INDUSTRIAL A N D ENGINEERING CHEMISTRY
2540
Vol. 43, No. 1 1 ACKNOWLEDGMENT
T A B LX. ~ MECHANICAL PROPERTIES OF SOMECHLORINATED RUBBERCOMPOSITIONS Chlorinated rubber, parts C1 content, yo M 0.parts DifJutylphthalate. parts Hycar OR-15 parts Molding tern;. O C. Tensile strength, k ./sq. om. Elongation at breaf Impact strength. cd.-?g.,sq. Flexural strength kg./sq. om Bending angle, ddgrees
... I . .
6
...
. ..
133
LITERATURE CITED
and 30 centipoises in a solution of 20% by weight in toluene a t 25” C., which is sufficiently low for use as a component for anticorrosive paint. Chlorinated rubber of lower chlorine content, as obtained by the direct onestage chlorination of latex, is generally insufficiently soluble in the conventional paint vehicles because of insufficient breakdown of the molecules or a slight amount of cross linking, On the other hand thie product can be obtained as a stable latex, which can be applied aa such-for instance, after adding a suitable amount of plasticizer and stabilizer, films can be cast directly from this latex. In this case it is an advantage that the original high molecular weight is conserved. The dry chlorinated rubber of lower chlorine content can be used as a thermoplastic molding material, in which case it is an advantage to add a stabilizere.g., magnesium oxide. If mixed with a rubberlike butadieneacrylonitrile copolymer, l i e Hycar OR-15, on the mill or in the form of latex, leathery polyblends are obtained. Table X lists some of the mechanical properties of various chlorinated rubber compositions. It is advisable to shorten the mixing period on the mill to the utmost, as the mechanical properties suffer from long milling under which process the molecules of chlorinated rubber break down. Chlorinated rubbers have sufficient chemical stability to withstand molding a t 110’ to 130’ C.without much harm.
m ’
The author wishes to thank the Rubber-Stichting for permission to publish this paper and G. Salomon for his interest and frequent discussions.
(1)Amerongen, G.J. van, and Koningsberger, C., J . Polymer S c i . , 5, 663 (1950). (2)Amerongen. G. J. van, Koningsberger, C., and Salomon, G.. Ibid., 5,639 (1950). 1 Bacon, R. G. R., and Evans, W. J. R., U. S. Patent 2,379,409 (1945). Baker, H. C., Proc. Rubber Technol. Conf. (London),1938, 209. Bloomfield, G. F., and Farmer, E. H., J. SOC.Chem. Ind., 53, 43T (1934). Boyer, R. F., J . Phus. & Colloid Chem., 51, 80 (1947). Fox, V. W., Hendricks, J. G., and Ratti. H. J., IND. ENO.CHEM., 41, 1774 (1949). Hopkinson, E., U. S. Patent 1,491,265(1924). Jordan, H.F., Brass, P. D., and Roe, C. P., IND. ENG.CHEM., ANAL.ED.,9, 182 (1937). Konrad, E., and Schwerdtel, F., Ger. Patent 616,364 (1935). McGavack, J., U. S. Patent 2,021,318(1935). Martin, G.,Davey, W. S., and Baker, H. C.. Brit. Patent 476,743 (1937). Morris, J. C., J . Am. Chem. Soc., 68, 1692 (1936). Nielsen, A.,“Chlorkautschuk,” Leipzig, Hirzel Verlag, 1937. Raynolds, J. W.,U. S. Patent 2,339,945(1944). Schidrowitz, P., Bull. Rubber Grower’s ASBOC., 16, 668 (1934); Rubber Chsm. and Technol., 8, 613 (1935). SOC.chim. du Caoutchouc, French Patent 793,607(1936). Veersen, G. J. van, Proo. 2nd Rubber Technol. Conf, (London), 1948. 87. R E C E I ~ ESeptember D 14. 1950. Communication No. 131 from the RubberStiohting, Delft, The’ Netherlanb.
l’ransmissionof Mechanical 0
0
Vibrations through Rubbers ROSS E. MORRIS, ROBERT R. JAMES, AND HAROLD L. SNYDER Rubber Laboratory, Mare Island Naval Shipyard, Yallejo, Calif. T h e purpose of this investigation was to determine how simple compression mountings for machinery compare wi Lh regard to transmission of mechanical vibrations when the mountings are made from different rubbers. The only difference between the mountings was the rubber contained; all mountings had the same physical dimensions, the same static modulus, and essentially the same hardness. Natural rubber was found to transmit less vibrational force than any of the other rubbers in the range of frequencies from 25 to 2000 cycles per second; moreover, its transmissibility was practically constant over this range of frequencies. Neoprene Type FR transmitted a little more force than natural rubber. Its transmissibility also remained practically constaut over the frequency range except for a rise at low frequencies. The transmissibilities of all other synthetic rubbers tested were higher and increased with rising frequency over the entire range.
The information obtained is of considerable value in the design of rubber mountings for isolation of machinery vibrations in the range of audible frequencies.
