INDUSTRIAL A N D ENGINEERING CHEMISTRY
262
Vol. 16, No. 3
Standard Atmospheric Conditions for Paper Testing' By E. 0. Reed* GOVERNMENT PRINTINGOFFICE, WASHINGTON, D. C.
NY atmospheric change, whether in temperature or relative humidity, which affects the moisture content of paper, and thus the physical properties of the cellulose, sizing, or other constituents of the paper, will consequently affect the physical properties of the paper as a whole. Yet, although it is fully appreciated that these atmospheric changes, either of temperature or relative humidity, produce marked changes in paper, careful consideration has never been given to the selection of a standard condition for testing. Paper testing has been conducted for a number of years at 70" F. and 65 per cent relative humidity by the very few laboratories that have taken any steps whatever to test under uniform conditions. In fact, the T. A. P. P. I., recognizing the need of standard uniform conditions, has recommended a relative humidity of 65 per cent and temperature of 70" F. It is doubtful if there can be found a logical reason for the selection of such a high relative humidity as 65 per cent. It must be borne in mind that in determining upon uniform standard atmospheric conditions for testing paper or any other material, a careful study should be made of the actual use and service required and under what atmospheric conditions the material must be satisfactory. In the selection of a 65 per cent relative humidity, however, this fact was undoubtedly not taken into consideration. The selection of 65 per cent relative humidity as a standard testing condition was due to the fact that it had been adopted by the German Imperial Institute for testing materials at Gross-Lichterfelde in about 1890.s The only explanation that can be found for their selection of this humidity is that since the average outdoor relative humidity is more than 65 per cent and the indoor humidity much lower it is easier t o maintain a high relative humidity. This surely cannot be considered a sound argument for the adoption of a standard relative humidity of 65 per cent, if we consider a t all the actual indoor conditions under which paper is used. Paper or any other material should be tested under conditions where a direct comparison may be made between the tests obtained and the actual serviceability. These conditions must be set so as not to favor the material, but to place it under average conditions, or possibly even under more severe atmospheric conditions than it must be exposed to and under which it must be handled. The average outdoor relative humidity in most localities in the United States is approximately 70 per cent. Wilson4 states that during the winter months the normal outdoor relative humidity over the more populous portions of the United States, especially east of the Missouri and north of the Ohio River, is 72 per cent, and that the average diurnal range is from 60 to 85 per cent. He determined the indoor relative humidity in heated buildings in several widely separated parts of the country, on the coast and in the interior, and states that it varies from 24 to 33 per cent, and that during that portion of the year when artificial heat is required in the United States it is lower than that of the driest climate known.
A
1 Received
December 20, 1923.
* Chief of Tests.
Mitt. K g l . tech. Vevsuchungsanslalten, 7, 2 (1889); 8 , 8 (1890). "Atmospheric Moisture and Artificial Heating," Proceedings the Convention of Weather Bureau Oficeus, 1898. 8
4
of
Personal observations on indoor atmospheric conditions extending over several years, a t Washington, D. C., and other points, very definitely show that the indoor relative humidity, especially in the winter months, is much lower than the outdoor relative humidity, and that, although the mean annual outdoor relative humidity a t the points covered in this investigation was about 70 per cent, the mean annual indoor relative humidity was between 35 and 45 per cent. During the greater part of the year-October through May-when artificial heat is required, the average monthly relative humidity indoors is not over 40 per cent, and during the five colder months averages less than 35 per cent. Frequently in winter the relative humidity is as low as 15 or 20 per cent, averaging approximately 30 per cent for December, January, and February. During June, July, August, and September the average indoor relative humidity is higher than 40 per cent, and may go as high as 70 or 75 per cent for a few days a t a time, and sometimes, but only for very short intervals, even somewhat higher than this. The averages noted for the month of June, however, were between 44 and 49 per cent, for July between 44 and 58 per cent, for August between 49 and 60 per cent, and for September between 55 and 65 per cent. I n no month did the average indoor relative humidity equal the existing average outdoor relative humidity, but averaged from 15 to 40 per cent less. The highest indoor relative humidity noted occurred in August and September. The colder the climate the lower was the average indoor relative humidity, owing to the longer requirement for artificial heat, and the warmer the climate the higher was the average relative humidity indoors. The largest quantities of paper are used indoors, where the average living and working conditions mentioned above are maintained throughout the year. It must be assumed, therefore, that a humidity approximating 45 per cent, and not one of 65 per cent, is more in harmony with the actual conditions under which paper is used and under which it should be tested. Giving careful consideration to these facts, the Government Printing Office has adopted a standard testing condition of 50 per cent relative humidity and a temperature of 70" to 75" F. For maintenance of these conditions a testing room designed by the chief of tests, which automatically operates day and night, was constructed about two years ago, and has continuously maintained a relative humidity of 50 per cent and a temperature of 70" to 75" F. This installation has been found entirely commercial and not unreasonable in cost. With the maintenance of any humidity for testing, it is also necessary to control the temperature within a certain range of from 70" to 75" F., and the cooling system necessary for this control will also be sufficient to de-humidify during the few short periods of the year when this is necessary. From information obtained from engineering companies it has been found that the cost of maintaining a 50 per cent relative humidity and 70" F. temperature is the same or less than that of maintaining 65 per cent at this temperature. This is contrary to statements heretofore made in defense of 65 per cent relative humidity. It is a well-known fact that papers in actual service do not stand up in direct proportion to laboratory test values a t 65 per cent relative humidity. This is especially marked in
INDUSTRIAL A N D ENGINEERING CHEMISTRY
March, 1924
-
regard to folding quality, which, when considered with the other tests, is most indicative of the serviceability, durability, and quality of paper. All papers exhibit a better folding quality at 65 per cent relative humidity, very few being unsatisfactory a t this unusual condition. Yet some papers that are satisfactory a t 65 per cent relative humidity are brittlo and even unserviceable at normal indoor conditions. It is paramount, therefore, that a testing condition be selected which will give results most indicative of the real properties of the paper when in use. From a wide experience with paper testing, especially as to folding quality, whether obtained by a folding endurance tester or by hand, it has been very definitely found that the effect of relative humidity or of temperature is not the same on all kinds of paper, or even on papers of the same grade or manufacture. This fact has been shown by other investigations to be true also of other physical tests. Hence, no factor caD be used in calculating any physical test from results at one condition to equivalents a t another or a standard
263
condition, and comparative ratings of papers a t one atmospheric condition will not be the same a t another condition. I n the report of the technical committee6 appointed by the Bureau of Standards in connection with the standardization of paper, the following statement was made: This Committee does not recommend a particular relative humidity a t which paper should be tested, but it does recommend t h a t this subject be considered and that some rather narrow limits of relative humidity be adopted, taking into consideration the average indoor relative humidity and the cost of maintaining constant humidity conditions in the testing laboratory.
