976
I
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
Vol. 6 , No.
12
ORIGINAL PAPERS
THE CEMENTING VALUE OF BlTUMINOUS BINDERS By LESTERKIRSCHBRAUN Received August 6 , 1914
During t h e past t e n years there has been a great development in t h e methods of valuating paving bitumens. Early a t t e m p t s a t determining these values were directed towards chemical determinations of properties or constituents which were thought t o have a n influence upon t h e quality of materials of this kind. Later developments have indicated t h a t , with few exceptions, t h e valuable properties of paving materials are included more directly in their physical characteristics. T h e present s t a t u s of t h e determination of paving values includes physical tests as t h e more imp o r t a n t ones, leaving a few chemical tests directed towards determining t h e permanency of these products, a n d t h e care with which t h e y have been prepared. It will be generally agreed t h a t a most i m p o r t a n t , if not primary physical property, which paving bitumens must possess is t h a t of cementitiousness. For a long time efforts have been directed towards devising a means of measuring cementitiousness of bituminous binders, b u t s o far nothing has been offered in this direction which has met with general approval. Among t h e tests commonly employed, t h a t of ductility has been assumed t o measure, or t o be, in a general way, a n indication of cementitiousness. While possibly there is some relationship between ductility a n d t h e cementing value of a given t y p e of bitumen, it will be a d m i t t e d t h a t no mathematically direct relationship, if a n y , exists a t all. So far, t h e n , a s our present means of determining cementing values are concerned, this most important feature must be arrived a t b y indirect interpretation of d a t a which afford no exact means of determining such values. T h e writer has for t h e past four or five years directed his attention t o t h e development of a method for directly a n d mathematically expressing t h e cementing value of plastic binders. Various means have been employed in t h e effort t o determine this factor, b u t all have been discarded a s i t developed t h a t t h e y fell short of measuring t h e properties sought for. T h e writer has, a t various times, devised bending or shearing tests; has investigated t h e tensile strength of briquettes of binders a n d mixture of same with mineral p a r ticles; has investigated t h e s t r e n g t h of briquettes joined together by films of binder, a n d has experimented with methods of determining adhesiveness a n d cohesiveness. These various a t t e m p t s , while in some cases giving valuable information, have failed, either through t h e inability t o devise a means of obtaining concorda n t results, or for t h e reason t h a t t h e properties actually measured in these efforts did not directly represent t h e cementing value property sought for. For example, in making tests of briquettes of nonbituminous material stuck together with a film of binder, it is not only difficult, if not impossible, to obtain concordant results on account of variations in
thickness of film, .etc., b u t results obtained do not measure binding value, b u t measure cohesiveness. When tests of this kind are made, t h e briquettes fract u r e with a cleavage, leaving a film or p a r t of a film on either end of t h e briquette. T h e property so recorded is cohesiveness or t h e ability of t h e material t o stick t o itself. Any method which measures strain endured by t h e film of asphalt in detaching itself without cleavage or fracture from a foreign surface measures a d h e s i v e n e s s . I n either case, t h e operation fails t o determine t h e cementing value or t h e ability of t h e material t o bind particles together under t h e conditions of service. If we analyze t h e results obtained upon a number of materials through a series of tests for tensile strength (meaning in this case t h e maximum strain endured in fracturing a briquette of bitumen) we shall find t h a t two different materials may sustain t h e same maximum stress a n d indicate t h e same cohesiveness, b u t t h a t one material will sustain this strain for b u t a short time before fracture, while another material will not only sustain t h e same maximum strain, b u t will endure it during a longer period or through a much longer distance of elongation. For example, a given t y p e of asphalt cement a t a certain penetration will, during application of strain, withstand a maximum of s a y , three units before fracturing. Another asphalt of certain penetration will withstand t h e same strain. I n t h e former case, however, after this maximum is reached, a n appreciable strain can be sustained for many times t h e elongation t h a t m a y be sustained with t h e latter material. Cohesiveness results in such cases would indicate equalitx, b u t a s a practical consideration, t h e l a t t e r material might be entirely unfit for paving purposes, a n d even from superficial observation might not indicate nearly t h e cementing qualities of t h e former. Such results, t h e n , not only become misleading a n d contrary t o practical observations, b u t fail entirely t o give us a n indication of t h e property sought for. T h e adhesiveness of bituminous binders m a y be determined b y means of a suitable apparatus. While t h e adhesiveness is a m a t t e r of importance in t h e effectiveness of bituminous application t o cold road metal by pouring processes, nevertheless, in considering hot mechanical mixtures, t h e adhesiveness appears t o play little p a r t in holding together t h e mineral aggregate. When asphalt pavement cracks or fractures or displaces, a n examination of t h e points of fracture discloses t h a t t h e films of bitumen coating t h e particles have fractured or cleaved, a n d not t h a t t h e bitumen has pulled off t h e mineral particles through lack of adhesiveness. I n other words, t h e adhesiveness is always greater t h a n t h e cohesiveness a n d t h e binding value. No mechanically measured results of adhesiveness need therefore be considered as a factor in this discussion, although t h e adhesiveness is a factor in considering pavements built by penetration methods.
