80
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
are interested primarily in the nature of the final product, while the technical man is much more concerned with efficiency in its bearing on cost of production, Consequently i t is being recognized more clearly every year t h a t a thorough grounding in organic and inorganic chemistry is merely a preliminary stage in the training of an industrial chemist and t h a t his chemical course is incomplete unless i t leads up t o and includes physical chemistry, which means the common-sense study of methods. Some fifteen years ago I announced a t one of the meetings of the American Chemical Society t h a t a proper training in physical chemistry was the best possible training for a man going into technical work. At the time the statement was considered as the unfortunate outburst of a misguided and unbridled
Feb.,
1912
imagination. To-day it is such a truism that i t is out of place anywhere except in an editorial. This is the more interesting because the real development has scarcely begun. Within the next three or four years we shall get colloid chemistry on a sound scientific basis, and when that time comes the field of the physical chemist will include: photography ; tknning ; brewing ; rubber; dyeing ; soap ; textiles ; artificial silk and other filaments ; cellulose; paper; celluloid and other plastics; starch; glues and cements ; paints, lacquers and varnishes ; lubricating oils and greases ; clay, fuller’s earth and putty ; inks ; ceramics; glass and enamels; milk, butter and caseine ; cooking; water purification and sewage disposal; food preservatives ; slags ; soils ; physiology, biology and medicine. WILDERD. BANCROFT.
ORIGINAL PAPERS SOME PROBLEMS IN CHEMICAL ENGINEERING PRACTICE.’ MANUFACTURE AND TESTING OF SHIPPING CYLINDERS FOR ANHYDROUS AMMONIA. B Y F. UT.FRERICHS. Received January 5 , 1912.
INTRODUCTION.
Encouraged b y some of my friends, I offer to-day what may be called a continuation of the address which I delivered in Chicago, six months ago, and which has been already published in the October and November (1911)issues of THISJ O U R N A L . For to-day, I have selected three problems in chemical engineering practice, the characters of which are widely different from the previous problems and which give additional proof of the statement that the field of chemical engineering is exceedingly varied and difficult’ t o define. Again, these problems are selected with a view of giving, t o the student of chemical engineering, an idea of the variety of questions with which he may be confronted but, a t the same time, they are chosen t o call attention t o the wide difference between scientific chemical research and the investigation of chemical engineering problems. Scientific research treats principally with reactions which produce definite substances and can be followed up b y chemical formulae. Chemical engineering problems are based on reactions also but they include cost, quality, yield, labor and safety of those engaged in the manufacture of goods, and provide for the convenience in handling t h e goods b y the consumer. In the manufacture of chloroform from bleaching-powder and ethyl alcohol, the process was known, but the object of the research was to find the conditions under which the maximum yield could be obtained b y a known reaction and to construct apparatus which would give these conditions with greatest economy. I n the 1 Address read a t the Annual Meeting of the American Institute of Chemical Engineers, in Washington, D. C., December 20, 1911.
construction of laboratory apparatus referred to, the saving of time effected by the improved apparatus was the leading element. In the manufacture and testing of shipping cylinders for liquefied ammonia gas, the safety of handling and the convenience t o the consumer was the object of investigation. These, together with the three problems treated, give only a few forms of the great variety of demands which are likely to be made on the chemical engineer, and I should feel highly gratified if I have given an ‘incentive to others of our members t o follow in this direction, b y giving the solution of additional cases, so that we may obtain a large variety of solved problems b y which the students in our profession may be guided in their work. MANUFACTURE
Ah*D TESTIh-G
O F SHIPPING
CYLINDERS
FOR ANHYDROUS AMMONIA.
All ,chemical manufacturers produce goods and these goods need packages in order to be marketed. The question of packages is often more difficult and requires more study and capital than the manufacture of the goods to be shipped in them and this is particularly the case if the packages are intended for compressed or liquefied gases, which can only be handled with reasonable safety if the containers in which the gases are confined are constructed and kept in such a manner, that danger in handling them as far as practicable is excluded. I t is often said that the manufacture of packages is not a chemical problem, and chemists should not meddle with questions outside of their line; that if questions of packages arise they should go t o men who make them and take their advice. But there are instances in which the chemical eAgineer is confronted with the necessity of going himself into the manufacture of containers, and in such cases the manufacture of packages becomes a chemical engineering problem. I experienced a case of that kind not long ago when I needed a large number of valves for ammonia shipping cylinders, and since the object.