ANY investigations of the dynamic properties of rubber have been reported. Outstanding contributions in thia field were made by Dillon, Prettyman, and Hall (2),Gehman, Woodford, and Stambaugh ( 4 ) , Mullins (7), Nolle (&IO), Stambaugh (II), and Witte, Mrowca, and Guth (12). None of the previous investigations, however, has been sufficiently comprehensive from the standpoints of the numbers of rubbers tested or the frequency range covered to serve as background for the development of rubber mountings for sound isolation. The present paper describes an investigation of the transmissibilities of natural rubber (Hevea) and various synthetic rubbers in vulcanized form for vibrations in the range of audible frequencies. These rubbers were corhpounded to have essentially the same
November 1951
I N D U S T R I A L A N D ENGINEERING CHEMISTRY
static modulus in compression and to have a hardness of a p proximately 50 Shore A. TESTING PROCEDURE
The testing procedure for determining transmissibility consisted basically of clamping a rubber specimen between two metal plates, causing one of the plates to vibrate through a small amplitude, and measuring the vibrational force transmitted through the rubber to the other plate. All vibration testing was done at room temperature, which ranged from 81' to 86' F. (27'to 30' C.).
2541
inch thick and 18/10 inch in diameter was inserted in the cavity in the end of each transducer. Intimate contact was assured by a film of grease. Each brass plate had a concentric cavity '/s( inch deep and a/, inch in diameter on its outer face for positioning the rubber specimen. The rubber specimen was inch thick and a/, inch in diameter before compression. The generating transducer W ~ caused R to vibrate in its longitudinal direction by applying alternating current a t 300 rootmean-square volts to the crystals. The current was obtained from a HewlettrPackard Model 205AG audio signal generator with a step-up transformer. The voltage t o the generating transducer was read from a Hewlott-Packard Model 400A vacuum tube voltmeter. The voltage produced in the detecting transducer by the vibrations was amplified by a Mama M-114B preamplifier and a Hewlett-Packard Model 450 amplifier, and indicated by a Hewlett-Packard Model 400A vacuum tube voltmeter in root-mean-aquare volts. As a separate experiment, the phase angle between the voltage in the generating transducer and the voltage in the detecting transducer was measured by connecting the Y axis of a DuMont Type 304H cathode-ray oscillograph t o the audio signal generator and the X axis t o the amplifier in the detector circuit (3). Measurements of transmission at various frequencies were commenced 1 hour after the specimen was compressed between the transducers.
Figure 1. Apparatus Used for Measuring Transmission of Vibration
A piezoelectric transducer was used to generate vibrations of any desired frequency in the audio range. An identical transducer was used as a detector. The rubber being tested was in the form of a right cylinder and was compressed axially to exactly 20% deflection between the opposing metal faces of the two tranaducers. The whole assembly was held together by a vise in the form of a heavy steel yoke. The threaded plunger of the viae, on which the detecting transducer was mounted, could be held at any desired position by means of a locknut. The assembly was isolated from outside vibrations by hanging it with rubber rings from a framework mounted on a bench. The essential features of this apparatus were suggested by A. W.Nolle of the University of Texas. Figure 1 is a photograph of the assembly and Figure 2 shows a dose-up of the rubber specimen clamped between the transducers. The transducers were actually accelerometers made by the Massa Laboratories, Inc., Hingham, Mass. (6). Each consisted of an assembly of several ammonium dihydrogen phosphate crystal plates mounted within a stainlem steel housing 1 inch in diameter and 1 7 / 1 ~inches long. The housing was closed with a metal cap which had on its outer face a concentric cavity 1 * / 1 ~ inch in diameter and S/U inch deep. In order t o position the rubber specimen between the transducers, a brass plate 6 / ~ (
Figure 2. Close-up of Rubber Specimen Clamped between Tramsducere
It waa found that measurements of transnlission could not be made a t higher frequencies than 2000 cycles per second becmse of mechanical resonances in the apparatus, which rail-ed great a p parent increases in transmission a t certain frequencies. These resonances were also observed when no rubber specimen was clamped between the transducers. A plot of the voltages a& the detecting transducer for the latter arrangement is given in Figure 3.
INDUSTRIAL AND ENGINEERING CHEMISTRY
2542
FREQUENCY
Figure 3.
OF VIBRATIONS, CYCLES P€R SECOND i l O t
Voltages a t Detecting Transducer
Rubber spaaimen not damped between transducers
The range of frequencies covered in this irivestigation was adequate for practical purposes; experience has shown that vibrations from ordinary machineiy usually have no components of significant intensity with frequencies higher than 2000 cycles per second.
