Sept., 1921
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
779
V iscosity-Temperature Curves of Fractions of Typical American Crude Oils'? By E. W. Dean and F. W. Lane BCREAUOF MINES,PETROLDCX DIVISION,CHDMICAL SECTIOS,PITTSBURGH, Pa.
Published information regarding the variation of the viscosity o f petroleum oils with temperature is, according to the observations of the authors, notably inadequate. Relatively few data are available in print, and in all cases there is a decided absence o f information regarding the origin and history of the samples on which tests were made. The investigation reported in the present paper was undertaken by the Bureau of Mines for the purpose of obtaining figures showing the viscosity-temperature relationships of a reasonably wide range of petroleum fractions of known origin and of accurately determined physical properties.
SCOPEAND GENERALMETHOD OF PROCEDURE The following three crude oils were selected for the present series of tests: 1--A sample of Pennsylvania crude petroleum obtained from a small refinery located in Pittsburgh, Pa. This oil was regarded as representative of the so-called "paraffin base" class. 2--A sample of crude petroleum from the Sunset Field, Kern County, Cal., which was regarded as representative of the class of so-called "naphthene bate" olli. 3-A sample of crude petroleum from the Salt Creek Field, Natrona County, \Tryo., which was regarded as representative of the class of so-called "intermediate base" oils. Analyses of these samples according to the Bureau of Mines procedure for crude oil appear in Table I. The viscosity determinations recorded in subsequent tables were made on fractions distilling at atmospheric pressure between limits of 100" to 125" C. (212' to 257" F.) 125" to 150" C. (257" to 302" F.), ete., u p to 250" to 275" C. (482" to 527" F.), and on all the " vacuum" fractions. These fractions are designated hereafter by letters and numbers indicating the crude from which they were derived, the pressure under n-hich they were distilled, and the upper (centigrade) distillation limit. Thus. for example, "W-V250" represents a fraction from Wyoming crude, separated between limits of 225" t o 250" C. (437" to 482" F.) in the '(vacuum " distillation at 40 inm. ; "C-A-175 " indicates a fraction from California crude, distilling between 150" and 175" C. (302" to 347" F.) at atmospheric pressure, etc. Viscosity determinations were made at the following temperatures: 0" C., 10" C., 20" C., 30" C., 40" C., 50" C., GO" C., 80" C., and 100" C. (32" F., 50" F., 68" I?., 86" F., 104" F., 122" F., 140" F., 176" F., and 212" F.). F o r obvious reasons, tests were not made a t all points on all fractions ; some of the fractions distilling a t atmospheric pressure have excessive vapor pressures a t 80" C. and 100' C., and some of the ('vacuum " fractions solidify partly or totally a t temperatures as high as 30" C. The determinations of viscosity-temperature curves f o r distillation fractions were supplemented by tests on the following oils : (1) The residuum (designated as " P-R ") from the distillation of the Pennsylvania crude, ( 2 ) a mixture (designated as '' P-mixt ") of this residuum and the Pennsylvania fraction " P-V-225," and ( 3 ) a mixture (designated as L L P-C-mixt "1 of " P-R " and the California fraction " C-V-250.'' Results in terms of kinematic viscosity (absolute viscosity divided by density) are given in full. Saybolt Universal equivalents are given for the majority o f figures representing kinematic viscosities in excess of 0.0142 (32 see., SavZPresented before the Section of Petroleum Chemistry at the 61st Meeting of the American Chemical Societ). Roche3ter, i X Y , April 26 to 29, 1921 2Publiqhed by permission of the Director, U. S.Bureau of Mines.
TABLE I-RESULTS O F ROUTIXB AiYALYSES O F SAMPLES O F CRUDE P E T R O L E U M FROM WHICHW E R D OBTAINED THE FRACTIOXB USED FOR VI~COSITY-TEMPERATURE MEASUREMENTS Sanwm No. 640 Pennsylvania Baume gravity--43.1 O Specific gravity-!l.809 Per cent water-Nil Per cent s u l f u r 4 . 1 6 DISTILLATJON, BUREAU O F hfINE.9 HEMPEL >IETHOD Air distillation, first drop, 24°C. (75°F.) Sp. gr. "BB. Viscos- Cloud Temperature Temperature Per Sum "C cent Per Cut Cut ity' Tpt "F. Cut Cent F. up to 50 1.5 1.5 0,639 89.1 Up t o 122 50- 75 2.8 4.3 0.662 81.5 122-167 75-100 3.3 7.6 0.702 69.4 T67-212 100-125 7.2 14.8 0.728 62.3 212-257 125-150 6.0 20.8 0.746 5i./ 25 I -302 150-1 15 6.5 27.3 0.760 54.2 302-347 115-200 5.5 32.8 0.713 51.1 34i-3'32 200-225 6.5 39.3 0.785 48.3 392-437 225-250 5.9 45.2 0,797 46.7 $37-482 250-275 7.0 52.2 0.810 42.8 482-527 Vacuum distillation a t 40mm. Cpto200 4.6 4.6 0.829 38.9 39.1 16 Up to392 200-225 5.2 9.8 0.834 37.9 44.5 36 392-437 225-250 5.2 15.0 0.842 36.3 53.5 55 437-482 250-275 5.2 20.2 0,850 34.7 71.2 74 482-527 275-300 6.0 26.2 0.8.59 33.0 109.5 90 527-572 Residuum 0.894 26.6 726.8 Residuum Carbon residue of residuum-2.3 per cent SAMPLENO. 474 Sunsst Field Kern County California SFecific yravity---0.878 Baumk gravity-29.5" Per cent w-ater-Nil Per cent sulfnr-0.73 DISTILLATION BUREAU OF MINESHEXFEIMETHOD Air distillation, first dropL26C". (79OF.) Temp:rature Per Sum Sp. gr. "Be. Viscos- Cloud Temptrature C. cent Per Cut Cutr itg Test F. Cut cent O F . 75-100 1.3 1.3 0.742 58.7 167-212 100-125 3.9 5.2 0.761 54.0 2 12-257 125-150 4.5 9.7 0.782 49.0 25'1-302 150-li5 4.0 13.7 0.805 43.9 302-347 175-200 3.7 17.4 0.818 41.1 347-392 200-225 4.8 22.2 0.844 35.9 392-437 225-2eo 5.5 27.7 0,860 32.8 437-482 250-275 5.9 33.6 0.874 30.2 482-527 Vacuum distillation at 40mm. Upto200 3.6 3.6 0,891 27.1 42.5 Up to 392 53.8 392-437 0.904 24.9 200-225 4.8 8.4 8 9 . 4 437-482 0.917 22.7 225-250 4.9 13.3 203.1 482-527 0,932 20.2 250-275 5.3 18.6 427.3 527-572 0.942 1 8 . 6 275-300 8.3 26.9 Carbon residue of residuum--16.4 per cent SAMFLENo. 561 Salt Creek Field Wyoming Baumb gravity-36.5' Natrona County Specific gravity-0.841 Per cent water-Nil Per cent sulfur-0.18 DISTILLATIOS, BUREAU O F MINESHEXPEL METHOD Air distillation, first drop-25'C. (77°F.) Tempyature Per Sum Sp. gr. "Be. Visoou- Cloud Temperature C. cent Per Cut Cut ityr Test O F . O F . Cut cent up to 122 Ur, t o 50 1.4 1.4 0,669 79.3 122-167 1.8 3.2 167-212 4.5 65.5 7.7 0.716 212-257 6.7 5 8 . 2 14.4 0.744 267-302 5.1 53.2 19.5 0.764 302-347 25.3 48.8 5.8 0,783 347-392 29.3 45.7 0.797 4.0 392-437 4 2 . 8 4.4 0.810 33.7 437-482 40.3 0.822 39.1 5.4 482-527 38.1 0.833 45.4 6.3 Vacuum distillation a t 40mm. 19 Up to 392 U~to200 4.1 4.1 0.849 34.9 40.1 392-437 39 45.2 0.852 34.3 260-225 6.1 '10.2 437-482 63 57.6 0,860 32.8 225-250 6.1 16.3 482-527 77 80.2 21.9 0.865 31.9 250-275 5.6 527-572 90 130.8 27.3 0.874 30.2 276-300 5.4 Carbon residue of residuum-6.1 per cent XViscosity figures are Sasbolt Universal see.. at 100°F.
