]\gt ; [ 279 ] VIII .
On the Whirting and of Shafts .
By ANLBY DUNKERLEY , M.Sc .
, Berkeley Fellow of the .
Owens College , nchester .
Communicated by Professor OSBORNE REYNOLDS , F.R. Recoived July 13 , \mdash ; Road Novembcr 23 , 1893 .
Revised June , 1894 .
TABLE 0F CONTENTS .
CHAPTER I. Art .
Page .
1 Introduction .
Object of investigation . . . . .
280 2-6 Experimental apparatus .
Method of making experiments .
Experimental shafts and pulleys . . . . .
, . .
281 CHAPTElli II .
7-8 General theory as given by Professor REYNOLDS ) . . . . .
285 rAPTER I Special cases.\mdash ; Unloaded Shafts .
9 CASE I.\mdash ; Owerhanging shaft fixed in direction at one end .
288 10-11 CASE II.\mdash ; Shaft resting on a support at each end .
Experimental results .
289 12-18 CASE III.\mdash ; Shaft resting on two bearings and overhanging on one side .
Experimental results . . . .
290 14 CASE IV.\mdash ; Shaft fixed in direct.ion at one end and resting freely on a bearing at other . . . . . . . . . .
293 15-16 CASE V.\mdash ; Shaft resting on three bearings .
Experimental results 295 17 CASE \mdash ; Shaft fixed in direction at each eud 298 CHAPTER : Special cases\mdash ; Loaded 18 Two methods of soiution .
First method . . . . . . . . . . . . .
299 19 CASE VII.\mdash ; Overhanging shaft fixed in direction at one end and loaded with ] pulley at the free end , the effects of both shaft and pulley being taken into account . . . . . . . . . . . .
299 20 CASE VIII.\mdash ; Shaft resting on a bearing at each end and loaded with a pulley , the 9ffects of both sbaft and being takert into account . . . . . .
301 14.8.94 Page .
Art .
21-22 Second method of solution .
Modification of general .
Investigation confined to effect of one pulley only , the effect of sbaft being neglecbed .
302 23-25 CASJ IX.\mdash ; Overhanging shaft fixed in direction at one end and loaded with a pulley at the free end .
Superior and inferior limits .
Natnral period of vibration .
ResuIts CASE X.\mdash ; Shaft resting on a bearing at each end and ]oaded with a pulley .
Superior and inferior limits .
period of vibration .
Theoretioal resuIts .
Experimcntal resuIts . . . . . . . . .
306 :31-34 CASE XI.\mdash ; Shaft , resting on two bearings and overhanging on one side , loaded with a pulley at end of owerhanging portion .
Superior and inferior limits .
Natural period of vibration .
Thooretical results .
Experimental results .
313 85-38 CAsf ] XII.\mdash ; Shaft , fixed in ection at one end and resting freely on a bearing at the other , loaded with a pulley .
Superior and inferior limits .
Natural period of vibratio Theoretical results .
Experimental results .
Comparison with Case X. . . . . . . . .
320 39-46 CASE XIII.\mdash ; Shaft resting on three bearings and loaded with a pulley on one of the spans .
Case of equal spans .
Superior and inferior limits .
Natural period of vibration .
Theoretical results .
Comparison with Case X. Experimental resuIts .
Remarks on formula . . . . . .
325 47-49 XIV.\mdash ; Shaft fixed in direction at each end and loaded with a pulley .
Superior and inferior limits .
Natural period of vibration .
Theoretical results .
with Cases XII .
and X. . . . . . .
336 50-54 CASE \mdash ; Sbaft supported on four bearings and loaded with a pulley on one of the outer spans .
Case of equal spans .
Superior and inferior limits .
Natul.al period of vibration .
Theoretical resuIts .
Comparison with Case XIII .
Stffiening effect of a fourth bearing . . . . . .
340 55-58 CARE XVI.\mdash ; Shaft supported on four bearings and loaded with a pulley on the middle span .
Case of equal span .
Superior and inferior limits .
Natural pel'iod of vibration , Theoretical results .
Comparison with Cases X. , XIII .
, and ening offect of a fourth bearing . . . . .
348 59 Caseof two or more pnlleys . . . . . . . . . . . . . .
355 60-62 CASE XVII .
Shaft lesting on a bearing at each end and loaded with two pulleys , the effect of the shaft being neglected .
Equations too complicated for actnal use .
Alternative method .
Formula for caleulabing the resulting whirling speed due to several causes , in terms of the whirhng speeds due to the several causes considered independently one another . . . . .
355 63 Concluding remarks . . . . . . . . . . . . . . . .
359 CHAPTER I.\mdash ; INTRODUCTION DESCRIPTION OF XPERIMENTAL A INTRODUCTION .
1 .
It is well known that every shaft , however nearly lanced , when driven at aparticular speed , bends , and , unless the amount of deflection be limited , might even break , although at higher speeds the shaft again runs true .
The particular or " " critical\ldquo ; speed depends on the manner in which the shaft is , its size and modulus of elasticity , and the size , weight , and position of any ulleys it carries .
AND VIBRATION OF SHAFTS .
The theory for the case of an unloaded shaft first received attention at lands of Professor RANKINE , obtained numerical formulae for the cases of an unloaded shaft resting freely on a bearing at each end , and for an overhanging shaft working in a shoulder at one end .
Professor GREENHILL has also obtainsd formulae for cases of an unloaded shaft resting on bearings at each end , and fixed in direction , each end .
The been further extended to the case of a shaft loaded with pulleys , by Professor REYNOLDS ; and the object of this investigation is to appIy that theory and so obtain formulae , and by experiment to verify them , giving the critical speed in terms of the diameter of the shaft , weights of pulleys , \amp ; c. , in particular cases applicable to the dffierenb conditions under which a shaft works .
In many cases , as might naturally be expected , the " " period of whirl\ldquo ; of the shaft is merely its natural period of lateral vibration when in a state of rest .
The two periods are coincident in the case of an unloaded shaft ( however supported ) , and for a loaded shaft on which the pulleys are placed in such positions that they rotate\mdash ; when the shaft is whirling\mdash ; in planes perpendicular to the original alignment of the shaft .
wiCh pulleys placed in any other positions , when the shaft is whirling , there is a righting moment , tending to straighten the shaft , which does not exist when it merely vibrates under the dead weight of the pulleys .
Hence , in an unloaded shaft , the period of whirl coincides with the natural period of lateraI vibration ; but , generally , in a loaded shaft , the period of whirl is less than the natural period of vibration , to an depending on the size and positions of the pulleys .
If , therefore , .
the period of disturbance ( that is , the period of one revolution ) be decreased , the shaft runs true until that period approximates to the natural period of vibration of the shaft ( assumed at rest ) under the given conditions .
If the shaft now receive any displacement , however slight , a violent agitation is set up , which will be most marked when the period of disturbance and the whirling period coincide .
As the period of disturbance is further decreased , the agitation becomes less , and , at a period of disturbance slightly less than the whirling period of the shaft , the shafb again run true .
As in the vibration of rods , so in the whirling of shafts , there are a series of periods at which the shaft whirls .
ExPERIMENTAL APPARATUS .
2 .
The experiments were made in the Whitworth Engineering Laboratory , Owens College , where the essential facilities for obtaining uniform rotation at any * RANKINE'S ' Machinery and , p. 549 .
'Proc .
of Insf .
Meoh .
Engineers , ' April , 1883 .
NDCCCXCIV.\mdash ; A. 2 MR. S. DUNKERLEY ON THE WHIRLING speeds wers afforded by of Professor REYNOLDS ' quadruple turbines under a constant head of 113 feet of water .
The essential parts of the apparatus by which the different formulae were consisted of ( see fig. 1 ) cast-iron bed plate , of stffiened channel section , 6 inches long and 4 inches wide , with its top and bottom faces planed parallel : a headstock which was inches high , 4 wide , and 4 inches long , with boltom face planed ; a headstock spindle ( which receives the motion ) , inch diameter , and provided with a shoulder at one end , a loose collar , and two speed pulleys , for directly receiving the motion , the other for transmitting the motion to a fngal fan indicatot ; which approximately indicates the speed of the headstock spindle , at any instant , by the height of a column of liquid forced by the fan a glass tube .
The scale of the indicator was graduated by accurately determining the speeds required to force the liquid up to two or three definite heights , and so ubtaining a formuh by means of which the heights due to certain speeds can be calculated .
The formula so obtained was where number of revolutions of headstock spindle per minute , AND VIBRATION OF SHAFTS .
and height of liquid , from level of still water , measured in inches .
The scale was graduated for every 100 revolutions per minute .
The bearings in which the experimental shaft ran consisted of castings of section with .
bottom faces planed .
They were bored at exactly the same height as the headstock , and the length of the bearings was about an eighth of an inch .
The deflection of the shaft , when whirling , was limited by the use of guard rings , whict ] consisted merely of ordinary bearing bored to a slightly diameter than the diameter of the shaft .
The motion was transmitted from the shoulder end of the headstock spindle to the experimental shaft by means of a piece of steel wire ( about inch long and 21 B.W.G. diameter ) , one end of which was dered into the end of the shaft , the other end being soldered into a piece of brass coned to fit into the headstock spindle .
By this means the shaft was subjected to very little constraint .
The headstock spindle was driven from a turbine which was 20 yards away from the experimentalist 's bench .
The motion was transmitted through 140 feet of quarter-inch cotton rope , the rope ascending vertically from the urbine and descending vertically to the headstock spindle .
The admission of water to the turbine was controlled by a hand-wheel close to the apparatus , by which .
an almost indefinitely fine adjustment of the speed of the turbine could be made from 200 to 2000 revolutions per minute .
By having speed-pulleys on the turbine shaft and headstock spindle , a range of speed of from 100 to 10,000 revoIutions per minute of the headstock spindle was obtained .
3 .
In taking the number of revoIutions to any period of whirl , an ordinary counter pushed into the end of the headstock spindle was used .
