]\gt ; th of Viscous Traction its Relation to Viscosity .
By .
T. Received Read February experiments are macle on he viscous flow of pitch and other nces of similar character , in form of rods or cylinders , by the torsional method , is that the rate of turning lmder torsion of these rods is not strictly proportional to couple .
Thus the rate of flow of the material under stress cannot be in simple proportion to stress .
If it is wished to exact the rate of flow with shearing forces , by means of the torsional lncthod , a complication is at once nlct with , alisin/ from5 the fact that the of flow in a rod is not of the same value evcryw]lere , but necessarily varies from nothing at the centre to a maximum at the surface of the rod .
lVith the view of , a nlorsuitftble way of the phenoJnenon , trials were made with different of observing the flow of bodies , un .
conditions in which the said ection does not apply .
The results obtained in these ways exbibit the same departures from linearity as was ested by the results obtained by the method of torsion .
types of flow examined were : ( 1 ) the flow produced in a rod or cylinder of material when under traction ; ( 2 ) when under axial con the flow of a freely descending strean of the material ; ( 4 ) the rate of bending of rod beam of material under its own weight when supported only at the ends .
The latter method , however , from the defect as the torsional method , namely , from an value a range from zero ) to a certain or bars of the material under tion were suspended from one end , while to other vcights were attached .
The rod was thus subjected to a force which continuously drew it out .
The rate of elongation per netre length of rod could then be determined .
The rods { 0 these ) by the method described in the already referled to .
For the purpose of the actional forces the rod was thickened { each end worked into all cubicaJ , as shown in , so as to fit a metallic ) or receptacle open to one vol. On th Coefficient of Viscous , etc. side , Oy )hich the rod could be from a support at the top and a attached at the lower end .
The ) in the box for the rod was slotted out to one side so as to admit the rod being slid into its place .
ting the stress in the rod we must add to the weight attached to the hook the weight of the box and of the pitch below the point of obsel.vation on the rod .
The weight of rod nnder observation produces an increase in the stress from the lower end upwards , its amount half } is obviously half that of the rod between the lower and upper mark .
For points above and elow t middle the stless is in defcct of the mean value by equal amounts , so that , linearity for the flow , the correction will l half the weight of the poltion of the rod in the observation .
The rate of elongation of the rod was observed by nteans of a cathetometer .
In observations with certain materials such as shoemaker 's wax the plan was adopted of under a liquid of the density as the material itself for the purpose of eliminating the stress due to its own weight , which much too great to adm-it of accurate otherwise made on the rate of flow .
lower end of the rod , in such cases , was made fast while the tractive applied at the top means of a cord and pulley .
The rods were thus drawn upwards .
This was done in order to be able the more easily to surround the rod with a liquid , having the same density as its material , and thus to eliminate the action of the rod 's own wei .
For the a widc vertical tube was fitted to the apparatus and surrounded the rod under examination .
Solutions of salt in water were used for the liquids .
of riments .
results obtained with pitch other stances showed the time rate of elongation per centimetre of a rod under tension is approximately proportional to the force of traction per square nctre cross section or , where is the force applied , the area of the cross section , the velocity at any point on the rod , and .
F. T. Trouton .
fictent of Viscous [ Feb. 12 , a constant for any iven material .
This we term the coefficient of vis( lction of the tcrial .
values found for for a few of the substances experimented with are iven to show the order of tho coefficient .
For an inar variety of pitch at C. ) For the same with a admixture of tar For shoemaker 's wax at C. . .
of \mdash ; The observations made on the rate of flow of rods of these substances show that it is faster immediately after the application of the force than afterwards .
As an example , the following table is given , in which the elongation is in arbitrary units and the in minutes and seconds to the nearest five seconds .
These observations are plotted in fig. 2 .
FIG. 2 . .
1906 .
] Traction and its to that of Viscosity .
It will be seen th the initial rate of flow is faster than the final rate .
This is similar to what was observed in the case of the viscous flow of rods under torsion .
* It was also noticed that there was a slow partial movement towards recovery on removal of the force of traction , which gradually fell to zero with time , just as had been previously observed in the case of torsional forces .
with Variation of Trnctive Force.\mdash ; Experiments made with erent values of the tractive force show that the rate of flow is not strictly proportional to the force .
The results of determinations made with a variety of pitch given in Table II an shown in fig. 3 are ypical .
Table II .
1oo 200 300 The ordinates represent rate of elongation , while the abscissae represent the force pplied to the in grammes .
It will be seen from the that except the origin the law is linear .
For forces above a certain value the rate of flow may be Jressed as .
The rate of elongation 'Phil .
Mag vol. 19 , p. 347 , 1904 .
Prof. .
T. Trouton .
Coefficient of [ Feb. 12 taken for the curve was in every case that after the initial had been passed .
