INDUSTRIAL AND ENGINEERING CHEMISTRY
364 T4BIE
11.
SMOOTHED T - k L l E S O F COAIPRESSIBI1,ITY
graphically for pure nitrogen as \vel1 as values obtained by interpolation from Sage and Lacey's tables (15). Eyperimental values for mixtures of 44.4, 51.6, and 78.9% propane are also given. At 300.00' F. esperimental measurements arc presented for 31.9, 71.3, and 83.4% propane. For convenience, Tahlr I1 is included with smoothed values.
FACTOR^
Pres~.
sure,
":;Ahs. ?. 250 ,500 750 1000 1500 2000 2300 3000 4000 5000 6000
% Propane a t 259.080 F.
% Propane a t 300.00° F.
44.4
51.6
78.9
31.9
71.2
83.4
0,981 0.962 0.946 0.932 0.907 0.900 0.918 0.951 1.022
0.976 0.954 0.930 0.903 0.870 0,864 0.883 0.916
...
...
0.987 0.975 0.967 0,960 0.953 0.955 0.963 0.980 1.034 1 101 1 167
0 . 954 0.911 0.863 0.817 0.752 0.732 0.747 0.788 0,897 1,000
0.936 0.862 0.785 0.719 0.628
0.991
0,926 0,897 0.822 0.760 0,682 0.681 0.748 0.823 0.973
1
.
.
...
...
...
1,114
Vol. 46, No. 2
ACKIVOW LEDGM EhT
The gas samples used in the 259.08" F. isotherm were analyded by F. A. Tatum of the Electrical Engineering Department of the college. Those used in the 300.00° F. isotherm were anal\zed by E, A. Hinkle, Monsanto Chemical Co. This assistance is gratefully acknowledged.
0.615
0.687 0.716 0.838 0.962 1.083
LITER4TURE CITED
fox analysis were taken after the first and last elrpansions. The composition %asconsistent within the error of the analytical procedure. The temperature was contiolled within 0.03" F. and was measured with a platinum resistance thermometer, nhich had been calibrated by the Sational Bureau of Standards, in conjunction iyith a Mueller bridge. The volume was measured indirectly by determination of the volume iatio of the expansion chambers, which had been determined previously. Because of the nature of the calculation of compressibility factors by the Burnett method, requiring two extrapolations to zero pressure, the over-all results are believed to be accurate to 3% for the 259.08' F. isotherm and 2% for the 300.0O0F isotherm. I t is hoped that they are more piecise, but because of the practical impossibility of reproducing gaseous mixture compo.;itions, the results could not be expeiimentally rechecked. For pure nitrogen the experimental results vxre compared with interpolated values from the tabulation by Sage and Lacey (15);the agreement was from within 0.1% for the lower pressures to 1% at the higher pressures. Cross plots indicated that the data are appai ently internally consistent to n ithin i 1 %.
(1) Bartlett, E. P., Cupples. H. I,.. and Tremearne, T. H., . I . A m Chem. Soc., 50, 1275 (1928). (2) Beattie, J. A , Ibid., 51, 19 (1929). (3) Beattie, J. A.. Proc. Am. Acad. A r t s Sei., 69, 389 (1934). (4) Beattie, J. A , Kay, W.C., and Kaminsky, 3 . . ,J, Am. C'hein. Soc., 59, 1589 (19371.
(5) Benedict, Ill., Ibid., 59, 2224 (1937). (6) Bridgeman, 0. C., Ihid., 49, 1174 (1927). (7) Burnett, E. S., Appl. Mechanics, 3 (a), 137 (1936). (8) Deming, W.I?., and Deniing, I,.S.,Pkys. Rea., 45, 109 (1934). (9) Hagenbach, TV, P., and C'omings. E. W., 1x11.Esc. CHEJI.,45. 606 (1953). (10) Keyes, F. G., P m .Am. A c d . A ~ t Sei., s 68, 505 (1933). (11) lieyes, F. G., and Tiuiks, 13. Q., %J. ,4m.Chem. Soc., 50, 1100 (1928). (12) Reamer, H. H., Sage, E. XI.. and Laccy, 15'. S . , 1x11. I h o . CHEZI.,41, 482 (1949). (13) Reamer, H. I%.,Selleck, F. T., Rage, B. FI., and Lacey, I T r , S . . Ibid., 44, 19s (1952). (14) Sage, B. H., Ibid.. 42, 631 (1950). (15) Sage, B. H., and I,ncey, W.E., "3Ionograph on X.P.l. Research Projert 3 7 . 'I'heniiodynaniic Properties of the Lighter I-Iyrlrocarbons and Xitrogrn," 1st ed., Ken, York, American Petroleum Institute, 1950. (IC;) Page, B. H., Schaafsnia, .J. G., and h c e y , 13'. X . , 1x1).F k c + . CHEX, 26, 1218 (1934). (17) Schneider, W. G., Can. J . Reamrcli, 27B,363 (1949). (18) Yelleck, F. T., Carmichael. L.'I?., and Sage, 13. H., lsn. I h o . CHEM..44, 2219 (1952).