Feb.,
1912
T H E JOURLVVAL OF I-YDL-STRI--lL. .-1Z;D E-YGISEERI-YG C H E M I S T R Y .
81
involved many thousands of dollars I went t o the trouble of studying the question thoroughly. I made designs and had models made a t great expense, some of which you see among the exhibits. When I had arrived a t a model which seemed to fill my requirements, represented b y Fig. I , I selected a shop where similar work had been done a n d contracted a t their price for 1,500 valves, equal t o model submitted. When deliveries u-ere begun I had t o reject every valve. When the defects were pointed out i t was admitted that the differences were there, but it was contended t h a t the valves delivered were just as good and answered the purpose. To this I took excepion. A s the proprietors of the shop were desirous of fulfilling their contract, I went into their shop to help things along, and investigated the cause of the unsatisfactory work which had been delivered.
to be invested in tools. The shop was unwilling to make such a n investment. Neither could they promise t o put only first-class machinists on the work, unless I would agree to pay a price which was prohibitive. In this manner, I was forced to undertake the work myself. I bought the best machine tools I could obtain, and arranged a shop containing two screw machines, two lathes, two drill presses, and one gas furnace for case-hardening. Furthermore, one power hack saw and emery wheel, all of which cost about $ 3 , 0 0 0 . I employed three good machinsits and two helpers, put one of my engineers in charge of the shop and turned out 180 good valves per week a t a cost which was much less than the price, whlch I had paid to the outside shop for unsatisfactory work. Compressed and liquefied gases are manufactured in very large quantities, and for their VALVE safe transportation b y common carriers, ~ O X h M M l l N / A C Y Il N D E R S rules have been adopted in many counTOP VIEW tries for constructing and testing conS ~ VIEW E tainers in which they may be accepted for shipment b y transportation companies. Containers used f o r , this purpose generally have the form of cylinders. One set of rules has been agreed upon by the railroads of Germany, Austria Hungary, Belgium, France, Denmark, Italy, Luxembourg, The Netherlands, Roumania, Russia, Sweden, and Switzerland, which rules are commonly referred to as the International Rules for European countries. Up to recent times there were, however, no regulations of this kind in the United States but the American Railroad Association, in conjunction with committees appointed b y Fig. 1.-a, '/a" pipe connection: b . 3/8" pipe connection: ,: plug: d . iron washer; e , rubber washer; f , valve s t e m ; 8 . stuffing box nut or gland; h. and h', iron washers; i, packmanufacturers of compressed gases, have ing; j , soft metal seat; k. pipe connection for dipper pipe, been working on such rules for several The drop forgings for the valve bodies were made years, and these rules were recently adopted b y the b y a reliable firm on my direct order. There was Interstate Commerce Commission of the United States, no trouble on this account and the valve stems and and went into effect, October I , 1911. the nuts were made from standard steel bars, so The general rules, applying now t o shipping cont h a t there could not be any fault with the material. tainers for compressed and liquefied gases, are as In the shop I found some good and some worn follows : machine tools, and there were some good machinists, " ( a ) Gases that may combine chemically must b u t the majority of help were second-class men. not be shipped in one cylinder. Cylinders purchased Since work was continuously coming in it was neces- hereafter for the shipment of compressed gases must sary t o shift my work from one place t o another to be made in accordance with specifications approved keep the shop going, and the result was unsatisfac- b y the Interstate Commerce Commission. tory work. " ( b ) By water jacket, or other suitable tests, each The valve stems and lock nuts had been sublet cylinder used for shipping liquefied gases must be t o another shop of good reputation a t a contract subjected a t least once in five years t o a uniform price, but they were turned out on a much used screw interior pressure not less than one and one-quarter machine, were therefore not accurately made, and times the interior pressure t h a t would result from were not satisfactory t o me. heating the cylinder uniformly, in its maximum For making good valves, perfect tools and good charged condition, t o a temperature of 130' F. Each mechanics seemed t o be necessary, and in order t o cylinder used for shipping under pressure not turn them out a t the rate of about 2 0 0 valves per exceeding 1,000 pounds per square inch nonliquefied week, b y my calculation, a sum of about $3,000 was gases or gases in solution, must be subjected a t least
i:
T H E J O U R N A L OF, I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