Th of Flow of rtictcs i'-out \mdash ; The mode of flow the particles of a rod while being drawn out is of interest .
Vith the view experimentally ascertaining if the particles lying in a cross section moved bymmetrically on taking place , that is to say , if half of the particles ttered uniformly out a cross section , are lefb ively behind , were of having approximately the same coefficient of viscou , traction , but of different colours .
Two different coloured rods were unite end to end and then drawn out , the surface of demarcation being atched .
of diffel'Gllt colouled( were tried for this purpose and , provide they ( approxitnately of the same fusibility , answered well .
thus mnde vere warnled and drawn out and the face where theJ joined observed .
This ined a plane ; sometimes did not remain a cross section .
tendency was noticed for the centra .
portion to a different rate to the peripheral parts .
The flow show .
itself to be cluite synlltetricnl the , as one wouId naturally expect Similar results were obtained with shoemakel.'S wax .
In this case tion of one end cted by the addition of a small quantity ol vermilion .
On -out , the line of demarcation be seen .
' of periments on the rate of a-vial contraction of cylinders under xial compression those obtained for the fraction of rods , as detailed above .
That is to say:\mdash ; ( l ) The rabe of contraction on first application if the stress is a little faster that finally reached ( see ; ( 2 ) there is a slow movement towards recovery on removal of the stress ; and ( 3 ) the final rate of flow increases at a rate with the increase in the applied stress when the 1 tter is above a certain value .
In the case of traction , the of the rod or cylinder be compared with its diameter , in most cases it between and 30 the but for compression on account of , it is well the to be not more than about three times the dianleter of the cylinder .
The plied by on a plate which j covered the top of the cylinder , which stood rate of depression of the cylinder observed by of a neter .
The coefficients obtained from on compression made in this were found to bout the halne in nitude as those obtained from traction .
For instance , with 1906 .
] its to of Viscosity .
a certain pecilnen of a rather soft pitch , the compressional coefficient was found to be , while the coefficient for traction of the same material to be FIG. 4 .
Time minubes .
of a Stream its own The stream of the material under examination was obtained by it to flow through a circular hole in the bottom of a wide tin vessel .
After a steady state is reached , the outline of the stream is of the character shown in fig. .
The stream gradually tapers down to a very fine thread , which breaks off intermittently from its lower end .
At any point distant from the top let be downward velocity of the material , then the time-rate of ation per centimetre of the material at this point is .
The tractive force is to the mass of the column below this point .
Let the cross section be , we have then , proyided the velocity is small , My .
Thus we may write , where is the volume of material supplied per second , and d the slope of the tangent to the surface at the point of the column where is the diameter .
Jfcthod of Dctcrmi'ning \mdash ; An optical method can bc adopted for determining the slope of the surface of the column at a definite point .
A beam of light a horizontal slit and lens is allowed to fall on the Prof F. T. Trouton .
Coefficient of Viscous [ Feb. 12 , column and is reflected into a telescope which is pointed upwards to receive it .
The of half the between the incident and reflected beams is evidently the slopc of the surface curvature .
FIG. 6 .
The value of can be got by off the column at the point in question , and wei(e .
The value varies , it is true , according as the column breaks off at its lower end or otherwise , but only to a ible extent .
The vaJue found for the coefficient of viscous traction by this method for a mixture of pitch and tar in the ratio of 7 to 1 was ; and for a mixture of the same materials in the ratio of 3 to 1 about .
These about the same values as obtained by the other methods .
of of a Strc its \mdash ; To determine the shape of the column we have the following considerations .
At any point we have seen that the force of traction is , ( 1 ) where A is the area of the cross section , so that const .
Also , ( 2 ) when is the density of the material .
1906 .
] Traction its Relation to of Viscosity .
If the rate of fall is small , the acceleration term may be ected , and we Substitute for from the relation , where is the mass of material falling per second ; and substitute for A. Then , after differentiating and , we have , where .
The solution of this equation is When is very small it represents a filament , such as in the present case .
The limiting solution when is .
This last expression was found , as described below , to fit experimental data with sufficient accuracy .
In order to experimentally examine the question , mixtures of pitch and tar were made sufficiently thick or glutinous for the flow to be slow enough to enable the acceleration term to be neglected .
The culvature assumed by the column was experimentally determined by observing the diameter of the column at various heights .
This was done in some cases by means of a cathetometer , in others by casting the shadow of the column from a distant source of light on a long vertical sheet of paper placed close to the column , and then tracing out the shadow with pencil .
The cathetometer telescope had a scale in the eye-piece , with which the horizontal breadth of the column at the different heights was observed according as the telescope was raised to various positions along the column .
From these readings , in conjunction with the height readings , the curve made by the pitch in falling could be plotted .
It was then possible to fit an equilateral hyperbola to it .
The following table ( p. 434 ) gives the results obtained with rather a thick mixture .