.
~
bolt Universal). Transformations were made by the use of the Bureau of Standards equation' vk=0.00220 fh--1 80 t* in which Vr is kinematic viscosity and ts Sapbolt Universal viscosity. EXPEKIMESTAL
The distillation methods used f o r preparing the various fractions are to be described in full in Bureau of Mines BuEZetin 207, which will be published at a slightly later 1W. H Herschel, "Standardization of the Saybolt Universal viscosimeter," Bureau of Standards, Technoloyir Paper IIZ (1919), 2d ed., 19.
780
T H E JOURNAL OF I N D U S T R I A L AND ENGINEERING CHEMISTRY date. The method involves distilling a 300cc. charge of oil in a s p e c i a l glass flask, equipped with a 6-in. Hempel column, until a vapor temperature of 2750 C. (527O F.) is reached. The flask and its contents a r e then allowed t o cool. The )cQg4' 0&3Cc
b
ggc
is removed and replaced by a "spray catching" device consisting of three inverted gauze cones. The flask is attached to a r P e r l y designed ('vacuum" System, and the distillation is continued at a reduced pressure of 40 mm. absolute until a vapor temperature of 3000 c. (5720 F.) is reached. Fractions'are separated at temperature limits that are multiples of 250 C. (The respective Fahrenheit limits aye 450 apart.) \-.-
1 CQPQ&
vbnui' /o C&
~
Viscosity determinat i o m were made by means of s p e c i a l l y constructed Ostwald viscosimeters, about a dozen of which were used. These instruments varied somewhat in the size of the upper bulb and the length and bore of the capillary. The approximate dimensions of two extremes of the set are shown in Fig. 1. Viscosimeters having relatively large bulbs and long, fine capillaries were, OP course, used for the less viscous fractions, whereas for viscous oils the small-bulbed instruments with shorter, wider capillaries were more convenient. Fig. 1-Form and approximate dimensions of two extremes of set of Ostwald viscosimeters used in present series of exweriments
METHOD O F FII,LIKG VISCOSIMETERS-The problem Of filling viscosimeters offered a little difficulty, as the conventional procedure of introducing a known volume of liquid from a pipet is obviously unsatisfactory in the case of the more viscous fractions. The scheme finally adopted involved using the lower reservoir of the viscosimeter as a measuring container. The process of ,filling was accomplished by setting u p a viscosimeter in a vertical position and introducing oil through the wider arm by means of a pipet o r pressure injector equipped with a long, slender tip. The oil was run in '(drop-wise " after the reservoir had become approximately three-quarters full and the flow was stopped as soon as the level of the liquid reached the junction of the capillary and the reservoir. This point was indicated very accurately by a shooting up " when the oil first entered the capillary. STBXDARIZATION O F VISCOSINETERS-Each viscosimeter was standardized by determining the efflux time with each of at least two liquids of known viscosity, the '(known " liquids being distilled water, and samples of oil for which kinematic viscosity figures were supplied by the Bureau of Standards. The selection of calibrating liquids was determined bv the " rapidity " or '' slowness '' of the viscosimeter under test. F o r the "slower" instruments, that is,
Vol. 13, No. 9
the ones with large bulbs and long, fine capillaries, water and one of the cess viscous "standard" oils were used. The more '' rapid " viscosimeters were standardized witli two oil samples. The method of Standardization involved the assumption of the following relationship : Vk=At (1) in which Vli is kinematic viscosity, t the efflux time and A a constant characteristic of the viscosimeter. The general equation is B Vb= At (2 1 t in which the term (the kinetic energy correction) is negt ligible if the rate of flow of liquid through the capillary is sufficiently slow. The employment of two liquids in the standardization of each viscosimeter indicated the negligible magnitude of the " kinetic energy correction " f o r when two liquids of differing viscosity give the same value for A in Equation 1 it is obvious that neither of them flows with sufficient rapidity to necessitate the use of Equation 2. THE " FILLIKG FACTOR"-In case a ViSCOSimeter is filled at one temperature and the efflux time is determined at some other temperature, it is obvious that some sort of a correction factor must be employed to compensate f o r the change in " head " brought about by thermal expansion or contraction of the oil. For purposes of operating convenience the authors usually filled viscosimeters at temperatures of either 25" C. (77" F.)o r 100" C. (212" F,). Actual determinations of viscosity were, of course, made a t temperatures ranging from 0" c. (32" F.) to 100" c. (212" F.). It was necessary, therefore, to use a series of so-called filling factors to compensate f o r the errors that would otherwise have been introduced by this detail of procedure. An average factor was determined f o r the whole set of instruments bv making viscositv tests on certain oils a t 100" C. (212" FY) with f h n g temieratures of 25" C. (77" F.) and 100" C. (212" F.). The average ratio f o r this difference in filling temperature was 1.023 a t 100" C. Using this figure as a basis, the set of "filling factors" given in Table I1 was calculated. 7
-
'
TABLD11-FIILING FACTORS UBED s HEN VISCO~IMETERR A R E FIILEDAT 25°C. (77' F.) O R 100" C. (212' F.) AKD VISCOSITYDETERMIKATJOSSA MADE AT OTHERTEMPERATURES Temperature a t which "Filling Factor" when "Filling Factor" when Viscosity Determination Pipet is Filkd at Pipet is Filled a t is Made 25' C. (77" F.) 100" c. (2120 F.) o _r . O_ W . 0 10
20
30 40
50 EO 70 80 100
32 50 68 86
104 122 1rO 158
176
212
0.992 0.995 0.999 1.002 1.005 1.008 1,011 1.014 1.017 1.023
.... ... ...