Dec., 1914
T H E J O U R N A L O F I N D C S T R I A L A N D ENGILVEERING C H E M I S T R Y
977 * t o briquettes a t t h e lower end of t h e runway with t h e box filled with water. At t h e lower end of t h e runway is permanently fastened a small boss or projection t o which one end of the briquette is attached, t h e other
W h a t , then, represents t h e cementing value of a material of this kind? Cementing value must mean t h e ability of t h e bitumen t o bind or hold together against rupture, particles of mineral matter which i t coats ob covers. Upon analyzing t h e conditions applying, i t becomes evident t h a t t o break a p a r t t h e bond between a n y two mineral particles held together b y a coating of plastic bitumen, a certain tension must be applied for a certain time or space of action. The application of a tension over a certain distance infers t h e necessity for t h e film of binder t o elongate at t h e point of contact of t h e t w o particles. T h e actual amount of elongation m a y be extremely small, b u t t h e relative a m o u n t in proportion t o t h e thickness of t h e coating m a y be very great. T h a t elongation does necessarily t a k e place must be admitted from practical experience, which has t a u g h t us t h a t a certain degree of plasticity or softness of our binders is essential t o prevent cracking of pavements. T h e effect of advancing t h e softness of t h e binders is t o favor their ability t o yield a n d elongate under strains. Cracking is therefore minimized by superior ability of a soft binder t o elongate over a hard binder of t h e same kind. If this FIG 1 ability t o elongate were not essential, a n d if i t actually did not t a k e place in t h e binder between t h e particles end being attached t o t h e rod connected with the of aggregate, i t is apparent t h a t t h e hardest binders dynamometer. The position of t h e bridge gives readwould be t h e best, as t h e y would be capable of sus- ings of elongation of t h e briquette, regardless of t h e taining greater strains t h a n t h e softer binders. It will movement of t h e carriage a n d dynamometer. The be seen therefore t h a t t h e ability t o elongate is a n form of briquette adopted is t h a t commonly known as essential feature, a n d must operate, otherwise our con- t h e "Dow" moulds which are used for making ductility clusions would lead us in t h e direction contrary t o tests. These moulds have a miminum cross section actual experience. of I sq. cm. Referring again t o our consideration of two particles I n working o u t t h e method of manipulation, i t was of aggregate bound together, i t is necessary, in order of course necessary t o determine a constant temperat o produce fracture or t o disrupt t h e bond, t o apply t u r e a t which t h e test should be made. The tema certain tension over a necessary distance. The prodperature selected was j " C. or 41" F. This was seuct of these factors is work done. T h e binding value lected for several reasons. I n t h e first place it was of a plastic binder is t h e n limited t o t h e a m o u n t of tension i t can sustain over a given distance, or is directly proportionate t o t h e work done in producing fracture or failure of a given unit of material. An a p p a r a t u s was accordingly devised for the.purpose of recording t h e factors above mentioned, namely, strain applied over distance. The apparatus in its final form is shown in Figs. I a n d 2 . It consists of a rectangular box insulated a n d lined with galvanized iron or copper. An inclined plane or runway is a t tached t o t h e box, t h e lower end of which reaches t h e b o t t o m of t h e box, a n d t h e other end projects a n equal distance outside. This runway carries a carriage through which passes a screw actuated b y a set of gears adjustable t o t h e desired speed. Upon t h e carriage is maintained a dynamometer, t o t h e end of which is connected a flat brass rod, which extends along t h e runway t o within a few centimeters of t h e end. a n d of t h e b o t t o m of t h e box. This rod is graduated in metric units, a n d passes under a n adjustable bridge a t which FIG. 2 t h e readings are recorded during travel of t h e rod. T h e runway is edged with guides of sufficit?nt width necessary t o obtain sufficiently large readings upon f o r t h e free passage of t h e briquettes. T h e arrange- t h e dynamometer. At normal temperatures, bitumiment a n d position of t h e carriage makes i t possible nous materials of this kind are ordinarily capable of t o apply a tension through t h e rod in a straight line withstanding b u t very little strain. Secondly, t h e
978
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
I
H
depending on t h e ability of various binders t o yield or elongate without fracture. It is apparent t h a t under t h e same conditions of volume change in a pavement a soft binder would yield a n d conform more easily t o t h e necessary change, thereby allowing much less strain t o be set up t h a n with a hard binder. Various trials were made a t different speeds in order toodetermine what speed was most adapted t o securing concordant results, a n d distinguishing t o t h e greatest degree t h e differences between various materials. Speeds between I t o I O cm. per minute were tried upon t h e same materials. It was found t h a t t h e higher speeds had a tendency t o effect too sudden application of strain, and, with t h e harder materials, t o produce fracture in the corner of t h e briquettes rather t h a n a t t h e point of minimum cross section. Again, a t t h e faster speeds, it became difficult t o take t h e readings near t h e time of fracture owing t o t h e rapid travel of t h e rod and the pointer of t h e dynamometer. After various trials i t was found t h a t a speed of travel of t h e carriage and dynamometer a t t h e rate of I cm. per minute provided t o t h e greatest extent €or t h e factors indicated above. It will be understood
12
t h a t in applying strain t o t h e briquette by means of a dynamometer traveling a t a uniform speed, t h e briquette has a t first slight tendency t o elongate under application of strain, until a maximum is reached, when t h e rate of elongation of t h e briquette exceeds t h e rate of travel of t h e carriage and dynamometer, a t which point the dynamometer begins t o register values below t h e maximum attained, until finally t h e briquette is either fractured a t a reduced cross section through elongation of t h e material, or t h e dynamometer returns t o zero without fracture of t h e briquette. An essential difference from t h e method of determining ductility is t h a t during t h e entire period of elongation, t h e material is under substantial strain. Ductility or elongation values t h a t include t h e distance traveled by a n extremely fine thread or filament of bitumen are misleading and immaterial. I n making t h e test, as finally adopted, t h e briquettes are prepared in t h e usual manner as for ductility test, and are placed in t h e test box a t 5' C. They are held for about three-quarters of a n hour a t this temperature before making t h e test. When ready, one end of
strains which produce fracture a n d cracking in a pavement in actual service are greatest during cold weather. Again, a low temperature being necessary, t h e temperat u r e of 5' C., was selected as being easy t o maintain constantly with ice and water. On account of t h e plastic nature of t h e material operated upon, a n d its tendency t o elongate under strain, it was necessary t o determine upon a uniform means of application of load. It was evidently impossible t o set a uniformly increasing load for t h e reason t h a t , as t h e material elongated, and its cross section became smaller, it would be necessary t o rapidly accelerate t h e rate of elongation or travel of t h e dynamometer towards t h e end of t h e operation, in order t o increase t h e strain. This would become impracticable as a matter of manipulation, a n d would be subject t o large variation through personal equation a n d through inability t o t a k e t h e required readings sufficiently rapidly. It was found necessary after many trials t o adopt a .uniform rate of travel of carriage a n d d y n a m o m e t e r . This is in accord with practical conditions, inasmuch as a n y strain set up in a pavement would be induced a t a variable rate of application,
Fie.3.
Vol. 6 , No.
