only 0.38.
The residue from the combustion consists then chiefly of potassic sulphate and carbonate and not of potassic sulphide, as is stated in most of the treatises on artillery and chemistry, this substance reaching scarcely one per cent.
B. Pulverulent Substances in the State of Smoke.
We have proceeded as follows for analyzing the pulverulent deposit held by the gas in the state of smoke and which was deposited in the long tube (e e) towards its open end. This matter formed a gray adherent coating and smelled strongly of ammonia. We dissolved it immediately in water, then separated the suspended charcoal by filtration and then divided the liquid into (eleven) equal portions which we employed separately for the following treatment.
1. Carbon. The carbon contained in the whole liquid weighed 0.08526 grms, and was free from sulphur. One part of the solution or 1-11 of the total mass contained then 0.00775 grms. carbon.
2. Potassic sulphide. The solution did not blacken paper which had been dipped in plumbic acetate and hence it did not contain any potassic sulphide.
3. Potassic hyposulphite. One of the portions of the solution required 11 degrees of the burette, containing a titrated solution of iodine, which corresponds to 0.02045 grms. of hyposulphite.
4. Potassic sulphate. One of the portions of the solution gave 0.3650 grms. of basic sulphate, corresponding to 0.2726 grms. of potassic sulphate.
5. Potassic sulphocyanate. One portion of the solution required 4.7 degrees of the normal solution which corresponds to 0.0023 grms. of sulphocyanate.
6. Ammonia. Two volumes of the solution gave us t1=27.0, t=23.9, a=0.05612, which corresponds to 0.0004373 grms. of ammonia sesquicarbonate or 0.0002445 grms. of carbon dioxide for one part of the solution.
7. Potassic nitrate. One portion of the solution gave t1=27, t=25.2, a=0.05612 which corresponds to 0.010374 grms. of potassic nitrate.
8. Potassic carbonate. Two portions of the solution gave us 0.0629 grins, of carbonic anhydride. If we subtract the carbonic anhydride of the ammonia carbonate found in experiment (6) we get for one portion of the solution 0.09803 grms. of potassic carbonate.
9. Potassic hydrate. Two volumes of the solution gave us 0.1169 grms. of manganous oxide which corresponds to 0.05845 grms. Mn3O4 for one portion of the solution. If we subtract from this the quantity of Mn3O4, (0.05466 grms.) which corresponds to the 0.03145 grms. of carbonic anhydride found in experiment (8), there remains 0.00379grms. of Mn3O4 corresponding to 0.00556 grms. of potassic hydrate.
As a result of these analyses we find one portion of the solution to contain:
Potassic sulphate 0.27258 grms.
Potassic carbonate 0.09803
Potassic hyposulphite 0.02045
Potassic sulphide 0.00000
Potassic hydrate 0.00556
Potassic sulphocyanate 0.00230
Potassic nitrate 0.01037
Ammonic sesquicarbonate 0.00044
Carbon 0.00775
Sulphur 0.00000
0.41748
One hundred parts of the solid substance which was dissolved contain then:
Potassic sulphate 65.29
Potassic carbonate 23.48
Potassic hyposulphite 4.90
Potassic sulphide 0.00
Potassic hydrate 1.33
Potassic sulphocyanate 0.55
Potassic nitrate 2.48
Ammonic sesquicarbonate 0.11
Carbon 1.86
Sulphur 0.00
100.00
In order to control these results we have treated a volume of the solution with sulphuric acid and we have thus converted all of the salts of potassium into sulphate. We have thus found 0.4286 grms. Of K2SO4; according to the preceding analysis we should have obtained 0.4345 grms. The accordance of these two numbers confirm the accuracy of the analysis. On comparing the results of this analysis with those which the analysis of the residue gave we arrive at the following conclusions:
1. The pulverulent matters which form the smoke have an analogous composition to those of the solid residue.
2. The principal difference consists in this, that the combustion of the sulphur and carbon is much more complete and that instead of finding a small quantity of potassic sulphide we have found notable traces of ammonic sesquicarbonate.
