16
THE JOURNAL OF INDUSTRIAL AND ENGINEERING CHEMISTRY.
Columns 9 , IO and rr.-Specific gravity at 30' C. without glycerine expansion correction and so on a s described for columns 3, 4 and
j.
SUMMARS
I n consideration of the importance connected with obtaining a true value for the thermal coefficient of expansion of glycerine, the greatest care was observed in every detail of the foregoing work. It may appear superfluous to report the determinations of the specific gravity to the fifth decimal place, b u t as they agreed so closely a t t h a t point when different pycnometers were used in doing duplicate determinations, i t seemed better to show the actual results obtained, which also shows the degree of accuracy t h a t entered into the work. From the foregoing results, i t is apparent t h a t the published values for the thermal coefficient of expansion of dynamite glycerine are not true between I 5.5' and 30' C. Taking the average coefficients of expansion from Table I1 as 0.000612 a t 2 0 ' C., 0.000617 at 25' C. and 0.000622 a t 30' C., which are in round numbers 0.00061 a t zoo, 0.000615 a t 25' a n d 0.00062 a t 3 0 ° , and applying them a s shown in Table IV, the determinations show very close agreement with the determinations made a t 15.jo C. The idea in using the round numbers 0.00061, 0.00061 j a n d 0.00062 a s the coefficients of expansion for glycerine is t h a t they are easily remembered a n d are accurate enough for all practical purposes. EASTERN LABORATORY, E. I. DUPONTDENEWOURS POWDERCo
,
September 1, 1909.
FLASH, FIRE AND EXPLOSION TESTS ON MIXTURES O F CARBON TETRACHLORIDE AND NAPHTHA. By EDW.A. BARRIER. Received September 2, 1909.
The frequent occurrence of fires, explosions, or both, resulting from the wide-spread use of naphtha in removing oil from textile materials or metal goods In factories, makes i t a matter of serious concern t h a t no suitable, non-inflammable substitute has a s yet been found. Of late years carbon tetrachloride has received some attention in this connection, b u t its comparatively high cost has barred its use except in a very limited way. Recent reductions in the price of this material, however, and the fact t h a t a certain percentage of naphtha can be added without rendering the mixture inflammable have brought the cost of such a mixed solvent down t o a point where i t can be used in many cases, especially, if the spent solvent is reclaimed b y distillation. T h e importance of this matter led to the follow-
Jan., 1910
ing series of experiments which were undertaken with the object of determining how great a percentage of naphtha, of various gravities, could be added to carbon tetrachloride, and still leave t h e mixture free from fire and explosion hazard. Four samples of commercial naphtha, labeled a s below, were obtained from a petroleum refiner for the tests. The specific gravities of these samples were found to be as follows:
.
. .....
I . . .... . . . 11.. . . . . . . 111... . . . . . . . . ,
.
IV.. . . , . . . . . . . ,
Labeled. 54' Be. 62OB6. Motor gasoline, 68-72' BC. 73-76' BC.
Sp. gr. found. 0.7556-55% a BC. 0.7251-63O BC. 0.6982-70%0 BC. 0.6811-75%0 Be.
Distillation tests of these naphthas were made with the following results: Gravity.
80-100°.
551/40
63'
.
I
.
.
110-120'.
120-130°.
.....
17.0 cc.1 2 2 . 0 ~ ~ . 18.0 cc.
.
12.5 cc.2
23 5 cc.
140-150°.
150-160°.
63' Gravity.
100-llOo.
.. . .. ..
551/&0
160-170°.
701/20 751/29 Gravity. 701/p0 7S1/2'.
50-60'.
60-70'.
2 . O cc.8 3.5 cc.
5 . o cc. 23.0 cc.
90-100°. 21 .O CC. 13 . O CC.
100-llOo. 1 1 .o cc. 5 . 5 cc.
20.5 cc. 8.0 cc.
170-180°. Above 180O.
1 1 . 0 ~ ~ . 9 . 5 ~ ~ . 7.0 cc. 16.0 cc. above 140°
1 7 . 0 ~ ~ .
Gravity.
130-140°.
70-80O. 19.0 cc. 24.0 cc.
17.0 cc.
80-900.
24.5 cc. 18.0 cc.
110-115°. Above 115'. 4 . O cc. 14.0 cc. 12 . O cc. above llOo
FLASH A N D FIRE POINT TESTS.