0
10
20
30
40
50
80
- POUNDS Figure 4. Load-Deflection Curve for Hycar OR-25 Test Speeimen
Vol. 43, No. 11
3 X lo-einch when 300 volts were applied to the terminals, and was relatively constant over the frequency range used. This very small amplitude is of the same order of magnitude as the amplitude of vibration of machinery supported on rubber mountings at frequencies above 100 cycles per second, according to unpublished information obtained at the Industrial Laboratory, Mare Island Naval Shipyard. The phase angle between the voltage applied to the generating transducer and the voltage produced in the detecting transducer wm used to calculate the relative internal viscosities of the rubber specimens. It waa assumed that the voltage applied to the generating transducer was in phase with the amplitude of vibration of the rubber specimen, and that the voltage produced in the detecting crystal waa in phase with the force of vibration. In order for these assump tions to be valid, it was necessary that the transducers and associated voltage meaauring circuits should not themselves introduce hysteresis into the system. That they did not do so to any significant extent wm demonstrated by performing a transmission experiment with a solid brass cylinder 0.75 inch in diameter and 0.40 inch long, clamped between the transducers. The aame clamping force, 62.6pounds, was used as in the case of the rubber specimens. As the brass cylinder had much higher mechanical impedance than the transducers, no mechanical work waa done on it; it merely transferred the force created by the generating transducer to the detecting transducer. Hysteresis losses, therefore, could occur only in the transducers themselves or in the voltage-measuring circuits. The phase angle found wing the brass cylinder was 2" over most of the frequency range. Thib phase angle could not be used as a correction for phme angles found when testing rubber specimens because the conditions of measurement were not the same. In the experiment described, the generating transducer was operated at much less vibrational amplitude and the detecting transducer was operated at much greater amplitude than when a rubber specimen was being tested. However, the small phase angle found in the present experiment demonstrated the low level of power losses in the transducers and electrical circuita
LOAD
PREPARATION OF TEST SPECIMENS
The rubbers tested were Hcvea, Neoprene Type FR, Neoprene Type GN-A, Neoprene Type RT, Neoprene Type W, GRS, X-453, X-454, X-485, Paraciil 18-80, Paracril 26-90, Hycar OR-25, Hycar OR-15, GR-I 25, and Thiokol ST. The X rubbers, obtained from the Office of Rubber Reserve, consisted of a
The output of the detecting transducer when the generating transducer was not operating did not exceed 80 microvolts at any time. This output, which waa due to background noise, either electronic or airborne, had a negligible effect on the output of the detecting transducer when the ,050 generating transducer was operating, because it was far below the voltage created by thr vibrating OR-1 25 !2 rubber specimen. P ,040 The voltage produced in the d e h t i n g transducer by the vibraiional force passing through the test HYCAR OR-I5 specimen was used as the basis for comparing the 0' .030 rubbers. It waa assumed that the amplitude of c vibration of the generating ti ansducer remained conI 2 ,020 HYCAR OR-25 stant over the frequency Iange and was the same PARACRIL 2 6 - 9 6 rz for all rubbers. I t was also msunied that the deTIIOKOL S T tecting transducer produced B voltage proportional ;,010 PARACRIL RACRIL 18-80 to the vibrational force acting on it. These assump U tions were based on two characteristics of the trans,000 ducers: their large mechanical impedance compared 0 200 400 600 800 1000 I200 1400 1600 IS00 2ooo to rubber, and their outstanding efficiency for conFREQUENCY OF VIEHATIONS, CYCLES PER SECOND verting voltage to mechanicid motion and mechanical force to voltage. Figure 5 . Vibrational Forces Transmitted by Test Specimens a t According to Massa (6), the amplitude of vibraVarious Frequencies tion of the generating transducer was approximately In terms of volts at detecting transducers
3 e
November 1951
INDUSTRIAL AND E N G I N E E R I N G
TABLE I. CHARACTERISTICS OF X R W B E RAND ~ GR-S But adiene-Styrene Charge Ratio
x-453 x-454 X-485
100/0 90/10 71/29 71/29
OR-S
2543
mens of this stock, 0.5 inch thick and 0.75 inch in diameter, similarly conditioned, were inserted between platens having cavities 1/84 inch deep and 0.75 inch in diameter, and each specimen was compressed by a different dead weight load. The deflections of the specimens were noted after l hour a t 82" & 5' F. A plot of load versus deflection waB made and the load t o produce exactly 20% (0.1 inch) deflection was read from the curve and found to be 62.6 pounds. The othei rubbers were then compounded with various amounts of SRF black, and R ecimens 0.5 inch thick and 0.75 inch in diameter were prepared {om each stock. These s ecimens were conditioned a t 82 f 5' F. for at least 3 days a n i then subjected to a 62.6-pound load in the dead wei ht apparatus. The deflections of the specimens after 1 hour a tm !t temperature were noted and a plot was made of deflection versus SRF black content in the stock. The SRF black content which corresponded t o 20% deflection waa read from the curve, and was used t o prepare the a ecimens for the vihration transmission measurements. A cieck was made of the deflection of one of these specimens under 62.6 pounds load. If the deflection was not 20.0 & 0.5%, further variations in black loading were made until a deflection within this tolerance was obtained.
butadiene polymer and several butadiene-atyrene copolymers and had the characteristics given in Table I. GR-S, a butadienestyrene copolymer, is included in Table I for comparison. Paracril 18-80, Paracril 26-90, Hycar OR-25, and Hycar OR-15 are copolpmera of butadiene with acrylic nitrile and have increasing acrylic nitrile content in the order listed.