C. Analysis of the Gases.
In order to resolve the third question, that is to say, to determine the nature of the gases which result from the combustion, we have employed the apparatus before described (Fig. 2, pl. I). On sucking with the mouth at the end of the tube ff the gases contained in the tube ee, which come from the burning of the powder we find that their taste is very nearly that of pure carbonic anhydride. This is verified by the fact that the nose does not detect the least odor of cyanogen, of sulphurous anhydride or of nitric oxide, but there are some barely recognizable traces of hydrosulphuric acid. Mixed with air it does not give any trace of reddish vapors. As the presence of even a few thousandth parts of cyanogen, sulphurous anhydride and nitric oxide can be detected by the odor and taste we can conclude that these gases are not contained in the mixture to be analyzed.
We have found, on the contrary, carbonic anhydride, hydro-sulphuric acid, traces of oxygen, of carbon protoxide, hydrogen, nitrogen and of nitrous oxide. The following are the methods which were employed in the analysis of the mixture.
The carbonic anhydride is absorbed by potassic hydrate and the hydro-sulphuric acid and oxygen by potassic pyrogallate. The remainder of the gas is then introduced into a eudiometer with an excess of oxygen and of detonating gas. After the passage of the spark the excess of oxygen is ascertained by mixing it with hydrogen and detonating again.
Let us designate the volumes of the different gases which exist in the mixture as follows.
Carbonic anhydride c
Hydrosulphuric acid s
Oxygen o
Carbon protoxide c1
Hydrogen h
Nitrogen n
Nitric oxide n1
Let A° be the volume of gas first submitted to analysis,
Ao=c+s+o+c1+h+n+n1
c1s and o are given immediately by absorption.
Let A1 be the volume which remains after the absorption. A1=c1+h+n+n1.
Let A2 be the fraction of the volume A1 introduced into the eudiometer to such a degree that we have A2/ A1=m
A2= mc1+mh+mn+mn1
Designating by C the volume of carbonic anhydride formed during the combustion and by D the diminution in volume then
mc1=C mh=(2D-C)/(3)
Let O be the volume of oxygen introduced into the eudiometer for determining the combustion of A2 and there will remain after the detonation and the absorption of the carbonic anhydride a volume O=(C+D)/(3) and if V represents the total volume which is obtained A2+O we should have
mn+mn1=V-O+(C+D)/(3)
In making the volume V detonate with an excess of hydrogen we observe a diminution of volume D’ which is due to the combination of a part of the hydrogen with the volume O-[(C+D)/(3)] of the oxygen and of the rest with the volume mn’ of the nitrous oxide. If from D' we subtract the diminution of volume 3O-D-C due to the oxygen we would have the volume of the nitrous oxide since one volume of this gas in combining with one volume of hydrogen gives one volume of nitrogen then
mn’=D’-3O+D+C
mn=V-D+2O-2/3(C+D)
Such are the calculations and the operations which have served for the analysis of the mixture of the gases produced by the powder. The numbers which have been obtained are,
1st, Gases absorbed.
| Volume | Pressure | Temp. | Vol at 0o under a pressure of 1 m. |
Original mixture | 136.2 | 0.7359 | 8o.8 | 97.102 |
After absorption of CO2+H2S | 69.1 | 0.6760 | 8o.0 | 45.382 |
After absorption of O | 68.7 | 0.6738 | 8o.6 | 44.877 |
2d, Determination of the Hydrosulphuric Acid absorbed by the potassic hydrate.
146 divisions of the test glass measured 30 cm3. One division of the graduated burette contained 0.001242 grms. of iodine. We employed for the potassic hydrate 1.8 divisions of the burette. Consequently in the volume 97.102 of the gas, at 0o and under a pressure of 1 metre, the volume of hydrosulphuric acid absorbed with the carbonic anhydride would amount to
(146/30)X(773/1.175)X(112.5/1588.7)X(0.76)X(.001242)X(1.8)=0.58 div.
3rd, Eudiometric Analysis.
| Volume | Pressure | Temp. | Vol. at 0o and under a pres. 1 met. |
Original mixture | 110.0 | 0.3569 | 10.2 | 37.846 |
After addition of O | 150.6 | 0.3974 | 10.3 | 57.674 |
After introducing the detonating mixture | 189.0 | 0.4350 | 10.3 | 79.230 |
After the detonation | 144.2 | 0.3915 | 9.3 | 54.595 |
After absorption of the CO2 | 135.7 | 0.3915 | 9.1 | 51.414 |
After introduction of dry H | 220.4 | 0.4753 | 10.9 | 100.740 |
After deton’tg (the dry gases) | 125.6 | 0.3917 | 9.0 | 47.629 |
With these given we have calculated the quantities which enter into the formulae given above.