I
RETURN
9
t h e briquette is hitched t o t h e fixed post a t t h e bottom of t h e box on t h e runway, a n d t h e dynamometer b r o u g h b i n t o position so t h a t t h e end of t h e rod attached t o same may be fastened t o t h e other end of t h e briquette. The apparatus is driven mechanically, a n d t h e carriage is started by closing t h e split-nut which brings it into contact with t h e screw. When t h e end of t h e dynamometer begins t o move over its zero mark, t h e bridge over t h e rod is adjusted t o t h e zero mark on t h e rod. Carriage a n d dynamometer continue t o travel a t t h e rate of I cm. per minute, a n d readings of t h e dynamometer are taken a t every half centimeter of elongation as shown by the rod. This is continued until t h e briquette fractures or elongates through its maximum back t o zero strain. The dynamometer carries a maximum pointer a n d is graduated in tenths of a kilogram, a n d t h e zero mark is taken as one-tenth kilogram, which includes t h e weight of t h e briquette and t h e frictional resistance of parts. The position of t h e rod and boss are such as t o lift t h e briquette slightly off t h e runway when t h e strain is applied, so t h a t no friction results on this account during t h e application of strain. Fig. 3 illustrates
De;., 1914
T H E J O C R - V A L O F I S D C S T R I A L A.VD EAITGIiVEERIXGC H E M I S T R Y
t h e position of t h e briquette a n d dynamometer a t various stages of test. T h e results of t h e d a t a obtained in these determinations may be graphically recorded. The distances elongated are recorded as abscissae, a n d t h e strain a s ordinates. Fig. 4 gives a typical example of results obtained upon t h e same material of different consistencies under this method of test. T h e areas en7 6
5 2 4
L 53
I
Cms. Elo ng a f i o n closed by these curves represent t h e product of t h e strain applied a n d t h e distance of its application, or t h e work done. T h e unit of value is expressed as kilogrammeters. This area may be obtained f r o m t h e graphical plot of t h e result, or may be secured much more quickly, a n d sufficiently accurately, by addition of t h e ordinates. T h e readings are t a k e n f o r every 0 . j cm. or 0.005 of a meter. Every unit of ordinate therefore represents 0.005 kilogram meters, a n d t h e s u m of t h e ordinates multiplied b y this factor gives t h e area or t h e work done in kilogram meters. Expressed mathematically, t h e formula f o r calculating t h e result is 2 Y X 0.005. This is sufficiently accurate for practical purposes. Referring again t o Fig. 4,i t will be noted t h a t t h e same material a t different consistencies requires widely varying a m o u n t s of work for failure. With t h e harder materials, a comparatively sharp curve is obtained, which reaches a certain maximum, a n d t h e n becomes less a s t h e material elongates faster t h a n t h e carriage travels. When t h e cross section is reduced t o a point a t which i t is unable t o withstand t h e strain induced, i t breaks. With t h e materials of t h e softer consistency J Cms. Elongation there is a tendency of t h e curve t o flatten a n d t o become larger- as t h e material is more plastic. Aconsistency is finally reached a t which t h e material is able t o progress t h r o u g h " t 0 its maximum a n d back t o zero (or 0 .I kg.) without fracture. All tests were obtained in duplicate, a n d t h e results indicate a h i t of accuracy of 0 . 0 2 kg. meter f r o m a n average, o n t h e highest results, t o a much closer agreement upon t h e lower values. Exception t o this accuracy
979
is noted upon t h e , h a r d brittle materials, which are unable t o elongate uniformly without setting up internal strain, due t o t h e corners of t h e briquettes. Having established a satisfactory method of determining t h e value sought for, it was desired t o outline a series of investigations covering t h e following points: I-Survey of t h e characteristics of commercial products with special reference t o determining t h e degree of differentiation possible with t h e various materials. 2-The possibility of valuating fluxes by this method. 3-Test of t h e commercial products, t h e chemical characteristics of which indicate inferior preparation. 4-The determination of a possible effect upon cementing value induced by improper preparation upon a series of products made known under known conditions. j-The determination of standards of value necessary for practical application. I n order t o t a k e complete information of t h e characteristics of commercial products, it is necessary t o collect such d a t a as would represent a wide range of consistency for each material. The refined asphalt was accordingly used a n d fluxed with t h e kind of flux ordinarily used in practice. Determinations were made upon each material a t a number of different consistencies, a n d these cementing values were graphically expressed as abscissae a n d t h e consistencies a's ordinates. T h e various materials examined were grouped into three classes according
C e m e n t i n g Vohe at 5°C. ( K i l o g r a m m e t e r s ) t o their origin. Record was made of the cementing value a s determined b y t h e d a t a presented before, together with t h e elongation a n d t h e maximum strain sustained. These d a t a are given ' in Table I (page 981). Group I contains asphalt cements prepared from solid n a t u r a l bitumens, a n d Groups 2 a n d 3 include products made f r o m natural liquid bitumens.
ot5'C.
980
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
Results obtained are given graphically in Figs. 5 , It was found desirable t o plot elongation in order t o complete t h e means of interpretation of t h e results. T h e graphical presentation of this d a t a
6 a n d 7.
Cms Elongotion atS°C indicates a relationship between consistency a n d cementing value on one hand, a n d consistency a n d elongation on t h e other. By referring t o plates, i t is possible t o compare t h e values of any of t h e materials a t t h e same penetration. It will be noted t h a t t h e ce' menting value increases generally as t h e consistency becomes harder, b u t when a certain hardness is reached, there appears t o be a breaking off in the values determined in this way, owing t o t h e brittleness a n d t h e inability of t h e material t o yield without internal
Vol. 6, No.
12
pull. Such procedure, however, would minimize t h e differences between these materials a t those consistencies a t which such differences are of greatest importance.