The whirling speed was taken to be at the commencement of whirl , that is to say , at the lowest speed at which the shaft definitely whirled .
Readings were taken , in each trial , over a period of fcom 3 to 5 minutes , the speed ( if it varied from some cause ) being kept constant by means of the valve regulating wheel .
The constancy of speed was shown by the steadiness of the liquid column of the icator .
In making any experiment three trials were made , and the mean of the results taken .
In all cases the theoretical speed was unknown when the actual whirling speed was obtained .
4 .
The headstock spindle was originally driven by hand .
This was accomplished by means of two cast-iron speed pulleys turning on pins bolted to the two ends of a cast-iron bracket , the bracket being bolted to the .
By running from a Iarge pulley on the hand-wheel to a small one on the second wheel , and from a large pulley on the second wheel to a small one on the headstock spindle , a very hi gh speed was att , ainable .
The motion was naturally unsteady , and available only for periods , whilst all additional erver was required .
By driving the shaft from bine a practically constant steady speed was obtained , and the increased of the trial considerably reduced the personal errors with the counter .
Moreover , an arrangement for ' the turbine valve at the bench , the action of the could be carefully observed whilst the speed was increased , and so personal errors determining the precise moment of reduced to a minimum .
5 .
The EXPERIMENTAL SHAET was of cast steel .
It was 3218 inches long , and 2488 inch diameter .
TlIe greatest variation in the diameter was of an inch .
It was turned by Mr. THOSE .
FORSTER of the Whitworth Engineering Laboratory , Owens College , Manchesber , to whom the author is indebted for the preparation of the greater part of the experimental apparatus .
The shaft weighed grms. , or .
The weight per foot run The determiuation of ( YOUNG'S Modulus ) , or rather EI ( I being the geometrical moment of inertia of the cross-section about a diameter ) , was accomplished as follows : experimental shaft was placed in bearings , 2 feet 8 inches apart , and loaded at the centre .
The deflection was measured by means of a micrometer , the distance measured being taken between the top of the shaft the bottom of a pin fixed in one of the guard castings .
The mean of the results so obtained gives for the Value of the gravitational system of units being employed .
[ NOTE.\mdash ; The value of expressed in pounds per square inch is 27,974,000 ] .
6 .
The LXPER1MENTAL PULLEYS were of brass and of the section ( fig. 2 ) .
rfhe moment of inertia taken ( for a reason which will appear later ) is where are the mass-moments of inertia about the axis of the shaft , and about a diameter passing through the centre of gravity of the pulley perpendicular to the axis\mdash ; both mom ) being expressed in units which , it may be remarked , are the ones adopted throughout the investigation .
AND VIBRATION OF SHAFTS .
The following table gives the dimensions and other necessary information .
In it the notation used is as follows:\mdash ; Weight of pulley .
I ' Moment of inertia .
\mdash ; B , where 's Modulus .
I Geometrical moment of ) ertia of cross-section of shaft about a diameter .
The pulleys were bored so as to fit the largest part of the shaft , bsing kept in position on it by rubbing bees-wax on the part of the shaft required , and heating the pulleys sufficiently to melt the wax .
On cooling , the wax was sufficient to firmly secure the pulley in its place .
It may be mentioned that PULLEY I. is the model of light pulleys generally used in workshops ; whilst PULLEY II .
is the model of a 3-feet belt ] , weighing about .
In designing the experimental pulleys , account has , of course , been taken of the differen .
sized shafts on which the actual pulleys run\mdash ; the pulleys being designed for weight and inertia .
The following are the actual sizes of the pulleys , of which I. and II .
are models:\mdash ; CHAPTER II.\mdash ; GENERAL THEORY , AS GIVEN BY PROFESSOR ]YNOLDS .
7 .
Take the axis of to be the original alignment of the shaft ; and that of perpendicular to it and revolvmg with the shafb .
S86 MR. S. DUNKERLEY ON THE WHIRLING bending moment at a a Let nt at a distance from the origin , and let the deflectio at this point be centrifugal force per ullit length of shaft .
geometrical moment of inertia of a cross-section of shaft about a diam. Modulus for the shaft .
angular velocity of shaft .
weight of shaft in lbs. per foot run .
weight , in lbs. , of any pulley which the shaft carries .
some moment of inertia of the pulley yet to be determined .
Neglecting the dead weight of the shaft , the ordinary equations of the beam give us . . . . . . . . . .
( 1 ) , and . . . . . . . . . .
( 2 ) , whence . . . . . .
( 3 ) , where Equation ( 3 ) holds between every pair of singular points , that is to say , between bearings and pulleys .
At a point of support , the difference of shearing force on the two sides must clearly equal the pressure , that is , . . . . . . .
( 4 ) , where and are the bending moments to the right and left of the support , and is the pressure on the support .
At a load consisting of a weight , this equation becomes ( neglecting the dead weight of the pulley ) . . . . . . .
( 5 ) .
A further equation may be obtained by considering the " " centrifugal couple\ldquo ; tending to straighten the shaft .
The moment of the centrifugal forces about a diametral line in the plane of the pulley and passing through its centre of gravity is where I ' and -moment of of pulley about an axis through its oenfre of gravity perpendicular to its plane , and AND VIBRATION OF SHAFTS .
-moment of inertia about a diameter through its centre of gravity perpendicular to the axis of the shaft .
Hence . . . . .
( 6 ) : 8 .
solution to equation ( 3 ) is well known to . . . .
( 7 ) .
The quantities are absolute constants betweell any two singular points ; but have not necessarily the same values between every pair of singular points .
If undashed symbols refer to those on the left , and dashed constants or symbols to those on the right of a singular point , then since the values of , are continuous , we have , at all singular points , whether points of supports or pulleys , ; whence sinh ( 8 ) , ( 9 ) .
But , at points of suppol.ts , ; whence . . .
( 10 ) , mx . . .
( 11 ) .
Also , since the bending moment is the same on both sides of a point of support , , whence ( 12 ) .
At a singular point , consisting of a concentrated load , we have , from equations ( 2 ) and ( 5 ) .
( A-A ' ) { ( 13 ) , and , from equations ( 2 ) and ( 6 ) , ( 14 ) , In to these equations we shall get equations according to the manner in which the shaft is supported at the ends .
If it merely rest on the bearing , so that the bearing exercises no restraint on its direction , the bending moment at that point is zero , that is , , and , therefore , ( 15 ) .
On the other hand , if the bearing be so long that it practically guides the directiou of the shaft , io other words , if the shaft be fixed in direction , then we have or . . .
( 16 ) .
It will be found that , in every case , equations ( 8 ) to ( 16 ) , incIusive , are sufficient in number to allow of the elimiuation of the ratios : \amp ; c. The resulting equation will give a relation between the whirling size and weight of the pulleys , diameter of the shaft , \amp ; c. , that relation depending on the manner in which the shaft is supported and loaded .
The proper value of has , of course , to be substituted , in the above equations , any particular singular point .
The values of the constants at the ends of a shaft are zero .
CHAPTER III.\mdash ; SPECIAL CASES\mdash ; UNLOADED SHAFTS .
I. 9 .
OVERHANGIITG , FIXED IN DIRI ?
CTION AT END .
Thus We have ( S7 , p. 286 , equation 3 ) , where , whence Taking the origin at the shoulder , we have , when and , when ( shearing force zero ) .
Hence we get AND VIBRATION OF SHAFTS . . . . . . . . . . .
( 2 ) , . . .
, ( 4 ) .
The elimination of , from these four equations , leads to either , or to i.e. , . . . .
[ ] .
The least value of which ] satisfies this equation is Case II .
10 .
SHAFT , LENGTH , MORELY RESTING ON A BEARING AT EACH END .
Thus\mdash ; We have ( S7 , p. 286 , equation 3 ) , whence mx. Taking the origin at the left-hand bearing , we have , when or , whence . . . . . .
, . . . . . . . . .
( 2 ) , A ( 3 ) , ( 4 ) .
* POISSON , ' Traite de Mecanique , ' vol. 2 , S 528 .
The elimination of : from these equations gives either , or , i.e. , 11 .
Experimental Results.\mdash ; For a description of the manner in which the experiments were made , see S3 , p. 283 .
The following are the mean resuIts\mdash ; the percentage errors being considered positive or negative according as tho observed is greater or less than the calculated speed .
The experiments show that , in the simple case of a shaft resting on two bearings in which the conditions reqnired by the theory can be very closely approximated to in practice , there is no appreciable difference between the observed and caIculated speeds .
Case III .
12 .
SHAFT SUPPORTED ON BEARINGS FEET APART AND OVERHANGING TO LENGTH ON ONE SIDE .
Taking the origin at , we have from to and , from A to ( S8 , p. 287 , equation 7 ) .
AND VIBRATION OF SHAFTS .
When and when Also , when ( shearing force zero ) .
Hence , we get A . . . .
( 1 ) , A . . . .
( 2 ) , . . . . . .
( 3 ) , . . . . . .
( 4 ) , . . . . .
( 5 ) , . . . . .
( 6 ) , . . . .
( 7 ) , . . . .
( 8 ) .
The elimination of : : : .
from these equations leads to the result ml . . . . .
[ A ] .
If , by dividing throughout by , the equation reduces to the equation already obtained for an overhanging shaft fixed in direction at one end ( Case 1 , S9 , p. 289 ) .
If , the equation [ A ] reduces to , i.e. , , the equation already obtained for a shaft resting freely on a bearing at each end ( Case II .
S10 , p. 290 ) .
The general solution to equation [ A ] is best obtained by assuming , wtar is less than tlnity , and expanding each term in asoending powel .
S of .
In this manner we get , to a sufficient degree of approximation the equation .
From this equation the following resuIts\mdash ; giving the of ml for different values of \mdash ; have been obtained .
If we assume !
, then the number of revolutions will be a maximum for a given length of shafting , when is a maximum .
From the above results the speed will be a maximum when the ratio is one-third .
Hence , for a shaft of given length running on two bearings , one being placed at the end , the best position for the other bearing is such that it divides the length of the shafting in the proportion of 1 In all cases that occur in practice the overhanging portion is small compared to the span .
