Investigations with the U. S. Bureau of Mines Modification of the Abel-Pensky and Pensky-Martens Flash-Point Testers
Journal of Industrial & Engineering Chemistry
I n the handling of lamp oils, particularly in mines and in buildings where open lights are used, it is important t h a t a relatively safe oil. i. e . . one with a high flash point. be employed. Equally important is t h e determination of t h e inflammability of all oils in common use. This paper, which is a brief of Technical Paper No. 49. published b y the Bureau of 1Iines in the interest of greater safety a t mining and metallurgical plants. does not aim t o establish definite flash points
... inite flash points for oils b u t describes the instruments and methods of testing t h a t can be used as a basis on which t o establish such flash points. T H E F L A S H P O I S T I n the various methods used for determining the flash point of a n oil the temperature a t which its vapor flashes is not a definite factor but is dependent on a number of physical and chemical conditions not under the control of t h e operator. I t is not a n indication of the value of a n oil for any particular purpose. I t is a n indication only of t h e temperature at which the oil gives off vapors in such proportions t h a t they form a n inflammable mixture with the air. If exposed t o an electric light or a flame the mixture will ignite and t h e flame extend throughout the mixture, t h a t is. a n explosive mixture of vapor and air is formed. T H E F I R E HAZARD The flash point, although probably the most important factor, will not, by itself, determine the fire hazard of any substance. The Ere hazard of a substance is also affected by the volatility. the boiling point. the vapor pressure, the vapor density, the diffusibility and tendency of the vapors t o travel and their explosive limits in air, its tendency t o chemical change, the quantity of heat liberated per unit of time and unit of volume, the temperature of the flame, the corrosive action and toxic properties of the substance and of its products of combustion, its behavior toward water both before and after ignition. and the tendency of the substance t o leak. .% requirement t h a t a n oil has the highest practicable flash point, tending thereby t o reduce the fire hazard t o a minimum, is praiseworthy and humane. T o require too high a flash point may impair the burning qualities of a lamp oil and also work a hardship on the r e h e r . However, the immediate profits of the refiner should be deemed less important t h a n the safety of the user and the protection of the public against possible loss of life and of property by fire, The fire hazard of lubricating oils is of importance Tyhen the lubricants are used in rapidly running machinery, as in spinning mills. wood mills, etc., or in factories containing combustibles and in compressors for air, ammonia, or other gases. Particular attention should always be given t o the Paper presented a t t h e 48th meetin4 of t h e A . C. S., Rochester, September 8-14. 1913, by permission of t h e Director of the U. S. Buresu of Mines. A4LVD E-VGISEERI-VG C H E M I S T R Y T'ol. 5, S o . 11 handling, storage, and inspection of all inflammable liquids. The Bureau of Mines in Technical Paper No. 49 has recommended the adoption of a modified form of the -1bel-Pensky and Pensky-Martens testers for the determinations of the flash points of oils flashing below 1jo0 C. The reasons for the recommendations of these testers are clearly set forth in Technical Paper No. 49. a brief abstract of which includes ( I ) the reasons for the necessity of exact determination of flash points of oils t o determine the fire hazard with their use; ( 2 ) t h e historical development of flash-point testing with descriptions of testers in use; (3) the different factors t h a t influence the obtaining of exact a n d concordant results; (4) a n d directions for the use of the two testers recommended. The modifications of t h e older forms of Xbel-Pensky and Pensky-3lartens testers consist of the following: each cup has been fitted with a n annular overflow cup into which t h e excess of oil due t o the expansion on heating flows through a small opening in t h e side of the cup on a level with t h e filling point. This assures the volume of t h e oil and the volume of the vapor mixture above t h e oil remaining constant. The Xbel-Pensky tester is provided with a paddle-wheel stirrer identical with t h a t of t h e Pensky-IUartens, while the automatic slide of t h e Abel-Pensky tester has been fitted t o the Pensky-Martens cup. However. in the last change it was found necessary t o raise the clock work on posts above the hot bath as the intense heat drew the temper from the spring. The two testers as recommended are identical with the exception of the use of t h e water bath for t h e testing of low flash oils and the metal-air bath for those of high flash point. The cups, the automatic opening devices and t h e stirrers of the two testers are interchangeable. An investigation undertaken in these laboratories indicates t h a t the Pensky-Martens tester can be used very satisfactorily for the testing of low flash-point oils though the temperature regulation is not as nice as with the water-jacketed Abel-Pensky tester. The same conclusion has been reached by other investigations.' Each tester is fitted with gas or oil igniti'on burners. T h e present paper includes results of a brief investigation with the two testers. A degree of accuracy greater t h a n 0 . j o C. for oils flashing below 100' C. and I o o C. for oils flashing above 100' C. was not attempted and any fractions of degrees reported in subsequent tables are derived as averages of several tests or from thermometer, barometer or instrument corrections. To obtain comparable results the two instruments were compared with a n Abel-Pensky tester standardized in 1908 by the Physikalische Technische Reichsanstalt, certificate No. S. & R. 31 14. Determinations of flash points of three different oils were made on the three instruments. Results .obtained are the averages of five tests agreeing within 0 . j" C. and are shown in Table I . 1 D r H W e b e , "Die Obere Brauchbarkeitsgrenze des Abel-Penskyschen Apparates s n < l seine l'ergleichung m i t dem Penskyschen Flammpunktpriifer, Petvoleurn. 1913, Band V I I I , S. 1061.