0.979 0.982 0.985 0.988 0.991 0.999 1.000
The equation actually used f o r the calculation of results therefore takes the form : Vk =Atf (3 ) in which Vic is kinematic viscosity, A the constant characteristic of the viscosimeter used, and f a properly selected filling factor from Table 11. It should perhaps be stated that the authors recognize the fact that their method of standardization and calculation involves certain approximations that would not be justifiable if the highest possible degree of precision were necessary. It has seemed, however, that the consumption of time necessary f o r increased accuracy is greater than is warranted through any possible advantage that could be gained. MISCELLANEOUS OPEIZATIXG DETArL+--Actna1 determinations were made with the viscosimeters immersed in constant temperature baths. A vigorously stirred mixture of cracked ice and distilled water gave a temperature of 0 " C. (32" F.), while temperatures from 10" C. (50" F . ) to 70" C .
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
Sept., 1921
.5-0
.45
.40
.35
.30
I + !
281
vania residuum and the two mixtures previously ref erred to. Fig. 2 shows graphically the majority of the data included in Table 111, A and B , and indicatef the similarities and differences of the viscosity-temperature curves for frae$ions from two extreme types of crude oil. It will be observed that the curves representing low boiling ranges and low viscosities are not unlike, but that there are impressive differences in the " higher " fractions. The curves for the California oil are much steeper than those f o r Pennsylvania cuts and indicate a greater variation of viscosity with temperature. Curves f o r the Wyoming oil are not shown. They have characteristics intermediate between those of the Pennsylvania and California curves and resemble the former more closely than they do the latter. While it is felt that the method of plotting the results adopted in Fig. 2 is the most effective way of illustrating the points emphasized above, another scheme of graphic representation has certain advantages.' I f the reciprocals of the kinematic viscosities as ordinates a r e plotted against the L* temperatures as abscis/*3 sas, the resulting lines a r e straight f o r the ~2 lighter fractions, and only slightly curved f o r 1.1 t h e heavier fractions. A plot so constructed permits easier interpo- .$ .9 lation t h a n t h e type of diagram shown i n Fig. o .B 2. For t h e g a s o l i n e 3 fractions, a straight .u .7 line drawn through two experimental p o i n t s Q .6 will represent satisfac.5 torily the viscosity ternPerature relationship k 4 while for the heavier .3 oils, three o r more points obviously will be required to determine 2 the curve. This subject is considered from a mathematical stand0L - J - - - l u point in a later connec40 $0 60 7U 80 Sa Mi?
/*
I I
2
./5
F
./a
.F
.Ob
,,
1
7
20
peruke, '6 Fig. Z--Viseosity-temperature curves of sets of distillation fractions derived from Pennsylvania and California crude petroleum
(158" F.) were maintained by the use of x a t e r batlis equipped with the conventional devices f o r thermostatic control. The baths f o r 80" C. (176" 17.) aud 100" C. (212= I?.) consisted of glass jackets through which benzene vapor and steam were p a s e d . Both of these baths were maim tained under pressures so regulated as to give tlic. prccisr temperatures desired. At least three determinations of the efflux tiiue w e ~ emadv f o r each oil at each temperature, and the viscosity figure.? reported were calculated from the mean of these espci i niental values
DISCL-SSIOS OF RESULTS The calculated figures f o r the kinematic viscosities arc given in Table 111. Table I V shows equivalent Saybolt Universal figures f o r the " vacuum " fractions, thc Pennsyl-
tion.
7empeerafur-e; Fig. 3-Viscosity-temperature curves Of several of the more viscous products d e rived from Pennsylvania, California and Wyoming crude petroleum.
Fig. 3 shows in detail a comparison of curves characteristic of viscous oils derived from Pennsylvania and California crude petroleums. The upper curve, marked '' P - R " represents the residuum from the combined " a i r " and " vacuum I' distillations of the Pennsylvania crude. This residuum is a product the properties of which approach those of so-called steam cylinder stock. I t s viscosity is high throughout the entire range of temperature, but it belongs to the same family as the lowest two curves of the diagram, marked " W-V300 " and u P-V-300," and representing respectively vacuum distillation fractions cut between limits of 275" to 300" C. (527" to 572" F.) from the Wyoming and Pennsylvania crudes. The intermediate curve, marked '' C-V-300," representing the highest vacuum fraction from the California crude, shows at 100" C. (212" F.) a viscosity slightly higher :The authors are indebted to Professor Robert E. Wilson, of the M a ~ m ohusetts Institute of Technology. for bringing to their attention the prnctiCa advantages of this method.
782
T H E J O U R N A L O F I N D U S T R I A L A N D ENGIAiEERI,VG C H E M I S T R Y TABLEIII-KIh'EMATIC
VISCOSITIES AT VARIOCS TEMPERATGRES OF
Description of Product Pressure Temperature of under which Limits Product Distilled of Cut Designation
P-A-125 P-A-150 P-A-175 P-A-200 P-A-225 P-A-250 P-A-275
Atmospheric Atmcspher?c Atmospheric AtmoEpheric Atmospheric AtmoEpheric Atmcspherio
'C. 100-125 125-150 150-175 175-200 2CO-225 225-250 250-275
P-V-200 P-V-225 P-V-250 P-V-275 P-V-300
40-mm.vac. 40 mm. vac. 40 mm. vac. 40 mm. vac. 40 mm. vac.
175-200 200-225 225-250 250-275 275-300
--
P-R
--
347-392 392-437 437-482 482-527 527-572
C-V.200 C-V.225 C-V-250 C-V-275 C-V-300
W-A-125 Atmospheric 1CC-125 W-A-150 Almcepheric 125-150 W-A-175 Atmotoheric 150-175 W-il-ZCO .4tinoepheric 175-ECO W-A-225 Atmcspheric 2CO-125 W-A-250 Almcapheric 225-250 T?i -A-275 Atrrc~pheric 2EO-275
212-257 257-302 302-347 347-392 Z92-437 437-482 482-527
W-V-200 4'0-mm. vac. W-V-225 40-mm. Yac. W-V-250 40-mm. vac. W-V-275 40-mm. vac. w-T'-3@040-mm. vac.
347-392 392-437 437-482 482-527 527-872
17E-200 200-225 225-250 250-275 275-300
TABLEIV-SAYBOLT Designation of Fraction
175-2CO 347-592 2CO-225 392-437 225-280 437-482 250-275 482-527 275--300 527-572 Rfsiduum RIixture P-R and P-V-225 ~
.. ..
175-2CO 247-592 c-V-200 , , , , , , C-V-225. , . , , , . , 2CO-225 592-437 C-V-250. ....... 228-250 437-482 C-1'-2 75 , , . , , , , 250-275 482-527 C-V-300. , . , , , , 275-3C0 527-572 P-C-Mixt , , , , , . Mixture P-R and C-1'-250
.
W-1'-200 . . . . . . . W-V-225 . . . . . . . W-V-250.. . . . . . W-V-275.. . . . . . W-V-300 . . . . . . .
