July, 1914
1
T I l E J O U R N A L O F I N D U S T R I A L A N D E-VGIAVEERING C H E M I S T R P
583
LABORATORY AND PLANT PITOT TUBES FOR THE MEASUREMENT OF GAS VELOCITIES‘ By AXDREWM. FAIRLIE
T h e need for a n accurate method of determining t h e velocities of moving gases in pipes a n d flues is receiving increased attention from the marine a n d mechanical engineers, a n d i t is a need which has long been felt by t h e chemical a n d metallurgical engineers, b u t has been somewhat neglected by t h e m . Gases are used as raw materials, or appear as either intermediate or final products, in many chemical a n d metallurgical processes of manufacture. Undoubtedly, in such processes, an accurate means of measuring gas velocities would lead t o the prevention of much loss a n d waste, through t h e substitution of known facts for guess-work. For example, wherever air is used under pressure, as in the copper blast furnace or t h e copper converter, a knowledge of the amount of air used, in comparison with t h e theoretical a m o u n t required, would in many cases lead t o a reduction of air, a n d so of power, wasted. I n t h e sulfuric acid industry, where gas is diverted t o t w o or more absorption towers connected in parallel, a knowledge of t h e velocity, a n d hence of t h e quantity of gas going t o each tower would lead t o a proper adjustment of dampers so as t o secure equality of distribution a n d more economical operation. I n suits for damages claimed t o have been caused by obnoxious fumes escaping from t h e flues of industrial works into t h e atmosphere, t h e substitution of knon-ledge as t o t h e quantity of escaping fumes, for estimates a n d guesses, would eliminate legal controversies. The business of selling natural gas a n d manufactured gas for power. heating a n d lighting, demands a n accurate means of measuring the velocity of flow, in order t o properly affix costs and selling prices. I n designing t h e equipment of chemical a n d metallurgical plants, some means of calibrating the capacities of blowers a n d ventilating machines, vacuum pumps. aspirators, etc., would enable engineers t o check up t h e claims of manufacturers of such equipment, and would prevent t h e installation of inefficient devices, as well as t h e financial losses involved thereby. Other examples of t h e value of a reliable means of measuring t h e velocities of gases will occur t o t h e reader. Georg Lunge statesz t h a t as early as 1866 Fletcher’s modification of Peclet’s differential anemometer was described. This is in effect a crude sort of pitot tube. T h e velocity of t h e gas was calculated on t h e basis of t h e difference between t h e pressures, as measured by a manometer, exerted by t h e current of moving gas on two tubes, one of which was straight, t h e other being bent a t a right angle, a n d turned so t h a t t h e current of gas would blow into it. I n using this anemometer. i t was t o be inserted into t h e air current t o t h e extent of about one-sixth of the diameter of t h e flue. 1 Presented at the 49th Meeting of the American Chemical Society, Cincinnati, April 6-10, 1914. 2 Geo. Lunge, “Manufacture of Sulfuric Acid and Alkali,” 3rd Ed., Vol I, Part 1, p. 563.
The velocity a t this point was assumed t o be nearly equal to t h e average, a n d t h e velocity was based on a reading taken a t this one point. Lunge admits t h a t the accuracy of t h e assumption is doubtful, and adds: “ T h e r e are no means a t present1 known of measuring t h e absolute quantities passing through a flue of a n y considerable sectional area with a n y degree of accuracy.’’ Since 1903 numerous investigators have published accounts of t h e results of their work on pitot tubes. Different forms of pitot tubes have been designed, each designer claiming for his form, perhaps, superior accuracy. With so many different forms of tubes, none acceptable as a standard, engineers were as badly off as if there were no pitot tubes a t all, and, in fact, all forms of pitot tubes were more or less distrusted. I n September of last year, t h e results of a n exhaustive series of comparative tests of t h e different forms of pitot tubes were published by W. C. Rowse.2 T h a n k s t o t h e painstaking work of this investigator, i t is now possible t o separate the wheat from the chaff, a n d a form of pitot tube may now be selected with some confidence. Those interested in t h e velocities of gases cannot do better t h a n read the whole of t h e paper by Rowse, which, with diagrams, tables, etc., covers a b o u t fifty pages. With the consent of t h e author, t o render his paper of still greater value, two errors in t h e statement of formulas which inadvertently crept into his work are here pointed out. On page 1343,paragraph 46 (lz) reads as follows:
‘‘ It appears t h a t a n approximate relation exists between the mean velocity head of a gas flowing through the pipe a n d t h e velocity head found by placing the t u b e a t the center of the pipe. For a 12-in. galvanized iron pipe results within 2 per cent may be expected f r o m using t h e formula ~
~
~~
where v = d ( 2 g ) (0.80) hc = velocity in feet per second. g = 3 2 . 2 ft. per second per second. h, = velocity head in in. of gasolene a t the center of the pipe obtained in a correct manner ” 2)
This is all correct except t h e value given t o hc, which should read: h‘
=
velocity head in feet of gas flowing at the center of the pipe, obtained in a correct manner.