A0=79.102 A1=44.877
A2=37.846 C=3.181
D=3.079 O=19.829
V=51.414 D1=53.111
On solving the equations, we find
c=51.140 s=0.580
o=0.505 c1=3.772
h=1.176 n=40.063
n1=-0.134
What is most noticeable in this result is the existence of free oxygen in the presence of a combustible gas. We cannot believe that this points to a fault in the analytical method employed, for its known accuracy and the care which we have taken in applying it will not allow us to suppose that such an error could be possible. We account for this fact, on the contrary, by supposing that, after the combustion of the sulphur and charcoal, the residue of the powder, dispersed in the state of smoke, still contains some saltpetre and that probably, while cooling, this evolves small quantities of oxygen, but the temperature is not sufficiently high to cause the burning of these gases when mixed with a very considerable volume of other incombustible gases.
If the powder in burning is converted, as the old theory teaches, into potassic sulphide, nitrogen and carbonic anhydride, these last two gases should be present in the proportion of 1 to 3. Our examination shows that, on the contrary, this proportion does not reach 1 to 1.5. Then the combustion of the powder ought to take place in a different manner from that which the theory accepted up to the present day supposes.
Chapter IV.
Relations between the Residue, the Substances Which Constitute the Smoke, and the Gaseous Products.
We propose now to answer the fourth question which we have propounded. What quantities of solid residue and of smoke on the one hand and how much gas on the other, can be obtained from a given weight of powder?
In order to solve this problem we have analyzed the mixture of residue and deposit which was produced by the quantity of powder that gave, in burning, the gas previously analyzed.
We began by dissolving an unmeasured portion of the substance to be examined in such a volume of water that the volume of the solution occupied 500 cm3. For each particular determination we took each time 45.474 cm3 measured in a gauged test glass. The data necessary for the calculation are, when they are not specially indicated, the same as those stated above in the analysis of the residue of the powder.
I. Carbon and sulphur. The 500 cm3 left a residue, consisting of carbon, sulphur and an incombustible residue, the whole weighing 0.2141 grms. 0.1758 grins. of this deposit gave 0.1749 grms. of bane sulphate, which gives by a proportion 0.0292 grms. as the weight of the sulphur contained in the 0.2141 grins. The deposit treated with nitric acid and ignited in an open crucible left an incombustible residue which weighed 0.0248 grms. For the volume of 45.475 cm3 of the solution used in the following investigation we would have
Carbon, 0.014561 grms.
Sulphur, 0.002656 grms.
Residue, 0.002256 grms.
II. Potassic sulphide. The solution (500 cm3) treated as before with cupric oxide yielded a precipitate of 0.9902 grms. of bane sulphate corresponding to 0.46787 grms. of potassic sulphide or to 0.1358 grms. of sulphur, which would give for the volume of 45.475 cm3 0.04255 grins, of potassic sulphide.
III. Potassic hyposulphite. One gauged volume required 35.1 divisions of the graduated burette which corresponds to 0.06525 grms. of potassic hyposulphite.
IV. Potassic sulphate. One measured volume gave 1.131 grms. Of baric sulphate, corresponding to 0.84463 grms. of potassic sulphate.
V. Potassic sulphocyanate. A measured volume of the solution required 12.5 divisions of the burette corresponding to 0.006105 grms. of potassic sulphocyanate.
VI. Ammonia. A measured volume gave a=0.06688, t1=46.6, t=17.3. This corresponds to 0.01645 grms. of Ammonia, or to 0.05709 grms. of 2(NH4)2O, 3CO2.
VII. Potassic nitrate. We first determine the quantity of ammonia contained in a quantity of potassic hydrate equal to that which is required for treating the solution. We obtained a=0.05612, t1=27, t=26.4. The quantity of ammonia corresponding is equivalent to 0.0034578 grms. of potassic nitrate, a quantity which it is necessary to subtract from the total quantity of potassic nitrate found in the solution. We found for the volume of solution employed a=0.05612, t1=46.6, t1=23.3, numbers which correspond to 0.077808 grms. of potassic nitrate, and, deducting the weight .003458 grins, we see that the solution contains 0.07435 grins. of potassic nitrate.