Cementing Value at5'C. (Kilogrammeters)
T h e effect of t h e presence of mineral matter upon t h e binding value and elongation is illustrated in case of material A - P , which was examined with t h e mineral matter in, a n d with t h e mineral matter removed. The presence of mineral matter in these materials lowers t h e result obtained for cementing value, although t h e cohesiveness or maximum strain recorded is greater for t h e same consistency with t h e mineral matter in. The apparent lowering of cementing value is due t o t h e lesser elongation produced under strain, a n d t o t h e fact t h a t t h e bitumen when examined with its contained mineral matter is actually softer t h a n indicated by penetration of t h e whole. T h e increase of binding value with hardness, a n d t h e increase of elongation with softness is quite in accord with practical observations. For example, if a material be laid under light traffic a t , say, 6 5 penetration, experience has shown us t h a t we must reduce t h e penetration under heavy traffic, in order t h a t the mineral particles may be bound together more solidly t o resist impact and displacement. On t h e other hand, when we lay a pavement a t 6 5 penetration for light traffic, we require greater ability Cementing Voiue at5'C. (Kilogrammeter) Cms.Elongation at5"C. in t h e material t o elongate in order strains induced by t h e form of briquettes. This is a t o resist cracking, t h a n we do under heavy traffic. factor dependent upon t h e speed of t h e pull, a n d in This is t r u e because, as is well known, heavy traffic all likelihood there would be a continuation of t h e tends t o knead t h e pavement and relieve t h e stresses curves beyond the present points a t a slower r a t e of set up by changes of temperature. It will be
Dec., 1914
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 TABLE I-GROUP FIG. 5 Asphalt
“
C ” a n d Flux 3
MaxPen. a t i m u m 770 F strain 35 5.7- 5.7 52 3.2-3.0 64 2.3- 2 . 2 78 1.4-1.1 98 1.1- 1.0
Average 5.7 3.1 2.25 1.35 1.05
5.1- 5 . 0 3.8-3.6 2.6-2.4 1.6-1.5 4.3- 4 . 3 6.4- 7.2 7.3.
5.05 3.7 2.5 1.55 4.33 6.8 7.3
Asphalt “ B ” and Flux 2 0.269-0.308 0.288 0,278-0.233 0.255 0,154-0.168 0.161 0,114-0.106 0.111 0,259-0.257 0,258 ......... 0,017 0.087, 0,087
6.4 4.3 2.85 2.3 2.C5 4.5
Asphalt “ A ” a n d Flux 2 0.194-0.202 0.198 0.215-0.186 0.201 0.119-0.152 0.136 0.111-0.125 0.118 0.107-0.109 0.108 0.197-0.172 0.185
41 51 64 80 45 11 22 37
50
59 72 82 46
..
6.34.22.82.32.04.5-
6.5 4.4 2.9 2.3 2.1 4.5
Cementing Values 0,199-0.198 0,134-0.126 0.101-0.101 0.077-0.063 0.060-0.063
Averaae 0,199 0,130 0.101 0,070 0.062
... .
6.5 8.75 10.0 12.0 7.1 0.22 1.2
6.5-6.5 10.0-10.5 13.0-15.0
6.5 10.25 14.0
Asphalt “ D ” and Flux 3 0,324-0.379 0,352 0.191-0.185 0.188 0,113 0,105-0.121
6.5- 7 . 0 9.0- 9.0 11.5-12.0
6.75 9.0 11.75
3.2- 3 . 4 5.5- 6 . 4 8.0- 7 . 7 15 .0-11 .O 9.0- 9 . 5 18.0-19.0 18.0-18.5 16.0-15.5
3.3 5.95 7.85 13.0 9.25 18.5 18.25 15.75
Asphalt “ F ” a n d Flux 1 0,074-0.149 0.097 7 . 4 0.415-0.382-0.441 0.413 7.2 0.463-0.394 0.428 5.1 0.354-0.344 0.347 0.2693.5 0.269 2.8 0.212- . . . . . 0.212 2.45 0.206-0.195 0.201 1.6 0,133-0.155 0.144 1.25 0.111-0.123 0.117
0.9- 1 . 7 6.5-5.7-6.1 7.0- 6.5 8.0-10.5 12 . O - . . . . 13. 0 - . . . . 14.5-15.5 16.0-17.0 17.0-18.0
1.5 6.1 6.75 9.25 12.0 13.0 15.0 16.5 17.5
6.0-6.1 2.4- 2 . 4 1.6- 1 . 6
34 55 81
5.8- 6 . 5 2.7- 2 . 6 1.3- 1 . 5
6.15 2.65 1.4
GROUP2-FIG. Asphalt
29 5.9- 7.2 34 7.3-7.2-7.8 36 7.5- 6 . 9 50 5.1- . . . 61 3.5- . . . 76 2.8- . . . 77 2.5- 2 . 4 109 1.6- 1 . 6 122 1.1- 1 . 4
6.0- 7.0 9.5-8.0 9.0-11.0 12.0-12.0 7 . 0 - 7.2 0.2- 0 . 2 5 1.2- 1 . 2
3.5 5.25 5.75 7.0
39 69 88
13.7-13.6 9.9- 9 . 6 6.6- 6 . 5 4.0- 4.3 5.0- 4 . 8 2.75-2.85 2.9- 2 . 8 1.2- 1 . 3
Averane 4.0 5.0 5.75 8.0 8.75
3.5 4.5 6.5 7.5 7.5 4.3
Asphalt “A-2’’ Purified 6.05 0.338-0.348 0.343 0.183-0.164 0.174 2.4 0.137-0.123 0.130 1.6
30 43 54 72 63 91 83 134
Elongation Cms. 4.0- 4.0 5.0-5.0 5.5- 6 . 0 7.5-8.5 8.5- 9 . 0
3.56.05.06.58.54.8-
8.0 4.6
6
“ E ” a n d Flux 1
13.65 9.75 6.5 4.15 4.9 2.62 2.85 1.25
0.382 0.486 0.450 0.353 0.395 0.229 0.257 0.101
6.8
.....
133
3.92.82.31.71.81.20.5-
4.0 3.05 2.35 1.8 1.8 1.2 0.5
Asphalt “ H ” a n d Flux 3.95 0.216-0.225 0.169-0.163 2.93 2.33 0.138-0.135 0.102-0.114 1.75 1.7 0.103-0.095 0.0938-0.0878 1.2 0.5 0.0517-0.0518
0.220 0.166 0.137 0.108 0.099 0.091 0.052
6.5- 6 . 5 6.5- 7 . 5 8.0- 8.0 9.0- 9 . 5 9.5- 9 . 5 11.5-1 1 . 5 13.5-14.0
41 53 61 69 80 136
3.82182.21.81.40.3-
4.1 2.9 2.1 1.7 1.5 0.3
Asphalt “ G ” a n d Flux “ G ” 3.95 0.221-0.229 0.225 2.85 0.165-0.158 0.162 2.15 0.137-0.134 0.136 1.75 0.124-0.105 0.115 1.45 0.093-0.103 0.098 0.3 0.036-0.033 0.035
7.0- 6 . 5 7.5- 7 . 2 9.5- 9 . 5 10.5-1 1.o 11.0-12.0 14.0-14.5
3.25 2.0
Asuhalt “I,” 0.199- . . . . . 0.146-0.122 0.077-. .
