IN DUSTR IA L A N D E N GI N EE R I N G CH E M ISTRY Bailey and Edwards, J. IND.ENQ.CHEM.,12, 892 (1920). Bartoli and Stracciati, Garr. chim. ital., 15, 417 (1902). Batelli, A t t i ist Veneto, 3, 1781 (1884). Bell, “American Petroleum Refining,” 2nd ed., p. 48, New York, D. Van Nostrand Co., 1924. Blacet, Leighton, and Bartlett, J . Phys. Chem., 35, 1933 (1931). Brame, in Redwood’s “Treatise on Petroleum,” 4th ed., Vol. 1, p. 279, Philadelphia, J. B. Lippincott Co., 1922. Bushong and Knight, J. IND. ENQ. CHGM.,12, 1197 (1920). Cragoe, Bur. Standards, Miscellaneous Pub. 97 (1929). Cragoe, International Critical Tables, Vol. 11,p. 151, New York, McGraw-Hill Book Co., 1927. Dana, Jenkins, and Burdick, Refrigerating Eng., 12, 387 (1928). Davis, Phil. Mag., 47, 1057 (1924). Eckart, Mech. Eng., 47, 535 (1925). Edgar, Cialingaert, and Marker, J . Am. Chem. Soc., 51, 1488, 1540 (1929). Ferry and Thomas, J . P h y s . Chem., 37, 253 (1933). Fitssimrnons and Bahlke, Oil Gas J . , 28, 154 (1929). Fortsch and Whitman, IND. ENG.CHEM.,18, 795 (1926). Gary, Rubin, and Ward, Ibid., 25, 178 (1933). Graefe, Petroleum Z., 2, 521 (1907). Heinlein, Motorwagen, p. 75 (Feb. 10, 1928). EN. C H m f . , Anal. Ed., Henderson, Ferris, and McIllvain, IND. 1, 148 (1929). Huffman, Parks, and Barmore, J . Am. Chem. SOC., 53, 3876 (1931). Huffman, Parks, and Daniels, I b i d . , 52, 1547 (1930). Huffman, Parks, and Thomas, Ibid., 52, 3241 (1930). International Critical Tables, Kew York, hlcGraw-Hill Book Co., 1926. Karawajeff, Petroleum Z.. 9, 1114 (1914); abstracted in J. SOC. Chem. Ind., 33, 128 (1914). Kelley, J . Am. Chem. Soc., 51, 2738 (1929). Kraussold, Petroleum Z., 28, No. 3 (1932); abstracted in J. I n s t . Petroleum Tech., 18, 108d (1932). Kremann, Meingast, and Gugl, Jlonatsh., 35, 1235 (1914). Kuklin, Ber. deut keram. Ges., 16, 949 (1883). Lang and Jessel, J.Inst. Petroleum Tech., 16, 476 (1930). Ibid., 16, 783 (1930-31). Ibid., 17, 572 (1931). Ibid., 18, 850 (1932). Leslie, Ibid., 13, 549 (1927). Leslie and Geniesse, IND.ENG.CHEM.,16, 582 (1924). Louginine, Ann. chim. phys., 13, 289 (1898). Msbery and Goldstein, Am. Chem. J.,28, 66 (1902). Mabery and Goldstein, Proc. Am. Acad. Arts Sei., 37,539 (1913). Marden and Dover, J . IND. E X + CHEM., . 9, 860 (1917).