Rubber
CHEMISTRY
Polymerization Temperature, a F. 41 41 41 122
All the rubbers were compounded with semireinforcing furnace (SRF) black. The quantity of black added to each rubber wm adjusted so that its vulcanizate had exactly the same static modulus in compression as the other vulcaniaates. The procedure for establishing the content of semireinforcing furnace black was as follows:
The load-deflection characteristics of the vulcanieatep, which were selected from the foregoing experiments, were determined with the specimens and dead-weight apparatus previously described. Deflection readings were taken after various loads had been applied for 1 hour. It was found that the load-deflection curves of the specimens were not straight lines up to 20% deflection, but were slightly concave toward the load axis, m illustrated by the curve for the Hycar OR-15 specimen in Figure
The SRF black content in the Hevea stock was adjusted so that the vulcaniirate had a hardness of exactly 50 Shore A when tested on a specimen 0.5 inch thick and 2 inches in diameter which had been conditioned for a t least 3 days a t 82" rL: 5" F. Speci-
TABLE 11. RECIPES,CURES,AND PROPERTIES OF VULCANWATES TESTED Recipe No. Smoked sheet (Hevea) OR-S
E-21-1040 100
.... .... .... ....
x-453 x-454 x-486
OR-I 25 Thiokol ST Pelletex (SRF) Zinc oxide Stearic acid Thermotlex A Tetramethyl thiuram monosulfide Tetramethyl thiuram disulfide Poly80 EQuinone dioxime eneothiazyl sulfenamide Diphenyl guanidine Sulfur Volume loading of carbon black per 100 volumes of rubber Cure min s clfio mivity X m i l e stren th, lb./sq. inch. stress at 200% elongation, Ib./sq. inoh. Ultimate elongation, % Shore hardness
0.6 0.4 3
1.8
15.4 25, 300' F. 1.084 3700
31.8 20, 310' F. 1.192 1890
650 220 50
.... 100 ....
..*. ....
.
43'
' '
6
1
0.3
0.3
I
....
.... ....
....
.. .. .. .. .
00 5
5
.... .
....
....
1 0.5
.t..
.
.... ....
....
.... ....
2
....
33.6 20, 310' F. 1.180 1820
30.6 20. 310' F. 1.201 1980
31.9 20, 310° F. 1.194 2180
E-21-1047
.... .... 100 ....
....
.... ....
0
810
..*. .... 100 . t . .
.... .... .... .
.
I
.
30 5
25,30.8 310' F. 1.108
I660
E-21-1095
E-21-1077
.. .. ,. ..
. .. .. ..
.... ....
100
....
.... ....
17.5 10
.... 20 ...
.... .... .
.
I
100
....
... 20
E-21.1093
.... ....
.... ..... ...
....
.... 32
100 5
.... ....
0.4
....
0.4
....
0.4
1.3
1.2
..l.
.... ....
0.5
....
0.5
1.4
23.5 40, 310' F. 1.172
22.2 40, 310' F. 1.179 1420
10.7 40, 310° F. 1.154 1690
12.1 40, 310" F. 1.562 3350
13.8 20, 310' F. 1.359 2770
20, 22.4 310' F.
.
....
650 360 51
....
510 700 50
....
530 640 50
29.7 26, 290' F. 1.424 1120 430 50
....
.
.... ....
600
920 370 49
720 490 51
430 53 E-21-1044
....
....
810 300 51
.... ....
i:s'
....
700 320 50
0.6 3
....
.... ....
1400
....
100 40
..*.
730 390 52
I
.... ....
....
2
.
.... ....
..... ... 1.9
820 390 51 E-21-1080
....
60
E21-1092
1.5
..I.
070 500 50
.
.... .... .... .,..
.... .... .... ....
....
I
0.3
0.3
.... .... .... .... .... 100
.
....
....
....
.
100
70 5 1
5 1
....
.... .... ....
.... ....
Ib./sq. inoh.
5 1
I
. . I .
.... ....
05
E-21-1088
.... .
.,.. 100
.... .... .... ....
00
E-21-1083
.... ....
100
.... .... .... ....
....
.
.... ....
....
....
E-21-1084
E-21-1085
.... 100
30 5 1 $1
E-21-1094 Recipe No. Paracril la80 100 Paraoril26-90 .... Hyoar O R 2 5 .... Hycar O R 1 5 .... Neoprene GN-A Neoprene RT .... Neoprene W Neoprene FR Pelletex 40 Zinc oxide 5 XLC magnesia .... Litharge .... Stearic scitl 1 Phenyl-1-naphthylamine . ... Tetramethyl thiuram monosul5de 0.4 2-Meroaptoimidazoline .... Sulfur 1.5 Volume loading of carbon black per 100 volumes of rubber 21.3 Cure min 40, 310' F. spechi0 gravity 1.142 Tensile stren th, lb./sq, inoh. 790 Stress at 200% elongation,
Ultimate elongation, % Shore hardness
E-21-1082
2
....