175-200 200-225 225-250 250-275 275-3G0
347-392 392-437 437-482 482-527 527-572
C.
Kinematic Viscosities Cal--culated or Interpolated f or Indicated Temperatures. 100OC. 37.8OC. 54.4'C.989OC 212O F. 100' F. 130' F. 210' F.
50°C. 122O F.
140° F.
80°C. 176O F.
0,00880 0,0111 0.0141 0.0199 0,0298 0.0454
0.00780 0.00957 0,0121 0.0166 0.0236 0.0345 0.0529
A-Pennsylvania Crude 0.00703 0.00643 0.00593 0.00542 0.00869 0,00780 0.00708 0.00649 0.0106 0.00946 0.00849 0.00769 0.0142 0.0123 0.0110 0.00964 0 .Ola6 0,0167 0.0145 0.0127 0,0277 0.0227 0,0191 0.0164 0.0407 0.0322 0.0265 0,0221
0.00508 0.00603 0.00702 0.00871 0,0112 0.0143 0.0188
0.00456 0.00524 0.00618 0.00737 0,00926 0.0115 0.0147
0.00467 0.00548 0.00642 0.00781 0.00958 0,0120
0,0729
30° C.
8 6 O F.
0.1296
0,0642 0.0489 . . . . 0,0884 0,1372 0,0962 0,0704 0,1533 0,1079 .. . .., .. .. . . . . .. . . . . . . ..,.. .. . . 0,1746 ...
40° C. 104O F.
GOo
........................ .................. .................. .................. .................. .................. ..................
0.0384 0.0544 0.0798 0,1226 0,2044
0,0315 0.0432 0,0609
0,0899 0.1444
0.0289 0.0366 0.0484 0,0694 0.1064
0,0195 0.0247 0.0327 0.0448 0.0654
0,0152 0,0186 0.0238 0.0312 0.0441
0.0400 0.0574 0.0841 0.1314 0.2245
0.0289 0.0153 0.0394 0.0188 0.0546 0.0242 0.0797 0.0318 0 . a 4 8 0 ,0450
..................
1.498
0,8064
0,5248
0,2581
0.1408
1.590
0 , 6 7 8 0 0.1512
..................
0.1869
0,1208
0,0918
0,0570
0.0398
..................
Bxalifornia O.OG902 0.00807 0.00735 0.0108 0.00957 0.00860 0.0138 0.0120 0.0106 0.0185 0.0158 0.0135 0 0263 0 0213 0 0180 0 0396 0 0312 0 0252 0 0666 0 0486 0 0375
Crude
, ,
0,1328 0,2825 0,8393
,
,,
0.0485 0.0803 0,1628 0,3877 0,8282
0,0383 0.0596 0,1122 0.2372 0,4612
0,0308 0.0459 0.0807 0.1572 0.2836
0,0217 0.0303 0.0482 0.0825 0,1306
........................
0,3228
0,2108
0.1485
0,0811
C-Wvoming 0.00927 0,00829 0.00741 0,00675 0.01141 0,00999 0,00890 0,00803 0.0113 0,0100 0.0151 0.0129 0,0207 0.0173 0.0148 0,0128 0.0299 0.0241 0,0200 0,0169 0.0231 0.0348 0 0281 0.0452 0,0327 0.0542 0.0414 0.0737
Crude
0,1432
,
,
............ ,
,,
0,2142
,,,
,,
0.0902
.. ,, .. .. 0,1717 0,4360 .................. .. . . . . , . . , ,
,,
0.0976
, ,
,
,
0.0698
0,0645 0,1136 0.2548 0,6926 , ,
., ..
0.0526
. , . , . . 0,1491 0.1026 0.0743 .. ,.. ., ,.. . . .. ,..., ,. , . . .0.1792 0.1227 . . . . . 0,2070 ........................ VISCOSITIES
0.00610 0.00565 0 .@0488 . . . . . . 0.00775 0,00711 0.00661 0.00570 0.00605 0.00942 0,00847 0,00773 0.00651 0.00568 0,0119 0.0105 0.00943 0.00781 0.00673 0,00961 0.00817 0.0154 0.0134 0.0120 0,0121 0.00987 0.0179 0.0154 0,0210 0.0124 0,0249 0,0209 0.0158 0.0302
0,00665
......
...... ...... ...... ...... ...... ...... ......
. . . . . . ......
...... ...... ...... ......
...... ......
0,0165 0,0219 0.0322 0.0502 0.0736
0.0510 0.0848 0.1766 0.4380 0.9363
0.0348 0.0529 0.0962 0.1961 0.3686
0.0817
.....
......
......
0.00489 0.00646 0.0054i
0 0114 o .oi46 0.0193 0 0268
0,00547 0.00628 0,00760 0,00919 0.0114 0.0146
0.00773 0.00935 0.0116
....
0.0412 0.0563 0 ,0892 0,1418 0,2501
0.0330 0.0442 0.0672 0.1033 0.1701
0,0199 0,0251 0.0356 0.0483 0,0723
0,0155 0,0188 0,0257 0.0334 0,0470
0.0433 0.0596 0.0955 0,154 0.274
.......
..... . . . . . . . . . .. . . . . . . 0.0303 0.0401 0.0600 0 .O905 0.146
.... ....
.... ....
71 . S 134.5 383.5
55.7 57.4 202.2
....
....
.... .... ....
75.9
58 4
48 6
....
....
0.0157 0.0101
0.0260
0.0340
0.0480
.4T YARIOUS
. . . . . . . . . . . . . . . . 105.1
0.0166 0.0222 0.0328 0.0514 0.0757
. . . . . . .. . . . . . . ....... .......
......
0.00567 0 00533 0.00669 0 00622 0.00808 0,00752 0 0101 0 00907 0 0127 0.0114 0.0143 0 0165 0.0191 0 0224 0.0274 0.0356 0.0528 0,0774 0.1230
...... ......
TEXPCRATURES O F .4 SERIES O F P R O D V G T S FRO31 CItCDE PETROLEUM (Fractions distilled at 40-mm. vacuum) Saybolt Viscosities O'C. 10cC. 20OC. 30W. 40%. 50OC. 60% SOW. 100°C. 37.8OC. 54.4OC. 32'F. 50OF. 68'F. 86°F. 104OF. 122OF. 140'F. 176'F. 212OF. 100'F. 130'F. A-Pennsulzania Crude 70.5 55.0 40.7 41.8 38.0 30.6 35.1 33.4 32.3 39.1 3j.9 .... 73.5 57.9 48.8 43.5 40.1 37.8 34.8 33.1 44.5 38.9 ........ 79.9 62.2 52.0 45.6 41.7 37.0 34.5 53.5 43.6 . . . . . . . . . . . . 88.6 67.8 55.6 48.4 40.5 36.0 71.2 52.0 . . . . . . . . . . . . . . . . 101.0 76.4 61.6 47.1 40.3 109.5 68.7 . . . . . . . . . . . . . . . . 682.2 368.8 241.6 123.9 77.3 726.8 310.8
cXIVERS.4L
Temperature Limits of Fraction OC. OF.