82
once in five years to a uniform interior pressure not less than twice the charging pressure for such cylinder, corresponding t o a temperature of 70’ F.; when the charging pressure exceeds 1,000 pounds per square inch, the test pressure must not be less than one and one-half times the charging pressure. A cylinder must be condemned when i t leaks, when the permanent expansion is due to local weakness, or when it is uniform and exceeds 5 per cent. of the total expansion. When the charging pressure is less than 300 pounds per square inch i t will not be necessary t o measure the permanent expansion in quintennial pressure tests, provided the cylinder in question has previously passed this test. “(c) The weight of gas charged into any cylinder must not a t a temperature of 130’ F. cause an interior pressure in excess of three-fourths of the elastic limit of the weakest part of the cylinder. “ ( d ) The manufacturer must not offer for transportation a cylinder filled with such charging density of any gas as would produce failure in the test prescribed for that gas. “ ( e ) After December 31, 1914, all cylinders must plainly be stamped with the date of last test-for example, ‘ 4-09’ for April, 1909-or otherwise durably marked to show compliance with this rule; and
Feb., 1912
inspected separately for defects inside and outside and then subjected t o the following tests: ( a ) To each crop end cut from the pipe from which the cylinders are t o be made, a flattening test must be applied with the weld 45’ away from the side which is subject t o the greatest bending stress, with knife edges of wedge shape converging a t an angle of 60’, the point being rounded off with a radius of one-half inch. In this test, the crushing of the walls must be to within four times the thickness of the metal, and a crop end must withstand the test without cracking. ( b ) An internal hydrostatic test of 600 pounds per square inch must be applied to each length of pipe under which it must not show signs of leaking a t the weld or elsewhere. 4. Each finished cylinder must be tested by internal hydrostatic pressure as specified in above table, paragraph I , the test to be conducted as follows : ( a ) After filling cylinder with water, the test pressure specified must be applied for the purpose of detecting leaks and rounding up the cylinder. ( b ) The cylinder, if tight, must be placed in a water jacket or other approved apparatus for measuring the expansion, and the test pressure again
TURNACE,
P OWE e
n Fig. 2.
b y December 31, 1911, not less than one-fourth of the cylinders in use for shipping purposes must be thus tested and marked. “ ( f ) Cylinders containing acetylene gas must be made of tough steel and must be completely filled with a porous material t h a t has been tested b y the Bureau of Explosives and approved by the Interstate Commerce Commision, and this material must be charged with acetone or its equivalent not to exceed 40 per cent. of the interior volume capacity of the cylinder. The pressure in cylinders containing acetylene gas must not exceed 2 5 0 pounds per square inch a t a temperature of 7 0 ’ F. ” I n addition t o these general rules, specifications for shipping containers are. adopted, each set of specifications being different for different gases. I. The specifications for lap-welded cylinders intended for shipping anhydrous ammonia are shown in Table I. 2. Cylinders must be manufactured from lapwelded pipe made of soft steel of the best welding quality, free from blisters, cracks, or other injurious defects. INSPECTION AND TESTING O F MATERIAL.
3 . The pipe intended for these cylinders must be manufactured by the best appliances and according t o the best modern practice, each length to be
applied. The permanent expansion under this test must not exceed I O per cent. of the whole volumetric expansion a t the pressure specified. GENERAL CONSTRUCTION AND INSPECTION.
5 . The manufacture of the cylinders must be completed with the best appliances and according to the best modem methods. All finished cylinders must show a reasonably smooth surface, and must have passed the above inspection and tests without showing any defect in workmanship or material likely t o result in any appreciable weakness in t h e finished cylinder. Each completed cylinder must be inspected for such defects a t the mill by the purchaser or his representatives before acceptance. These rules and specifications are the results of extensive tests made with cylinders taken from the present equipment of the various manufacturers and with new cylinders made for the specific purpose of testing. I a m much pleased to say right here that tests have shown the shipping cylinders of old equipment generally t o be of greater strengths than required b y the regulations, from which fact it may be concluded that manufacturers always have been anxious t o make shipping cylinders of ample strengths 6 0 t h a t accidents might be prevented.