Hence , we may say that if a shaft , span , overhang a distance less than one-fifth the span , then 13 .
Experimental Results .
The following are the mean results , the calculated speeds being obtained according to the formulae in the eceding article ( p. 292 ) , when the value of is taken .
These results show that the ated speeds are less than the observed speeds provided ( that is ) be less than one-fifth ( which is always the case in practice ) , and in excess for greater values of .
In ouly two cases is the percentage error greater than 3 per cent. , thus amply verifying the theory .
The maximum observed speed is when ) 1/ 3 , a result which has been shown to follow immediatelv from the equations .
Case IV .
, LENGTH RESTING FREELY ON A SUPPORT ONE EiND AND IN DIRECTION AT THE OTHER .
Thus\mdash ; We have S8 , p. 287 , equation 7 ) Taking the origin at , we have , when ; and when Hence , . . . . . .
( 1 ) , . . . . . .
( 2 ) , riosh . . . .
( 3 ) , . . . .
( 4 ) .
The elimination of : : : from these equations leads to To solve this equation , draw the curves of ml and .
The points of intersection of nd with will give values of which satisfy the equation From the diagram , it will be seen that the first value of is less than by a small quantity .
It may be shown that , to a sufficient degree of approximation , AND VIBRATION OF SIIAFTS .
Case 15 .
SHAFI SUPPORTED ON THREE SUPPORTS , FEET APART RESPECTIVELY , SUPPORT BEING AT EACH END .
Thus\mdash ; Take the origin at , and let dashed letters refer to the ight of , and undashed letters to the left .
Then we have ( S8 , p. 287 , equation 7 ) from to and from A to mx. When , .
; when ; when Hence we get . . .
( 1 ) , . . . .
( 2 ) , . . . . .
( 3 ) , . . . . .
( 4 ) , A . . . .
( 5 ) , A . . . .
( 6 ) , . . .
( 7 ) , . . . .
( 8 ) .
By elimirlating : : : : : : : from these equations we obtain the results , that either or First consider the solution It follows that are all zero , and that , Dsin Hence , in addition to the solution the equations ( 1 ) are satisfied when simuItaneously , and being in tegers .
Hence , if be a mulbiple of , that is : if a multiple of , one speed of whirl is clearly that of the shorter span when the longer span is result , of course , identical with the vibration of strings in segments .
Secondly , consider the solution If or , we get The physical interpretation of this equation is that the shaft in the one case is horizontal at the middle bearing , and in the other at the end bearing .
In other words we get Case IV .
, S14 , p. 294 , which has already been solved .
[ It should be noticed that the case when comes under the first solution .
] The solution to the general equation may be performed by putting where is the ratio of the spans , being always less than unity .
By expanding we obtain the equation AND VIBRATION OF SHAFTS .
( ml)4 The following are the results obtained from this equation , the value of having been calculated for different values of The formuIa is not sufficiently approximate if When is very small the result closely approximates to the result obtained for a shaft working in a sleeve at one end , .
: .
( S14 , .
294 .
) 16 .
Results .
The following are the mean resuIts , the calculated speeds being obtained according to the formulae in the preceding article when the particular value of is taken .
It will be noticed that in some of these experiments the observed speed is and in others less , than the calculated speed .
With the exception of Experiment , the experiments amply verify the theory .
From Experiment 39 it would appear that when the ratio of the spans is one-half , calculated speed is less than the observed speed .
It , therefore , errs on the right It appears from Experiment 40 , that the same formula as for holds , to a sufficient degree of approximation , until the ratio of the spans is equal to 17 .
, LENGTH IXED IN DIRECTION AT EACH END .
Thus\mdash ; Taking the origin at , we have ( S8 , p. 287 , equation 7 ) when , or ; whence .
( 1 ) , . . . . . . . . .
( 2 ) , . . . .
( 3 ) , . . . .
( 4 ) .
The elimination of : : : from these equations leads to The least value of ml which satisfies this equation is , ' Trait6 de Mecanique , ' vol. 2 , S 528 .
AND VIBRATION OF SHAFTS .
CHAPTER IV.\mdash ; SPECIAL CASES\mdash ; LOADED 18 .
In considering shafts loaded with pulleys two methods may be adopted .
First .
The period of whirl may be calculated taking both the shaft and into account together .
Second .
The period of whirl may be first calculated for the shaft , neglecting the pulleys , and then for the pulleys , neglecting the shaft .
By eans of an approximate , the period of whirl , taking both shaft and pulleys into account , may be calculated from the separately calculated periods of whirl .
FIRST OF SOLUTION .
Investigation shows that the first method leads to equations whioh are nob solvable , so as to give results in a form convenient for actuaI The following two simple cases will illustrate this .
19 .
OVERHANGING SHAFT , LENGTH FIXED DIRECTION AT ONE END , AND LOADED WITH A PULLEY , WEIGHT AND MOMENT OF INERTIA , AT ITS FREE END , THE COMBINED EFFECTS OF BOTH SHAFT AND PULLEY BEING EN INTO ACCOUNT .
Thus\mdash ; Fig. 10 .
we have ( S7 , p. 286 , equation 3 ) for every point between the bearing and the puUey , ; whence Taking the origin at the shoulder , we have at a ] point consisting of a concentrated load , ; ( S7 , p. 286 , equation 5 ) .
whence , when , since , we get Again , from equation ( 6 ) , p. 287 , we have ; wherefore , when ( 2 ) .
Again , when ; whence . . .
( 8 ) , . . . . . . .
( 4 ) .
The tion of from the four marked equations leads to [ A ] .
If we assume the pulley to be removed , that is , if we put , I ' in equation [ A ] , we obtain the same as that obtained in Case I. , p. 288 .
The equation [ A ] can only be solved by ming some relation between coeflicients ; in other words , we cannot obtain a general sohltion which could be readily applied in any actual case .
AND VIBRATfON OF SHAFTS .
Case VIII .
20 .
SHAFT LENGTH , MERELY RESTING ON A SUPPORT AT EACH END , AND LOADED WITH A PULLEY , WEIGHT AND MOMENT OF INERTIA I ' AT DISTANCES FROM THE Thus\mdash ; Taking the origin at , we have ( S8 , p. 287 , equation 7 ) from A to and from to where dashed lettersrefer to the right , and undashed letters to the left of the pulley .
At the pulley , when , we have ( S8 , p. 287 , equation 13 ) ( A\mdash ; A ' ) ( 1 ) .
Again , from equation 14 , p. 287 , we have mc ( 2 ) , when whence ( A-A ' ) ( 3 ) .
( 4 ) .
, when , or whence ( b ) , . . . . . . . . . . .
( 7 ) , . . .
( ) .
The elimination of : : : from these equations leads to ml ml mc [ A ] , where Equation [ A ] is , of course , symmetrical with respect to If we imagine the pulley to be removed by and I ' the equation A reduces to i.e. , a result already obtained in Case II .
, p. 290 .
As in Case VII .
, S19 , we cannot obtain a general solution to [ A ] , which could be readily applied in any actual case .
SECOND METHOD 0F SOLUTION .
21 .
The formulae obtained by considering the effect of ) pulleys and the shaft combined have thus been shown , even in simple cases , to be absolutely useless for practical purposes .
By the second method of solution the irling speed of the pulley neglecting the shaft is first obtained .
The general theory ( Chapter II .
) will have , therefore .
to be slightly modified .
Since AND VIBRATION OF SHAFTS .
equation 3 , of S7 , becomes . . . .
( 1 ) , and , therefore , .
( 2 ) .
If ; as before ( S8 ) , undashed symbols refer to the symbols or constants on the left , and dashed symbols to those on the right of a singular point , then ( as in Chapter II .
, equations 7-16 ) we shall have precisely the same differential } holding at the specified singular points , the only difference being that when those differential equations are integrated , the forms of the resulting equations are altered from a ical ( in Chapter II .
) to an algebraic form in the present case .
22 .
It is now proposed to investigate some of the cases , commonly occurring in practice , according to the second method of solution .
Whatever be the manner in which the shaft is supported , the effect of the shaft is neglected , and the shaft supposed to be loaded with one pulley only .
The effect of the shaft , and of more than one pulley , will be considered in SS59--62 .
Case IX .
23 .
OVERIIANGING SHAET , LENGTH , FIXED IN DIRECTION AT ONE END , AND LOADED WITH A PULLEY , WEIGHT , AND MOMENT OF INERTIA , I ' AT ITS END .
Thus\mdash ; Fig. 12 .
We have ( S21 , equation 2 ) , Taking the origin at the shoulders , we have , when whence .
( 1 ) , . . . . . . .
( 2 ) . .
S. DUNKERLEY ON THE WHIRLING When , that is , at the pulley , we have ( S7 , equation 6 ) , ( S7 , equation 5 ) .
whence and or and Let so that where , having the value assigned to it in S7 .
The elimination of from the four.equations marked leads to . . .
whence . . .
24 .
tion [ may be put in form If be or is negative , and therefore the equations do not hQ1\amp ; HeIlce , for whirling to be at all possible , must be and ; that must lie between 3 and 12 .
The speeds which these values give for any value of may be termed the superior of the speed .
The values of corresponding to theselimits are zero and infinity .
In other if the shaft whirl at a speed which satisfies or 12 , the effect of the inertia of the pulley is either zero or nity .
In the first case AND VIBRATION SHAFTS .
should have zero righting moment , and in the second , an infinite righting moment .
In other words , in the one case there would bs no tendency to make the pulley deviate ffom its natural plane of rotation , and in the other , any such tendency would be met by an infinite moment tending immediately to right it .
In either case , therefore\mdash ; assuming whirling to take place at the speeds given by the limiting values of \mdash ; it would whirl in such a manner that the still rotates in a plane perpendicular to the original alignment of the shaft .
In fact , the period of , corresponding to the inferior limit of the speed , identical with the natural period of vibration of the light shaft under the given onditions .
This may be easily proved independently .