VIII. Potassic carbonate and potassic hydrate. With the volume employed we obtained 0.1124 grms. of CO2 and 0.2240 grms. of Mn3O4. From this weight of carbonic anhydride it is necessary to deduct that which belongs to the ammonia sesquicarbonate, which gives us 0.3531 grms. of potassic carbonate. In this weight of 0.2240 grms. of Mn3O4 0.19539 grms. belong to the potassic carbonate. It follows then that 0.0286 grms. of the Mn3O4, corresponds to 0.03515 grins, of potassic hydrate. But the potassic sulphide has furnished 0.03635 grms. of potassic hydrate. We may then conclude that there is no potassic hydrate in the substance analyzed.
One volume of the solution (45.475 cm3) contains then:
Potassic sulphate, 0.84463=0.45676 of potassic oxide.
Potassic carbonate, 0.25279=0.17233 of potassic oxide.
Potassic hyposulphite, 0.06525=0.03232 of potassic oxide.
Potassic sulphide, 0.04255=0.03637 of potassic oxide.
Potassic hydrate, 0.00000=0.00000 of potassic oxide.
Potassic sulphocyanate, 0.00611=0.00296 of potassic oxide.
Potassic nitrate, 0.07435=0.03464 of potassic oxide.
0.73538
Carbon, 0.01456
Sulphur, 0.00266
Ammonic sesquicarbon’te 0.05709
1.35999
The same volume gave 1.380 grms, of potassic sulphate corresponding to 0.7463 grms. of potassic oxide, a number which agrees with that which we have deduced from the previous analysis as closely as can be desired.
The powder employed and the products of its combustion have then the following composition.
A. | B. | C. |
Powder | Solid Products | Gases |
Saltpetre, 78.99 | Potassic sulphate, 62.10 | Carbonic anhydride, 52.67 |
Sulphur, 9.84 | Potassic carbonate, 18.58 | Nitrogen, 41.12 |
Charcoal Carbon, 7.69 | Potassic hyposulphite, 4.80 | Carbonic protoxide, 3.88 |
Charcoal Hydrogen, 0.41 | Potassic sulphide, 3.13 | Hydrogen, 1.21 |
Charcoal Oxygen, 3.07 | Potassic sulphocyanate, 0.45 | Hydrosulphuric acid, 0.60 |
| Charcoal, 1.07 |
|
| Sulphur, 0.20 | Oxygen, 0.52 |
| Ammonic acid sesquicarbonate, 4.20 | Nitric Oxide, 0.00 |
100.00 | 100.00 | 100.00 |
|
|
|
All of the potassium in the powder ought to be found in the solid residue. We should be able then by the aid of analyses A and B to calculate the weight of the residue which corresponds to 1 grm. of powder. This weight of powder contains then, from analysis A, 0.3055 grms, of potassium; the weight of residue which contains the same quantity is 0.6806 grms.
This weight of residue contains a certain weight of nitrogen easy to calculate; on deducting it from the total weight of nitrogen contained in one grm. of the powder we should have the weight of nitrogen contained in the gas produced by the combustion of one grm. of powder. The weight of gas corresponding to this weight of nitrogen is, according to analysis C, 0.3138 grms. We should have then for the products of the combustion of one gram of powder;
THING
In the analyses we have usually compared the total weight employed with the sum of the different products obtained as a check. This check is impossible here since we have operated upon an undetermined quantity of the solid residue. We can however use another species of check. The weight of potassium, nitrogen, sulphur, carbon and oxygen contained in one grm. of the powder ought to be entirely found in the products of the decomposition. We found thus
Powder before the combustion: K 0.3050, N 0.1096, S 0.0984, C 0.0769, O 0.4057
Weight of the elements after the combustion: K 0.3050, N 0.1096, S 0.0989, C 0.0780, O 0.3936
The weights of potassium and of nitrogen are precisely the same in both cases, which shows that the preceding calculation has been made without error. The slight difference which exists between the weight of sulphur, carbon and oxygen demonstrates the accuracy of the method pursued.
We see by the preceding table that one gram of powder, in burning, produces 193.1cm3 of gas. The theory adopted up to the present time requires 330.9cm3, or nearly three times as much.
Chapter V.
Temperature of the Flame.