T H E J O L ' R S A L O F IA'DL-STRIAL A N D EATGINEERIAVGC H E I M I S T R Y .
P e b . , 1912
83
TABLEI -SHIPPIKG CONTAINER SPECIFICATION No. 4.-(See
Outside diameter of cylinder. Inches.
Over-all length of cylinder.
Nominal thickness of cylinder wall. Inch.
_ _ & _ _ ~
Ft. 3 4 7 3 4 7 7
10 10 10 10 10 10 12
in. 10 0 0 10 0 0 0
0.19
0.19 0.19 0.27 0.27 0.27 0.21
par. 1822(a), P. 4 1 . ) Lapwelded Steel Cylinders for Anhydrous Ammonia. Revised January 1 , 1912. Effective March 31, 1912. Table of Details as to Dimensions. Test Head. Mouth. pressure 7 DiamLow High per square eter. limit. limit. inch. Depth. Inches. Inches. Inches. Pounds. Inches. 3I5/is 71/2 73/8 1000 3 3/r 315/16 71/2 7%/a 1000 33/4 315/16 71/2 73/8 1000 33/* 1500 315/is 71/2 i3/n 33/4 3'5/is 71/2 73/8 33/4 1500 3I6/i6 71/2 73/8 1500 53/4 315/,, 9 9 1000 33/,
-
--
By these regulations, containers for liquefied gases must be safe a t pressures t o which the fully charged cylinders will be exposed a t 130' F. and the following table gives the pressures corresponding t o various temperatures for some liquefied gases most commonly handled. TABLE 11.
-
--
Test pressure, absolute. Internation:
Absolute pressure a t
---".
Carbondioxide Nitrous oxide. Ammonia .... . Chlorine.. . Sulphurdioxide
.. .
768 732
104 85 40
1160 1112 215 156
1227 1180 340 250 147
?s. absolute.
Lbs. Lbs. Atm.' Lbs. Atm: 2454 2793 190 1132 77 2360 2646 180 1103 75 425 441 30 1690 115 312 323 22 1374 9 3 . 5 184 176 12 1160 7 8 . 9
ature. 'C. OF: 3 1 . 1 88 3 5 . 4 96 1 3 0 . 0 266 1 4 5 . 4 294 155.4 312
'
.. From this table it is seen t h a t the pressures vary greatly with the different gases and, consequently, shipping cylinders used for different gases must be of different strengths. From the table it is also evident t h a t , in a general way, the International and American Rules agree fairly well, and we will now have t o see in which way the shipping cylinders must be constructed in order t o comply a t the same time with the rules of the Railroad Companies which are considered b y them necessary t o secure safe hanciling; with the interests of the mills, which require t h a t the specifications be inside of the possibility of manufacture a t a minimum cost; and with the interests of manufacturers and consumers of the goods, which make it desirable t h a t the shipping eylinders be as light as possible in order to facilitate the handling and securing the lowest freights. In designing shipping cylinders, the first question is: which metal can be used and what must be the strength and thickness of the material of the container to withstand the pressure to which the cylinder will be subjected? The answer to this question is found by the following considerations : Materials used for construction of shipping cylinders must be indifferent t o t h e gases, which are t o be shipped in them. All materials have constant properties, among them, ductility, the tensile or breaking strength and elastic limit. A bar of any material, having a cross section of one square inch, subjected to
Weight of cylinder. 7
High limit. Inches. 75/s 7 5/8 75/a 7"s 75/8
91/, 75/8
Low limit. Pounds. 87 90 150 113 117 203 220
High limit. Pounds. 112 115 175 141 145 237 250
a moderate strain, will elongate and spring back t o its initial length when the strain is released. This indicates the elasticity of the material. If, in repeating the experiment, the strain is increased, a point will be reached a t which the bar does not spring back t o its original length, and the strain a t which this takes place is the elastic limit of the material. If the strain is increased beyond the elastic limit, the bar will be disrupted and the strain a t which disruption takes place is the breaking strain of the material. The toughness or ductility of a material is tested b y bending a strip of i t , and looking for cracks on the bend. I t is accepted t h a t a cylinder made of material which complies with this bending test will not break up into flying fragments if i t explodes by influence of internal gas pressure. From the values of the breaking strain and the elastic limit of materials, the dimensions of shipping cylinders may be computed. In a cylinder, filled with compressed gas, the strain is twofold, namely, in the direction of the longitudinal axis of the cylinder and a t right angles t o this direction. Given a cylinder of ten inches diameter, filled with
Fin. 3.