* The superior limit is double the inferior limit .
The inferior limit may be taken as a first approximation to the period of whirl .
25 .
Referring to equation [ B ] , S23 , by giving different values likely to be met with in practice , we get , for each value of , a relation between , the angular velocity of whirl , and , the overhanging portion .
Knowing , therefore , the particular value of , the value of may be readily calculated .
The following are the results obtained in this manner from equation * This may be seen as follows:\mdash ; If be the weight of the pulley , and the force necessary to deflect it one foot , then ( the time of lateral vibration ) is .
To get , if be the load acting at a distance from the shoulder , as in fig. 12 , the bending moment at a distance from the shouldcr , then where and I have the same meaning as in the text .
Hence , where A and are constants of integration .
When , and ; whence , and The deflection , therefore , at the weight is , and when this is nity .
Hence and , therefore , natural period of vibration .
Whence .
NDCCCXCIV .
Vcdues of in the equation , being the of the Pulley from the Shoulder .
Case 26 .
SHAFT , LENGTH , MERELY qTING ON SUPPORT AT EACH END AND LOADED WITfl A PULLEY , WEIGHT AND MOMENT OF INERTIA , AT DISTANCES FROM THE SUPPORTS .
Thus\mdash ; Fig. 13 .
Let the origin be taken at the left-hand bearing .
We have ( S21 , equation 2 ) between A and , and between and C. When AND VIBRATION OF SHAFTS .
Therefore .
( 1 ) , . . . . . .
( 2 ) .
When ; therefore ( 8 ) , . . . . .
( 4 ) .
At the pulley , when ; therefore ( 5 ) , . . . . .
( 6 ) .
Again , when ( S7 , equation ( 5 ) ) , and \mdash ; ( S7 , equation ( 6 ) ) ; whence A\mdash ; A ' ( 7 ) , and . . .
( 8 ) .
The elimination of the seven ratios ffom the equations marked leads to the equation [ A ] , in which ( see p. 305 ) .
Hence , so that , for whirling to be at all possible ( see Case IX .
, S24 , p. 804 ) , must be and If be equal to the first or eecond of these quantities , the corresponding value of is the inferior or superior limit the ) respectively .
Moreover , the period of whirl corresponding to the ferior limit of speed is the period of vibration of the light shaft under the given conditions .
* The superior limit is the inferior limit multiplied by some function of the position of the pulley .
With the same pulley on the same shaft the superior limitinferior limit .
$ This be seen as follows:\mdash ; If be the weight of the pulley , and the force necessary to deflect it one foot , then time of lateral vibration ) is ) .
To get , if be the load acting at distances from the bearings , as in fig. 13 , the bending moment at a distance from the shoulder , then ( fig. 12 ) from , and Bto C. Hence , from A to , from to and I the same meanings as in the text .
We get , therefore , , from A to . . . . .
( 1 ) , , from to . . . . .
( 2 ) .
When ; therefore . . . . .
( 3 ) .
When ; therefore . . . . . . . . . .
4 .
When , and ; therefore . . . . . . . . . . . .
( 6 ) .
From equations ( 6 ) , and ( 6 ) , we get AND VIBRATION OF SHAFTS .
If , that is to say , if the pulley be placed in the middle of the span , the superior and inferior limits coincide , and the pulley , at all speeds , revolves in a plane perpendicular to the original alignment of the shaft .
Whatever be the size of the pulley , the period of whixling is the same as the " " natural period of vibration The solution to equation [ A ] may be put in the form [ B ] , in which ratio of the distance of the pulley from the uearer bearing to the radius of gyration , and ratio of the distance of the pulley from the nearer bearing to the whole span .
Assuming certain values for and , results might be obtained relations between .
In the equation [ B ] , the ratio fixes the position of the pulley on . . . . . .
( 7 ) . . . . . . . . . . .
( 8 ) , whence , substituting in ( 2 ) we get Hence , when deflection This is unity when , wherefore Hence The corresponding value of in text is therefore or as in text .
the shaft ; and that being determined upon , the ratio will fix the size of the For the same value of , therefore , we should have different values of The following are the results obtained , in this manner , from equation [ B ] .
{ ical columns give the values of for different values of , the value of being Axed } whilst the rows denote the values of for different values of , the value of being kept the same .
27 .
Vcdues of in the , being the of from the nearer bearing .
It may be pointed out that when is very large , and the pulley near the bearing , sa that is very small , the limit for the case of the overhanging shaft ( Case S25 ) is the superior limit for the case of pulley on a shaft resting on two hearings .
the value of being the same in both cases .
The superior limit varies ffom times the inferior limit to equality with it ; and as the pulley is removed ffom AND VIBRATION OF SHAFTS bearing to the centre of the span , the limits betwee , which whirling is possible approximate more closely to each other .
28 .
Experimental Results .
The resuIts , as given by equation [ A. ] , page 308 , merely take account of the effect of one pulley , the effect of the shaft and of all other pulleys which it carries being neglected .
If be the separate speeds of whirl of the shaft pulley , on the assumption the effect of one is neglected when that of the other is under con , then it is shown in S6 page 357 , that the resutting speed of whirl due to both causes combined be to be of the form .
If the resulting speed given by this formula does not sufficiently to the observed speed , then , by the introduction , in the terms of the , of constcmt multipliers ( which determined by experiment ) , it will be shown , as occasion arises , that the speed by the may be made to sufficiently approximate to the actual speed in all cases .
, however , the resulting speed given by the will be found to give ?
.esults sufficiently accurate for purposes .
29 .
The following are the mean results of experiments made with pulleys I. and Il .
( see Chapter I. , S6 , page 285 ) in different positions .
The shaft , the ] has been investigated in SS10 , 11 ; whilst the calculated speeds for the pulleys alone have been obtained from equation [ A. ] , page 308 .
The resulting calculated speed is taken to be where are the separate speeds of whirl of the shaft and pulley .
In all cases the distances are measured from the centre of the to the centre of the pulley .
PULLEY II .
It will be noticed that in no case does the error materially exceed 3 per cent. of the observed speed .
From erimen 4 , 46 , and 54 it would appear that with the pulley near the bearing the pulley stiffens the shaft .
That is to say , the shaR would whirl at a lower speed without the pulley than with it .
The resulting calculated speed given above must , of necessity , be less than either or .
If , however , the resulting calculated speed be taken to be , where is some constant determined from the experiments , then when is large compared to , the resulting speed is , and if be less than unity this would be greater than .
In this way all the calculated results could be made higher than those given , so that in the experiments on PuUey I. ( since the observed is greater than the calculated speed throughout ) the observed and calculated results could be made to differ very slightly from one another .
however , the errors in the experiments on Pulley II .
are sometimes positive and sometimes negative , the resulting speed given by is sufficiently near the actual speed for practical purposes .
30 .
The following are the mean results of experiments with both pulleys ( and II .
) on the shaft at the same time .
It is shown in Case XVII .
, SS60 , 61 , that the only way to deal with two or more pulleys is to consider separately and then obtain the resulting speed of whirl by a formula similar to that in SS28 or 62 .
The case or the shaft only is considered in SS10 .
11 ; whilst the calculated results for each of pulleys ( considered separately ) are obtained from the preceding article .
lculated speed is to be where the speeds of whirl for the shaft , Pulley I. and Pulley separately .
The span in all the experiments was 2 ' 8 AND VTBRATION OF SHAFTS .
PULLEYS I. and II .
These experiments show that the method adopted in calculating the final resulting speed gives resuIts very little different from the observed resuIts .
In all cases the percentage error is positive , so that the resulting speed , as calculated above , is slightly below the actual speed , and , sequently , errs on the light 31 .
SHAFT RESTING ON TWO SUPPORTS , FEET APART , AND OVERHANGING ON ONE SIDE FEET , LOADED WITH A PULLEY , WEIGHT AND MOMENT OF INERTIA I ' AT ITS END .
Take the origin at B. Then from A to we have equation NDCCOXCIV .
and from to When whence . . . . . . . . . . .
( 1 . . . . . . . . . .
( 2 ) , . . . . . .
3 , . . . .
At , where therefore . . . . . . . . . .
( 6 ) .
When ( at the pulley ) , equation ( 5 ) ) , and ( S7 , equation ( 6 ) ) ; whence ( 7 ) , and The elimination of the seven ratios\mdash ; from these eight equations leads to in which and ) ( see S23 , p. 305 ) .
AND VIBRATION OF SHAFTS .
If , in equation [ A ] , we put , we get the equation already obtained for the of an ovel.hanging shaft fixed in directio1l at one end .
( Case IX .
, p. 304 .
) From equation [ A ] we get so that ( as in S24 , p. 305 ) for whirling to be at all possible must be and .
If be equal to the first or second of these quantities ) corresponding of gives the inferior or superior limit of the speed respectively .
Moreover , the of whirl corresponding to the inferior timit of speed is identical with the of vibration of the light shaft under the given The superior limit is the inferior limit multiplied by some function of the ption of the pulley , that is , some function of and .
The superior limit inferior limit If , as in Case X. , p. 308 , we put .
ratio of overhanging portion to the radius of ation , and ratio of overhanging tion to the then the solution to the equation [ A ] is [ B ] .
32 .
The following are the results obtained from this equation by assuming certain values for and , as in Case X. , S27 , p. 309 .
The vertical columns give the value of for different values of , the value of being fixed , whilst the rows denote the value of for different values of , the value of being kept the same .
MR. S. DUNKhRLbY ON WHIRLING Values of in the ) , being tength of the por.tion .
Comparing these results with those in Case X. , S27 , it will be noticed that when the is very long and the pulley near the bearing , so that very small , the whirling speeds in the two cases are the for the values of , whether the pulley be placed between the bearin , or overhang an equal distance on one side .
For any other value of , the superior limit in the present case is greater , and the inferior limit less , than the corresponding limit in Case X. , the values of and the in both cases .
In the present case the superior limit varies from times the inferior limit twice that limit when shaft orks shoulder .
33 .
Results \mdash ; The same remarks apply as in S28 , page 310 .