0.199 0.134 0.077
4.2- . 4.5- 5 . 0 6.0- , . .
3.9 2.4 1.55 1.25 0.95
Asphalt “ K ” 0,150-0.173 0.129-0.121 0.077-0.082 0.085-0.082 0.059-0.054
0.162 0.125 0.080 0.084 0.057
4.5- 5 . 0 7.5- 7 . 5 9.0- 9 . 0 10.5-10.5 9.5-10.5
3.9 2.2 1.4 0.9
Asphalt “ M ” a n d Flux 1 5.75 0.263-0.284 0,273 3.85 0.213-0.202 0.208 2.4 0.167-0.141 0.154 L35 0.112-0.113 0.112 0.9 0.088-0.086 0.087
6.2- 6 . 8 8.5- 8 . 5
“N” 0.104 0.076 0.065 0.059 0.047
2.52.52.53.03.0-
2.5 2.5 2.5 3.0 3.5
2.5 2.5 2.65 3.0
P” 0.059 0.023 0.016 0.047
1.52.42.52.0-
1.5 2.5 2.5 2.0
1.5 2.5 2.5 2.0
42 51 60 69 77
85
38 51 69 51 71 94 109 137
5.5- . . . 3.4- 3 . 1 2.0- . . . 3.82.41.51.31.0-
4.0 2.4 1.6 1.2 0.9
5.5
.. .
GROUP3-FIG. 31 40 55 70 97
5.5- 6.0
3.82.61.30.9-
“
H”
..
;:;
g:;5 9.5 11.5 13.75 6.75 7.s5
1;::5 11.5 14.25 4.2
z:i5
;:z5 10.0
7
39 52 57 72 94
4.73.32.82.21.8-
4.6 3.4 2.8 2.3 1.7
Asphalt “N” a n d Flux 4.65 0.105-0.103 3.35 0.075-0.077 3.8 0.068-0.062 2.25 0.058-0.060 1.75 0.045-0.049
37 100 118 65
4.31.20.92.5-
4.3 1.2 1.0 2.7
Asphalt “ P ” a n d Flux 4.3 0.059-0.059 1.2 0.025-0.023 0.9 0.015-0.016 2.6 0.045-0.050
5.5- 5 . 3 12.5-13.0 14.5-15.0
2:;
18.5 2,25 14.75
3.25
“
981
noted t h a t t h e cementing value is not directly proportionate t o t h e elongation, a n d t h a t in some cases t h e binding value of t w o different materials a t t h e same penetration is very similar, whereas their ability t o elongate a t t h e same penetration is appreciably different. Likewise, particularly a t t h e lower penetrations, some materials elongate t o practically t h e same extent, yet their cementing values are widely different. It is evident from t h e s t u d y of this graphical d a t a , t h a t i t is not only necessary t o consider t h e elongation as a factor in t h e cementing value, b u t i t is also necessary t o consider i t a p a r t from t h e cementing value. If this were not so, our pavements would be less liable t o crack when laid a t t h e harder consistencies. It appears, therefore, t h a t while t h e cementing value indicates t h e ability of t h e material t o bind aggregate together, a minimum ability t o elongate is also necessary t o avoid cracking. This is well illustrated with material “ N ” which, when laid, has sufficient cementing value t o bind t h e mineral particles together for light traffic, but which has a very pronounced tendency t o crack. I t s binding ability a t no consistency would be sufficient for conditions of heavy traffic. I t will further be noticed t h a t t h e general tendency of t h e elongation curves is t o straighten upward beyond a certain consistency, a n d in some cases, even t o retreat. This indicates t h a t beyond a certain a m o u n t , fluxing not only results in less binding value, b u t results in reduced ability t o elongate, a n d t h a t nothing would be gained, so f a r as plasticity is concerned, b y further fluxing. I n t h e case of materials “ P ” a n d “N,”there would hardly be a n y consistency a t which t h e material might be handled which would overcome t h e tendency t o crack. This i n a practical way has been actually found to be t h e case. Attention m a y be called, at this point, t o t h e fact t h a t various materials showing t h e same maximum strain or cohesiveness possess widely differe n t cementing values. F o r example, asphalt “G” at 53 penetration shows 2 . 8 j kg. maximum strain. T h e same value for cohesiveness is shown b y asphalt “ E ” at 83 penetration, yet t h e cementing value of this latter greatly exceeds t h a t of asphalt “G.” This is a typical example which indicates t h e abilitv of __ this test t o differentiate beyond a n y ordinary tension tests. Table I1 is given showing t h e results of investigations made in t h e effort t o determine t h e possibility of valuating various fluxes. I n this series of tests t h e same asphalt was used throughout, a n d was fluxed with different fluxes, t h e cementing values a n d other d a t a being determined as indicated before. These are given in T a b l e 11. Graphical results are shown in Fig. 8. It will be noted t h a t t h e cementing values of these materials a r e progressively greater a s t h e fluxes t e n d tow-ards asphaltic base, with this exception t h a t t h e paraffine combination shows slightly better values t h a n t h e light-semi-asphaltic. This is due t o t h e fact t h a t much less paraffine base fllax was used t o produce t h e consistencies noted t h a n i n t h e case of t h e light semi-asphaltic flux. When, however, t h e elongation
982
T H E JOU-RNAL OF I i Y D r 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
curves are examined, a progressive increase without exception is noted in ability t o elongate as t h e fluxes used partake more of a n asphaltic character. I t will TABLE11-FLCX TESTS-FIG. 8 R . A. and Paraffine Flux Elongation Cementing AverPen. at Maximum Aver77OF. age age Cms. strain Values 2.5- 2 . 3 0.082-0.083 0,083 45 3.9-4.0 3.95 0,130 4.6-5.0 0.115-0.144 53 3.0-3.1 3.05 0.074 5.5- 5 . 5 1.7 0.072-0.075 88 1.7-1.7 0.094 5.5- 5 . 5 2.1 0.093-0.095 76 2.1-2.1 R . A. and Light Semi-Asphaltic Flux 39 3.8-3.9 3.85 0.092-0.114 0.103 3.0- 3 . 7 0.096 8.2- 8.0 68 1.5-1.5 1.5 0.096-0.096 83 1.2-1.1 1.15 0.077-0.067 0.072 8.5- 8 . 5 0.165 7.4- 5.5 0,188-0.141 44 3.G3.0 3.0 R. A. and Heavy Semi-Asphaltic Flux 34 5.8-6.5 6.15 0,325-0.379 0.352 6.5- 7 . 0 55 2.7-2.6 2.65 0.191-0.185 0.188 9.0-9.0 81 1.3-1.5 1.4 0.105-0.121 0.113 11.5-12.0 R . A. and Asphaltic Flux 130 0 . 4 - . . . 0.4 0.043- . . . . . 0.043 14.0- . . . . 79 1.4-1.3 1.35 0,119-0.125 0,122 13.0-14.5 59 2.7-2.6 2.65 0,246-0.198 0.222 13.0-11.0 45 4.1-4.2 4.15 0.316-0.327 0.322 9.0- 8.8 35 6.0-5.8 5.9 0.282-0.246 0.264 5.0- 5.5 51 3.6-3.6 3.6 0,254-0.249 0.252 8.5- 8 . 5 33 .,.. . .,. , .. . . . . . . 0.210 ,..,.
