Apr., 1913
T H E JOL-RA’AL OF I N D U S T R I A L A N D EA-GINEERIATG CHE;LIISTRY
349
1
NOTES AND CORRESPONDENCE AN APPLICATION OF THE ELECTRIC RESISTANCE FURNACE TO THE DETERMINATION OF OXYGEN IN IRON AND STEEL. A CRITICISM Editor o j the Journal of Industrial and Engineering Chemistry: In THISJOURNAL, 5, 123, appears a n article on the application of the electric resistance furnace t o the determination of oxygen in iron and steel. The writer’s’ modification of the Ledebur method is referred to, and the double electric furnace as well as certain other modifications are proposed as improvements. The laboratories under the writer’s supervision have had probably a larger and wider experience with the determination of oxygen in iron and steel than any other in this country. In our opinion the modifications proposed by McMillen present no improvements over the process used by us, and are open t o criticism in respect to several important points. In a routine laboratory where a large number of oxygen determinations have t o be made, the time element is of great importance. With the gas blast furnace, using a silica, tube i t is possible t o cool down immediately by turning on the cold air blast, so t h a t accurate results can be obtained in one-half hour. The double electric furnace modification runs the test for a full hour, and if the analysis is started with a cold furnace, it will require a n hour t o reach the specified temperature of g50° C., whereas t h e gas blast furnace can be brought to temperature in a few minutes. In our practice we often exceed this temperature and have never found t h a t we obtain higher results with the higher temperatures. McMillen proposes the use of a porcelain boat instead of the nickel boat used by us. We especially avoided the use of porcelain, owing to the tendency of t h a t material t o adsorb moisture rrom the air. The most important criticism of McMillen’s modification t h a t we have t o offer is with respect to weighing the phosphoric acid absorption tube filled with hydrogen. We believe this to be bad practice, and numerous tests made in this manner have shown discordant results owing to the tendency of hydrogen t o diffuse through “air tight” stoppers. The McMillen method is almost identical with the one published by the writer, with the exception of the points discussed above, and for the reasons stated i t does not appear t h a t any contribution has been made t o the older method. McMillen again points out the well-known fact t h a t the Ledebur method will not determine total oxygen, owing t o the fact t h a t certain oxides and silicates are not reducible by hydrogen. Walker and Patrickz have proposed a method for determination of total oxygen, by making use of a vacuum electric furnace of t h e Arsem type. The installation for this method is very expensive and a t best i t could hardly become a routine determination in any chemical laboratory. On the other hand, i t should not be forgotten t h a t in most cases i t is only the determination of oxygen combined with iron t h a t yields the data desired, and i t is this information which can be readily obtained by t h e Ledebur method. The amount of oxygen combined with silica and other oxides unreducible by hydrogen has no bearing on the question as to whether iron or steel is properly deoxidized in the process of manufacture. ALLERTON S. CUSHMAN INDUSTRIAL RESEARCH WASHINGTON, FEE. 26, 1913
INSTITUTE O F
AN
ACCURATE
VISCOSIMETER WITH PRACTICAL QUALITIES Editor of the Journal o j Industrial and Engineering Chemistry: The statement of W. F. Faragher concerning “A Standard Viscosimeter” in THIS JOURSAL, 5 , I 6 7 , is, unfortunately, too true to arouse the least hint of denial. Yet we can hardly expect to obtain consistent results with viscosimeters many of which, stripped of all external equipage, are fundamentally dependent on little more than a hole in a can. This statement, which perhaps may seem altogether facetious, will be verified by a glance a t the Engler, Redwood and Saybolt types and others in common use. Where then can we get an instrument of less faulty construction for our standard? Is it necessary t o turn to Germany? Let us first look over what our own chemists have to offer. In 1894 the English chemists Thorpe and Rodger published a paper‘ describing a viscosimeter which introduced the foundation principle involved in a very neat and accurate instrument developed later by Bingham and White.z As used by them, however, i t was scarcely a n implement for practical industrial work until further simplified by White. In an article on the viscosities of blood and blood serum, which appeared in the Biochemische Zeitschrif, 37, 482 ( I ~ I I ) , the latter gives a detailed description of his viscosimeter, the main points of which are noted below. The instrument consists of two vertical limbs of glass tubing fused into the ends of a horizontal capillary tube of about 0.I millimeter inside diameter and of length depending upon the viscosity of the substance used; i. e., varying from about IO centimeters for liquids of small viscosity like water to 2-3 centimeters for the more viscous oils. The limbs are marked so that, by falling from the higher to the lower mark, 3 or 4 cc. of a liquid have passed through the capillary. The volume of the tube between these marks is carefully determined before fusing in the limbs. One limb is also furnished with a trap by means of which the volume used can be regulated. The fluid is forced through the capillary under a known pressure and the viscosity calculated by means of the Poisseuille formula corrected for kinetic energy as shown: ed nr ‘ t t
T i = 8vl
THIS JOURNAL, 3, 372 J b i d . , 4, 799.
8ntl
The volume ( v ) having been determined as indicated above, the constants
nr4
and
li, ~
snl
are easily obtained by calibrating
with a liquid of known viscosity, such as water. The pressure is measured with a manometer and the temperature regulated by immersing t h e instrument in a bath. With regard t o the accuracy of the instrument, a few data will serve better than words of praise. The following is taken from work done on fish and vegetable oils at Wood’s Hole by White and Thomas:S VISCOSITY OF DOGFISH OILS Temperature Degrees 30 50
70 go
Right limb 0.4143 0.2113 0.1247 0.08046
Left limb 0.4140 0.2110 0.1247 o ,08063
Average 0.4142 0.2112 0.1247 0,0805s
In order t o show t h a t accuracy is attainable not by the inventor only, there are noted below a few results obtained by the writer working on butter fat with a n instrument of this style. Having demonstrated the consistency of results obtained, i t remains t o be seen whether there are enough points in favor of 1
1 2
--
2
Phil. Trans., 186, 397 (1894). 2. ghysik. C h . .80 670 (1912).
a THIS JOURNAL, 4, 378 (1912).