AND VIBRATION OF SHAFTS .
The following are the results of experiments made ( 1 ) with Pulley I. ( p. 285 ) , and ( 2 ) with Pulley II .
at the end of the overhanging portion , the ratio of the erhanging portion of the span being made to vary .
The shaft without the been investigated in SS 12 , 13 ; whilsb the calculated speeds for the puileys alone have been calculated ffom equation , S31 , page 313 .
The calculated speed obtained from the formula where are the } ) arate speeds of whirl for the shaft and pulley , gives results , in every case , much lower thah the observed results .
To bring the lculated results more in with the observed results , the resulting ealculated speed is tnken to be , the of a determined from Experiment 64\mdash ; chosen the observe of the whiring speed for the slmft alone are practically the same ( see S13 , page 294 , ExperinJent 2\mdash ; being885 .
In this expression , is the multiplier of the greatest term iu the denominator .
resulting when and when ld to a sufficient degree of approximation , whatever be the of the overhangin portion to the span , or whatever be the size of the pulley .
Nloreover , as in S29 , p. 31 ] , when the pulley is near the .
bearing the shaffi is stiffend by the pulley , and the lighter the pulley the further the distance which it might be from the bearing before this stiffenin action ceases .
( Compare S29 , Experiments 48 , 46 , 54 , and present article Experimenb 63 , 64 , 62 .
) 34 .
The following the mean results of experiments with both pulleys , I. and Il .
, on the shaft .
It is shown in Case XVII .
, SS59-62 , pp. 76-80 , that the only way to deal with two or more pulleys is to consider each separately and obtain the resulting whirling speed from a formula of the type , whel.e , are the separate speeds of whirl due to the several causes on the assumption that each cause is neglected except the one under consideration .
In the present series of experiments Pulley I. was kept at the end of the overhanging portion , whilst Pulley II .
was placed in different positions between the bearings , the distance between the bearings being also altered .
The notation used in the following results will be understood from the diagram .
To get the resulting calculated speed , the resulting speed for the shaft , and for Pulley 1 .
is obtained , as explained in the eceding article .
Let this be called .
The whirling speed for ] II .
alone is obtained from equation A. , S26 , p. 308 .
Let this be .
Then the resulting speed for both the pulleys and the shaft combined is taken to be AND VVIBRATIO OF SHAFTS .
MR. ON THE WHIRLING These experiments show that the method adopted in calculating the final resulting speed gives results approximating very closely to the observed speeds .
When the overhanging portion is small , and both pulleys are near the bearings , the shaft is stiffened ) the pulleys ( Experiment 83 , 93 ) .
fn all other cases the resulting speed as calculated above is slightly below the actua ] speed .
The consequently errs on the right side .
Case XII .
35 .
IIAFT , , RESTIN FREELY ON A SIJPPORT AT ONE END AND FIXED IN DIRECTION AT WITH A , WEIGHT AND MOMENT OF INERTIA I ' , AT A DTSTA THE SHOULDER END , AND FROM THH } We have equation , taking the the shoulder end A to the pulley , and from the pulley to B. WheJl x ; whence O. . . . . . . . . .
( 1 ) , . . . . .
( 2 ) .
When ; whence ( B\mdash ; B ' ) -\mdash ; 12 . . . .
( 4 ) .
AND VIBRATION OF vWhen whence ( 5 ) , ( 6 ) .
When ( at C ) , we have ( S7 , equation ( 5 ) ) , and ( S7 , equation ( 6 ) ) ; whence we obtain , putting as before ( S23 , p. 305 ) , and , where ( - ' ) . . . . .
( 7 ) , . . .
( 8 ) .
The elimination of the seven ratios from the eight equations marked leads to [ ] , a quadratic in , which is not , of course , symmetrical with respect to If , then , and the equation reduces to the equalion already obtained for the case of an overhanging shaft working in a shoulder .
Case IX .
, S23 , p. 304 .
) From equation [ A ] we get so that for whirling to be at all possible ( see Case IX .
, S24 , p. 305 ) we must have CCCXCIV .
and .
If be equal to the first or second of these quantities , the corresponding of gives the inferior or superior limit of the peed respectively .
The values of corresponding to these limiting values of , are zero and infinity , and if the whirl at the speeds given by them it will do so in such a manner that the puUey still rotates in a plane perpendicular to the original alignment of the shaft .
, the period of whirl corresponding to the inferior limit of speed is identical with thoe period of vibration of the light shaft under the given oonditions .
The superior limitinferior limit Let ; that is and are the ratios of the distance of the pulley the shoulder end of the shaft to the radius of gyration of the pulley and to the span respectively .
Ako let be the corresponding ratios when the distance of the pulley is measured from the free end of the shaft ; that is and Then the solution to equation [ A ] , p. 320 , may be expressed in either of the forms [ B ] , or ( C ) .
As in Cases X. and XI .
( SS27 , 32 ) , by assuming certain values for , or the corresponding values of or can be found , and so , for any particular value of or , the value of readily calculated .
Two sets of results have thus been compiled .
The first set ( obtained from equation [ B ] ) gives the values of for of and , and is applicable when the pulley lies between the shoulder end and the centre of the span ; whilst the second set ( obtained from equation [ C ] ) gives values of for different values of and and is applicabk when the pulley lies between the free end and the centre of the span .
AND VIBRATION OF SHAFTS .
36 .
of in the , lrhen the pulley lies between he shoulder end and the centre of the span , arul of pnlley from houlder end .
37 .
of in the equation , when the pulley lies betuJeen the free end and the centre of the , and of ) free end When the span is very long and the pulley is near the shoulder , so that may be taken to be very small , a comparison of the results in SS36 and 25 shows ihat the effect of the free end is nil ; in other words , the speeds are the same as if the shaft merely overhung .
If the pulley be near the end of the span , so that may be taken to be very small , a comparison of the results in SS87 , 232 shows that the effect of the shoulder is precisely the same as that of a free bearing .
These results might , of course , have been anticipated .
38 .
Comparing these results with those obtained in Case X. , S27 ( that is , with the of a on a shaft merely resting on a support at each end ) , we see that in the case one end is fixed in direction , the calculated speed for the pulley AND VIBRATION OF SHAFTS. .
alone exceeds that in the case of a shaft free at both ends , in a certain that ratio depending on the position and size of the pulley .
Considering bhe superior limits in each case , the increase of speed due to the shoulder is 100 per cent. at the shoulder end , decreasing to 91 at one-third the span from the shoulder end , 58 at the centre of the span , and zero at the free end .
Considering the inferior limits in each case , the increase of speed is 225 .
cent. Ilear the shoulder end , 51 at the centre of the span , and 19 per cent. near the free end .
Case XlII .
39 .
SHAFT SUPPORTED ON THREE BEARINGS , AND FEET APART RLSPECTIVELY AND LOADED WITH A PULLEY , WEIGHT AND MOMEiNT OF INERTIA I ' ON THE SPAN OF LENGTH , THE PULLEY BEING DISTANT FEET FROM THE MIDDLE BEARING AND FEET FROM THE END BEARING .
Thus :\mdash ; We have , taking the origin at the middle bearing ( S21 , equation 2 ) , , from Ato , from to and from to C. When ; whence we obtain . . . . . . . .
( 1 ) , ( 2 ) , . . . . . .
( 3 ) , . . . . . . . . . . .
( 4 ) .
When ; whence ( 5 ) , - . . . . . . . .
( 6 ) .
When ; whence . . . . . .
( 7 ) , . . . . . . . . .
( 8 ) .
At the pulley ( D ) we have , when ( S7 , equation 5 ) , ( S7 , equation 6 ) ; whence ( 9 ) , . . .
( 10 ) , . . . .
11 , . . .
( 12 ) , where , as in S 23 , p. 305 , and , where .
The elimination of the eleven ratios : : : : from the twelve equations marked leads to . . . . .
[ A ] , a qnadratic in which is not symmetrical with respect to If in equation we put , it reduces to AND VIBRATION OF SHAFTS .
the equation already obtained for the case of a shaft resting freely on two suppol.ts at the ends , and loaded with a pulley distant from the bearings ( Case X. , S26 , p. 308 ) .
If , the equation reduces to the equation already obtained for the case of a haft resting freely on a support at one end and working in a shoulder at the other ( Case XII .
, S35 , p. 320 ) .
If the equation reduces to the equation already obtained for the case of a shaft , span , and overhanging a distance , the pulley being at the extremity ( Case XI .
, S31 , p. 313 ) .
40 .
In the case of two spans , one of which is loaded , it is , of course , useless to completely solve the many cases which might occur .
The three cases which at once suggest themselves for full investigation are\mdash ; ( 1 .
) Unloaded span zero .
( 2 .
) Unloaded span infinite .
( 3 .
) Unloaded span equal to loaded span .
It has been shown that the first two cases have been already investigated ( Cases XIL and X It onty to solve the third case when the two spans are equat .
If equation [ A ] becomes .
[ B ] , from whioh we immediately get so that for whirling to be at all possible we must have Case , and If be equal to the first or second of these expressions , the onding value of gives the inferior or superior limit of the speed respectively .
Moreover , the period of tuhirl corresponding to the inferior limit speed is with the period of of the light shaft under the given conditions .
The superior limit inferior limit Let and that is , and are the ratios of the distance of the pulley from the middle bearing to the radius of gyration of the pulley and to either span lespectively .
AIso , let be the corresponding ratios when the distance of the pulley is measured from the bearing ; that is and Then the solution to equation [ B ] may be put iu either of the forms [ CJ ; and [ D ] .
As in Cases X. , XI .
, and .
( SS27 , 32 , 36 , 87 ) , by assuming certain values for , or , the corresponding values of , or can be found , and so , for any AND VIBRATION OF SHAFTS .
particular value of or , the value of readily calculated .
Two sets of results have thus been compiled .
The first set ( obtained from equation [ C ] ) gives the values of for different values of and , and is applicable when the pulley lies between the middle bearing and the centre of the span ; whilst the second set ( obtained from equation [ D ] ) gives values of for different values of and , and is applicable when the pulley lies between the end bearing and the centre of the span .