The flash and fire point tests were made in the open cup of the New York S t a t e Board of Health tester, and also in some cases, in a six-inch evaporating dish. No difficulty was experienced in determining the flash points of the various mixtures, b u t in no case could a definite fire point be found. On raising the temperature from the flash point, the liquid would usually flash and burn for one or two seconds, and then go o u t of itself. A further raising of the temperature resulted in some cases in lengthening the time of burning, a n d , in other cases, in decreasing the time of burning. Repeated tests of the same mixture, however, gave reasonably concordant results. The mixtures containing the limiting amounts of naphtha, which are later stated to be reasonably safe, were tested in a n evaporating dish, in addition to the open cup, a s i t was thought t h a t this form of container with a large evaporating area was more comparable t o the ordinary open b a t h frequently found in works. On account of the absence of a true fire point the temperature was gradually raised from the flash point to the boiling point, a n d the observations recorded below were taken. The results are given in the following table: 1 Distillation 2
3
'.
began a t 123 Distillation began a t 8 5 O . Distillation began a t 5 5 O .
BARRIER O N FLASH, FIRE A N D EXPLOSION TESTS. Observations.
Kind of naphtha and flash-point.
Mixture tested.
I
55% 80
i O B &
c.
I
50 per cent. naphtha and 50 per cent.
tetrachloride..
....................
Flash-point of mixture.
.....
..
Did not flash a t any temperature.
c.
.....
..
Did not flash a t temperatures above the flash point.
.....................
350
70 per cent. naphtha and 30 per cent. tetrachloride. . . . . . . . . . . . . . . . . . . . . .
28' C.
450 550
c. c.
4 6
Above 5 5 O C . the would not burn.
24' C.
350
c. c.
5
Increasing the temperature appeared to have no appreciable effect.
80 per cent. naphtha and 20 per cent.
tetrachloride.
.....................
450
i 1
40 per cent. naphtha and 60 per cent.
tetrachloride.
.....................
38'
C.
.....
50 per cent. naphtha and 50 per cent.
tetrachloride.
1
E;:.'
.....................
18' C.
5 5 per cent. naphtha and 45 per cent.
tetrachloride.
.....................
100
I 60 per cent. naphtha and 40 per cent.
tetrachloride.
Remarks.
.......
60 per cent. naphtha and 40 per cent.
tetrachloride.
Temp.
-
Time of burning, see,
.....................
00
c.
c.
450
c. c.
25'
C.
350
.. 3 5
1
350
c.
5
250 350
c. c. c.
3 10 20
450
i
8
70' C.
..
2.5' C. 350 c.
2
mixture
On raising temperature mixture w-ould not flash.
Further increase of temperature decreased the time of burning.
Further increase of temperature appeared to have no appreciable effect.
Burned continuously until extinguished.
30 per rent. naphtha and 70 per cent.
tetrachloride.
~
70% Be. Below -180
.....................
-
5OC.
1
At higher temperature would not burn
40 per cent. naphtha and 60 per cent.
tetrachloride.
.....................
1 C.
-1
25' C.
359
c.
00
c.
10 15
Further increase of temperature appeared to decrease time of burning.
Below 50 per cent. naphtha and 50 per cent.
tetrachloride.
7 5 3 Be. Below -180 c.
I
.....................
C.
-18'
c.
2
2.5' C. 350 c.
10
c.
1 2
30 per cent. naphtha and 70 per cent.
tetrachloride.
.....................
-1
1
C.
100 2 5 O C. 350
c. 4 5 0 c.
..
5 8
Burned continuously until extinguished.
Further increase of temperature appeared to decrease time of burning.
Below 40 per cent. naphtha and 60 per cent.
tetrachloride., ....................
-18'
C.
00
c.
25' C . 350 c.
i
450
E X P L O S I O N TESTS.
T h e explosion tests were carried o u t in a specially constructed box, IO"^ I "I 15" (see Fig. I ) , heavily built, and made air-tight. T h e cover was provided with a tongue on the under side, which fitted into a channel placed on the upper, inside edges of the box. T h e channel was partly filled with water, thus making a n air-tigh't seal. The box was also provided with a n inlet pipe a t the top of one end, and a n outlet pipe a t the bottom of the other end, both of 1/4'' bore. Two copper bars passing through
c.