Again, on page given which reads:
1372
(paragraph 76) a formula is
‘‘h = 144SfJH, where it = velocity head in feet of air. s = sp. gr. gasolene. p = weight of water in pounds per cu. in. taken from Chart G. H = velocity head in in. of gasolene as given in columns 18 or 19.” 11903. J . A m . SOC.Mech. Eng., Sept., 1913.
2
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
584
This formula should read
where w =
weight of
I
cu. ft. air under existing conditions, and
p , and H have the values assigned above.
h, s,
I have been requested by the author t o refer also to the third error in t h e paper, Z J ~ Z . ,Chart A, Appendix KO. 3. Inclined dotted lines should be designated as follows, beginning a t the bottom: 1/10; 6/100; j / ~ o o ;4 / 1 0 0 ; 3/100; 2 / 1 0 0 ;
I/IOO;
1/200.
From this paper i t appears t h a t the most accurateform of the pitot tube is the standard tube of the American Blower Company, which was developed by Chas. H . Treat. After the publication of Rowse’s report, the American Blower Company issued a special bulletin, entitled “ T h e Pitot T u b e and F a n Testing.”’ This Bulletin, which can be obtained gratis from the American Blower Company, Detroit, Mich., gives a dimensioned sketch of the pitot tube adopted b y t h e m as a standard, a n d reviews to some extent the work of Rowse. Some formulas for calculation are also given. The Bulletin is of some value, b u t the calculations and formulas given must be accepted with caution, as a number of errors have been noted, particularly in the first edition. Curiously enough, one of these errors is again in t h e value of “h” (see above), the velocity head in feet of air. On page 2 5 of the Bulletin*the second equation given reads : h =
‘(
45 IZW
where p
velocity pressure in inches water gauge. w = weight of one cu. ft. of air, under existing conditions.” =
The radical sign in the equation should be omitted On the same page i t is stated (second line from the bottom) t h a t w = velocity of air in feet per second, as applied t o t h e various equations given on t h a t page. I n t h e fifth a n d sixth equations, however, the value of ZJ is, in fact, the velocity of air in feet per minute. I n t h e fifth a n d sixth equations, therefore, the symbol “ v ” should be replaced by “v,,,,” t o represent the velocity of air in feet per minute. There are also evidently some errors in the formulas given on page 26 of the early edition of the Bulletin. Reverting now t o t h e consideration of the work of Rowse, the most important of his conclusions are: I-The pitot tube as a means of measuring gases is reliable within approximately I per cent when the static pressure is correctly obtained a n d when all readings are taken with a sufficient degree of refinement; in order t o obtain this degree of accuracy the pitot tube should be preceded b y a length of pipe 2 0 t o 38 times t h e pipe diameter in order t o make the flow of gas as nearly uniform across the section of the pipe as possible. 2-All t h e methods of obtaining the dynamic head used in his experiments gave accurate results. Of Bull. 36, Series 1, January, 1914. It is understood t h a t in the later editions of the Bulletin, most of the errors have been corrected. I
2
Vol. 6, No. 7
the methods of obtaining t h e static pressure by means of the pitot tube, the most reliable and accurate is by means of a very small hole in a perfectly smooth surface, as in t h e standard tube of The American Blower Company. 3-It appears t h a t a n approximate relation exists between the velocity head found by placing the pitot tube a t t h e center of the pipe, a n d the mean velocity head of t h e air flowing. (See formula for a 12-inch pipe on page j83.) From these conclusions it is apparent t h a t there are limitations t o the uses t o which t h e pitot tube can be put. T o extend these limits,further work will be required. There are many pipes already constructed in which i t would be desirable t o measure t h e speed of the moving gases, but which do not present a length twenty times the pipe diameter. How shall we determine t h e velocities in such pipes? Rowse has given a convenient formula (conclusion “3” above) for arriving a t approximate velocities from a single pitot tube reading taken a t the