gas under 100 pounds prcssurc t o the square inch, the total prcssure on each of the circular ends is equal to the area of a ten-inch circular plane, expressed in square inches, multiplied by I O O pounds, which is equal t o 7584 pounds. Since this stress must be taken u p by the sides of the cylinder in t h e circumferencial scam. between the bottom of the cylinder and its sides, each inch of the circumfer-
Fib. 4.
encc (which measures 3 1 . 4 inches) must withstand a stress of 7584 divided by 3 1 . 4 , equal to 241 pounds. The stress in the material at right angles to this direction is equal to the charging pressure on a plane which is placed through the longitudinal axis anti confined by the sidcs of the cylindcr. Assuming again the charging pressure t o be 100 pounds per square inch, then the strain on two opposite sections in the sides of the cylinder, each scction one inch long and of thc thickness of the material, is equal t o the same pressure or 100 pounds on a plane of one inch width and in length equal t o the diameter of the cylinder. This is in our case T O O times I O , cquals io00 pounds. Onc-half of this stress is taken u p on each of the ends of the plane or what is thc same in each line of the cylindrical part of the container which is parallel to the axis of the cylinder and one inch long, which in our case is j o o pounds. This stress, bcing greater than the stress in thc direction of the axis, it must govern in the calculation of thc cylinder. If the cylinder is made of steel and assuming the tensile strength of the material t o be 50,000 pounds to the square inch, with an elastic limit of 30,000 pounds, thcn the tubular part of the cylinder must have a thickness, expressed in inches, equal t o 500, divided by 39,000,or of a n inch, in order to withstand an interior pressure of T O O pounds t o the square inch and straining the sides not beyond t h e clastic limit of thc material. For the heads, the required thickness of the material is the same as for the sides, if the heads have the form of a convex hcmispherc. If the heads are flat or concave, the t.hickness is not easily figured, since it depends upon the form of the hcad and the bending strength of the material, which in thc samc material is subjected to considerable variation. There are two Icinds of pipc steel manufactured
on bending, for which reason its manufacture was abandoned. I f the necessary strengths of shipping cylinders for the various gases arc figured from the values givcn in Table 11, it will bc found t h a t carbonic acid, nitrous oxide and oxygen (undcr ~ , j o opounds pressure) require cylinders made of steel of high tensile strength, while cylinders for ammonia and compressed air (uiidcr 600 pounds pressure) for railroad cars and for many other gases, can be made from steel of less strength. Bearing in mind t h a t steel of high tensile strength cannot be welded, while stccl of less strength has welding qualities, the manufacturc of steel cylinders in this country is carried out in two different ways. Cylinders made from high-grade stcel are made by the seamless method. This class of cylinders is represented by carbonic acid cylinders and cylin-
I
!
--,'! I
I & &i...-
~ t ! -.t
.+, '
i:
{ i~ i
1
f+I
/.!
r:
~-7.: i;f .$:
.
I
::$--. -'
,
8
.
-2.0
I I ,
,
T H E JOL-RA’AL OF I S D C S T R I A L A N D E N G I S E E R I S G C H E M I S T R Y .