The first column gives the height in centimetres ; the second one half the observed diameter ; the third column gives the calculated value of the radius derived from the formula The values of the constants used were and .
This last is the height above the bottom of the vessel at which ) horizontal asymptote is situated .
The divisions of the scale of the eye-piece corresponded to cm .
This was subdivided by eye , so that the crreement is quite within the limits to be expected .
There was some difficulty at times in ding the diameter , especially towards the lower end , as the column would sometimes sway slightly about , even it was placed inside a case with of the end breaking off .
to shelter ieasure tffecGP fCoefficient oiscous [ Table by Tcthod.\mdash ; From the value of the constant in the coefiicicnt of viscons traction may be calculated .
Thus as found above , we was at a nJean temperature of C. A rod from the same mixture of tar and } ) for by the traction method the value at C. Another rod from the same mixture boave at C. reement will be considered sufficiently to confirm the theory when the chal.acter of the material is Unfortunately , though very materials of any desired viscosity , these mixtures suffer from the that when they heated for the manufactule of the test rods they lose some of their more volatile constituents , becoming more viscous in consequence .
bjj nlodification from the perbolic form , which the of material undergoes when it is not so viscous to render the inertia term ible , may ) appreciated by noting when } } ) is hyperbolic tho acceleration third of the fallen , so that at fJrst t ) incltia tetll 1nay bc quite itlnay at lower points become and sensibly reduce the tction effect .
In this the contour at oints assumes of the hypelbolic and approximates the c3 In limit viscosity is ) ) ] tcly ided the ) 1906 .
] Traction and its Relatio to that ofViscosity .
drops due to surface tension .
Comp ring the two cases , for points and , where , we have for the non-inertia case ) , for the non-viscous case ) .
Where both causes act , some intermediate shape will be by the materisl .
Descending columms of viscous liquids , such as the familiar one of honey falling from a spoon , fornl instances of this .
of a If a rod of pitch is laid across between two horizontal supports it will be funnd to continuously downwards .
rate at which this occurs the consistency of the material .
To find how the rate of depends on the coefficienlt of viscous traction , can resolve the stresses in t'ne material at any cross after the manner usual in the cases of stressed bealns .
This ives compressional forces above and tractive below a tain point in any cross section .
Taking this as being at the central ontal line of the cross section , we have for the moment of the force about this line where , is the rate per unit of elongation , or the contraction as the case may be , at any point situated at distances from the central line , and where is the breadth there .
This value for may be approxinlately expressed in terms of the rate at which a plane in the material at the point rotates at the moment when it is at right to the axis ; thus , }here is this rate of rotation at any cross section , so Now , the rate of sagging at the celltre , is given ; and , recollecting that \mdash ; , where ? ?
is the mass of the rod between the supports and its length , we get , after arranging and intewhere I is the moment of inertia of the cross section of the rod .
One of the methods employed to test this formula was to coulpare the rate of tging rods different quantities involved ) unaltered .
A rod of pitch of circular ) section ) laid between } supports which could be placed at various distances } ) , and the tinltl of the same distance in each obserYcd .
itll cases the F. T. Trouton .
of Viscous [ Feb. 12 , initial rate of not included in the measurements .
To allow this to done the rod was made with a camber and the observations only just beforc position .
It will be seen on comparing rods of erent 1 , if represents the time , at any iven span , to the standard distance , that COllst .
The following table gives in the first column the temperature the time of the experiment , in the second column the span , in the third the time taken to fall through a standard distance , and in the fourth the value found for this constant in each case .
Table The last two experiments were made on a different day from the others , and were made at a slightly temperature .
This may in for the smaller value found the constant .
The curve obtained by against is shown in fig. 6 .
1906 .
] fraction and its to that of Viscosity .
Initial Rate of \mdash ; As other cases previously dealt with , so in the case of beams , the initial rate of flow is greater than subsequently .
The table , obtained from experiments with a certain variety of pitch , illustrates this .
In the first column are given the distances fallen from zero by the central point of a rod or beam of pitch .
In the second column the time taken to each point .
The curve obtained by plotting these is shown in fig. 7 .
FIG. 7 .
Time in secs .
Table dimensions of this rod , which was circular , were follows : between supports , mean radius , mass between supports , final rate of .
This , using above formula , .
The same rod by the traction method A number of other lnixtures of pitch and tar of proportions were experimented with .
These gave approximately the same value for coefficient of viscous traction as that found by either column method or by the direct traction method .
Some of the mixtures of pitch and tar were too soft to deal with as a beam in air , so they were experimented with under water or brine .
In this way the downward 1noment could bc very reatly reduced , or , if desired , could be even changed in , when , of course , the V0L .