vol. 27, KO.1
Pagliani, Atti accad. sci. Torino, 17, 97 (1881). Parks and Huffman, J. Am. Chem. Soc., 52,4381 (1930). Parks, Huffman, and Thomas, Ibid., 52, 1032 (1930). Redwood, B., “Treatise on Petroleum,” 4th ed., Vol. 1, p. 277, Philadelphia. J. E. Lippincott Co., 1922. (45) Redwood, I. I., ‘‘*MineralOils and Their By-Products,” p. 200 (1897). (46) Regnault, Ann. chim. phys., 73, 5 (1840). (47) Ibid., 9, 349 (1843). (48) Regnault, M e m . acad. sci., 26, 262 (1862). (49) Rey, Ann. mines, 8, 68 (1925). (50) Richards and Wallace, J . Am. Chem. Soc., 54, 2705 (1932). (51) Scheel, “Thermische Eingenschaften der Stoffe,” 1926. (52) Scheller and Georghui, Petroleum Z.,8, 533 (1913). (53) Schiff, Z . physik. Chemie, Stochiometrie und V e r w a n d t s c h f t s Zehre, 1, 378 (1887). (54) Schlamp, Ber. Oberhess. Ges. - V a t u w . Heilk., 31, 100 (1895). (55) Schlesinger, Phys. 2. vereinigt. m i t dem Jahrbuch der Radioaktivitat und Elektronik, 10,210 (1909). (56) Schmits, Mat. grasses, 58, 3005 (1913). (57) Sheppard, Henne, and Midgley, J. Am. Chem. Soc., 53, 1948 (1931). (58) Sherman, Dansiger, and Kohnstamm, J . Am. Chem. Soc., 24, 269 (1902). (59) Siivola, Oversikt Finska Vetenskaps-Soc. Fdrh., 56, No. 8. 7 (1913-14). (60) Spaght, Thomas, and Parks, J. P h y s . Chem., 36, 882 (1932). (61) Sullivan, McGill, and French, IND. ENG.CHEW,19, 1040 (1927). (62) Swann, Fuel, 11, 113 (1932). (63) Syniewski, Z. angew. Chem., 11, 621 (1898). (64) Texas Co.. unpublished compilation on “Physical Constants of Principal Hydrocarbons.” 1929. (65) Timofeev, Zswiestja Kiew polyt. Inst., 1 (1905): Diss., Kiew, 1905. (66) Trehin, Ann. phys.. [91 15, 246 (1926). (67) Vogel, Z . physik. Chemie, Stochiometrie und Verwandschftslehre, 73, 429 (1910). (68) Ton Reis, Ann. Physik u n d Chemie, 13, 447 (1881). Eso. CHEM.,6, 727 (1914). (69) Wales, J. IND. (70) Watson and X’elson, Ibid., 25, 880 (1933). (71) Weir and Eaton, Ibid., 24, 211 (1932). (72) Williams and Daniels, J . Am. Chem. Soc., 46, 903, 1569 (1924) (73) Wilson and Bahlke, IND. ENG.CHEM.,16, 115 (1924). (74) Zeitfuchs, Ibid., 18, 79 (1926). (41) (42) (43) (44)
I
R ~ C E I V EAugust D 27, 1934. Presented before the Division of Petroleum Chemistry at the 88th .Meeting of the American Chemical Society, Cleveland, Ohio, September 10 t o 14, 1934.
Lubrication as Affected by Physical Properties of Lubricants ROBERTC. WILLIAMS,The Ironsides Company, Columbus, Ohio
I
N h STUDY of lubrication as related t o the process of
drawing wire. the utility of various solid lubricants mas investigated. The present report deals with a certain class of solid lubricants, which, when subjected to a shearing action between surfaces under considerable pressure, reduce friction to a remarkable extent. The results presented here are somewhat related to earlier work by Hardy and Doubleday ( 1 ) . They found the coefficient of static friction to decrease to zero a t the melting point of solid lubricants and believe that a film of rather appreciable and critical thickness was responsible for this action. The finding, in itself, was not of great interest to Hardy, and he dismissed this acpect of his work
EXPERIMENTAL PROCEDURE The apparatus used in this investigation was identical with that described previously (4)and consisted briefly in an experimental set-up for drawing wire where the total pull on the die was measured. The total pull equals the pull
required to produce deformation of the wire during the reduction in diameter plus the pull required to overcome friction. The die pull n-as practicalIy independent of the speed of drawing within limits as reported previously (4),and these experiments were carried out with a drawing speed of 30 cm. (11.8 inches) per minute. In order to simplify the calculations of the coefficient of friction and the pull required t o overcome friction, no back pull was applied to the wire as in the earlier work. Care was taken, however, t o provide axial extrance of the die by the wire. The lubricants were applied to the wire (previously cleaned with acetone) by gently rubbing it with a piece of the lubricant or by swabbing it v-ith the liquid or molten material. Wire which was properly cleaned could not be drawn without seizure through a clean die. The results were reproducible within * 5 per cent. In a study of lubricants and friction, one ordinarily presents the results in terms of the coefficients of friction. Lewis ( 2 ) , in a series of articles on the wire-drawing die, develops a
January, 1935
INDUSTRIAL AND
ENGINEERING CHEMISTRY
plausible equation relating the total die pull to the pull required t o reduce the mire and t o overcome friction:
1' [l -I- @Otan o/2) (cf.)l mean yield point of the wire and is considered to be the ultimate Of the wire before drawing A - a = difference in cross-sectiona~area before and after drawing cotan el2 cotangent of the "half-angle" of the die (the Of the throat Of the die is bisected to obtain the half-angle) cf. = coefficient of friction Die Pull = (m. J'. p.1 ( A -
where m. y. p.