Feb., 1 9 1 2
is made from steel of less strength which has welding qualities, and this type of cylinders is used for liquefied ammonia gas and compressed air under less than 600 pounds pressure. Their mode of manufacture is as follows: A sheet of steel, about 25 feet long and of a width equal t o about 3 1 / ~times the diameter of the tube which is t o be made, is heated in a furnace (Fig. 2 ) to welding heat, whereupon it is drawn over a core through a round die, forming a tube of such diameter t h a t the edges of the sheet overlap one another and b y pulling the sheet with great force over the core and through the die, the weld is pressed together so that i t does not exceed in thickness the original plate. Cutting off the crop ends from the tube, a length of about 2 2 feet is obtained which is put into the testing bench, in which both ends of the tube are closed b y suitable pistons. This done, water under about 600 pounds pressure is applied to the interior of the tube, whereby leaks may be detected and whereby the entire tube is rounded out into a uniform shape. If the tube has proved sound, i t is cut in suitable lengths, which serve for the manufacture of shipping cylinders. Circular plates are cut for the heads and while hot are dished and flanged in a suitable press. A boss is welded t o the center of one head for each cylinder, to give sufficient metal for the hole receiving the \ralve and the heads are pressed into the ends of the cylinders, heated t o a welding point, pressed together and flared between conical rollers, as shown in Fig 3, whereupon the cylinders are ready for testing. The testing includes an inspection for visible defects, the ascertaining of their length, diameter and weight, a test of the material for ductility and strength, and the test of the finished cylinder b y pressure test. For the test of ductility of the material, the crop ends may be used which constitute the waste in cutting tubes for cylinders t o length. The test for ductility is made b y bending test strips t o four times the thickness of the material, whereupon no cracks must be visible on the bendsq(Fig. 4 ) . The test for strength of the material may be made b y cutting test strips from the crop ends, finishing them as indicated in Fig. 5 and pulling them in a testing machine t o destruction. If waste material has not been saved in cutting pipes, a number of cylinders, say one out of 2 0 0 , are cut up for the purpose of obtaining test pieces. Cylinders, the measures and weights of which are beyond the limits or the material of which is deficient or which, upon inspection, show deformities or defects, are rejected without’further test The finished cylinders are subjected to the following pressure tests: Prelivniriary Test -A hydrostatic pressure of I ,000 pounds t o the square inch is put on to round up the If, upon inspection cylinders and to detect leaks after this test, the cylinders do not show any leak, Y
85
local weakness or any visible defect, they are subjected t o the Water-jacket Test, t o ascertain their elasticity and permanent expansion. The test is made in the apparatus illustrated in Fig. 6. A steel cylinder, B, closed a t one end and of sufficient size to receive the cylinder A, which is to be tested and having a cover of a design similar
e
A
4
t o C, in Fig. 6 , is set in an upright position, and is connected with a gauge tube, D, which in turn is provided with a scale divided in inches and tenths. The apparatus is operated as follows: Vessel B (called the jacket) is filled with water and cylinder A, also filled with water, is lowered into jacket B , care being taken that all the air is removed by means of a bent pipe from the cavity in the lower head of the cylinder. Cover C is put on, fastened with suitable clamps, and b y screw-
86
T H $ \$O URlVAL OF I N D U S T R I A L A N D ENGIiVEERING C H E M I S T R Y .
ing u p the hand wheel a , the joint is made watertight between the rubber gasket b and the head of the cylinder. By opening valve c, water is ad-
Feb., 1912
the water column in tube D rose 11.1 inches and this value represented the total cubical expansion of the cylinder under the prevailing pressure, which,
1
Fig. 7.
mitted t o B which drives out all the air from the apparatus through the escape valve d. After closing d the water rises t o the top of the gauge glass D, and sufficient water is allowed t o run from the tube t o remove all the air from the apparatus, whereupon valve c is closed. By manipulatingvalve e , the water column in D is adjusted t o zero on the scale, and the apparatus is ready for the test. Water under high pressure is now admitted through tube E, t o the interior of cylinder A, the pressure being observed on the gauge F. With increasing pressure, cylinder A expands and drives an amount of water equal t o the cubic expansion of the cylinder into the 'gauge glass D, where its volume is read off by the scale. Fig. 8 . In a specific instance,
in this instance, was
1,000 pounds to the square inch. Upon relieving the pressure, cylinder A contracted to near its former volume, the water column in D a t the same time receding towards the zero mark. In the instance referred to, the water receded to 0.35 inch on the scale, which represented the permanent expansion of the cylinder caused b y the application of 1,000 pounds pressure, and
the quotient
0.35
x
IO0 =
11.1
3 . 1 7 ~indicated the per-
manent set, expressed in per cent. of the total expansion. If the permanent set is less than 10% of the total expansion, the cylinder is considered safe for use and is accepted. These are the tests prescribed by the regulations. For the purpose of investigating how cylinders behave under other conditions, some additional experiments were made, which may be briefly mentioned. TEST O F
A CYLINDER,
WALLS. Pressure.