41 .
Values of in the when the pulley lies between the middle bearing the centre of distance from NDCCOXCIV .
42 .
of in the when the pulley lies between the free end the of the distance of putley frcnn free end .
By a comparison of these two sets of results it will be noticed that the same pulley placed at equal distances from the middle bearing and end bearing of the shaffl whirls at speeds , those near the middle bearing being higher than those near the end bearing .
Moreover , if the span be very long the pulley be the bearing so may be taken to be very small , it wiIl be seen that , whilst the limits in the two cases are the same , the ratio which the inferior limit bears to the superior limit is less when the pulley is near the end bearing than when it is near the middle bearing .
Also the superior limits when the pulley is near either bearing are the same as those obtained in Case X. , S27 , Case XI .
, S32 , and also in Case XII .
, provided the pulley is near the free end of the span .
The superior limit in any AND VIBRA OF SHAFTS .
these cases is the inferior limit obtained in Case IX .
, S25 , and also in Case XII .
, S36 , provided the pulley lie near the shoulder end of the span .
43 .
Comparing the results in SS41 , 42 with those obtained in Case X. , S27 ( that is , with the case of a pulley on a shaft merely resting on a support at each end ) , we see that in the case of two equal spans the calculated speed for the pull alone exceeds that in the case of a single span ( equal in length to either of the two equal spans ) in a certain ratio\mdash ; that ratio depending on the position and size of the pulley .
Considering the superior limits in each case , the increase of speed due to the extra span is 10 per cent. when near the middle bearing , 24 ( maximum advantage ) when one third the span from the middle bearing , 19 at the centre of the span , and zero at the end bearing .
Considering the fferior limits in each case , the increase of , ed is 35 per cent. when near the middle bearing , decreasing to 18 at the centre of the span and 8 per cent. near the end bearing .
44 .
imental Results .
The same remarks apply here as in S28 , p. 310 .
The following are the mean results of experiments made with different spans and with different positions of pulleys I. and .
( p. 285 ) on those spans .
The shaft without the pulley has been investigated in SS , whilst the calculated speeds for the pulleys alone have been calculated from equation [ A ] , S39 , p. 325 , or , in the case of equal spans , from equation [ B ] , S39 , p. 326 .
PULLEY I. PULLEY II .
AND VIBRATION OF SHAFTS .
These experiments show that the formula used for calculating the resulting speed , where are the separate speeds of whirl\mdash ; holds , to a sufficient degree of approximation , whatever be the ratio of the spans or the position and size of the pulley .
When one span is small compared to the other ( Experimeuts 119-125 and 98-104 ) , the conditions approximate to those required in Case XII .
, S35 .
In this case the error is sometimes positive and sometimes negative , and the percentage error , with one exception , is under 3 .
The average error .
In Experiments 105-108 and 126-129 , in which the ratio of the spans is one-fifth , the error is practically positive throughout .
The mean error is .
The calculated results couId be made to approximate more closely to the actual speeds by using the formula instead of , in which case the errors in Experiments 126 , 127 , 128 , 129 , 105 , 106 , 107 , 108 , would be , 4 , per cent. respectively , giving a mean error per cent. But the speeds , as obtained by , are sufficiently near the actual speeds for practical purposes .
When the spans are equal ( Experiments 109-11S)\mdash ; which is the mos6 important case\mdash ; one span being loaded , the error is sometimes positive and sometimes negative , but in only two cases ( Experiments 111 , 117 ) does it slightly exceed three per .
cent. The mean error is per cent. The experiments , therefore , amply verify the theory .
45 .
The following are the mean results of experiments with the Pulleys I. and II .
on the shaft at the same time .
The spans have been each taken to be 16 inches .
The first series include these experiments wibh Pulleys I. and II .
on different spans , and the second with them on the same span\mdash ; the positions in the two series being similar .
The notation used to determine the position of the pulleys is the followare the distances of Pulley I. and of Pulley II .
from the middle and outer bearings .
The resulting calculated speed is taken to be , where are the separate speeds of whirl for the shaft , Pulley I. , and Pulley II .
( see Case , SS 59-62 ; also SS 30 , 34 ) .
The calculated speed for the two pulleys neglecting the shaft is given , having been calculated from the formula .
For the three causes together the resulting speed is , which is equivalent to AND VIBRATION OF SHAFTS .
The formula by which the resulting speed is calculated , viz. :\mdash ; , gives , of course , the same calculated speed whether the pulleys be on different spans , or similarly placed on the same span .
The experiments show that , with the pulleys on different spans , the observed speed is higher ( with one exception ) than when pulleys are similarly on the same span .
In Expeliments 138 and 148 the observed speed is the same in each case .
Moreover , with the pulleys on different spans , the observed speed is , with one or two exceptions , in excess of the caJculated speed ; whilst , when on the same span it is , without exception , less than the calculated speed .
In the fornler case , the average error is about per cent. , and in the latter , about per cent. , giving a mean of \mdash ; 1 per ce , nt .
Either one or other of the separate errors ( Experiments 130-140 or 141-149 ) could be reduced by the introduction of a constant in the denominator of the expression , as in SS33 , .34 , but whilst reducing one it would also increase the other .
Considering , however , the complexity of the probtem the eceding results justify , to a degree , the ssumptions thave to be made in the course of the investigation .
The experiments made with the pulleys on different spans are very instructive as showing how one pulley affects the other in regard to whirling .
For example , Experiments 130 , 131 , ] 35 , and show that , when the two pulleys are both taken into account , the calculated speed is much too low .
Hence we may infer that if Pulley I ( which is the lighter of the two ) be placed on the far side of the centre of its span from the middle bearing , its effect on the whirling speed is very small .
The whirling speed may , in fact , be taken as that resulting from the combined effects of the heavier pulley and the shaft .
On this assumption , the calculated whirling speeds in the above four experiments would be ( see S44 , Experiments 110 , 109 ) 3056 , 3056 , 2600 cind 2600 respectively , and the percentage errors would , instead 46 .
The discrepancies between the observed and calculated results are accounted for by the fact that the empirical formula\mdash ; viz. , \mdash ; upon which the resulting calculated speeds are based , is not strictly accurate .
In the case , however , of two or more equal spans with pulleys on each span , that formula gives calculated results less than the observed results and , therefore , erring on the right side .
This is apparent from Experiment 130-140 , but it might also be proved by considering the case of two equal spans with a pulley placed in the centre of each span .
If the two pulleys be of the same size and weight they will have , separately , the same whirling speed .
Let that whirling speed be .
Then , from S41 or 42 , we have , about , where length of a single span .
Using the ordinary formla , the resulting MR. S. DUNKERLEY ON THE WHIRhlNG whirling speed , due to both pulleys , will be ( see S62 ) , so that the whirling speed for the two pulleys will be proportional to .
But , since the two spans are equal and similarly loaded , it is clear that there is no bending moment on the middle bearing .
Consequently , as in the case of an unloaded shaft , S , the spans will whirl independently of each other , and the actual speed of whirl will bherefore be proportional to . . . . . .
( see S27 ) .
where and have the same values as before .
Hence the whirling speed , as given by the formula used in the investigation , is only 84 per cent. of the actual whirling speed of the pulleys .
When the shaft is also taken into account the difference between the two calculated will be decreased by an amount depending upon the relation between the whirling speed of the shaft , taken separately , and the whirling speeds for the pulleys as calculated above .
Reasoning in a similar way , we may conclude that when the spans are not similarly loaded , the whirling speeds as obtained in the investigation will be less than the actual whirling speeds .
In other words , the formula used to determine the resulting speed of whirl errs on the right side .
Case , WEIGHT , AND MOMENT OF INERTIA I ' , AT DISTANCES FROM THE SHOULDERS .
Thus\mdash ; Take the origin at A. Then from A to we have ( S21 , p. 304 , equation 2 ) and from to AND VIBRATION OF SHAFTS .
When ; whence ( 1 ) , . . . . .
( 2 ) .
When ; whence ( B\mdash ; B ' ) ( 3 ) , ( A\mdash ; A ' ) . . . . .
( 4 ) .
whence .
( 5 ) , . . . . . .
( 6 ) .
When ( at the pulley ) , ( S7 , equation 5 ) , and ( S7 , equation 6 ) , wtlence we obtain , putting as before ( S23 , p. 305 ) , and , where , . . . . .
( 7 ) , ( \mdash ; ' ) . . . .
( 8 ) .
The elimination of the seven ratios the eight equations marked leads to . . .
[ ] , a quadratic in which is symmetrical with respect to NDCCCXCIV .
2 X If , then and the equation reduces to the equation already obtained for the case of an overhanging shaft working in a shoulder ( Case IX .
, S23 , p. 305 ) .
From equation [ A ] we get so that for whirling to be at all possible ( see Case IV .
, S24 , p. 305 ) , must be and .
If be equal to the first or second of these quantities , the corresponding value gives the inferior or superior limit of the speed respectively .
Moreover , the of whirl corresponding to the inferior limit of speed is identical with the of of the light shaft under the.qiven conditions .
The superior limitinferior limit ; that is , and are the ratios of the distance of the pulley from the nearer shoulder to the radius of gyration of the pulley , and to the whole span respectively .
Then the solution to equation [ A ] , p. 337 , may be expressed in the form [ B ] .
As in Case X. ( S 27 ) , by assuming certain values for and , the values of can be found , and so for any particular value of the value of calculated .
rfhe following are the results obtained from equation [ B ] .
The vertical give the values of for different values of , the value of being fixed ; whilst rows denote the values of for different values of , the value of being kept same .
AND VIBRATION OF 48 .
of the of pulley from the shoulder .
It will be noticed that when the span is very long and the pulley near the shoulder , so that may be considered very small , the whirling speeds , for the same sized pulleys , are the same as those obtained in Cases IX .
and X. , SS25 and 36\mdash ; the value of being the same in all three cases .