3 12 15
Burned continuously unti' ex tinguished.
..
the outlet end of the box served as terminals for making connections with a n electric circuit. The vapor was introduced into the box b y connecting the outlet pipe with a suction pump, and the inlet pipe with a n Erlenmeyer flask, fitted with a rubber stopper a n d two glass tubes, one of which dipped into the mixture to be tested. The amount of vapor sucked into the box was regulated b y using a constant flow of water through the suction pump, and varying the time suction was applied. When the desired amount of vapor
T H E JOURNAL OF INDUSTRIAL A N D ENGINEERING C H E M I S T R Y .
1-8
had been drawn into the box, the suction p u m p a n d flask were disconnected, a n d the terminals were wired to a n electric circuit. T h e charge was then fired b y “ s h o r t - circuiti n g ” a one ampere fuse placed on the inside of the box. B e f o r e the final series of tests were run a number of preliminary experiments w e r e Terminals I/! made with pure naphtha, in which the time o f application of the suction -’ \Oh. was varied, in FIG.\. W.M.N Jr order t o determine within what range the maximum explosive
-
I
-
Jan., 1910
effect could be obtained. This was found to be from 30 seconds t o four minutes, a n d this range, except in certain cases where n o explosion could be obtained, was accordingly used in the series of tests which appear in the following tables: TABLEI. 55%’ BB. NAPHTHA.TEMPERATURE ZOO. Time suction Height cover was applied. was lifted. 1 min. Did not move. 2 min. 4 min. ........ 8 min.
Mixture. Pure naphtha..
...... ........ ........
Remarks. No explosion.
TABLE11. 63’ BB. NAPHTHA.TEMPERATURE ZOO. Mixture.
Time suction Height cover was applied. was lifted. 1 min. Did not move. 2 min. 4 min. 8 min.
...... ........
Pure naphtha..
Remarks. Norexplosion.
........ ........
All of the above tests with the 7 0 l / ~ O n a p h t h a were repeated using a temperature of 3 8 O , with the idea of simulating conditions which occur in summer. Practically the same resultsTwere obtained as with the lower temperature, with t h e exception t h a t the same explosive effect was obtained in a shorter time. From a n examination of the tables, i t will be
111. 701/20 Bi. NAPHTHA TEMPERATURE 21 ’. Time suction was applied.
Mixture
................................... .....................................
pure n a p h t h a . ,
1 min. 2 min.
.......... ............
5 0 per cent. CCl, and 50 per cent. naphtha..
............ 40 per cent. CClr and 60 per cent. naphtha.,
..........
............ ............ 30 per cent. CCl, and 70 per cent. oaphthr.. .......... ........... ........... Iv.
Remarks.
No flame visible. Considerable flame, evident. excess of vapor.
7 ft. 21/,
ft.
Did not m o r e
No explosion.
2 min.
4 min.
1 min. 2 min. 4
min.
1 min. 2 min. 4 min.
Very slight movement
I/*
ft.
‘/4
ft.
Green flame, evident excess of vapor.
.
751/20 Bi, NAPHTHA TEMPERATURF 21 Time suction was applied:
Mixture. pure naphtha
1 min
Height cover was lifted.
................................... .....................................
..................................... .....................................
.......... ............
60 per cent. CCI, and 40 per cent. naphtha..
............ ..... 50 per cent. CCL and 50 per cent. naphtha.. ...........
30 sec 1 min. 2 min. 4 min.
Height cover was lifted. Did not move. 6 ft. 2 ft. 1 ft.
Did not move.
4 min. 1 min.
............
Moved slightly. 2 inches.
.......... ............
1 min. 2 min. 4 min.
3 inches 9 inches 11/2 f t .
1 min.
4
2 min.
2‘/2
............
.......... ............ ..........
10 per cent. CCI, and 70 per cent. naphtha..
No flame visible. Some flame visible Considerable flame, evident excess of vapor.
1 min. 2 min.
2 min. 4 min.
$0 per cent. CCll and 60 per cent. naphtha..
Remarks,
4 min
inches fL 2 ft.