center of a pipe 1 2 inches in diameter. Who will develop similar relations for pipes of other sizes? T h a t such relations for different sizes of pipes are desirable is evident t o any who have gone through the laborious work of taking 2 0 pitot tube readings across two diameters of a pipe in order t o obtain a single mean velocity head. Some have attempted a short cut method by using, for obtaining the dynamic pressure, a tube long enough to reach entirely across t h e pipe containing the gas whose velocity is t o be measured, with a number of small holes (centers all in t h e same plane) bored in the tube, each hole being a t the center of one of t h e annular zones of equal area into which t h e cross-sectional area of the pipe is assumed t o be divided, a n d all of the little holes being turned so t h a t the current of gas will blow directly into them. Another short cut device which is sometimes used for obtaining the dynamic pressure consists of a number of tubes of different lengths, each with the inside end bent a t a right angle towards t h e gas current, t h e inside end of each being located in t h e center of one of the several annular zones into which the cross-sectional area of the pipe is mentally divided, a n d t h e outside end of each being connected t o one common manifold pipe, where the inequalities in the various pressures exerted on the several tubes are neutralized. With either of these devices i t is assumed t h a t the dynamic pressure tube gives a mean dynamic pressure for the entire cross-sectional area of t h e pipe, from which, b y deducting t h e static pressure, t h e mean velocity head of the gas can be obtained. However, since t h e velocity of a current of gas does not vary directly as the velocity head, i t is difficult t o see how such instruments, which give as net result a straight average of the velocity heads a t t h e different points in the pipe tested, can be expected t o produce data on which t o calculate a correct mean velocity. From the formula v = 427h the velocity varies as t h e square root of the velocity head. T o obtain a correct mean velocity, a correct
July, 1914
6
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
mean velocity head is needed; a n d t o obtain a correct mean velocity head i t is necessary t o average t h e square roots of all t h e velocity head readings taken, throughout t h e cross-sectional area of t h e pipe, a n d then t o square t h e average. There is a t present no royal road t o obtaining accurate velocity measurements b y means of t h e pitot tube. Investigators, t o obtain accurate results, must have recourse t o t h e painstaking methods adopted b y Rowse. We are told t h a t , for a 12-inch pipe, results within 2 per cent of correct may be obtained b y using 0.8 of t h e velocity head in feet of gas a t t h e center of t h e pipe. Lacking similar factors for pipes of other sizes, velocities must be calculated from readings taken a t ntimerous points in t h e cross-sectional area of t h e pipe. I n recapitulation, it m a y be observed t h a t : I--As a result of t h e work of Rowse, engineers m a y now select a t y p e of pitot t u b e which may be used, under certain conditions, with confidence. 11-Further investigation needed, t o render t h e pitot tube more generally available as a means of measuring gas velocities, includes: I-A means of determining accurately t h e velocity of gases i n pipes whose length is less t h a n 20 times t h e diameter. a-The establishment of definite relations between t h e velocity head a t t h e center of a pipe, a n d t h e mean velocity head, for pipes of various sizes and shapes. NOTE-I a m indebted t o t h e author of t h e original paper which constitutes t h e basis for this one-Mr. W. C. Rowse-for his courtesy in reading this manuscript, a n d for valuable suggestions offered b y him a n d adopted herein. COPPERHILL, TENNESSEE
THE NON-UNIFORMITY OF DRYING OVEN TEMPERATURES By LORIN H. BAILEY Received April 13, 1914
While testing t h e accuracy of t h e heat control of a new electric drying oven, i t was observed t h a t there
585