Having completed our investigations into the nature of the products of the decomposition of the powder during its combustion we come now to consider the fifth question that we proposed to ourselves at the commencement of this work, viz., the determination of the temperature of the flame. This determination made we shall be in a position to estimate the theoretical value of the work produced during the combustion of the powder. It is important to first define the nature of the flame which the powder produces. Imagine that 1 grm. of powder be inflamed at once throughout all its mass, it will thus produce c units of heat, which will serve to raise the products of combustion to the temperature c/s, s being the specific heat of the constituents. This temperature c/s is evidently determined by the aid of the thermometric unit which was adopted in selecting the unit of quantity for heat. But in practice the temper attire of the flame diminishes constantly through radiation and conduction, and, as it does not remain equal to c/s only during a very short period, it is impossible to measure it by the ordinary thermometric methods. The combustion of a mass of gunpowder takes place under similar conditions. In this case, it is true, a nearly constant temperature equal to c/s exists but only in the very thin layer where the combustion is going on and the heat produced is dissipated towards the point of the flame by conduction and radiation. In order to obtain the true temperature of the flame produced by the combustion of the powder and to avoid the influence of exterior circumstances c and 8 should be determined separately.
We have employed the following apparatus (Fig.4) for determining c.
A is a brass tube filled with a known weight of fine powder firmly packed. In the mouth a of this tube, which is slightly widened, a small neck of glass b is cemented, to the surface of which two platinum wires c c are fixed and joined to one another by a very fine platinum wire which touches the powder. The small apparatus A is placed in the glass tube B, which is closed at the bottom and open at the top and the whole is finally placed at the bottom of the large tube C in which lateral openings d d are pierced for the two wires. When the two wires c c have been passed through these openings they are closed by the aid of the lamp, and the top of the tube e is also closed. The tube C can be kept vertical by the aid of an appendage which has been blown on the bottom and which penetrates into a cork E; and it is placed in a larger jacket D whose walls are made of very thin brass. In this species of calorimeter a circular agitator is placed which can be raised or lowered by the aid of the very fine wires g g. The weight of the powder and of the glass, brass and platinum, which enter into the construction of this apparatus, having been determined it is filled with a known weight of cold water, placed in a case of wood and deposited in a place where the temperature remains nearly constant. There it remains until all of the parts have attained nearly the same temperature.
The quantity of heat furnished by the powder is determined by observing at different moments the temperature of the bath by the aid of a thermometer, placed in a small appendage on the side, which reads to the hundredths of degrees.
We begin by observing the rise of the thermometer before the pow der is lighted. Then note the moment when the powder is ignited by the aid of the electric current, then when the thermometer attains its maximum temperature, and, finally, follow its descent during some time. Throughout the duration of the observations be careful to agitate the water in the calorimeter. The following are the numbers obtained in one determination made with the greatest care.
| Time | Temperature observed |
| 0’ | 19.86 |
| 5’ | 19.83 |
| 6’ | 19.83 |
Combustion, | 7’ |
|
Maximum temperature, | 16’ | 21.10 |
| 26’ | 20.98 |
| 56’ | 20.60 |
The weights of the different parts being
Weight of glass, 79.14grms.
Weight of brass, 132.11grms.
Weight of platinum, 3.50grms.
Weight of powder, 0.712grms.
Weight of water, 376.40grms.
The value, in terms of water, of the different parts of the apparatus, is 404.7 grms, and the heat evolved by the combustion of the powder raised the temperature 1.14°. The heat produced by the powder which we used, that is to say, the rise in temperature which a given weight of powder could produce in the same weight of water is 643.°9.
It is necessary to apply a slight correction to the number that we have found. The combustion takes place in a tube which is in effect hermetically closed and filled with air. The combustible gases produced by the combustion of the powder would burn and produce a small quantity of heat which is foreign to that which is produced by the powder itself. We find from table D that 0.7125 grms. of powder would yield
Carbon protoxide, 0.00669 grms.
Hyrdogen, 0.00014 grms.
Hydrosulphuric acid, 0.00128 grms.
We will adopt from Favre and Silbermann the numbers 2403, 34462 and 2741 as representing the heat disengaged by the combustion of these gases. The heat produced under these circumstances would raise the temperature 24.°4 which should be deducted from the 643.°9 obtained above. The actual quantity of heat disengaged by the powder will then be 619.°5. The heat produced by the augmentation of the pressure, which manifests itself in the glass tube after the combustion of the powder, is so slight that it may be neglected.