Flame visible, evident acess of vapor,
L E A V I T T ON MOISTURE DETERMINATION A S APPLIED TO CEREALS. seen t h a t no explosion could be obtained with t h e 5 s 1 f 4 0a n d the 63' naphthas. This is in line with what experience has generally shown, namely, t h a t there is b u t slight explosion hazard connected with the ordinary use, at the usual room temperatures, of naphthas having gravities around 60' or below. CONCLUSIONS.
As a result of these tests the following conclusions seem warranted : ( I ) T h a t a certain percentage of n a p h t h a can be added t o carbon tetrachloride a n d still leave the mixture free from fire and explosion hazard. ( 2 ) T h a t the percentage which can be safely added varies with the gravity of the naphtha. (3) T h a t 55' n a p h t h a a t ordinary room temperatures is practically free from explosion hazard, b u t in order to be reasonably safe from fire hazard, it should contain at least 30 per cent. of carbon tetrachloride. (4) T h a t 63' n a p h t h a at ordinary room temperatures presents b u t slight explosion hazard, b u t in order to be reasonably safe from fire hazard, i t should contain at least 45 per cent. of carbon tetrachloride. ( 5 ) T h a t 70' n a p h t h a in order to be safe from explosion hazard should contain a t least 50 per cent. of carbon tetrachloride, a n d to be reasonably safe from fire hazard should contain 60 per cent. (6) T h a t 76O n a p h t h a to be safe from explosion hazard should contain a t least 60 per cent. of carbon tetrachloride, and to be reasonably safe from fire hazard should contain 70%. T h e above statements apply, naturally, only t o naphthas which show approximately the same results on distillation, especially with respect t o the lower boiling fractions, a s those tested. T h e expression " reasonably safe from fire hazard" is used advisedly, since the above mixtures, while possessing b u t little fire hazard in open containers, will burn if spread o u t over a considerable area, on a n y other readily combustible material such as cotton waste. T h e percentages of n a p h t h a , stated in each case, are limiting ones, and good practice would call for a n additional 5 per cent. of carbon tetrachloride in order to allow some factor of safety .
THE MOISTURE DETERMINATION AS APPLIED TO CEREALS. B y SHERMANLEAVITT.
Received October 5, 1909.
The object of this investigation was to find out if a variation of one or two degrees from the standard temperature of one hundred degrees centigrade in
'9
the determination of moisture content of cereals would have a n appreciable effect on the percentage of moisture obtained under otherwise similar conditions. These conditions were that all the determinations be made in a partial vacuum with a gentle current of air passing through and that the drying be continued until the sample ceased to lose and commenced to increase in weight, it being generally assumed that cereals are dry when they begin to increase in weight on further heating. I n the fall of 1907, the writer had occasion to make moisture determinations on a large number of samples of wheat. Quite a number of these samples, unknown to the writer a t the time, had been accidentally duplicated by another chemist. The results obtained by him on the same wheats were found to be in general about two per cent. higher on the basis of the fresh substance than ours. As this difference in results was fairly constant and was material in amount, the present investigation was started to determine, if possible, the cause for this divergence. A cylindrical, heavy, brass, vacuum oven which was constructed with a water jacket was used. The oven was provided with a vacuum gauge and, a thermometer, standardized by the Bureau of Standards, for which the correction was known. During the drying process, a slow current of air, dried by passing through strong sulphuric acid, was drawn over the samples, maintaining meanwhile a vacuum of about twenty-four inches of mercury. The samples were different varieties of wheat, ground to pass a one millimeter sieve, being well mixed and kept in tightly stoppered bottles. The experiment was devised and carried out so as to have as nearly as possible only one variable, i. e . , the temperature of the interior of the oven. As stated before, the end point of the drying operation was when the samples ceased to lose weight and showed a tendency to increase in weight. This was generally accomplished in five hours, the time being taken as soon as the oven had reached the desired temperature in the partial vacuum and current of air. The temperature was changed by the following simple conditions: When the oven was unprotected by asbestos covering and heated by a n ordinary bunsen burner, with the oven jacket partially filled with tap water, a very even temperature of 97' Centigrade was maintained. By using a larger gas burner of the gridiron type and asbestos covering for the oven and flame, temperatures of 98' and 99' could be obtained under the standard vacuum conditions of twenty-four inches of mercury and two or three bubbles of air to the second passing through the oven. It is interesting to note a t this point that the temperature of the interior of the oven, as indicated by the standard thermometer, never reached 100' by