The superior limit varies from twice the inferior ( when the span long and the pulley near the shoulder ) to equality with it ( when the pulley is at the centre of the span ) .
49 .
Comparing the results contained in the article with those ) Case XII .
, SS36 , 87 ( that is with the case of a shaft working in a shoulder at one end and resting on a bearing at the other ) , we see that in the case where both ends work in a shoulder the calculated speed for the pulley alone exceeds that in the of a shaft free at one end and working in a shoulder the other , in a certain that ratio depending on the position and size of the pulley .
Considering the superior limits in each case , the increase of speed due to the two shoulders is zero at the shoulder end , increasing to 27 per cent. at the centre of th6 span , and 100 per cent. at the free end .
Considering the inferior limits in each case , the increase of speed due to the two shoulders is 2 per cent. .
the shouIder end , increasing to 32 at the centre of the span , and 180 per cent. near the free end of the shaft .
Again , comparing the results obtained in the present case those obtained in Case X. , S27 ( that is , with the case of a shaft merely resting on a bearing at each end ) we see that , considering the superior limits in each case , the increase of speed due to the two shoulders is 100 per cent. , whatever be the position of the pulley ; whilst considering the inferior limits the increase of speed near the bearing is 233 per cent. , decreasing to 100 at the centre .
50 .
SHAFT SUPPORTED ON FOUR BEARINGS , , AND FEET APART RESPECTIVELY , AND LOADED WITH A , WEIGHT , AND MOMENT OF INERTIA I ' , ON THB OUTER SPAN OF LENGTH \mdash ; THE PULLEY BEING DISTANT FROM THE AND FEET FROM THE OUTER BEARING .
Thus\mdash ; We king the origin the bearing , ( S21 , equation 2 ) .
, from A to , from to from to and to AND VIBRATION OF When ; whence . . . . .
( 1 ) , . . . . .
( 2 ) , . . . . .
( 3 ) , . . . . .
( 4 ) .
When ; whence ( 5 ) ( 6 ) .
When , ; whence . . . .
( 7 ) , .
( 8 ) .
When ; whence . . . .
: ( 9 ) , - . . . .
( 10 ) , S . . . . .
( 11 ) , . . . . .
( 12 ) .
Again , at the pulley , when iiou 5 , ( .S7 , equation 6 ) ; MR. S. DUNKERLEY ON THE WHIRLING hence , as before , therefore , where , . . . .
( 14 ) , ( 15 ) , ( 16 ) .
The elimination of the fifteen ratios from the sixteen equations marked leads to . . . . . .
a quadratic in which is not symmetrical with respect to If in equation ] we put , it reduces to which is identical with the equation obtained independently , but not reproduced of two spans , one of which is loaded , the outer end of the shaft on the loaded merely resting on a bearing , whilst the outer end of the unloaded span works in shoulder .
'fhus\mdash ; AND VIBRATION OF SHAFTS .
If , in addition , in the above equation we put ( i ) and ( ii ) , it further reduces to the two equations already obtained in Case XII .
, S ' 35 , p. 320 , and Case X. , S26 , p. 308 .
By making and each equal to infinity , the equation further reduces to which is the equation for a single span overhanging at one end and working in a shoulder the other , the pulley at the end of the overhanging portion .
Thus By putting , in this equation , we obtain the equation already obtained in Case IX .
, S23 , p. 305 .
If in equation [ A ] we put , it immediately reduces to the already obtained for the case of a shaft of two spans , the shaft merely resting on the bearings at the ends , ]oaded with a on one of the spans .
( Case XIII .
, S39 , p. 325 .
) Again , if in equation [ A ] we make and each equal to infinity , it reduces to which is the equation for an overhanging shaft loaded at the end , and having two spans on one side .
That is , for the case of By making last equation , it becomes identical with that already investigated for the case of a single span one side ( Case S31 , p. 313 ) . .
In the case of three spans , one of the end ones of which is loaded , the three cases which at once suggest themselves for fnll investigation ( 1 ) .
The two nnloaded spans zero .
( 2 ) .
The two unloaded spans infinite .
( 3 ) .
All the three spans equal .
It has been shown that the first two cases have been already investigated ( Cases XII .
and X It only remains to so the third case when the spcms equed .
If equation [ A ] on p. 341 , becomes [ B ] , which we immediately get so that for whirling to be at all possible ( S24 , p. 305 .
) must be and AND VIBRATION OF SHAFTS .
If be equal to the first or second of these expressions , the corresponding value of gives the inferior and superior limit of speed respectively .
Moreover , the period of whirl to the inferior limit of speed is identical with the natural period of vibration of the tight shaft undethe given conditions .
The superior limitinferior limit .
Let and ; that is , and are the ratios of the distance of the pulley from the inner bearing to the radius of gyration of the and one of the spans respectively .
Also , let be the corresponding ratios when the distance of the pulley is measured from the end bearing ; that is , and Then the solution to equation [ B ] may be put in either of the forms .
[ C. ] .
[ D. ] As in Cases X.-XIV .
( SS 27 , 82 , 36 , 37 , 41 , 42 ) , by certain values for , or , the corresponding values of , or can be found , and so , for particular value of or , the value of is readily calculated .
Two sets of results have thus been compiled .
The first set ( obtained from equation [ C ] ) gives the values of for different values of and , and is appIicable when the pulley lies CCXCIV .
between the inner bearing and the centre of the span ; whilst thS ( obtained from equation [ D ] ) gives values of for different values of and is applicable when the pulley lies between the end the span .
52 .
Vcdues of in the when the putley lies the bearing the centre of the span , and of puiley bearing .
AND VIBRA TION OF SHAFTS .
53 .
Values of in the equation when th lies between the outer bearing and centre of the span , of pulley from outer bearing .
The same pulley placed at equal distances from the inner and outer bearings whirls at different speeds , those nearer the inner bearing being considerably higher than those near the end For further remarks , see those made in S 42 , p. 329 , which apply also to the present case .
54 .
Comparing these results with those obtained in Case XIII .
, SS41 , 42 , that is , with the case of two equal spans ( one of which is loaded ) , we see that in the case of three equal spans , an outer span of which is loaded , the calculated speed for the pulley alone exceeds that in the case of only two spans in a certain atio , that ratio depenffing on the position and size of the pulley .
MR. S. ON THEi WHIRLING Considering the superior limit in each case , the increase of speed due to the additional span of length equal to the length of either of the two spans ( the two unloaded spans being on the same side of the loaded span ) is 11 per cent. when near the inner bearing , when one-third the span from the inner ( maximum advantage ) , at the centre of the span , and zero at the outer bearing .
Considering the inferior limits in each case , the increase of speed is per cent. when near the inner bearing , decreasing to at the centre of the span and per cent. at the end bearing .
We thus see , in the present , the effect of the second span from the one , in increasing the speed at which the pulley will cause the shaft to whirl , be such to the increase in the whirling speed to exceed 8 per cent. of tculated on the ssumption the effect of second span is cdtogether neglected .
When the effect of the shaft is also taken account , the increase in the whirling speed due to the third span will be less than 3 per cent. ( S62 ) .
Case 55 .
SHAFT SUPPORTED ON FOUR BEARINGS , , AND FEET APART PECTIVBLY , AND LOADED WITH A PULLEY , WEIGHT , AND MOMENT OF INERTIA , ON THE MIDDLE SPAN OF LENGTH \mdash ; THE PULLEY BEING DISTANT FEET FROM THE BEARINGS .
Thus\mdash ; We have , taking the origin at the bearing ( S21 , equation , from A to , from to from to from to AND VIBRATION OF SHAFTS .
When ; whence . . . . .
( 1 ) ; . . . . . . . . . .
( 3 ) , . . . .
( 4 ) .
When ; whence ( 5 ) , ( 6 ) .
When ; whence . . . . .
( 7 ) , . . . .
( 8 ) , . . .
( 9 ) , . . . . . . .
( 10 ) .
When ; whence ( 11 ) , . . . . . . . .
( 12 ) .
Again , at the pulley , when ( S7 , equation 5 ) , ( S7 , equation 6 ) ; ence ( 13 ) , ( 15 ) , .
( 16 ) , where , as in S 23 , p. 305 , , and , where The elimination of the fifteen ratios from the sixteen equations marked leads to . . . . . . . .
[ A ] , a quadratic in , symmetrical with respect to and .
If in equation [ A ] we put , it reduces to which is identical with the equation obtained independently , but not reproduced here , of two spans , one of which is loaded , the er end of the shaft on loaded span in a shoulder , whilst that of the uuloaded span merely rests on the support .
Thus AND VIBRATION OF SHAFTS .
If we further put , the equation reduces to that already obtained for the case of a shaft working in a shoulder at each end ( Case XIV .
, S47 ) .
If , instead of , we obtain the equation for the case of a shaft working in a sleeve at one end and merely resting on a bearing at the other ( Case XII .
, S35 ) .
If in equation [ A ] we put , it reduces to that already obtained for two spans , ons of which is loaded XIII . .
If in addition we obtain Case X. , S26 .
56 .
In the case of three spans , the middle one of which is , the three cases which at once suggest themselves for full investigation are\mdash ; ( 1 .
) the two unloaded spans zero , ( 2 .
) the two unloaded spans infinite , ( 3 .
) all three spans equal .
It has been shown that the first two cases have been already XIV .
and X It only to solve the case uhen the equal .
If , equat , ion [ A ] reduces to . . . . .
[ B ] , from which we get so that for whirhng to be at all possible we must have ( see Case IX .
, S24 , p. 305 ) , and If be equal to the first or second of these expressions , the corresponding value of gives the inferior or superior limit of the speed respectively .
Moreover , the period of whirl to the inferior limit of speed is identical with natural period of vibration of the light under the given conditions .
The superior limit inferior limit .
Let and that is , and are the ratios of the distance of the pulley from the nearest the radius of gyration of the pulley and to the span respectively .
Then the solution to equation [ B ] p. 350 , may be expressed in the form .
[ C ] .
As in Cases X. and XIV .