=
=i
The product of the first two numbers of the equation represents the component of the pull which does the useful work of deforming the metal. The final member represents the frictional component of the pull, and, when the die has a cylindrical bearing, Lewis has devised a modification of the cotangent value to include friction in the cylindrical bearing. ON DRAWING WIRE TABLEI. RESULTS
LUBRICANTS
THROUGH AN
These data have been selected from a large number of experiments involving waxes and waxlike substances. Beeswax was found to be the most effective lubricant under the experimental conditions outlined above, Copper oleate, ceresin wax, and various blended waxes were in the same range of effectiveness as paraffin and beeswax. When the temperature of the die was artificially raised above the melting point of any wax in question, the coefficient of friction rose to values closely corresponding t o those for mineral oils or other liquid lubricants a t the same temperature. It is apparent that the lubricant must enter the region of great pressure in the die as a solid rather than as a liquid, as in the case of the heated die, if enhanced lubrication is t o be effected. The waxes of higher melting point became effective a t temperatures approaching their melting points; carnauba wax, for example, a t a die temperature of about 70" C. (158" F.) was comparable in effectiveness with beeswax under ordinary conditions. Halowax, a chlorinated naph-
0.0225-INCH DIAMETER TUNGSTEN CARBIDE DIE AT 30 INCHES) PER MINUTE
0.0253-INCHSOFTCOPPERWIRE ReCoefficient work lost quired of due to pull friction friction Pozlnan
.-
o/,
0.21 57.3 8.5 Mineral oil (Saybolt viscosity, 100 seconds a t 100' F.) 0.28 63.7 10.0 Mineral oil (Saybolt viscosity, 160 seconds a t 210' F.) 0.23 59.7 9 0 Oleic acid, U. S. P. 48.1 0.15 Stearic acid,a c. P. 7 0 37.4 0.095 Paraffin wax (m. p.. 57-58' C. or 134.6-136.4' F.) 5 s 35.2 0.086 Beeswax (m. p., 60.5-62° C. or 140.9-143.6° F.) 5 6 47.4 0.14 Halowax 10145 (m. p,, 133-139O C . or 271.4-282.2° F.) 6.9 Aroclor No. 1254b (Saybolt viscosity. 46-47.7 aeconds 63.7 10.0 0.28 a t 210' F.) 0.49 76.6 14.9 Petroleum asphalt (m. p . . 77' C . or 170.6' F.) 0.58 78.6 17 None (wire may break) a Lubricants marked a were a p lied when molten and were allowed to cool before drawing. b A chlorinated diphenyl, liquif a t room temperature; the Halowax is a waxlike chlorinated naphthalene
In these experiments the die used had no cylindrical bearing of any consequence. The die angle of 18" (the half-angle is therefore 9") tapered gradually through the desired gage into a relief angle a t the end. The cotangent used in these calculations is therefore that for a 9' angle (6.314). Soft copper wire and 18-8 stainless steel wire were drawn. The tensile strength value used form. y. p. of the soft copper wire used here (0.0253 inch in diameter), as determined experimentally, was 17.4 pounds. Calculated to the basis of one square inch, the value is 34,600 pounds which compares favorably with the value of 35,000 pounds * 5,000 for annealed copper as given in the National Metals Handbook ( 3 ) . The reduction of the copper wire was to a diameter of 0.0225 inch, so that the value for ( A - a) was 0.000105 square inch. The tensile strength of the 18-8 wire of 0.0250 inch diameter mas 43.75 pounds. This corresponds to a value of 89,200 pounds per square inch. This result indicated that this wire was not completely annealed, as the handbook ( 3 ) records a range of 70,000 to 80,000 pounds per square inch as the tensile strength for annealed 18-8 wire. The reduction of the 18-8 wire was also to 0.0225 inch in diameter so that (A - a ) equals 0.0000925 square inch. The experimentally determined values for tensile strength were used in the equation for m. y. p. The deformational work in foot-pounds required for drawing one pound of wire may be calculated by multiplying the product of (m. y. p.) and ( A -a) by the number of feet of wire in one pound. The fraction of the work lost due t o friction is found by dividing the difference between the observed die pull and the calculated deformational pull by the observed die pull. The results are presented in Table I. Data for some conventional lubricants are included for comparative purposes.