In calculating the quantity of heat which the powder would produce if all its combustible elements united directly with oxygen we obtain, if we employ the numbers given by Favre and Silbermann, for the combustion of the sulphur, the carbon, and the hydrogen, the number 1039.1 The combustion of the combustible elements of the powder with the oxygen of the saltpetre gives them much less heat than when they combine with free oxygen. This should not surprise us if we consider that the nitrogen contained in the potassic nitrate, and which approaches nearly two thirds of the weight of the combustible elements, ought to absorb a great quantity of heat in its conversion into gas.
We can obtain the temperature of the flame of the powder directly, or rather the temperature which would exist if there were no loss by conduction or radiation, by dividing the number 619.5 by the specific heats of the products of the combustion. In order to obtain this quantity we must multiply the weights of each of the bodies produced by the combustion by its specific heat and add together the products thus found. We have neglected in this sum the potassic hyposulphite, the potassic sulphocyanate and the ammonic sesqui carbonate which are present in very small quantities and the hydrosulphuric acid whose specific heat is unknown. The numbers obtained from these very small quantities would only influence the unit place of the number sought, and can be neglected without appreciable error.
| Weight | Specific heat | Product |
Potassic sulphate | 0.4554 | 0.1901 | 0.08656 |
Potassic carbonate | 0.1362 | 0.2162 | 0.02944 |
Potassic sulphide | 0.0229 | 0.1081 | 0.00248 |
Potassic nitrate | 0.0401 | 0.2388 | 0.00957 |
Carbon | 0.0079 | 0.2411 | 0.00190 |
Sulphur | 0.0015 | 0.7026 | 0.00031 |
Nitrogen | 0.1075 | 0.2440 | 0.02623 |
Carbonic anhydride | 0.2167 | 0.2164 | 0.04692 |
Carbon protoxide | 0.0101 | 0.2479 | 0.00251 |
Hydrogen | 0.0002 | 3.4046 | 0.00073 |
Oxygen | 0.0015 | 0.2182 | 0.00033 |
| 1.0000 |
| 0.20698 |
On dividing 619.5 by .207 we obtain for the temperature of the flame of the powder burning freely in the open air the number 2993° C. If the powder burns in a closed space where the gases are not allowed to expand freely the temperature of the flame will be different. We can find this by dividing the quantity of heat due to the combustion by the specific heat of a constant volume. The following is a table of the calculations.
| Weights | Specific heat | Product |
Potassic sulphate | 0.4554 | 0.1901 | 0.08656 |
Potassic carbonate | 0.1302 | 0.2162 | 0.02944 |
Potassic sulphide | 0.0229 | 0.1081 | 0.00248 |
Potassic nitrate | 0.0401 | 0.2388 | 0.00957 |
Carbon | 0.0079 | 0.2411 | 6.00191 |
Sulphur | 0.0015 | 0.2026 | 0.00031 |
Nitrogen | 0.1075 | 0.2440 | 0.01846 |
Carbonic anhydride | 0.2167 | 0.2164 | 0.03426 |
Carbon protoxide | 0.0101 | 0.2479 | 0.00177 |
Hydrogen | 0.0002 | 3.4046 | 0.00048 |
Oxygen | 3.0015 | 0.2182 | 0.00023 |
|
|
| 0.18547 |
The temperature of the flame is then (619.5/.18547)=3340o in the case when the combustion takes place in a closed vessel and the gases are not able to expand freely.
If the flame of the powder consisted only of the combustible gaseous products, as the calorific capacity of the bodies is constant, as shown by the researches of Reguault and the theoretical work of Clausius, whatever the temperature may be, this value c/s could be determined with great precision. But as the specific heat of the solid bodies augments with the temperature we must consider the numbers 2993° and 3340o as only approximate values, yet they ought not to be very far from the real values considering the slight increase of the specific heat with the temperature. Since s increases with the temperature c/s as found is evidently too large; to this it is necessary to add that the flame is always cooled by radiation and conduction. We can then state with certainty that the numbers 3340 and 2993 form a maximum limit which the temperature of the flame will approach more or less but which it will never attain. With this given we can estimate the pressure exerted by the powder burning in a space which it completely fills.
Chapter VI.
Work Produced by the Combustion of the Powder.