EXPERI>lENTAI, DATA
The experimental data are presented in Figures 1 and 2 by plots of the compressibility factor of the mixture against pressure at different compositions. T h e data are also given in Table I. For the 259.08' F. isotherm experimental values are presented
RECEIVED for review ilpril 28, 1953. ACCEPTED September 2 5 , 1953. Presented in part before t h e Seventh Southrvest Regional Meeting, AicexrC A S CHEXICAL BOCIETU, Ailstin, Tex., December 1951.
ielectric onstants ethylsilicones J
8.B. YOUNG AND CR4RLES E. DICKERJIAN Southern Illinois University, Curbondale, I l l . LTHOUGH the polysiloxanes constitute a relatively ncw family of synthetic compounds, research into the chemical characteristics of thwe polymers has already produced a considerable body of published literature. Liquid dimethylpolysiloxanes have received much of the a b tention directed toward the polysiloxane group. These materials are distinguished by quadrivalent silicon atoms, each of which is linked to two methyl groups and two oxygen atoms. The oxygen-silicon linkages form a chain molecule analogous to the molecule formed by the carbon-carbon linkages of the hydrocarbon oils. This investigation was undertaken as part of a joint research agreement between the Physics Department of Southern Illinois
University and the Sanganio Electric Go., Capacitor Divieion, a t Ordill, Ill. Its aim was to obtain information conccrnirig the physical properties of dimethylsilicone oils which would be of use in determining the advisability of utilizing these silicones as impregnating compounds for oil-impregnated capacitors Oils of three viscosities, 20, 50, and 350 cs., were selected because they offer more promise than oils of other viscosities for use as impregnating compounds. PROCEDURE
The silicone oils used are manufactured by the Dow Corning Corp.. Midland, Mich., and are marketed as now Corning (or
February 1954
INDUSTRIAL AND ENGINEERING CHEMISTRY
365
values for the dielectric constants over the range 145" to 160" C. and -73" to 25' C., hornever, represent, as far as available sources are concerned, new information. Two sets of data were taken for each oil, one over the temperature range above room temperature, and one over the lower tclmperature range. A small overlapping of the two sets of data was obtained by preheating the test cell before i t was placed in the cold chamber. The agreement between the two sets is good for 50- and 350-cs. silicone. However, the disagreement between the two sets for 20-cs. oil is only a difference of about l yo for the two points most widely separated-a difference well within the maximum possible difference of 2% (Figure 2)
00
ma TEMPERATURE, O C .
IW
Figure 1. Variation of Dielectric Constant with Temperature for 50-Cs. Dimethylsilicone 0 Experimental
0 Baker,
Barry, and Hunter (2)
1X) 200 fluids. A11 measurements were made a t 1000 cycles per second, because information released by the manufacturer indirated that there was no appreciable change in the dielectric constant with variations in frequency below frequencies of the order of lo8 cycles per second, frequencies beyond the range of most high frequency laboratory equipment and far outside t'he range of standard equipment (8). The temperature range investigated extended from -73" to 160" C. The apparatus consisted of a General Radio Type 716-C capacitance bridge, with a General Radio Type 722-M precision capacitor, a General Radio Type 1302-A oscillator as source, and a General Radio Type 1231-B amplifier and null-detector. Measurements below room temperature were taken with dry ice and a dry ice-acetone mixture as cooling agents. A capacitance test ccll, constructed according to standards set up by the American Society for Testing Materials, v a s used for the capacitance measurements (1). The substitution method was used with the bridge. The bridge is first balanced without the unknown capacitor. Then the unknown is connected in parallel with the balancing capacitor of the bridge, and the bridge is balanced again. The capacity of the unknown is given by the difference between the first and second settings of the balancing capacitor of the bridge. The maximum possible error of the bridge when the mbstitution method is used is 2 micromicrofarads. However, with the use of B correction chart, this may be reduced to 0.5 micromicrofarad, a maximum error of 0.8%. If a conservative estimate of the maximum error introduced by a temperature measurement error is added to this, the total maximum possible error is 1%. I n determining the dielectric constant of the oil, the capacitance of the test cell, filled with oil, is divided by the capacit>ariceof the cell, with air as the dielectric. (The capacitance in air may be taken as the capacitance in vacuum, because the dielectric constant of air is 1.0006 under standard conditions of temperature and pressure.) If there were an error in the determination of the value of the capacitance of the test cell in air, this error would show u p in the experimental curves as a constant vertical displacement.
TEMPERATURE,
OC.
Figure 2. Variation of Dielectric Constant with Temperature for Dimethylsilicone Oils 0 20 cs.
$
50 CY. 350 CY.
The curves show a smooth variation over the entire temperature range studied. This is an interesting result, because most polar capacitor impregnants, such as chlorinated biphenyl and the chlorinated polgphenyls, which are n-idely used as impregnating compounds for oil-impregnated capacitors, show a marked change in dielectric constant in the regions of their "freezing" temperatures (4). The freezing or solidification point of dimethylsilicone oil varies with the viscosity, but for the oils investigated it lies xithin the -40' to -60" C. temperature range. I n conclusion, from the point of variation of dielectric constant with temperature, the silicones seem promising as impregnating compounds for oil-impregnated capacitors for use in installations subjected to >Tide ranges in temperature. Their superiority to mineral oils has not been established. ACKNOWLEDGMENT
The authors are grateful for assistance given by Kenneth McGee, chief engineer, Capacitor Division, Sangamo Electric Co., and his associates. LITERATURE CITED
(1) Am. SOC.Testing Materials, Standards, 1946, Pt. IIIB, Appendix
111,pp. 678-9.
DIELECTRlC CONSTANTS
The experimental values for room temperature agreed well with the published values, in fact, well within 1%. There was a small amount of data available over the temperature range 25' to 145" C. These data also checked well with the experimental results ( 2 ) . This check was made on 50-cs. oil (Figure 1). T h e
(2) Baker, E. B., Barry, A. J., and Hunter, M.J., IND. ENQ.CHEM., 38,1117 (1946). (3) Dow Corning Corp., Midland, Mich., "Dow Corning Silicone Notebook," Fluid. Ser. 3, 19 (1948). (4) Monsanto Chemical Co., 8t. Louis, Mo., "Aroclors," Application D a t n Bull. P-115,p. 8. RLCEIYEDfor review January 22, 1953.
ACCEPTED October 22, 1953.
.