E-21-1097
.... ....
.... .... .... .... ....
100 25
.... ....
10
....
....
1.385 3370 900 400 51
1
10.3 40, 310' F. 1.343 1770
600 -360 47
INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY
2544
P
,454 x453 GR-S
.015
e’ NEOPRENE W NEOPRENE RT NEOPRENE GN-A NEOPRENE FR
HEVEA
E
--0
Vol. 43, No. 11
RECIPES AND MISCELLANEOUS PROPERTIES
020
200
400
800
FREQUENCY
e00
1000
1200
1400
1800
1800
zoo
OF VIBRATIONS, CYCLES PER SECOND
Figure 6. Vibrational Forces Transmitted by Test Specimens at Various Frequencies In terms of volts at detecting transdueer
FREQUENCY
OF VIBRATIONS, CYCLES PER SECOND
Figure 7. Transmissibilities of Synthetic Rubber Specimen6 Relative to Hevea Specimen
The recipes for the stocks which were selected for the vibration transmission measurements ape given in Table 11. Included in the table are the cures, specific gravities, tensile properties, and hardnesses of the vulcanizates. The loadings of semireinforcing furnace black in volumes per 100 volumes of rubber ranged from 12.1 (Neoprene Type GN-A) to 36.6 (X-454 j. The vulcanizates prepared from Hevea and all the neoprenes except Neoprene Type FR had good tensile properties. The vulcanizates prepared from Neoprene Type FR, GR-I 25, and the butadiene polymer and copolymers except Parwril 18-80 had tensile properties which were lower than the aforesaid but were adequate for some mounting applications. The tensile properties of the Thiokol ST and Paracril 18-80 vulcanizates were 80 low that these vulcanizates were probably unsuitable for any mounting application. Manifestly, Borne of these rubbers would have to be compounded differently if they were to be used in practical applications, particularly if they were required to withstand shock loads or be serviceable at low temperatures. The Shore hardnesses of the vulcanizates ranged from 47 to 53, in spite of the fact that the stocks had been carefully compounded to have equal compression a t equal load. The explanation for the divergence in hardness valuee rests on the differences between the duration of the Durometer test and the duration of the loaddeflection test as well as on the inherent inaccuracies of the Durometer test. The deflections of the rubber specimens under the pressure of the Durometer indenter were not fully accomplished in 10 eeconds because the temporary creep had not run ita coulge in this interval, whereas the deflections of the rubber specimens under the 62.6-pound load were practically complete in 1 hour.
I n decibels at various frequencies RESULTS OF WBRATION TESTY
4. However, the secants to the curves between 0 and 20% deflection had the same slope, as all the specimens attained the same deflection (within tolerance) when subjected to a 62.6pound load. The static Young’s modulus calculated from the slope of the secant was 709 pounds per square inch. The Shore hardnesses of the vulcanizates used for the transmissibility tests were measured on specimens 0.5 inch thick and 2 inches in diameter with an A Durometer. The readings were taken 10 seconds after applying the Durometer to the specimens.
TABLE 111. FORCE IN VOLTSTRANSMITTED BY RUBBER S P E C l M E N S AT 25, 1000, AND 2000 C Y C L E S P E R SECOND Rubber Vulcanirate Hevea Neoprene F R Neoprene R T Neoprene QN-A h’eonrene W Paracril 18-80 x-485 Thiokol S T Paracril 26-90 GR-S Hycar OR-25 x-453 x-454 Hycar OR-16 OR-I 25
Force Transmitted, Volts 25 c.p.8. 1000 C.p.8. 2000 C.P.9. 0.00305 0.0035 0.003 15 0.0047 0.0046 0.0036 0.00635 0.0039 0.0057 0.0045 0,0058 0.0061 0.0052 0.0087 0.0079 0.0106 0.0067 0.0098 0,0084 0.0130 0.0118 0.0083 0.0136 0.0126 0,0075 0.0167 0.0132 0.0145 0.0134 0 0092 0.0188 0.0142 0.0071 0.0156 0.0146 0.0103 0.0166 0.0106 0.0154 0.0328 0.0217 0,00635 0.0445 0,0292 0.0069 I
The vibrational forces transmitted by the test Specimens, in terms of volts a t the detecting transducer, are plotted against frequency in Figure 5. As some of the curvea tended to lie over others, it was necessary to separate them in two graphs. The curve for Hevea appears in each graph for comparison. To give an idea of the relative ratings of the vulcanizates, they are listed in Table I11 in order of increasing transmission a t 1000 cycles per second. This table also gives the transmissions at 25 and 2000 cycles per second for comparison.