P-V-200.. . . . . . . P-Y-225.. . . . . . . P-V-250. . . . . . . . P-V-275.. . . . . . . P-V-300.. . . . . . . P --R... . . . . . . . . . . Y-Mixt . . . . . . . .
loo C.
13, No. 9
CRUDE PETROLEU1\1
20°C. 6S0 F.
0.0103 0,0124 0.0162 0 ,0223 0.0332 0.0523 0 ,C935
Mixture of Pennsylvania residue and C-V-250
P-C-Mixt.
PRODUCTS DERIVEDFROM
50° F.
OOC.
Residuum
40-mm. vac. 175-200 347-392 40-mm. V L C . 200-225 392-437 40-mm. t-ac. 225-250 437-482 40-mm. vac. 250-275 482-527 40-mm. .iac. 275-300 527-572
OF
32' F.
P-Mixt. Mixture residuum and fraction P-V-225
C-A-225 AtmcEpheric C-A-250 Atmcspheric C-A-275 Atmospheric
SERIES
Kinematic Viscosities Determined a t Indicated Temperatures
OF. 212-257 257-302 302-347 347-392 392-437 437-482 482-527
A
1701.
46.8 64.3 122.5 317.5
....
.... 43.0
. . . . 7 8 . 3 60.2 50.1 . . . . . . . . 90 5 67.8 . . . . . . . . . . . . 102.1 ................
than those of the corresponding Pennsylvania and Wyoining products, whereas at the lower temperatures it approaehes more closely the viscosity of the Pennsylvania residuum. Fig. 4 shows viscosity-temperature curves for typical kerosene fractions ( ' I P-A-225," '' C-A-225," and '' W-A225 ") from the Pennsylvania, California, and Wyoming crudes. Viscosity differences are o f the same type as those shown f o r higher boiling, more viscous fractions, but are of a lesser order of magnitude.
85.5 67.1 B-California Crude 38.6 41.7 45.2 52.2 63.8 83.8 1E0.7 114.9 213.5 378.7 152.1 103.7 C-Wyoming Crude 39.5 37.1 44.2 40.3 55.3 47.6 75.3 60.5 120.5 86.7
98.9OC. 210'F. 32.3 33.2 34.6 36.7 40.0 79.1
56.2
44.3
39.0
....
....
36.4 40.9 52.3 81.5 135.0
33.9 36.3 41.6 53 .O 70.8
32.6 34.0 36.8 42.2 49.9
42 5 53.8 89.4 203.1 427.5
37.6 43.1 57.9 07.5 172.3
32 , 6 34.1 37.0 42.8 50.6
77.2
52.5
42.8
....
....
....
355 37.8 43.0 51.2 68.0
33.5 34.8 37.8 41.6 49.4
32.3 33.3 35.0 37.2 41.2
40.1 45.2 57.6 80.2 130.8
36.3 39.1 45.3 55.8 77.0
32.4 33.3 35.1 37.4 41.5
..
Fig. 5 gives interesting indications regarding the vjscositytemperature curves of mixtures. It has just been pointed out (refer to Fig. 3 ) t h a t the viscositS-temperature curye of the Pennsylvania residuum resembles the curves for the distillates from the same crude. I n F i g 5 the curve marked " P-mixt " represents a blend of Pennsylvania residuum and the Pennsylvania vacuum fraction " P-V-225." It will be noted that this curve is entirely similar to those obtained from the fractions " P-V-300 " and ('P-V-275," between which it lies. The curve marked P-C-mixt " represents
T H E JOGKiC'AL O F I N D C S T K I A L AiVD EXGINEERILVG CHEMISTRY
Sept., 1921
w
I
I
l -Oo6L
i 10
do
3;
4;
i
o!
Zmprafure, 'C
I4,
7L
50
I 20
Fig. 4-Viscosity-temperature curves of typical kerosene fractions derived from Pennsylvania, California and Wyoming crude petroleum
a mixtme of the Pennsylvania residuum and one of the lighter vacuum fractions of the California crude. It will be noted that its slope is riot as steep as that of the adjacent curve '' C-V-275," representing a '' straight " California distillate. The indications furnished by these comparisons point t o the fact that tlie viscosity-temperature curves of the irnportant group of commercial lubricating oils made by blending " paraffin base " distillates and " paraffin base " residuums show the same characteristics as the corresponding curves for " straight " distillates. When blends are composed of oils derived from different types of crucle petroleum the characteristics of the curves are intermediate between those of the constituents.
GENERAL CHARACTERISTICSOF VISCOSITYTEXPERATURE CURVES From the data presented in Tables I11 and IV, and in Figs. 2 to 5, inclusive, two conclusions may be drawn. I n
175-200 200-225 225-250 250-27Fj 275-300
347-392 392-437 437-482 482-527 527-572
Pressure under whioh Distilled .4t mospheric Atmospheric Atmospheric Atmospheric Atmospheric .4tmospheric Atmospheric 40-mm. vacuum 40-mm. vacuum 40-mm. vacuum 40-mm. vacuum 40-mm. vacuum
c
h
113.2 90.19 70.30 49.60 33.50 21.37 12.97
7.21 4.43
1.88
+0.810 -4.0'0
I
I
I
1
I
70
80 T€MP€P?TUR€, %.
60
I
I ?GO
the first place, the work done apparently indicates the nonaistence of any simple rule by means of which the variation of viscosity with temperature can be predicted. The ratio between viscosities a t different temperatures varies with the type of crude petroleum froin which the oil is derived and also with the physical and clieniical properties of fractions from a given crude. For example the kinematic viscosity at 100" C. (212" F.) of the California fraction "C-V-300" is 8.9 per cent of tlie figure determined for 40" C. (104" F.). A corresponding ratio for the Pennsylvania fraction "P-V-300" is 21.6 per cent. It may be noted further that for the Pennsylvania fraction '' P-V-250 ') the ratio between kinematic viscosities at the same two temperature points is 29.8 per cent. Similar variations may be discovered when comparisons are made of products derived from different crudes and having equal viscosities at any given temperature. This matter is discussed in more detail in a later connection. TEM11ER.4TORE CCRVES O r
- ----
Pennsylvania Fractions' A B 4.00200 1.498 1.300 4.000691 1.250 -0.0012E 1.072 4.0000380 0,865 +O ,000794 $0. 000810 0.751 0.583 0.00130 0.400 0.252 0.182 0,0994 0,0522
I
I
Fig. 5-Showing t h a t the viscosity-temperature curve of a mixture of Pennsylvania distillate and Pennsylvania residuum has the same form as curves f o r unblended P e ~ n s y l v a n i adistillates, whereas the curve for a mixture of California distillate and Pennsylvania residuum varies appreciably from the form of a curve f o r unblended California distillate. 1 Compare Basler chemische Fabrik, D. R . P. 205,3ii, a n d 211,696.