As stated, the Government Printing Office Testing Section has selected 50 per cent relative humidity and 70" to 75" F. temperature as a fair testing condition, and under these conditions our papers tested have shown the very properties in service that were brought out by laboratory tests. It is recommended, therefore, that full consideration be given to the acceptance of these conditions as standard for paper testing. 6
Paper Trade J.,75, 38 (1922).
Composition of Natural Gas Gasoline' By R. P. Anderson and A. M. Erskme UNITED
HE lower boiling constituents of crude petroleum have been carefully separated and identified by Young2 and his eo-workers, but, as far as can be learned, the exact nature of the compounds present in the gasoline from natural gas has not been determined. The purpose of the work described in this article was t o study the chemical composition of natural gas gasoline by separating the constituents by fractional distillation in a sufficiently pure state to make possible their identification.
T
NATUEAL G A S CO.,OIL CITY, PA.,
Received October 1, 1923. Young and Thomas, J. Chem. SOC.(London), 71, 440 (1597); Young, Ibid., 73, 905 (189s); Francis and Young, Ibid., 73, 920 (1898); Fortey, Ibid., 73, 932 (1898). 2
HAMILTON COLLEGE, CLINTON, N. Y .
A sample of natural gas gasoline was separated by fioe fractionations into four portions with a boiling point range of about 20" C. each, in the neighborhood of the normal parafin hydrocarbons pentane, hexane, heptane, and octane. Three fractionations of the pentane portion with a regulated temperature stillhead resulted in the isolation of isopentane (27.8" to 28.3" C.) and normal pentane (35.5' to 36.0" C.). From the hexane portion by a series of four fractionations with the same apparatus isohexane, 2-methylpenfane (60.0' to 61.0" C.), and normal hexane (68.0' to 69.0" C.) were isolated. Xsoheptane, 2-methylhexane (90.0" to 91 .O" C.), and normal heptane (98.0' to 99.0" C.) were separated in a fairly pure state from the heptane portion in twelve fractionations, the last two with the regulated temperature stillhead. The approximate quantitatioe composition of this gasoline is: Propane and butanes 20 per cent (by colume), isopentane 13 per cent, n-pentane 17 per cent, isohexane 9 per cent, n-hexane 15 per cent, isoheptane 8 per cent, n-heptane I2 per cent, octane 4 per cent. and absorption oil 2 per cent. The specific gravity-boiling point relations indicated the presence of traces of benzene and toluene.
EXPERIMEKTAL The material used was a sample of absorption gasoline taken July 3, 1922, a t the Strong gasoline plant of The Mars Company. A series of five preliminary fractionations was made, which resulted in a fairly satisfactory separation of the gasoline into four distinct portions, each with a boiling point range of about 20 " C. in the regions of the normal paraffin hydrocarbons pentane, hexane, heptane, and octane. The fractions in the pentane region were then treated separately in a series of three fractionations, using a regulated temperature stillhead of high efficiency similar to that employed by Young. A similar series of four fractionations was carried out on the fractions in the hexane region, using this same type of apparatus. The fractions in the heptane range were treated in ten fractionations using ordinary distilling 1
AND
columns, followed by two fractionations with a regulated temperature stillhead.
PRELIMINARY FRACTION-
o ~ s - I n Preliminary Fractionation I a short Hempel column of 2.2 em. (7/* inch) internal diameter filled with perforated glass beads to a depth of 13.3 em. (5l/4 inches) was used. In the second fractionation a long Hempel column with a diameter of 1.9 em. (3/4 inch) and a 51 em. (20 inch) depth of beads waB used below 150" C., while above this temperature the Hempel was replaced by a Vigreux column of 1.9 em. diameter and with indentations for 22.9 cm. (9 inches) of its length. I n the other three preliminary fractionations this change of columns was made a t 85" C. A copper tube condenser of the standard type for oil distillations was used with the Hempel columns, while a glass Liebig condenser in the vertical position and cooled in the usual way was used with the Vigreux column. The copper tube condenser was cooled by cracked ice for fractions below 85" C. and by running tap water for fractions above 85" C . Ordinary glass graduates fitted to the end of the condenser by a cork were used as receivers. These were surrounded by ice when the latter was used for condensation. The distilling flasks were the round-bottom, short-necked type, of capacities adjusted to the size of the fraction being distilled. High-grade Anschutz thermometers totally immersed in the vapor were used for the measurement of all vapor (distillation) temperatures. The original sample, with a volume of 3000 cc. and a specific gravity of 0.6667 (15.5"/15.5" c.),was distilled and ATI