DYNAMIC MODULUSOF RUBBER TABLEIV. RELATIVE SPECIMENS AT 25, 1000, AND 2000 CYCLMPER SECOND Rubber Vulcaniaate Hevea Neoprene F R Neoprene R T Neoprene GN-A Neoprene W Paracril 18-80 x-485 Thiokol S T Paracril 26-90 GR-S Hycar OR-25 x-453 x-454 Hycar OR-15 GR-I 25
Relative Dynamic Modulus25 c.p.8. 1000 c.p.8. 2000 C.P.E. 1.00 1.15 1.03 1.51 1.18 1.54 2.08 1.28 1.87 2.00 1.48 1.90 1.70 2.59 2.85 2.20 3.47 3.21 4.26 2.75 3.87 4.46 2.72 4.13 6.47 2.46 4.33 4.76 4.39 3.02 6.16 4.65 2.33 6.11 4.79 3.38 5.44 3.47 LO5 10.8 2.08 7.11 14.6 2.26 9.57
November 1951
INDUSTRIAL AND ENGINEERING CHEMISTRY
2515
Figure 6 compares the Bynthetic rubbem to Hevea in regard to relative transmissibdity in decibels over the frequency range. Here, again, it 8 I I2 was necessary to use two graphs to prevent the 2 curves from lying over one another. The points on these curves were calculated from the following $ IO NEOPRENE W equation: Relative decibels transmissibility = Y 20 logtoVs/Va where Va = voltage produced, in detecting transNEOPRENE RT NEOPRENE ON-A ducer a t a even frequency by a vibrations awing through the synthetic rulber specimen NEOPRENE FR VR voltage produced in detecting transducer a t the same frequency by v i b r a t i o n s passing through the Hevea specimen The foregoing information on transmission of 0 ~ 4 o o M x , 8 o o a a , ~ l 4 o o l s o c c e o o 2 o o o vibrational force haa considerable practical interest because it s h o w the relative e5ciencies of the FREQUENO OF VIBRATIONS. CYCLES PER E C O W l rubbers for isolating machineryvibrations. Several Figure 8. TransmissibilitiesOf Synthetic Rubber SWdmene Relative however, should be kept in in this to Hevea Specimen regard. First, the relative efficienciea hold only In deaibels at TAU frequencies for the particular vulcanhates tested. If the rubbers were compounded differently, they would probably have different transmissibilities. Secondly, Hevea transmitted less force than any of the other vulcanizates the relative efficiencies were determined with the specimens over the entire frequency range; moreover, ita transmission waa in compreasion. Whether or not the same relations exist beessentially constant except for a small rise at low frequencies and tween the vulcsnizates when they are supported in shear is a m a l l drop a t high frequencies. Neoprene Type FR, a copolynot known, although it is believed that the same relations would mer of isoprene and chloroprene, transmitted a little more force hold. At audio-frequencies and normal temperature, the shear than Hevea, an isoprene polymer. The transmission of Neoprene modulus for rubber is one third of its Young's modulus (9). VELOCITY Third, the relative efficiencies do not hold near the frequency of resonance of the m m of the machine with the compliance (inverse of stiffness) of the mounting. Usually such resonant frequencies are below 20 cycles per second; thus they are below the range of frequencies studied here. DYNAMIC MODULUS. The vibrational forces transmitted by the vulcanizates in terms of volts can be regarded aa relative meaaures of the dynamic moduli of the vulcanizates. This follows from the considerations given below.
g
X-454
bl
'4
X-453 OR-S X-485
c 4
5
I
-
[Figure 9. Vector Representation of Vibration Phenomena
Type F R also remained practically constant over the frequency range except for a rise a t low frequencies. The transmission of
Dynamic modulus, E
MIA
AH/U
where aF = maximum value of sinusoidal force applied t o compreseed specimen to cause vibration the other vulcanizates increased with rising frequency over the A cross-sectional area of specimen entire range; the rate of increaae waa greater a t low frequencies aH amplitude of vibrbtion H height of compressed specimen in all cases. The greatest increases in transmission over the frequency range were obtained with Hycar OR16 and GR-I 25. Surprisingly, these vulcanbates did not have outstandingly high transmission a t 25 cycles per aecond, although their transmiasions of vibrational force at 2000 cycles per second were, respectively, 10.4 and G R - 1 25 14.1 times that of Hevea. HYCAR O R - U It is noteworthy that the neoprene8 had leas transmissibility than the butadiene polymer and copolymers. HYCAR O R - 2 5 Paracril 18-80, a butadiene-acrylic nitrile copolymer, PARACRIL 26-00 had the least transmissibility of the latter vulcankates. The transmissibilities of the four butadiene-acrylic THIOKOL ST PARACRIL IO-OO nitrile copolymers were in the same order aa their acrylic nitrile contents-i.e., Paracril 18-80 with the lowest acrylic nitrile content had the lowest transmissibility and waa followed in order by Paracril 26-90, Hycar HEVEA OR25, and Hycar OR15. Thiokol ST had more 2 3 4 5 s 6 1 0 14 20 K W transmissibility than any of these copolymers up to 200 FREWENCY OF VIBRATIONS, CVCLES PER SECOM i m o cycles per second; beyond 700 cycles per second this vulcanizate had less ,transmissibility than any of these Figure 10. Relative Internal Viscosities of Rubber Spacimw at Varioue Frequencies oopolymers except Paracrill8-SO.