T A B L E T'-\'ALCES O F THE CALCVLATED CONPT.4STS h.1 A AND R REPREEEXTIXG THE T-JSCOSlTY a r V E D FROM PENIiSPL\-ANIA AND CALIFORNIA CRCDE PETROLEUM
'TemFerature Limits of Fraction "C. F". 100-126 212-267 125-150 257-302 160-176 302-347 175-200 347-392 200-.225 392-437 225-250 437482 250-275 482-527
I
783
O.OOlS9 0 00243
0.00219 0.00213 0.00175
ri
97.60 P0.74 61.32 44.14 29.57 i8.61 10.29 4,242 1.254 0,550 1.070 1.967
THE
D~FTIILATIOX I.'R.~CTIOXS DH-
California FractionsA 1.287 1.204 1.108 0.989 0.853 0 661 0.48s
7
R
$ 0 . 000688
4,000298 +O 000450 $0.000623 0,000813 0 00169 0.00217
0.309
f0.178 + O . 0319
-4.0629 4,109
Residuum +0.881 -4,0468 i o ,00106 lThe viscosity temperature curves can be represented in terms of an equation of the general form 1 V k = ~ + ~ t + ~ t l n which Vk is kinematic viscosity, t the rentigrade temperature and h,A , and B constaptg chararterietic of each individual fraction from each oi:.
0
'
0.00259 0 00269 0.00274 0.00252 0.00225
Vol. 13, No. 9
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
784
Secondly, an outstanding similarity is found in the general form of the various viscosity-temperature curves. These can be represented by equations of the form: 1 V k= (4) K +At +Bt2 in which v k is kinematic viscosity, f the centigrade temperature and K, A and B, constants characteristic of each individual fraction from each crude.' The values of the constants for the products derived from the Pennsylvania and California c&des have been calculated by the method of least squares and are shown in Table V. The excellent agreement between experimental and calculated values is illustrated by Table VI, which includes figures f o r several typical fractions. TABLE VI-AGREdMENT
.
0.003
0.002
o.uu/
B
0.000
aoo/ 0.002
B ~ T W E EEXFERIMENTAL N AND CALCULATED t'ALUB5
OF h I N E M A T I C VISCOEITY O F SlLECTED FRACTIONS 1'ROM PENNBYLvANI.4
CALIFORNIA CRUDEPETROLEUM TemFraction A-150 Fraction A-215 Fraction V-225 perahinematic hinematic hinematic tye Viscosity Viscosit Viscosity C. EXF. talc. ~ x c . &IC. ~ x p . Cnlc. AND
Fraction V-300 hinematic Visconity ~ x p . calc.
A
Pennsylvania Fractions
0 10 20 30
o.ii2i o.i38k
40
50
€0
SO 100
0.0962 00701 0:0544 0.0432 0.0356 0.0247 0.0186 California Fvactione . . . . . . 0.0935 0.0972 0.0656 0.0650 0.2825 .0.0486 0.0418 0.1717 0.0375 0.0372 0.1136 0.0302 0.0300 0.0803 0.0249 0.0249 0.0596 0.0209 0.0211 0.0159 0.0158 0.0158 0.0303 0.0124 0.0128 0.0219
.~
0 10 20 30 40 50 60 80 100
0.01239 0.0107c) 0.00950 0.00858 0.007i9 0.00713 0.00658 0.00571 0.00505
0.0956
00708 0:0543 0.0433 0.0353 0.0249 0.0186
0:3030 0.1700 0.1111 0.0791 0.0593 0.0163 0.0306 0.0215
1::: .. ..
:::I
0.202 0.1444 0.1064 0.0654 0.0441
0.2048 0.1633 0.1061 0.0651 0.0441
. ..
12 0
..... ' .'.' .' . .... . . . .... ....
100
,
0.8282 0.4612 0.2836 0.1306 0.0736
0.5230 0.4651 0.2826 0.1306 0.0735
80
x
It will be understood that the constants f o r the viscositytemperature equation of any oil can be calculated from the experimentally determined viscosities at three suitable temperatures. The method of calculation involves the solution of three simultaneous equations, and does not require the use of the more accurate but decidedly tedious method of least squares. Obviously, the values f o r the constants are useful in calculating viscosities at temperatures, within the experimental range, f o r which actual determinations are lacking. The reliability of such calculated viscosities can be accepted without question; but the use of the constants at temperatures considerably below or above the extremes represented by the experimental data is not to be recommended on account of a possible change in state of the oil, or a change in the viscosity-temperature relationship. While the equations would seem to offer an effective method f o r estimating solidification temperatures (at which the kinematic viscositv becomes infinite), it is felt that the significance of such calculations is open to question because of the fact that partial solidification so often takes place. I n the authors' opinion. it is equally unjustifiable t o attempt the calculation of viscosities for temperatures a t which the oils have appreciable vapor pressures, and hence are changed in properties through partial distillation. As a n indication of the validity of the equations at temperatures considerably above those covered by the foregoing experimental data, two viscosity determinations were made this is in effect the Slotte equation (See A. E. Dunstan and F. B. Thole "The Viscosity of Liquids," London, 1914,4; Schlotte, Wied. Ann., 14 (1551), 13, Beibl. 16 (1892), 182; Thorpe and Rodger, Phil. Trans.A , , 1'85 (1894), 397, which is usually employed to indicate the relation between abaolute viscosity and temperature but nhich the authors have found equally useful as applied to kinematic viscosity.
60
40 20 00
20/i5/h /)3
u
2L.5 L O Z!'5 ZOO 228 250 275 300
Upper cenfiqmde d&i%ttikmUpper cenfi' rude
hh75 of 02 fracfiom
d/'sf//ut;on 8m;f.s of Vacuum fiUCl/Ol;iS
Fig. %Showing similarities in curves obtained by plotting against upper distillation limits the values of the Constants K, A a n a B from the equations representing viscosity-temperature curves of the series of fractions from Pennsylvania a n d Cali fornia crude petroleum
at approximately 180" C. (356" F.)and compared wit5 l l i c ~ calculated figures based on the experimental range 40" to 100" C. (140" to 212" F.). Table V I 1 ( p a r t A ) shows t h e agreement to be satisfactory in both cases. TABIE
T'II--CALCUIATETJ AND E X P E R J M E S T A L TIVDLY HIGH TIVPCRATLTRE~~
VISCOSITY
FIGURES r O R
R ti,k-
~
PRODUCT Temperature P-I2 c-v-300
"C.