Pounds.
1 0 INCHES BY 7 FEET, 167 POUNDS, 0.19-INCH 7 2 PARTSON THE SCALE= 100 c c .
Exp. on scale.
400 5 00 600
4.4 5.4
700
7.7 8.9 9.9 11.1 11.7 12.5 13.5
800
900 1000 1050 1100 1150
6.6
Set per cent. of exp.
Set on scale.
2.3
0.1 0.2 0.2 0.25 0.25 0.25
3.7
3 .O 3.2 2.8 2.5
0.35
3.1 4.3
0.5 0.8
6.4 17.8
2.4
L
yield point
Feb., 1 9 1 2
T H E JOC'R.\'AL
OF I S D L - S T R I A L A N D ELVGISEERIAVGC H E M I S T R Y .
I . Subjecting a cylinder to water jacket test and ascertaining expansion and set a t various pressures : The experiment was made in apparatus Fig. 6 and the results are entered in the preceding table.
87
other cylinder of the same description, lying on the ground: The experiment was arranged as sketched in Fig. 9. Cylinder X was put in a horizontal position
B
Fig. 9. 2. Subjecting a cylinder to hydropneumatic pressure t o destruction, done for the purpose of ascertaining whether cylinders made from certain materials will break up into flying fragments if they are exploded by gas pressure: The arrangement of the experiment was as shown in Fig. 7 . Cylinder A was set in pit B, which was dug into the ground and protected b y heavy timbers. On a near-by river bank, cylinder C (having five times the size of A ) , of heavy material and previously tested by internal hydrostatic pressure at 3,000 pounds'to the square inch, was set a t an angle of 4 5 O , the top of this cylinder being connected with cylinder A b y a I/,-inch extra strong pipe. Both cylinders were filled with air a t 1 0 0 pounds pressure by valve a, whereupon valve a was closed. Subsequently water was admitted under high pressure by valve b into cylinder C, increasing gradually the air pressure and driving the air from cylinder C into cylinder A until explosion occurred. An exploded cylinder is represented b y Fig. 8. It was burst open, not sending out any fragments of metal in its destruction. 3. Subjecting a cylinder t o drop test for' the purpose of ascertaining whether cylinders of certain materials charged with compressed gas will break if they are dropped in a loaded condition from a platform, wagon or car, landing crosswise on an-
into pit B, connected up to cylinder C and filled with air of the desired pressure. A drop weight, G, having the weight of a cylinder filled with ammonia, was suspended over the center of the cylinder and was dropped b y releasing the lock. Two pictures (Fig. IO) show the indentations made b y the weight. There being no fracture, the conclusion was drawn t h a t the cylinders could have dropped from an equivalent height, without causing a serious accident b y explosion, PRESSURE
GAUGES
USED F O R
T H E TESTS A N D E X P E R I MENTS.
Attention is called to the unreliability of commercial pressure gauges under high pressures. Variations as high as 107~were repeatedly found, and i t is necessary t o make frequent comparisons with dead-weight testing apparatus to ascertain the correctness of the instruments,
Fig. 10
88
1.
T H E J O U R N A L OF I N D U S T R I A L A N D E h ; G I N E E R I S G C H E M I S T R Y .
The apparatus used for controlling the gauges in the described experiments was of the design as offered b y the American Steam Gauge and Valve Manufacturing Company, of Boston, Mass., the accuracy of which was controlled b y the United States Bureau of Standards and found t o be correct within a fraction of one per cent. ’ (Continued i n the next number.)
THE MANUFACTURE AND TESTING OF CARBONIC ACID CYLINDERS.’ BY JOHN C. MINOR, JR. Received Dec. 26, 1911.