-
. I
*
--
INDUSTRIAL AND ENGINEERING CHEMISTRY
quency range, whereas the dynamic moduli of the synthetic rubbers were w.much as 3.47 times greater at 25 cycles per second and 14.6 times greater at 2000 cyclea per second than the modulus of Hevea a t 25 cycles per second. The magnitudes of these differences are surprising, in view of the fact that all of the rubbers had essentially the same static modulus. INTERNAL VISCOSITY. It is customary to use the equation for sinusoidal motion of a perfect spring and dashpot in parallel to calculate the internal viscosity coefficient and real Young's modulus of rubber in vibration ( 1 ) . When the spring and dashpot are compressed by a static force, F, to a height, H , and then subjected to a vibratory motion, the equation has the following form:
4
3 2 ism* 6 5 4
3 GR -S x -485 x -454 x-453
2 1.10-6 8
5
Vol. 43, No. 11
HEVEA
2
3
4
5
6
OF VIBRATIONS,
FREQUENCY
8 1 0
15
20
where A = cross-sectional area of specimen
CYCLES PER SECOND f tW
Figure 11. Relative Internal Viscosities of Rubber Specimens at Various Frequencies
y
= coefficient of internal viscosity, referred to hereafter as internal viscosity Ah = displacement from static position a t any instant
t
5
time
E1 = real Young's modulus Af
=:
force causing displacement from static position at any instant
d(Ah)
A"t -2-
FR
OF
VIBRATIONS.
CYCLES PER SECOND
Y
pres8 &e resilient component
4Ah) is 90" out of phase with the The velocity term, --, dt displacement term, Ah; therefore the velocity term ie equal t o 2 ~ v ( A h ) where , Y is frequency in cycles per second. The force term for the viscous component is equal to the projection of the force vector, Af, on the velocity vector, as shown in Figure 7.
/OO
Figure 12. Relative Internal Viscosities of Rubber Specimens at Various Frequencies
Also Af
All specimens had the same shape and were compressed to the same extent; therefore they had the same values for A and H. As a constant voltage was applied to the generating transducer in all experiments, the specimens were vibrated through the same amplitude, AH. Thus, AF remained m the only variable, and
E = k( A F ) where k = a constant AF is the vibrational force applied to the specimen by the generating transducer. This force is equal to the force transmitted to the detecting transducer, inasmuch as the height of the specimen wag short compared to the wave length of sound in the rubber. (If this had not been the case, resonances due to standing waves in the specimen would have been found in the range 25 to 2000 cycles per second). The voltage, Y , produced by the detecting transducer is directly proportional to the vibrational force exerted on it: AF = k'V
where k, = a constant It follow8 that
B = kiV where tb2 = a constant Thus, the rubbers can be compared for dynamic modulus by taking the ratio of the respective voltagcs produced in the detecting transducer. The data in Table I11 were recalculated and are presented in Table IV on a relative basis with the force transmitted by Hevea at 25 cycles per second being made unity. The dynamic modulus of Hevea changed very little over the fre-
force required a t any instant to compresll the viscous component
A K ( A h ) = force re uirement a t any instant to com-
W RT GN-A
FREQUENCY
e
-
2 m A ( A h ) y = .If sin 9
AE( Ah)
where E = dynamic Young's modulus. y = - E sin 6 Therefore 2zv
I t was previously shown that the voltage, V , p r o d u d by the detecting transducer is proportional to the dynamic modulus of the specimen; consequently y
P
koV - sin 9 V
As the constant, ka, was the same for all the vulcaniiates, relative values for the internal viscosities of the vulcanizates can be calculated from the known quantiti-s: V , sin 4, and Y . The value I' sin for each vulcanirate is p l o t t d agaiwt frequency in Fiepre 8 using log scales. The curvea are separated in three graphs to avoid having the curves foi one c h of rubber lying over the curves for another class of rubber.
'
The absolute values in Figure 8 have no signficance, but the relative values and shapes of the curves present a valid comparison of the internal viscosities of the respective vulcanizates and the effect of frequency on these viscositiee. The most noteworthy features of these plotted data are the differences between the relative internal viscosities of the various vulcanizates and the downward trend of the relative internal viscosities with increasing frequency. At any frequency in the range studied, the vulcanizates ranked in about the same order as for relative dynamic modulus in Table IV. This was not unexpected, as all the vulcaniaates had essentially the same static
November 1951
INDUSTRIAL AND ENGINEERING CHEMISTRY
modulus, and therefore ditrerences in dynamic modulus must have been due to differences in internal viscosity. Hevea and Neoprene Type FR had the lowest, and GR-I 26 and Hycar OR16 had the highest internal viscosities. The downward trend of internal viscosity with inoreasing frequency has been observed by many investigators. This behavior, of course, is a major point of difference between the coe5cients of internal viscosity of rubbers and the coefficients of viscosity of ordinary liquids. The latter coefficients are not d e c t e d sensibly by frequency (velocity). LITERATURE CITED (1)
Ballou, J. W., and Smith, J. C., J. Applied Phys., 20, 493 (1949).