Aver. age 2.4 5.0 5.5 5.5 3.35 8.1 8.5 6.45 6.75
1:::s 14.0 13 75 12.0 8 9 5.25 8.5 4.4
be noted t h a t t h e ability. t o elongate would become even a more sensitive indication of t h e character of t h e flux itself t h a n t h e cementing value. T h e practical necessity for considering this ability t o elongate is a p t l y illustrated here. This refined asphalt used with paraffine fluxes has proven unsatisfactory i n practice. Under light traffic, pavements laid with this have shown sufficient binding qualities t o hold t h e mineral particles together, b u t t h e tendency t o crack is very decided. On t h e other hand, this refined
asphalt with light semi-asphaltic flux has shown, i n a practical way, t h a t t h e cracking tendency is greatly reduced in laying a t a sufficiently high penetration, a n d it has been much more satisfactory t h a n t h e previous combination, even though i t s cementing valuk is not greater. T h e material with t h e asphaltic flux a n d t h e heavy semi-asphaltic flux have been eminently successful under service conditions. It is evident from t h e above t h a t t h e method described furnishes a means of determining t h e value of various fluxes a n d their suitability in asphalt combinations. During t h e course of commercial practice, various
Yo\.6 . No.1 2
materials have been encountered which gave indication, from analysis, of having been poorly preparedso much as t o warrant their rejection for use. A f e w of these samples were on file, a n d were available f o r cementing value tests. These materials were generally refined asphalts of from 30 t o 40 penetration, reduced from liquid albumen. T h e y were subjected t o t h e cementing value test a t t h e consistency a t which t h e y were obtained, a n d t h e n fluxed with t h e same asphaltic flux t o a consistency corresponding t o paving penetration, i n order t o afford comparison with normal products of t h e same type. T h e determination of cementing value a n d elongation are given in Table 111, together with t h e principal analytical characteristics. These results are graphically shown in Fig. 9, a n d are given i n connection with a curve showing t h e quality of a n average acceptable commercial sample of material prepared from similar crude. I t is evident from t h e s t u d y of these results t h a t t h e chemical indications of inferior preparation shown b y t h e high fixed carbon a n d per cent insoluble in carbon tetrachloride are confirmed by actual lowering of t h e cementing qualities, a n d elongation values. I n order t o confirm t h e indications of lowering of cementing qualities with accompanying evidences of poor preparation, products were made in t h e laborat o r y from t h e same asphaltic base crude oil under known conditions. These were made in a small still, a n d were prepared a t temperatures of j25', 700' a n d 8 2 5 ' t o 850' F. at t h e end of t h e distillation.. Four runs were made, t w o of which were conducted at t h e last named temperature. I n t h e first three runs, t h e materials were not pushed t o a hard penetration, b u t of t h e last two r u n s at 850' F., one was pushed t o a very h a r d consistency. I n order t o obtain a n indication as t o t h e comparative extent of cracking which took place, composite samples of t h e digtillate were tested for gravity. I n t h e first t w o runs, samples were t a k e n from t h e still a t various points of consistency indicated hereafter. I n t h e fourth run, t h e material, as stated, was pushed t o a very hard penetration a n d fluxed back with well prepared asphaltic base flux from t h e same crude. All of these samples were tested for cementing value, a n d those samples most nearly corresponding in consistency or penetration were examined for their main chemical characteristics. These results are given in Table I V (page 9S4) a n d are shown graphically on Fig. I O . On account of t h e insufficient number of points of elongation secured, no a t t e m p t was made t o draw a smooth curve covering these points. Inspection of t h e above tabulated results shows t h a t very little change or decomposition has occurred under 700' F., i n running down t o t h e consistencies noted. T h e material prepared a t 850' F. gives evidences of substantial decomposition. This is indicated by a lighter
Dec., 1914
T H E J O U R N A L OF I N D I ' S T R I A L A N D ENGINEERING C H E M I S T R Y
gravity of distillate a n d falling off in solubility in carbon disulfide. T h e gravity of distillate in runs T-3 a n d T-4 hard are, of course, not strictly comparable owing t o t h e different consistencies t o which these were pushed. As a consequence of t h e decomposition t h a t has occurred, there follows a substantial increase of fixed carbon over n o r m a l , a n d t h e conversion of p a r t of t h e bitumen t o a form insoluble i n carbon tetrachloride. T h e differences between products obtained in t h e t w o T-4 runs show t h a t not only is temperature a factor in decomposition, b u t also t h e degree t o which t h e operation is pushed, or t h e degree of "concentration" resulting. This latter
t o note t h e persistency with which material prepared from this crude retains its cementing value even when badly decomposed; a n d although showing evidences in this direction which would cause its rejection under s t a n d a r d specifications, i t s cementing value has not been reduced below a n acceptable minimum. A s t o whether or n o t materials so prepared, a n d possessing t h e characteristics of t h e last named product would r e t a i n t h e cementing value indicated, is a matter requiring investigation before a n y conclusion can be drawn as t o t h e effect of t h e features noted. I t will be seen, in referring t o commercial products prepared from crude of this ,character, t h a t t h e loss of cementing
TABLE 111-FIG. 7199
h-0.
tlxd.