I n t h e above tests only those thermometers were used which had been standardized b y t h e Bureau of Standards. With ovens having glass doors t h e thermometers were placed on t h e shelves a n d t h e temperatures read b y making t h e observations through t h e glass in t h e door without opening the oven. With t h e other ovens t h e following scheme was adopted: Six 50 cc. Erlenmeyer flasks were filled with clean, dry sand, stoppered, and through t h e stoppers t h e thermometers were inserted so t h a t t h e bulbs were held in t h e middle of t h e flasks. These flasks with their thermometers were then placed in t h e various positions i n t h e ovens and after having remained there long enough t o come t o equilibrium they were removed a n d t h e thermometers read as quickly as possible. While this method is not absolutely accurate, i t is sufficiently so t o indicate whether or not there is any great variation in temperature. After testing these various drying ovens t h e writer's attention was called t o an article t o R . G. Grimwoodl on t h e "Analysis of Crude Glycerine b y The International Standard Methods, 191I . " I n this article t h e author mentions difficulty when using a drying oven which complies with t h e International Standard Methods specifications and which shows a maximum variation of 16" C. on one shelf and between t h e two shelves a maximum variation of 4 j o C. H e then describes a special electric drying oven which has proved satisfactory for drying glycerin. This oven showed a maximum variation of z . z o C. from 160' C., t h e temperature desired. The results shoa-n b y Mr. Grimwood are very much better than t h e writer was able t o obtain with a n y oven not surrounded b y boiling water a n d steam. The results in t h e above table show t h a t elaborately designed a n d expensive ovens are no more reliable t h a n t h e most simple a n d inexpensive ones. Of t h e ovens tested. only those surrounded b y boiling mater a n d steam are capable of maintaining even approximately uniform temperatures. I wish t o t h a n k Dr. J. A. LeClerc for his interest
Reading of Maximum range of temperature on Maximum thermometer . variation Approximate inserted through ___* top of oven top shelf lower shelf on both shelves inside dimensions in. O C OC O C OC No. Kind of oven 107 99-114 90-108 24 12 X 12 X 14 1 Electrically heated and controlled.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 89-100 88-103 15 12 X 13 X 18 2 Electrically heated and controlled. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 94-104 89-102 15 9 x 9 x 1 5 3 Electrically heated and controlled.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 8892 92-102 14 11 X 12 X 16 4 Gas heated porcelain lined., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 94- 96 105-118 24 8 X 10 X 11 ... 5 Gas heated air jacketed.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 9699 9598 4 13 X 15 X 15 6 Gas heated constant level water and steam jacketed., . . . . . . . . . . . . 100 101-101 101-102 2 15 X 8 in. diam. 7 Steam jacketed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 100-100 100-100 0 15 X 8 in. diam. 8 Gas heated constant level water and steam jacketed vacuum oven. .
was a wide range of temperature i n different parts of t h e oven, not only between different shelves b u t also between different positions on t h e .same shelf. Ordinarily t h e temperature recorded on t h e thermometer inserted through t h e t o p of a n oven is taken as t h e temperature a t which t h e drying is made, b u t i t was seen t h a t such could not be done with this particular oven. Following this observation a number of different types of drying ovens were tested as t o their uniformity of temperatures throughout t h e drying chambers. The variations i n temperature are shown in t h e accompanying table.
a n d suggestions in connection with t h e testing of these various ovens a n d t h e writing of this report. LABORATORY OF PLANT CHEMISTRY, BUREAU OF CHEMISTRY u. s. DEPARTMENT O F .kGRICULTURE, m ' A S H I N G T O N
A MANOSTAT FOR USE IN GAS ANALYSIS By HARVEYN. GILBERT Received April 14, 1914.
I n t h e combustion of gases confined over mercury in a combustion pipette, more or less difficulty is always experienced b y t h e operator i n avoiding a difference of pressure due t o t h e difference of level between t h a t 1
J. SOC.Chem. I n d . , 38, 22.
