65435 str (114)

> tac rtciprocal of thc density: the heavicr tlu bętad, tac fiaster the snrface falls. On aceottat rj **•    to spc^jc of the kmematic Yiscosity

te&oed as viscosi ty /density. The unit of kmematic viscosity is the stokes, and its hundredth part, th_e centistokes. Its dimensions naturally are Mass x Length^-1 X Time^-1 _T

-Mass _x Length^-*-^{= Length x Time}

= cm.*/secs.

In order to distinguish it from kinematic viscosity the trne viscosity is sometimes referred to as " dynamie viscosity.”

In a number of the " technical ” viscometers that have been designed, the hydrodynamics involved are very complex, and the results can only be considered as being of relative value. Several viscometers pf this type are in generał use in working with oil. In the Redwood, which is most commonly used in England, the time of flow through standard agate jets of a known volume of oil (50 ml.), at constant temperaturę, is measured. The dynamics of such flow is very complex. Attempts have been madę to use rape oil as a standard, and standardising against absolute viscometers shows a fairly good correlation of the type :—

Tj/p (kinematic viscosity) = AT_R — B/T_R| where T_R is the Redwood time of flow, and A and B are constants. The National Physical Laboratory has done extensive work on the standardisation of jets, and now issues certificates for these instrumentu. Another instrument, the " Number II Redwood,” has been designed for materials with a.łńgher rótoaty (tj greater than 5 poises). This, of conrse, pres difiemst values of A and B.

The Saybott viscometer is the American equivaleEt

COMMERCIAL VISCOMETERS 53

of the Redwood, and is very similar in principle. Herschel has examined the Saybolt, and derives eąuations relating flow times to true kinematic yiscosities. The Engler, or German version, differs little from this, the principal difference being the method of ensuring the proper height for the materiał in the bowl. An eąnation of the same type as that ąuoted for the Redwood has been derived for the Engler, and Vogel has also proposed a morę complex one

ijfp = (Engler time) x (o-0760)

Orifice viscosity methods have been largely used for rubber. Dillon and Johnson get surprisingly good Bingham curves at high pressure, which is curious for a materiał like rubber; but it only happens for fairly heavily compounded mixtures filled with caibon black. Lightly compounded mixtures give curvilinear curves for which the Herschel-Bulkley eąuation (to be discussed later) is applicable.

It is proposed to deal in detail with the Bingham-Murray method, because it is applicable to a very large number of problems, especially where materials are not truły fluid. Consider the case of an empty capillary tubę, of length L and radius R, gradually filling up with fluid supplied at a constant pressure P (dynes/cm.²), as in Fig. 12.

Fig. 12.