&
Pent. a t i i 0 F 32 Duct. a t 7 7 O F . . . . . . . . . . . . . . . . . . . 60 cm. 5 hr. loss . . . . . . . . . . . . . . . . . . . . . . . 0.4 % Per cent hardening., . , , , . , , , , , . 25.0 CSz soluble.. . . . . . . . . . . . . . . . . . . . . 9 9 . 7 Mineral. . . . . . . . . . . . . . . . . . . . . . . . 0.2 Difference . . . . . . . . . . . . . . . . . . . . . . 0.1 CClr soluble.. . . . . . . . . . . . . . . . . . . 9 7 . 0 Fixed carbon . . . . . . . . . . . . . . . . . . . 1 6 . 6 Cementing value a t 5' C . . . . . . . . . . 0 . 2 8 9 Elongation a t 5 C.. . . . . . . . . . . . . 5.0
63
.
0.141 5.5
7222
tlxd.
b
32 4 . 5 cm. 9.6 % 23.0 98.9 0.2 0.9 90.0 18.6 0.085 1.2
60
Cms €/ongafion o f 5 2
M u n i c i # d Engineering, 36, 349
46
0.096 2.9
0xd. 68
0 . 8 75 36.9 99.0 0.2 0.8 94.4 17.7 0.126 2.5
7260
flxd.
A -
40 i.0 cm.
65
99.0 95.9 0.082 3.0
0.136 2.5
0.lli 3.75
-
7324
41 1 , 2 per cent 65 cm. 43.9 99.7 0.2 0.1 97.9 16.9 0.258 6.5
tlxd
71
0 141 9.i5
value is much greatef with slighter evidences of decomposition t h a n is obtained f r o m t h e products prepared in t h e laboratory. h-0 explanation of these variations is offered, aside from t h e indications given b y these commercial products of having been prepared by other methods, t h a n those involving straight distillation. It is evident t h a t this method of determining cementing values is capable of distinguishing between products prepared with more or less care, and t h a t a reduction in cementing properties follows as a consequence of decomposition occurring during preparation.
Cementing Value 0752 (M/of rommefers)
There is marked loss in cementing value in t h e material pushed t o hard penetration a n d fluxed back, and-2the evidences of decomposition are very decided. I n fluxing this hard material (T-4) back with well prepared flux, approximately equal quantities of flux were required t o bring i t back t o 6 2 penetration. T h e effect of t h e introduction of so large a q u a n t i t y of well prepared flux has, of course, minimized t h e loss of cementing value. Nevertheless, i t is astonishing 1
--
9 7223
7 . 5 cm.
is i n confirmation of t h e investigations made b y t h e writer some time ago.' It will be observed from t h e inspection of t h e graphical d a t a t h a t t h e most cementitious materials were produced a t t h e lower temperatures, a n d t h a t t h e cementing values decrease as t h e conditions of preparation become more severe. While t h e differences in cementing value between materials produced a t j 2 j O a n d 700' F. are slight, there is a substantial difference between these results a n d those obtained with t h e materials produced a t 850' F., a n d between these latter, prepared a t substantially t h e same temperat u r e , b u t pushed t o a varying degree of hardness.
983
I n t h e previous discussion, various commercial asphalts which have been used in paving work have been examined, together with their appropriate a n d inappropriate fluxes. There has also been observed t h e effect on asphalt derived from liquid bitumen, of improper conditions of preparation. It remains, therefore, t o analyze t h e d a t a obtained, in connection with practical results, a n d t o determine, if possible, what cementing characteristics appear t o be necessary for t h e successful binder. Discussion of this
984
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
phase of t h e m a t t e r will be confined t o only one form of construction, namely, sheet asphalt pavements. F r o m long experience with various materials observed i n this discussion, a n d i n segregating those binders which, in t h e writer's observations, have been successf u l a n d unsuccessful, certain tentative values may be set which should differentiate t h e successful or acceptable cement from t h e non-acceptable. Carefully considering this m a t t e r in connection with practical results, t h e lowest cementing value which is considered acceptable under conditions of light traffic is
Vol. 6 , No.
12
terials laid under traffic, calling for a penetration of 35 t o 40, with materials of t h e first group, t h e writer would prescribe for such traffic a minimum cementing value for t h e bitumen used of a b o u t 0.24 kilogrammeters, a n d elongation limits between four a n d six centimeters. Maximum limits of elongation are set as well a s minimum limits for t h e reason t h a t above t h e maximum limits of elongation, is t h e accompanying higher consistency, which results in too soft a cement under conditions of hot weather. These limits are, of course, suggestive, a n d will be modified according
T A B LIV-FIG. ~
....................... ............... ................... ............. ........................ .......................... ..........................
T- 1
Temperature s t i l l . . 522' F. 21.2 Gravity of distillate at 60' F Pushed to penetration.. 50 Fluxed to penetration at 77' F . . 59 115 cm. Ductility at 7 7 O F.. Fixed c a r b o n . . . 5 . 5 per cent 99.6 Soluble in CSa.. 99.6 Soluble in CCl4.. 0.478 Cementing value a t 5 C . . .................. 7 . 7 cm. Elongation a t 5' C......................... Melting point.. ................................
.........................
+
10 T-2 700' F. 21.2 42 66 115 cm. 7 . 0 per cent 99.7 99.7 0.439 1 1 . 5 cm.
+
.....
set a t 0.08 kilogrammeters. This value, if t a k e n alone, will include some materials which, although able under light traffic t o bind t h e mineral particles in a sheet asphalt pavement together, are prone t o develop cracking. As s t a t e d previously, i t is necessary not only t o consider t h e cementing value alone, b u t also t o consider t h e ability of t h e binder t o elongate, in order t o prescribe those which have t h e property of holding together mineral aggregate without undue tendency t o cracking. For light traffic, this elongation value is set between t h e limits of 8 cm. a n d 14
Cms. Elongation 075~6
T-3 850° F. 26.2 66
....
102 cm. 13.1 per cent 98.9 97.5 0.311 1 7 . 0 cm.
.....
T-4 hard 850' F. 25.6 2
T - 4 fluxed
.... ....