LXXVII. .
21 Prof F. T. Trouton .
Coefficient of Viscous [ Feb. 12 , was gave the value when operated under water .
ends of the beam hadtobehelddownyfno .
rticular mixture density : The value previously found for it the column method was .
The velocity in this case was perhaps too great to negect the inertia in the column method , and account for the highervaluee of thecoefficientn obtained ) it .
, an accoun or betwccn of Viscons of Viscosity .
It is obvious that there must be an intimate relation between the coefficient of viscous traction and the coefficient of viscosity as ordinarily defined .
The tractional force applied to a rod may be resolved , is usual in questions of elasticity , into two equal shears , which are situated at angles to each other and at to the direction of traction , with a uniform force of dilatation .
The value of either shearing stress , and also of the dilatation stress , is in each case one-third of of the tractive stress .
In the first instance on the application of the tractive force , the resolved effects produced corresponding to these resolved stresses will consist of a dilatation and of shearing strain .
It car only be to the flow resulting from the latter that the elongation of is due .
similar can take place in the of the stress of dilatation , which only can have an initial effect .
The continued application of each shear will produce a flow , criven in each case by , where is the stress , the coefficient of viscosity , and ) the rate of of direction of any line in the matelial in the plane of the shear as it passes through the direction normal to the shearing stress .
The resulting flow in the di1ection of the axis is obtained by adding the resolved components of the two flows in that direction ; so that the two effects , adding the components , and the axial ation to that per umit length , we find that Sin ce , and , where is the tractive force netre , we get , so that the coefficient of viscosity is to onethird of the coefficient of viscous traction .
* terms of the mole usual of cous f constant stress-modulus , the ument would take the following form : onsider i viscous cylinder tion at rate ; if is but slightly coulpressible , it mtlbt at the contraction at rate all ausverse directionb .
If the osity the of tension of intensity there is tension of intensity 1906 .
] and its to that of Viscosity .
In order to compare the coefficient of viscous traction with that of viscosity for the same material , two distinct plans were adopted .
One was to select a material sufficiently viscous to allow the coefficient of viscosity to be determined by means of the torsion of a rod made of it , also which allowed the coefficient of viscous traction to be found by directly drawing out the rod , or by the method of the horizontal beam .
The second plan was to select a lnaterial sufficiently fluid to admit of the coefficient of viscosity being determined by the rate of flow a tube under a pressure head , while at the same time not so fluid but that the coefficient of viscous traction could be observed by the method of the beam or by the method of the column descending under its own weight .
The following are the results obtained for the value of and in the case of several materials of wide in the value of the constants .
It will be seen that the value of is , generally speaking , in fair reement 1-ith three times the value of , the viscosity .
A variety of pitch which gave by the traction method was found by the torsion method to have a viscosity .
Another variety of pitch gave by the traction method and by the beam method , while the viscosity was found to be by the torsion A material made by adding a little tar to pitch by the traction method and by the torsion method .
A similar material containing a little more by the bion method by the torsion method .
A specimen of shoemaker 's wax by the tlaction method and ' by the torsion method .
For making a comparison by the tube method a ) of pitch and tar of about three to one was used .
This passed sufficiently freely through a tube to enable the coefficient of viscosity to be determined .
This was found to be , while the coefficient of viscous traction was found ) the beam 1nethod to be .
Another mixture of somewhat similar proportions , but better filtered , gave by the tube method and by the descending column method .
acting on the surfaces of the cylinder amount in all to a uniform hydrostatic pressure ether with a of intensity .
Of these the pressure is entirely neutralised by the reaction arising from the slight compression of the materials which it produces ; while the .
tension , having au intensity-coefficient , alone remains to operate in other ways , as in the text .
'Phil .
, p. 34 1904 . .
W. H. Dines .
Temperature [ Nov. 10 , These results collected in Tablo , where it will be seen that the coefficient iscuus traction is times that of viscosity .
Table The Gradients on the Coast of Scotland at Oxshott , Surrey .
By W. H. , F.R.S. ( Received November December 7 , In a paper by Dr. Shaw and the author read before the Royal Socievy on an accoullt of an investigation into the conditions of the upper air over the sea in the ohbourhood of inau , on the West Coast of Scotland , was viven .
Since that time two fresh series of observations in the same locality have been obtained , the lesults of which are submitted .
In each case bservations of temperature and were made by -recording ments sent up by of one or more kites , which wele flown from the deck of a steam vessel .
The pense has been met by a grant of X200 made by the Government Grant Committee , a grant of by the British Association at the port ] of f40 at the ; and also by ill atloltynlous contribution of by ellow of the Soci hese gtants ) not becll used entireIy for the vatious , but havG q neans of on meant 1 al , too , paratus for a ation crried out by .
SiInpson on ) Sca has ovided .
For vatious inan in a tug was hired , Lords Commissioners of ' ' , vol. 202 , ) .
Quart .
, No. 13