181.7 182.5
Part . 4
hinematic Viscositv Exnerimental Cal'culated 0.0363 0.0366 0.0171 O.Oli5
Part 5
P ~ o n r - c r -kinematic Tiscosity a t Temperature Indicatod-Experimental Calculated 100°C. (212°F.) 200°C. (392'F.) 300°C. (572°F.) 0.1468 0.0295 0.0122 P-R P-v-300 0.0441 0.0124 0.00578 C-V-300 0.0/36 0.0148 0.00582
Assuming the validity of the equations for temperaturci considerably above those of the experimental range, it is interesting to consider the values so calculated as indicating whether viscosity actually becomes constant at higher temperatures. Table V I 1 ( p a r t B ) shows the calculated values a t 200" C. (392°F.) and 300" C. (572" F.) as compared with the experimental figures at 100" C. (212" F.). These figures seem to indicate that viscosity continues to decrease even at temperatures as high as 300" C. (572" F.). The constants for the equations may also possess addi-
785
T H E JOURNAL O F I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
Sept., 1921
TABLEVIIT-COMPARATIVE SAYBOLT UXIYBRSAL VISCOSITIBIS AT 210'F., 130"F., AND 100°F.F O R OILS DERIVBD FROM PnNNaYLYANIA AND CAI IFORNIA CBUDEPETROLEUM E-Figures for Pennsybania Oils Onhi Gaybqlt Saybolt Saybolt Visccsity Viscosity Viscosity at 130°F. at 100°F. at 210"E.
71 72 73
291 306 321 337 353 369 385 401 419 437 455 473 $91 010 529 548 567 587
75
...
56 57 58 59
60
61 62 63 64 65 66
67 68 69
io
74
76
... .
I
.
17 i8
79 80
'-30
50
70
90
//O
/30 150
470 /YO 2/0
Saybo/f hscode a? / o o * f Fig. 7-Showing curves obtained by plotting Saybolt viscosities at 2100F. against aSybolt viscmities at 100OF. f o r the series of products obtained from Pennsylvania and California crude petroleums.
tional theoretical or practical significance, and the authors have under consideration several possibilities that will be investigated as soon as further experimental data' are available. F o r the present, attention is called only to an interesting relationship indicated by the curves i n Fig. 6 , which were obtained by Flotting values of the constants f o r the series of fractions from Pennsylvania and California crudes against the upper (centigrade) distillation limit f o r each cut. An interesting similarity mill be noted in tli' .form of the curves representing values of the constants 1; and A f o r equivalent fractions from each of the two types of crude petroleum. The similarity becomes equally striking in the curves f o r the constant B if, instead of adhering so closely to the plotted points as has been done in Fig. 6, more general curves a r e drawn, RELATIONBETWEEN SAYBOLT VISCOSITIESAT DIFFERENT TEMPERATURES It has already been stated that no simple general rule has been discovered which may be applied to estimate viscosity at one temperature from determinations made at some other temperature.' The nearest approach, of practical value, to a relationship of this order is indicated by the fact that a smooth curve may be constructed to represent the viscosities at any two selected temperatures f o r the series of fractions derived from any given crude. Fig. 7 illustrates this possibility and shows curves obtained by plotting Saybolt Universal viscosities at 210" F. against Saybolt Universal viscosities at 100" F. f o r the series of products obtained from the Pennsylvania and California crudes. Table VI11 includes a series of Saybolt viscosity equivalents f o r the commonly used temperature points, 100" F., 130" F., and 210" F. This table may prove of practical use in .The conclusions of Oelschlager ("The Viscosity of Liquids," 2. Verein. I w . . 62 (l9lS),422, Science AEslracte. 21-E (1918).320) are obviously untenable, at least in part, in view of the exFerimenta1 evidence included in the present paper. The logarithmic relationship developed by this investigator (p. 424) is undoubtedly of considerable value. but his constants are not applicablezto lubricating oils from all types of crude petroleum. deut.
case any oil under consideration is known to be of the same type as either of the two crudes studied by the authors. Such a happy coincidence is, unfortunately, not likely to be a frequent occurrence, and the table is of greatest significance as inclicating possibilities of variation when it is attempted to estimate viscosity at one temperature from figures determined a t some other temperature. It appears, f o r example, that Pennsylvania and California products, having at 210" F. a viscosity of 50 see. Saybolt Universal, have at 100" F. respective viscosities of 209 and 396 see. It is possible that a proper use of figures f o r other physical properties of oils might permit reasonably accurate interpolation between limits such as have just been indicated. The data at present available have not, however, seemed adequate to permit formulating any rule f o r interpolation that would be satisfactorily reliable.
SUNNARY 1-Viscosity-temperature curves have been determined f o r series of products derived from samples of Pennsylvania, California, and Wyoming crude petroleum. 2-The California fractions show a greater change in viscosity with temperature (the curves a r e steeper) than the corresponding Pennsylvania products. Differences are most impressive f o r fractions of high viscosity. Viscositytemperature curves for the Wyoming fractions show characteristics which are intermediate between those of the California and Pennsylvania curves; but they resemble the latter more closely than they do the former. 3-The viscosity-temperature curve of the undistilled residuum from the Pennsylvania crude is of the same type as the curves f o r the distillation fractions. &The viscosity-temperature curve f o r a mixture of Pennsylvania residuum and Pennsylvania distillate shows the same characteristics as the curves f o r unblended distillates. Furthermore, the curve for a mixture of Pennsylvania residuum and California distillate apparently averages the characteristics of the curves of the two constituents. 5-The viscosity-temperature curves f o r the Pennsylvania, Wyoming, and California kerosene fractions show the same qualitative relationships as those f o r the lubricating fractions. 6-Viscosity-temperature curves can he represented b y an equation of the general form : 1 Vk= K+At+BP
Values of the three constants. K, A, and B,have been determined for the products from the Pennsylvania and California crudes. An interesting similarity in form may be noted in the curves obtained by plotting values o € these constants against the upper distillation limits of the respective cuts from the two types of crude. 7-Present results have failed to indicate any reliable genera1 method of calculating viscosity-temper~,.‘,urecurves of oils for which less than three experimentally determined points are arailable.
Binoham. “Criticism of Some Receu; Yixo . Inyescignxions.” J Chem. Soc., 103 (19131, 950. Thole, Mussel and Dunstan s c a s i t y I L a x i p a snci Their Ioterpretation,” J . C l i e v ~ . ~ , S ~103 . c . , (1913), 11 Flowers. Viscosity lleasurrmPnt, a n d a h-em T-isrosirneter,” Pnrc. .47n. SOC.Test. Alalerialsz,, 14 (1914 Anonymous,,, hbaolute ter,” EnGineerinR 100 (131.5) 254. Dubrisay. Method of the ‘iiscosity oi LubricotiT$ Oils,” J . SOC.Cheni. I n d . , 36 (1917) Hersrhel. “Determinaticn of Absolute Visco by tho Saybolt Universal and the E n g l ~ Visco5imeters,” r Proc. .4m. S o t . T F C ITInteriaZs, ~. 17 (19171, 11,551, Lidstone. “A Mercurial 1-iscosimeter,” J . Soc. Chem. Ind., 36 (1917).
.