65
CENTIMETERS
(11.8
0.0250-INCHANNEALED1s-8 S T A I N L E S ~
Re-
wired Pull
STEELWIRE Coefficient Of
friction
Pounds
Work lost due to friction 9% .59.5 64.6 68.0 39.8 46.4 30.7 53.9
20.4 23.3 25.8 13.7 15.4 11.9 17.9
0.23 0.29 0.34 0.10 0.14 0.07 0.19
22.9
..
0.28
64.0
40
o:i1
79:4
thalene, being tough instead of exceedingly brittle at room temperature, was quite effective under ordinary conditions. It appears that the lubricant must be transformed into a fairly mobile liquid under the conditions of temperature and pressure a t the mire-die interface. Asphalt, as shown in Table I, was rather poor as a lubricant, and it is not very mobile somewhat above its flowing point. A very uniform, visible film of asphalt covered the wire as it came out of the die. Several low-temperature (60" to 100" C. or 140" t>o 212" F.) thermoplastic resins were tested, but the wire could not be drawn because of breakage unless the die or wire as heated. Good adhesion of the wax to the wire or die is apparently of great importance, for, when beeswax was applied to a wire that was previously wetted with a mineral oil, the coefficient of friction obtained was practically the same as that for the oil. On the other hand, when mineral oil was applied t o the waxed wire, there was little effect on the coefficient of friction if the test was carried out immediately. Strong adhesion apparently permits a film of wax of relatively great thickness t o remain on the wire until the region of great pressure is reached in the die where the wax melts and, being prevented from exuding a t the entrance of the die by the solid wax, acts mainly as a "fluid film" lubricant.
COSCLUSIOKS These considerations lead to a t least three conditions which must apparently be satisfied t o observe this enhanced lubrication in wire drawing: (1) The lubricant must adhere strongly to a t least one of the surfaces-the wire or the die; (2) the lubricant must be solid prior to being subjected to the relative shearing of the surfaces to be lubricated; and (3) the lubricant must melt or be transformed to a fairly mobile liquid between the surfaces while shearing takes place. The action appears t o be analogous to that occurring when
INDUSTRIAL AND ENGINEERING CHEMISTRY
66
a sled or skate slides on ice. Polar explorers have observed that sledges are pulled with considerable difficulty during the coldest polar weather, because the frictional heat and pressure are insufficient to melt the ice under the runner. An observable greater die pull is registered when the w i r e drawing machine is started after it has been idle for some time when using a wax lubricant. The coefficient of friction with a solid lubricant that meets the conditions outlined above depends largely, it is believed, ilm of lubricant rather on the viscosity and thickness of the f than on the chemical nature of the lubricant or surfaces. Points of asperity a t which conditions of boundary lubrication do exist probably account for not obtaining lower frictional coefficients. The lack of exact parallelism between the re-
Vol. 27, No. 1
sults for copper and stainless steel is believed to be due mainly to differences in adhesion of the lubricant for the metal, to the influence of the temperature attained a t the interface, and to specific effects a t points of asperity. LITERATURE CITED (1) Hardy, W. B., and Doubleday, I., Proc. ROY.SOC.(London) 101A, 487 (1922). (2) Lewis, K. B., Wire & Wire Products,8, 197, 234, 266, 331 (1933). (3) National Metals Handbook, 1933. (4) Williams, R. C., J. Phys. Chem., 36,3108 (1932). RECEIVED September 21, 1934. Presented before the Division of Colloid Chemistry at the 88th Meeting of the American Chemical Society, Cleveland, Ohio, September 10 t o 14, 1934.