(2)
Dillon, J. H., Prettyman, I. B., and Hall, G. L., Zbid., 15, 309 (1944); Rubber Chem. and TechnoE., 17, 597 (1944).
2s47
(3) DuMont Laboratories, Inc., Allen B., “Manual for Cathode-Ray (4)
Osaillograph.” Gehman, 5. D., Woodford, D. E., and Stambaugh. R. B., IND. ENQ.CHEM.,33, 1032 (1941); Rubber C h . and Technol., 14. 842 (1941).
Massa, F., Instruments, 21, 1012 (1948). (6) Maasa, F., privata communications. ( 7 ) Mullins, L., Trans. Inst. Rubber Znd., 26, 27 (1950). (8) Nolle, A. W., J. A m s t . Soc. A n . , 19, 194 (1947). (9) Nolle, A. W., J. Applied Phgs., 19, 753 (1948). (10) Nolle, A. W., J. Polymer Sci., 5, 1 (1950). ENQ.CEEM.,34, 1358 (1842); Rubber (11) Stambaugh, R. B., IND. (5)
C h a . and Technol., 16, 400 (1943). (12)
Witte, R. S., Mrowca, B. A., and Guth, E., J. Applded Phys.. 20, 481 (1949); Rubber C h a . and Technol., 23, 183 (1950).
RECEIVED Marah 2. 1951. Presented before the Division of Rubber Chemistry, AMERICANCHEMICAL SOCIETY, Washington, D. C.
Quantitative Separation of West Virginia Petroleum into Several Hundred Fractions ISOLATION AND PROPERTIES OF C, TO C,, NORMAL PARAFFINS AND OTHER HYDROCARBONS A. J. W. HEADLEE AND R. E. McCLELLAND West Virginia Geological Survey, Morgantown, W. Vu.
a
-
Petroleum is a complex mixture of hydrocarbons concerning which more detailed quantitative data are needed. Recent recovery engineeripg and refinery practice requires more complete data on the composition and properties of petroleum. The development of more efficient laboratory fractionators and sorption techniques has made it possible to separate petroleum into many of its components. This study was made to supply detailed data on the petroleum produced in Test Virginia. This report gives the properties and an extensiveanalysis of the hydrocarbons found in a representative sample of high grade paraffin-base crude oil. Many of the fractions 80 obtained were made up of more than 50% of one known
compound. The properties of the fractions fit into three groups representing the paraffin-naphthenes, aromatics, and sulfur fractions. The paraffins show concentration maxima periodically a t the boiling points of the normal paraffins. The normal paraffins make up approximately one third of the original petroleum. These data can be used to determine aut points, to blend fractions, and to calculate yields of narrow cuts having special properties. The properties of and the quantitative data on the waxes, resins, and color bodies are of value in reservoir engineering. These data will aid in interpreting processes related to the origin, migration, and accumulation of petroleum.
HE natural hydrocarbons in petroleum may be grouped as T p a r a G s , naphthenes, aromatics, and aaphaltics. The aaphaltics are not aa clear-cut a group aa the other three, and in this paper are referred to as sulfur fractions or color bodies which probably consist of asphalts, and oxygen-, sulfur-, and nitrogen-containing compounds. They make up a small fraction of the crudes under investigation. When this study waa initiated, it was decided to attempt a much more efficient separation of the constituents in petroleum than is obtained by the generally accepted routine methods of analysis by Hempel distillations (6,6). The samples were quantitatively separated into three groupsparaffins and naphthenes, aromatics, and color bodies. These groups were then fractionated in a high efficiency still into 2’ C. cuts and a few properties determined on these cuts. The samples were collected from representative wells from 113 reservoirs throughout the oil fields of West Virginia. They
were collected from the lead line to the storage tank at the wells in most cases. INITIAL DISTILUTION.Two hundred milliliters of each sample were fractionated (1, 8 ) in two 13-mm. Heli-grid packed columns with fractionating effectiveness of 80 to 120 theoretical plates. Fractions were collerted so that each included the normal paraffin and its isomers boiling above the next lower normal compound. The normal paraffins from hexane to decane were kept separate from their isomers. The straight-run distillations were carried out so that the pot temperature never exceeded 350’ C.t o prevent cracking. When the pot temperature reached 350” C., the column waa operated under a decreasingly lower pressure until the fraction corrected to a boiling point of 412’ C. at 1 atmosphere was distilled over without decomposition. Cub with the same boiling range from each of the 118 samples were accumulated in the pame bottle and further separated.