Up to the present time it has been generally supposed that the residue held in the vapors during the combustion of powder exerted a strong influence over its mechanical action. Though we cannot deny absolutely that this volatilization takes place, nevertheless at the temperature of the flame of the powder the tension exerted by these vapors does not reach one atmosphere. In order to show this we have fused a small globule of the residue, placed on the end of a platinum wire in a hydrogen flame. The substance was insensibly volatilized in the air but without ebullition, and consequently the tension of its vapors was less than one atmosphere. Now Bunsen,* in his method of measuring the gas, assigns 3259°C as the temperature of the hydrogen flame. We may then neglect completely the pressure of these vapors in the combustion of the powder which is effected between the temperatures of 2993° and 3340° C and calculate thus the maximum tension exerted by the gas produced by the combustion of the powder in a closed space.
Let Gp be the weight of the powder employed and Sp be its gravimetric density; Gr the weight of solid residue and Sr its specific gravity at the temperature of 3340°C; and finally let V be the volume of the gas produced taken at the temperature of 0o and under a pressure equal to one metre, and designating by to the temperature produced by its combustion in a closed space we should have in order to determine the pressure po exerted by the gas the formula.
po={[V(1+0.00366to)]/[(Gp/Sp)-(Gr/Sr)]}
In this equation there is only one quantity, the determination of which presents much difficulty, and this is Sr the specific gravity of the solid residue fused at the temperature of the explosion. We have determined this specific gravity by a method which is not yet known but which one of us has employed for studying the expansion and volatilization of rocks fused at very high temperatures and which renders it independent of the expansion of the vessels which contain these products. The researches made by this method have given us only approximate values, it is true, but sufficiently accurate for the end that we have in view. We have found, by this method, for the specific gravity of the solid residue.
at 18o 2.35
at 2808o 1.52
Consequently we have obtained by interpolating to 3340o Sr=1.50. The quantities which enter into the preceding formula are then
Gp=1.000 Sr=1.50
Sp=0.964 V=193.1 cm3
Gr=0.6806 t=3340o
And substituting these quantities in the formula we have
p=4373.6
If in calculating the pressure we take for the specific gravity of the powder residue the number which corresponds to the ordinary temperature (2.35) we have for po the value 3414.6. In the pressure of 4374 atmospheres found there are consequently about 1000 which are due to the expansion of the residue of the powder on account of the elevation in temperature.
A powder, whose composition is the same as that which we have used, in burning in a cannon behind a projectile, in consequence of the inevitable loss of heat, cannot then exert a greater pressure than 4500 atmospheres, if we admit that the decomposition takes place as we have indicated. If; on the contrary, the decomposition goes on in an essentially different manner when the powder burns freely in a cannon from .that which takes place when it burns under strong pressure we could determine it only by collecting the residue formed and the gas developed under those circumstances. If then we discover that under those circumstances the decomposition remains sensibly the same it would be necessary to conclude that some of the estimates of the pressure of powder gases in cannon which are in general acceptance are based upon erroneous principles. Many of the writers upon artillery have stated this pressure as high as 50,000 and some even at 100,000 atmospheres.
The preceding researches furnish the means for calculating the maximum mechanical effect, or the theoretical work which the powder can produce, when the gas which it furnishes, taken at first under a pressure corresponding to its primitive volume expands in a given space without loss of heat. Let a1, a3, a3, a1, (Fig. 5) be the volume occupied by the weight of powder Gp; a2, a3, a3, a2, the space occupied by the residue whose weight is Gr, and finally let a1, a2, a2, a1, be the space occupied by the gas at the moment of the combustion. This volume vo=(Gp/Sp)-(Gr/Sr) serves for calculating the pressure p as we have seen above. Finally let a, a2, a2, a, be the space which the gas fills when, in consequence of its expansion, the pressure becomes p. Assuming c, a2, a2, c, as the amount of expansion of an infinitely small portion of the gas at the commencement, under the initial pressure of po, the work done will be pod v and that which will be produced by the complete expansion of the gas.
∞
∫ p d v
vo
Assuming that we have lost no heat through the walls of the envelope and considering po as the pressure corresponding to the volume vo we should have p=po(vo/v)k being ratio between the specific heats of gases under a constant pressure and under a constant volume. The value of this definite integral becomes in this way (povo/k-1).
One gramme of the powder which we have employed gives for vo the value (Gp/Sp)–(Gr/Sr)=0.5836 cm3 and for po the value 1029.9x4373.6 grms. k was calculated from the composition of the gases and was equal to 1.39. Consequently one kilogram of this powder produces, when decomposing as we have stated, a theoretical work of 6741.0 kilogram-metres.