Glycol-Water Mixtures Vapor Pressure-Boiling Point-Composition Relations H. M. TRIMBLE AND WALTER POTTS, Oklahoma Agricultural and Mechanical College, Stillwater, Okla.
T
HE vapor p r e s s u r e s of The pressure-boiling point-composition relarefractometer which was calitions f o r glycol-water have been inbrated with solutions of glycol the components of glycolwater solutions are of imand water of accurately known vestigated from 0 to 100 per cent glycol and u p to portance, since these solutions composition by weight. The find m a n y applications. The aPProximateb' aimosPheric Pressure* The boilrefractometer was held a t 25" ing point-pressure relations f o r t w i o u s comC . by running water t h r o u g h only i n f o r m a t i o n along this positions are given in lhe f o r m of the Young its chambers from a large therline in the literature seems to be equation, the constants varying wilh the cornmostat* in the form Of a curve g i v e n by Lawrie ( 3 ) . T h i s r e l a t e s Corrections to be applied to position. the thermometer had been dethe boiling Doints of the soh- . - These mixiures obey Raoult's law fairly closely. termined by comparing it with tions t o t h e p e r c e n t a g e by a s t a n d a r d thermometer calivolume of glycol in the liquid and vapor phases, presumably for atmoqpheric pressure. It was brated by the Bureau of Standards. Corrections were made believed worth while to extend the data t o include the whole for the emergent stem. range of pressures from 0 to 1 atmosphere. Boiling in the still was promoted by introducing boiling tubes of glass. There was no bumping, even a t the lower APPARATUS AND METHOD pressures, and the readings of the thermometer showed no In this work at the higher temperatures, a form of the irregular variations. It is believed that superheating was small or absent. At the start of a run the temperature rose, Othmer still (4), modified as shown in Figure 1, was used: The still was made from a Kjeldahl flask, and heating was because of preferential loss of water from the boiling liquid. carried out by means of a Cenco electric heater with a 1.25-inch When distillate began to run back from the collecting cham(3.1Scm.) opening in the Transite top instead of by an immersed ber, the temperature fell slightly but soon became constant. coil. Tubes sealed into the bulb of the flask served, respectively,t o The heating was stopped when the temperature had remained hold the thermometer and to introduce liquid and remove samples. constant for 30 minutes to an hour, the still was cut off from The return tube from the condensate collecting chamber was made of 3-mm. tubing to avoid undue mixing of condensate the vacuum train by closing a stopcock; when pressure had been released, samples from the still and the collecting chamwith the solution in the still when boiling was stopped, pre ara tory to takin samples. The addition tube at the top o f the ber were taken and cooled if necessary, and their refractive column was, of course, eliminated. A few runs showed that there indices were determined. By reference to a curve the peris much refluxing in this column, even when it is well lagged with asbestos. To prevent this, a heating coil supplied from the centages of glycol by weight were read. laboratory current was incorporated in the lagging. A thermoVapor pressures a t the lower temperature (25" C . ) for couple with one junction buried in the lagging served to measure glycol-water mixture were studied by one of the authors the temperature. The column was at all times maintained about (Potts). The Walker method as modified by Pearce and 30' C. above the temperature of the boiling liquid. Thus it was assured that all the vapor evolved in boiling passed over to the Snow (5) was used. The current passed to generate the condenser. The thermometer slipped snugly into the long tube electrolytic gas was measured by means of a gas coulomb which held it and was fastened in place by a piece of rubber tubing meter, using alkali of the same concentration as that in the which was wired t o thermometer and tube. Other openings were closed by means of ground-glass sto pers. Thus the liquid electrolyzers, and placed in series with them. Nickel elecand vapors did not come into contact witg anything except glass. trodes were used in both, and efforts were made to maintain As an added precaution against loss of vapors, a condenser was the same conditions in the electrolyzers and in the coulomb sealed to the tube leading to vacuum. The bulb of the ther- meter throughout each experiment. The current passed was mometer dipped directly into the boiling liquid. measured by loss in weight of the coulomb meter. A number The barostat and accompanying apparatus used are de- of tests showed that the current passed, as thus measured, scribed by Daniels, Mathews, and Williams (1). Evacuation was the same as that measured by a copper coulomb meter in series. was accomplished by means of an oil pump. In determining the quantities of glycol and water passing Analyses of the solutions were made by means of an Abbe ~
~