( SS27 , 48 ) , by assuming certain values for and , the corresponding values of can be found and so , for any particular value of , the value of readily calculated .
The following are the results obtained from equation [ C ] .
The vertical columns give value of for different values of , the value of being fixed ; whilst the rows denote the value of for values of , the value of being fixed .
AND VIBRATION OF 57 .
of in the equation pulley from the The superior limit thus varies from times the ]imit ( when the pulley is near the bearing ) to equality with it ( at the centre of the span ) .
Moreover , when the span is very long , and the pulley near the bearing , so that may be taken to be very ] ] , no whirling can place provided the radius of gyration is less that the distance of pulley from the bearing .
( See also SS27 , 32 , 37 , 41 , 42 , 58 .
Comparing these results with those obtained in Case X. , S 27 ( that is , with the case of a single span ) , we see that in the case of three equal spans , the middle one of which is loaded , the calculated speed for the puUey alone exceeds that in the case of a single span in a certain ratio\mdash ; that ratio depending on the position and size of the pulley .
onsidering the superior limits in each case , we see that the speed due to two additional spans ( each equal in length to the first span ) , one on each side , is , as regards the superior limits , 10 per cent. near the bearing , and per cent. at the centre of the span ; and , as regards the inferior limits , it is per cent. near the bearing , and per cent. at the centre of the span .
Again , comparing the case under discussion with Case XIV .
, S48 , in which the two end spans are zero ( that is , the sha , ft works in a shoulder at each end ) , we see that the increasB of speed , due to the two shoulders , is , as regards the superior limits , 100 per cent. near the bearing , and per cent. at the centre of the span ; and , as regards the inferior limits , 135 per cent. near the bearing , and per cent. at the centre of the span .
Comparing the present case with the resuIts obtained in Case XIII .
, SS , 42 ( that is , with the case of two equal spans , one of which is loaded ) , we see that the increase of speed due to the span ( the loaded span being in the middle ) is , as tegards the superior limit , zero when the pulley is near the inner bearing , increasing to per cent. at the centre , ( maximum advantage ) at one-third the length of the span from the free end , and decreasing to 10 per cent. when near the free end ; and , as regards the inferior limits , the increase of speed is 5 per cent. when near the middle bearing , at the centre of the span , and per cent. at the outer bearing .
Finally , the results in the present case with those obtained in Case XV .
, SS 52 , 53 ( that is , with the case of three equal spans , one of the end ones being loaded ) , it will be noticed that the results in the latter case are the higher when the pulley is near the bearing , as regards the superior limits , but less as regards the inferior limits .
By further referring to what was proved in S 54 , p. 346 , we infer that if , in the present , cm dditional espan be added on side middle , the effect of those , in increasing the speed which the pulley will to whirl , will never be such as to the in the whirling speed to exceed one or two per cent. of that calculated on the assumption that the effects of the two additional spans altogether neglected .
When the effeet of the shaft is also taken into account , the increase in the whirling speed , due to the two additional spans , will be still further reduced ( S62 ) .
This result , and that obtained in S54 , are extremely important , as they practically limit any problem to the case of three spans .
In other words , in the case of a continuous shaft , supported on bearings , placed at equal distances apart , and Ioaded with a pulley on one of the spans , the whirling speed , due to that obtained by considering the loaded span , and bhe span or spans immediately adjacent to it on either side , is sufficiently accurate for practical purposes .
AND VIBRATION OF SHAFTS .
Case of two or more Pulleys .
59 .
So far ( Cases IX.-XVI .
, SS 23-58 ) we have only fully investigated the effect of one pulley on a shaft supported in different ways , the effect of the shaft being neglected .
It was shown in SS 19 , 20 that , even in simple the equations obtained by considering the shaft and a single pulley together were too complicated to allow of a soIution in a form convenient for practical application .
The following case show that , even in the simple case of a shaft freely supported at the two ends , the equations obtained by considering the effect of the two pulleys together\mdash ; the effect of the shaft being together neglected\mdash ; are also too complicated to allow of a solution which can be readily applied to any actual case .
Case 60 .
SHAFT , LENGTH , MERELY RESTING ON A SUPPORT AT EACH END AND LOADED WII'H TWO PULLEYS , iHTS W , AND MOMENTS OF INERTIA , PLACED AT DISTANCES AND RESPECTIVELY , FROM THE BEARINGS .
Thus\mdash ; Taking the origin at the bearing , we have ( S21 , equation 2 ) , from A to , from to from to B. Wheu ; whence . . . . . .
( 1 ) , . . . . . . . . . . . . . . .
( 2 ) .
MR. S. DUNKERLEY ON THE wmRLING When ( S7 , equation ( 5 ) ) , ( S7 , equation ( 6 ) ) ; whence , ( 3 ) , . . . . . .
( 4 ) , ( 5 ) , ( 6 ) .
Similarly , when , we have . . .
( 7 ) , . . . .
( 8 ) , . . . .
( 9 ) , . . .
( 10 ) .
Again , when whence . . . . .
( 11 ) , ( 12 ) .
tting and , where , and , where AND VIBRATION OF SHAFTS .
the elimination of the eleven ratios leads to [ A ] , in which If the second pulley be supposed removed , that is , if we put and each equal to zero in equation [ A ] , we get a result , of course , identical with that already obtained ( Case X , S26 , p. 308 ) .
It be seen at once that the equation [ A ] is practically useless unless some special relation be assumed between the dimensions of the pulleys , \amp ; c. , and even then it would be impossible to coml ) a table which could be used except few cases .
Cases , other than the above , in hich a shaft is supported in a certain manner and carries two pulleys ( for example , a span with an overha1lging portion on one side and supporting one pulley between the bearings and another at the end of the overhanging portion ) have been investigated , and in each case the result obtained was too complicated to admit of any practical assumption .
61 .
The only alternative method is to consider the effects of the atever its mode of support ) each of the pulleys be their , position , size ) , and so obtain the whirling speed for on the ssumption that the others are neglected .
By of an the whirling speed , the effects of the shaft and of the pulleys into account , may be from the separately calculated whirling speeds .
62 .
The particular form of the } ) irical formula was found as follows:\mdash ; If a weight be supported by a spring which requires pounds to stretch it one foot , then the number of vibrations which that weight makes per second is .
The number of vibrations which a second weight ( attached at the same of the spring as the weight ) makes is ; and the number which the combined weight would make is In the same manner this formula would be strictly accurate in the case of a rod , .
however supported , provided that any concentrated loads which it might carry could be supposed concentrated at the same , point .
For example , if three loads be concentrated at the same point of the rod ( the effect of the rod being neglected ) , and if the number of vibrations which each makes per second , when assumed independent of the others , be , then the number of vibrations of the three together will be and so on for any number of Ioads .
If , however , the loads be concentrated at different points , the above formula will not be strictly true ; for , in addition to the number of vibrations varying inversely as the square root of the weight , the value of will vary with some function of the distance of the weight from the point of support .
In the same manner , in the whirling of shafts , if be the whirling spee due to three pulleys when each is considered independently of the remaining two , have ( SS25 , 27 , 32 , 36 , 37 , 41 , 42 , 48 , 52 , 53 and 57 ) , since therefore as , where diameter of shaft , AND VIBRATION OF SHAFTS .
where are constants depending on the position and size of the pulleys ; and ( assuming all the pulleys to be so near together that each affects the others ) , the formula . . . . .
( A ) will only be correct provided Even when these relations do not hold , it is shown in SS 29 , 30 , 33 , 34 , 44 and 45 , that the formula ( A ) is , with certain modifications and restrictions to suit particular cases ( SS33 , 34 , 45 ) , sufficiently accurate for practical purposes .
The formula ( A ) may be extended to any number of disturbing elements .
If , for example , there were four , and the speeds corresponding to them be then the resulting whirling speed is and so on .
Considering case of two disturbing elements , if their speeds of whirling , taken separately , be each equal to , the resulting whirling speed due to two causes combined is If there were three urbing elements , and if their speeds of whirling were all equal to , the resulting whirling speed would be Of two disturbing elements , if the whirling speed for one of them be four times that of the other , the resulting whirling speed is not more than three per cent. less than the smaller whirling speed .
Concluding 63 .
In conclusion , it should be noticed that in finding the speed at which a continnous shaft of given diameter , supported on bearings placed , equal distances apart , and loaded with pulleys on any or all of the spans , will whirl , the first step is to find the span which will have the biggest whirl ( that is to say , the span which carries the heaviest and most advantageously situated pulleys as regards whirling ) , and to consider this span and the spans immediately adjacent to it on either side .
The spau in question can , in general , be determined on , at a glance , from the consideration of the weights , sizes , and positions of the pulleys which each span carries .
Having fixed upon the three spans , the next step is to find ( by the formula for each case ) the whirling speed for the shaft and each of the pulleys on the three spans in question , on the ion that the effect of every , except the one under discussion , is neglected .
The resulting whirling speed may then be obtained by an empirical formula of the form given in the preceding article .
The speed : obtained will be less than the actual speed of whirl ( see S 46 , p. 384 ) .
A nearer approximation to the actual speed might be obtained by considering only those which lie near the centres , or between the centres of the side spans and the bearings of the middle span ( see S 45 , p. 334 ) , neglecting the effect of those pulleys which lie beyond the centres of the side spans .
In doing so , however , there a danger of the calculated speed exceeding the actual , whilst , by taking the pulleys on the two sides into account , the calculated speed wiIl be less than the actuai speed ( see S46 , p. 334 ) .
[ The above method of solution and the consideration or only three adjacent spans .
is based on the results arrived at in SS54 , 58 , pp. 347 , 353 .
It has been verified , not only by experiments made with the experimental apparatus , but also by experiments made on actual cases of shafting carrying heavy pulleys .
] In the case of a continuous shaft of equal spans which are all similarly loaded , each span whirls independently of the rest , and the problem , therefore , reduces to the case of a shaft loaded in a given manner and merely resting on a bearing at each the distauce between the bearings being the same as between those of the continuous shaft .