9713 ?,I
_.”(
I
Sheppard. “The Measurement of the Absolute Viscosity of Very Viscous Media,” J . I n d . f l n g . Chem. g (191i) 523. Bingham. “Variable Presswe ?;lethod for the Measurement of Viscosity,” Proc. ilm. S O C .Teef. Mate&k, 18 (1916>,11, 3 i 3 . Ringham and Jackson. “Standard Substances for the Calibration of Viscoaimeters,” Bureau of Staqdards, Scientific Paper, 298. Dunstan and Thole. Relation between Viscosity and the Chemical SHORTBIBLIOGRAPHY Constitution of Lubricating Oils,” Petroleum Res., 38 (1918), 24;: 267. Herschel. “The Standard Saybolt, Cniuersal Viscosimeter, Proc. Am. Girard. Memoirs de l’acodemie d f s Sciences, 1816. Soc. Test. Materials, 18 (1918): 11, 363. Poiseville. Reczieil des Savants Eiratzgers 1842;A n n . chim. p h y s . , 131 7 A-icolardot and Baume. The Dubrisay Method of Examining Lubricat(1843), 50 Couette. “Studies on the T-iscosity of Liquids,” Q ? I L chin,. phyi.., (61 ing, Oils” Analyst, 43 (1918), 226. Nicolardot and Baume. Contribution t o the Study of the Viscosity of 21 (1890), 433. Lubricating Oils,’:,Chimie & i n d u s t v i e , I i1918), 265. Thorpe and Rodger. “Relation between the l’iscosit,y of Liquids and Oelschlager. The Viscosity of Oils, 2. V e r . deut. I n g . , 62 (1918), 422. Their Chemiral Sature, Phil. Trans. -4., 185 (1894), 397. Thorpe. “Viscosity of Pure Liquici?,” Science Progresr, IZ (1918), 583 Thorpe and Rodger. “The Viscosity of Mixtures of Nisrible Liquids,” Faust. “Visrosity Measurements 2. physik. Che,nz., 93 (1919),, i58. J . Chem. Soc., ,?I (18971, 360. Herschel. “Standardimtion of the Saybolt liniversal Viscosimeter,” Beck. Beitrage cur relatiyen Innern Reibung,” 2. p h y s i k . Chem., Bureau of Standards TechnoZgic Paper I I Z (1919). 58 (1907), 425. Herschel. “Viscosity of Gasoline,” Bureau of St>andards, Tcchnolgie Dunstan and RTilson. “Relation between Molecular Weight and 7%Paper I25 (1919);‘ , cosity of a Series of Compounds,” J . Cliem. S o c . , g1,(1907j,90. 1.awaccek. \iscosity and I t s Meesur@ment,” 2. I e r . deut. Ing., 6.3 Dunstan and eo-workers. “Viscosity of Liquid Mixtures,” J . C h e m (1919), 677. Sac., 85 (19041, 817; 87 (1905), ?,l;91 (19071, 83. Stan;y. “Determination of the Absolute 1-isc0sitic.s of Liquids at High Dunstan and co-workers. Relation between Viscosity xiid Cheniicnl Pressure E?zgineering, In8 ,(1919), 520. Constitution,” J . CRem. SOC. 93 (1908),,1918; 95 (190?), 15#?6. Herkhel. “The MacRlichBel Torsion Viscosimeter,” J . I d . E n g . Chem., Findlay. Viscosity of ‘Binary hIixtures a t Their Boiling Points,“ Z. 12 (19201, 817. physik. Chem., (1909), 203. Herschel. “Saybolt Viscosity of Blends,” Bureau of Standard@,TeciiBingham. Viscosity a n d Fluidity,” A m . C h r m . J . . 35 (1906), 195; 40 nica2 Paper 364 (192G). (1908), 277; 43 (19101, 2%. Schwedhelm. “The Viscoaity of Oils and Other Liquids a s a Function Dunstan and Strerens. “The Viscosity of Lubrirating Oils,” J . S O C . of the Temperature,” Chem.-PlQ.. 4 5 (1321), 41. Chem. Ind., 31 (19121, 1063.
9,
The Effect of Chemical Reagents on the Microstructure of Wood’’2 By Allen Abrams RhSE4RCH LABORLTORI OF
APPLlBD CHE\IISTRY,
11 9SSkCHLSETTS
I n connection with research work on paper being carried by this laboratory for the Mead Research Company of Dayton, Ohio, it became desirable to secure an insight into the changes occurring in the structure of wood during various chemical treatments. Consequently, an in.r.estigation was undertaken, as a result of which a method was developed and made use of in carrying out such treatments. A number of reagents have been studied, and while the results here presented must be regarded as preliminaq and incomplete, it is hoped that they may stimulate further work along similar lines by other investigators. The most obrious method f o r studying the effects of reagents on the microstructure of wood would be to treat blocks of the wood with these reagents a t the desired temperatures and pressures for specified lengths o € time. The wood might then be sectionecl and studied either under the microscope o r by means of photomicrographs. Unfortunately, however, the mechanical difficulties of sectioning such wood are so great that, even with the most skilful technique, it is apparently impossible to make sections without altering the anatomical structure. F o r this reason it has been necessary t o develop a new procedure, which consists essentially in first making thin sections of wood and treating these with reagents under the proper conditions. In order that the changes produced by these chemical treatments may be comprehended, it is first desirable to note the three planes in which wood may be sectioned-a cross or tl-awwerse section is one cut perpendicular to the axis of the t r u n k ; a radial section is one cut longitudinally along the radius of the trunk; a tangential section is one cut longitudinally, tangent to the rings of growth and there1 Presented before the Section of Cellulose Chemistry a t the 61st Meeting of the American Chemical Society, Rochester, N. T...4pril 20 to 29. 1921. ‘Published as Contribution X o . 34 from the Research Laboratory of Applied Chemistry, Massachusetts Institute of Technology.
I N B T TLTE OF
T B C H N O L OCi\IBR C~,
DOE, X I i S S I C H L S E T T S
fore, perpenaicular to the radius of the trunk. Figs. 1,Zy 3 are, respectirely, cross, tangential, and radial sections of pine.
CELL STRUCTURE The process of plant growth is essentially that of cell division whereby each cell in the growing region (the ‘‘ cambium ”) is split into two daughter cells. The partition separating these cells is known as the “ middle lamella ” (Fig. 4 ) . As growth continues, other walls are laid clown adjacent to the middle lamella and nearer Ihe hollow interior (‘( lumen ”) of the cell. The complete facts connected with cell growth are extremely complicated. To the paper-maker, however, the important fact is that any chemical process f o r making paper should have for its primary object the separation of individual cells by dissolving out the middle lamella with as little effect as possible on the remainder of the cell wallll. To be sure, another very important action is that of decomposing compound celluloses, such as lignocellulose, and thus producing a pure cellulose. The history of the cell walls and the study of their chemical structure have been the subjects of extensive research, conducted usually, however, by men with botanicaI rather than chemical training. The -results of these investigations3 show that the middle lamella should be regarded as the primary partition wall, serving to bind the tracheids together; but at the same time, i t must be understood that this layer has a complicated history in which it undergoes changes in form, mass and chemical composition. Allen‘ is certain that the middle lamella differs chemically from the later walls. He believes that the first formed cellular wall consists essentially of pectin-like substances 1 C.E. Allen, Bot. Ga
., 33 (19011, 1.