The manufacture of liquefied carbonic acid was begun in this country about thirty years ago and has expanded t o an industry of considerable magnitude. While its uses are of a varied nature, it is as the essential ingredient of carbonated beverages t h a t carbonic acid has become an important article of commerce. , Until very recently there were no traffic regulations in this country covering the shipment of compressed gases, nor was there-unless perhaps in isolated cases-any effort on the part of manufacturers of such gases t o procure gas cylinders of any distinctive type or particular quality. I t was considered sufficient t h a t the cylinders had-by a custom inherited from Germany-passed a hydrostatic test of 3700 pounds per square inch. I n 1896 a British Parliamentary Committee made a lengthy investigation of questions relative t o the safe transportation of compressed gases, and certain regulations were recommended in its report but did not become official. They mainly summarized what was in most respects the prevailing British practice and when later they were largely embodied in the British railroad transportation requirements but little difficulty was found in complying with them. I n Germany there were a t t h a t time Governmental regulations regarding the filling and testing of such cylinders, b u t only within comparatively recent years were these extended t o cover the manufacture of the cylinders. I n this country the transportation of all dangerous or inflammable articles is now carried on under the regulations of the Bureau of Explosives of the American Railway Association as approved b y the Interstate Commerce Commission. The Bureau, through its chief, Col. B. W. Dunn, first took up the matter of compressed and liquefied gases early in 1909. There had already been some careful and extensive work done on this subject b y Prof. R. T. Stewart, of the TJniversity of Pittsburg. I n his paper, “ T h e Physical Properties of Carbonic Acid and the Conditions of its Economic Storage for Transportation, ” read in’December, I 908, before the American Society of Mechanical Engineers-and from which the writer has freely drawn-are to be found in condensed and Paper presented at Annual Meeting, American Institute of Chemi-
cal Engineers, Washington, December, 1911.
Feb., 1912
usable form valuable data relative to the physical properties of carbonic acid. The National Tube Company of Pittsburgh had also instituted a preliminary investigation of the subject and when, a t the suggestion of the Bureau of Explosives, the manufacturers of liquefied gases determined upon a n investigation of the cylinder question, every facility was offered them b y t h a t company t o make it complete and t o i t is due the credit of much of the cylinder data here presented. Before considering the cylinders themselves several facts regarding carbonic acid may be noted. This product may appear in transit either in liquid form, or as a vapor, or in both conditions, together. Its critical temperature is 8 8 . 4 ’ F., above which i t cannot exist in liquid form. At any temperature below this, all of the cylinder may be filled with the liquid gas or a part of i t may be filled with the liquid and the remainder with the saturated vapor. Just as long as there is saturated vapor present, the pressure will be constant for a fixed temperature without reference t o the proportion of liquid present, but when the cylinder becomes entirely filled with the liquid the pressure will no longer be constant for a fixed temperature It is obvious t h a t the condition, whether gaseous, liquid, or both, that obtains below the critical temperature within the cylinder and the pressure thereby exerted on its walls is dependent on the relation existing between the total capacity of the cylinder and the amount of carbonic acid which i t contains. This relation is expressed b y Professor Stewart in his table appended hereto, as the density and represents the proportion of filling in pounds of carbonic acid per pound of water capacity, being for each condition tabulated, the combined densities of the liquid and its saturated vapor and being equal in each case t o their combined weights divided b y the weight of their combined volumes of water a t its maximum density; t h a t is, a density of 0 . 7 5 would mean t h a t a cylinder which could hold I O O lbs. of water had 75 lbs. of carbonic acid in it, or a cylinder with a water capacity of 80 lbs. contained 60 lbs. of carbonic acid. Reference to this table of ProfessorStewart will show for any given proportion of filling or density, the pressure exerted by carbonic acid within the cylinder for any given temperature and which i t must under all conditions safely resist. These pressures are increased b y about 60 lbs. per sq. in. for each per cent. of air present. Manifestly, strength t o resist pressure is only one feature of a properly designed cylinder and ability t o withstand rough handling in transportation and t o resist fragmentation on fracture are other important elements in the problem of producing a safe cylinder. Carbonic acid gas cylinders are made in this count r y b y three different processes. I n the plate process, a circular plate, of a thickness and diameter corresponding to the size of cylinder t o be manufactured, is heated and punched into a half spherical cup (Fig. I). This cup is then pushed