.... ....
2 6 . 4 per cent 85.3 67.9
....
2 1 7 ' P.
.
62 5 7 . 0 cm. 15.2 per cent 93.3 87.3 0.212 13.75
.....
t o t h e individual experience a n d ideas of those who might employ these tests. It is probable t h a t a n empirical formula m a y be devised giving t h e relationship between traffic units m d cementing-elongation values in such a way as t o indicate t h e necessary characteristics applicable for a n y intensity of traffic. There remains another factor which may influence t h e valuating of materials in accordance with these characteristics, a n d t h a t is, t h e possible r a t e of loss of cementing value which will be shown upon tests carried o u t over a prolonged period of time. As is well
Cementing Value ~7752.(Kilogram m eters)
cm. f o r t h e pure bitumen. Below these values are found, in t h e writer's experience, unsuccessful materials for t h e class of construction referred to, a n d corresponding t o these values are found those which have demonstrated satisfactory use under light traffic. AS previously indicated, pavements under heavy traffic require stronger a n d more cementitious binders with less need of these binders t o elongate. It is .difficult t o describe accurately conditions which would be understood by all t o be designated a s heavy traffic. Keeping in mind, however, service records of ma-
known, various bituminous materials harden with age a n d suffer marked loss of ductility upon standing, sometimes for comparatively short periods of time. It is likely t h a t t h e materials here examined would likewise show variable r a t e of loss of cementing value a n d elongation with age, particularly those materials showing t h e effects of severe t r e a t m e n t in preparation. D a t a of this kind covering a period of t i m e or indicating t h e same effect of accelerated tests will be of considerable importance in fully determining t h e relative values of various t y p e s of binders.
Dec., 1914
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 B E M I S T R Y
It is t h e intention of t h e writer t o t a k e up these phases of t h e subject, b u t owing t o t h e time necessary t o compile complete d a t a , i t is considered advisable t o present t h e foregoing in t h e expectation t h a t t h e methods employed m a y be improved upon b y others interested, a n d t h a t t h e features indicated m a y be simultaneously investigated b y independent observers. It is suggested t h a t this test n o t only covers all of t h e d a t a heretofore furnished b y t h e ductility test, .but in addition furnishes direct information as t o t h e cementing values of t h e materials examined. T h e elongation values determined during t h e cementing value tests afford a more accurate a n d truly representative indication of ductility t h a n t h e test usually employed. T h e elongation values determined by this method have a n advantage of being t a k e n a t a temperature a t which such properties come into play most effectively in pavements, a n d are recorded only while t h e binder is capable of Withstanding a n appreciable strain. These values are not clouded b y t h e fine hair-like filaments into which bitumen is drawn a t normal temperatures, a n d which h a v e n o significance in determining a n y valuable property. I n addition, it is difficult, if not impossible, b y t h e usual method, t o differentiate between materials having a ductility i n excess of certain limits fixed b y t h e devices usually employed, which cannot be extended much beyond I O O centimeters with a n y significant results. CHICAGO PAVING LABORATORY 160 NORTH 5 T H AVENUE,CHICAGO
HYSTERESIS TESTS FOR RUBBER B y EARLEI L. D A ~ I Z S Received September 2 5 , 1914
The expert usually judges a piece of rubber b y means of a crude hysteresis test which he performs b y stretching a small strip with his fingers. Experience enables him t o judge closely, b u t b y no means accurately, small differences between t w o samples; i t does n o t enable him t o standardize his tests, nor t o make his results available t o others. Several machines have been devised t o perform a n d record these tests graphically, b u t t h e y have not come into their full usefulness, d u e t o t h e difficulty encountered in translating t h e graph into terms which are intelligible a n d comparable. T h e object of this paper is t o point out a n d explain some of t h e relationships between t h e mathematical equation for t h e curve a n d t h e properties of t h e rubber being tested, a n d t o point o u t t h e close relationships which some of t h e tests, made in this laboratory, show between t h e theoretical curve a n d actual curves made b y t h e machine. T h e accompanying figure shows t h e typical form of curve produced b y t h e machine. T h e ordinates represent tension and t h e abscissae, stretch. Cheneveau a n d Heim’ have shown t h a t this curve has t h e equation: x = cy 4- a sin2 b y (1) a n d t h a t when O B is drawn t a n . t o t h e curve OD at 0, c = tan. YOB. From ( I ) a sin2 by, = x1 - cy1 “Sur 1’ extensibilitb du caoutchouc vulcanise,” Comgt. rend., p. 320, Feb. 6, 1911; also, “The Rubber Industry,” (1911).
ml and YI
Since
c = -
21
98s
= xl - ml
then 21 = X I - cy1 or 21 = a sin2 by, I n like manner it may be shown that z2 = a sin2 nby, From
(2)
sin2by1
=
a
- 2A 221 = I - COS 2 by1
Since then
2 sin?il = 1
COS
a
or From (3) Since then
sin2 zbyl = cos 2A = I
and from (4)and ( 5 ) Whence Since From
(2)
From (6) and (7)
(“-“y
cos2 Pbyl =
I
22
-
a
- sin1 A - sin2 by1 = I - 22 ~
a
a
4212
a=-
4 1 -22
- sia*A 21 = I --
cos2A = 1 COS’
by1
cos2 by1 =
a
22 -
(7)
(8)
41-
Whence
by1 = c o P d $
(9)
T h a t these terms are not so mysterious as t h e y appear, becomes evident after a n analysis of t h e above proof. T h u s we find t h e constant “c” is dependent solely upon t h e initial resistance of t h e rubber t o stretching; i t will be largest for a pure gum stock a n d
Y PZ
x
, become relatively smaller a n d smaller a s t h e stock is more heavily compounded. From t h e relationship in Equation 6 we can readily see t h a t in t h e case of a pure gum stock where t h e stretch is long a n d uniform, ( m l ) a n d ( m z )having been large “a” will be relatively large, in comparison with a “tread” stock which will be strong a n d cause t h e entire curve t o be fairly steep, b u t will be small in comparison with a “whiting stock,’’ which is characterized b y a n initial stiffness after which i t offers comparatively little resistance t o stretching. Equation g shows t h a t for pure g u m or tread stocks cos b will be small a n d “b” consequently large. T h e ease with which these values could be standard-
