Illinois institute

of Technology

Libraries

AT 285

Cooper, Howard Friction tests of lubricating oils-influence

FRICTION TESTS OF LUBRICATING OILS —INFLUENCE OF RATE OF FLOW

A THESIS

PRESENTED BY

HOWARD COOPER CLIFFORD MILTON LARSON

TO THE

PRESIDENT AND FACULTY

OF

ARMOUR INSTITUTE OF TECHNOLOGY

FOR THE DEGREE OF

BACHELOR OF SCIENCE IN MECHANICAL ENGINEERING

HAVING COMPLETED THE PRESCRIBED COliRSE OF STUDY IN

MECHANICAL ENGINEERING MAY 29. 1913

aiNOiSlNSTiTUTEOFTECHNOLOGV PAUL VGALVIM LIBRARY

35 WPST33RD STREET

CHICAGO, IL 80616 ^^

INDiiX.

Page

Introduction -------------- i

Testing L.'achine ------------ 3

Feeding Apparatus ----------- 4

I cthod of Procedure ---------- 6

Determining of Iv:oment - - - - 6

Running of Test ------- 10

Detern";ining characteristic

of oils --------- 13

Theory ----------------- I6

Sample 1;

Physical characteristics - - - 19

Data ------------- 20

Curves ------------ 25

Sample 2;

Physical ch.-riict^ristics - - - 31

Data ------------- 32

Curves ------------ 37

Calculations ;

foment ------------ 43

Runs - ----------- 44

Discussion --------------- 47

Conclusion --------------- 50

ILLUSTRATIONS.

Page

Photograph of I.cchine ------------ Frontiiipiccc

Arre.ngement of ^-aciiine for Deter.i.ining

I'oment ------- ---------- 8

Arrangement of i.achine During Runs ----- 9

Viscosimeter ---------------- 14

Flash Test Apparatus ------------ 15

Drawing of ft-achine ------------ Appendix

FRICTION TESTS OP LUBRICATING OILS -Il^LUMCE OF RATE OF FIOW-

With the view of determining the influence of the different rates of flow of oil on the bearing friction, the following series of comparative tests, ranging from the low restricted rates of feed to the "flooded" or oil hath method, were carried out on a railroad luhricant testing machine. This machine was selected for the tests because of the wide vari- ation of hearing pressures, and of the promtitude of adoption, which could he obtained by its use. Then too, the condition of constant speed was afforded by the variable speed motor used to drive this machine.

The most serious loss encountered in the mauH- facturing world is that of the waste power caused by friction. Prof. Peabody states that 5 per cent and often greater than 15 per cent of the indicated horse pov/er of a steam engine is expended in overcoming the frictional resistance; whereas Archbutt says that from 40 per cent to 80 per cent of the 10,000,000 h.p. used in Great Britain is consiimed by friction;

' ' *to ■:: 'lift •il*

«. ^, ^ ■L«t

i.l t

^.> -'i

:ift€Nd aC> ^ci;«

rri •'•Iff . ,

probably in the United States this fraction would "be much greater due to the cheapness of fuel.IIuch of this needless waste caused by friction as well as that due to neglect in reclaiming the oil is gradually being reduced to a minimiun by the in- stallation of oiling systems.

The lubrication of bearings, guides, and all external rubbing surfaces may be performed in e number of ways. These parts may be given an inter- mittent application of oil; they may also be supplied restricted rates of feed; or they may be flooded with oil. The intermittent feed, v/hich is effected by the occasional use of an oil can, is mostly limited to moving parts carrying light pressures which do not easily permit the use of other systems. The restricted feed, or method of lubrication by means of cups from which oil is fed to the rubbing surface by drops is in majority in the average plant. The flooded arrange- ment is effected by allowing a continuous flow of oil, which is forced to the various parts either by gravity or pressure from a pump, to completely"flood" the bear- ing. In the latter system the oil is used over and over again, that lost by leakage and depreciation being replenished by the addition of new oil.

t al

•* rv t ^«J^f^r^**f«rl •if t^firt K*njA'Yrr!c; "ftfrfifrfirt T«rf«*^f<

»ti\l iiA a«v'	■■ iq »«i .';^«)« le '
» XMf	rq« t;
r^ A T "i»r! -	••nA^ "n »!6""^T "^ptn 'T^t'
■.,^«o«ri» e<i i	.^ci iitf;- .110
' J0« nl	4 !« env lactoluaooo
■^f> {'nf!^ nt^Txr^.mri	tan 5?*tan "irrtvon

As was mentioned before, the machine employed in the series of tests was a railroad-lubricant testing machine, This machine, ao shov.Ti in the enclosed drawing-, consists of a shaft, which has a pulley, carried between two bearings, and driven by a belt from an overhead line shaft. The shaft of the machine extends beyond one of the bearings so that on the overhanging part is a pendulum which contains the test brasses. To insure that all of the surface does actually rub against the shaft, the brasses were so cut thrt instead of the width of the projected area being that of the diameter of the shaft, it was approximately three-quarters of the diameter. To these brasses a pressure is exerted by two heavy springs placed one inside the other. When these springs are compressed between the lower end of the pendulum, which is a pipe fit- ted with a castiron pltig, and a nut on the upper part of the springs, the reaction on the plug la transferred through the bolt to a plunger that is pushed against the lower Journal bearing. The mag- nitude of the total pressure on the bearing is in- dicated by a pointer attached to the nut, which moves along a graduated scale on the pendulum. The graduations were laid out according to the manu-

yj

.-?

aX t-

+ f.rvtp'^-

facturer's calitration curve for these spring-s.The deviation of the pendulum is measured by a gradua- ted arc fastened to the frame of the machine and a pointer which Is attached to the upper prrt of the pendulxun. The bearing temperature is given by a thermometer which is placed in a hole v/hich is drilled through the bearing into the babbitt metal. This hole is filled with oil so that the bulb of the ther- mometer may be completely submerged.

To make the machine more sensitive, that is, to increase the deflection for the same amount of friction, a large cast iron ball was added above the bearing that the moment due to the necessary heavy construction of the pendulum might be reduced. In case the friction becomes too excessive, the counterbalance can be lowered on its rod and securely held in position by means of a set screw. To pre- vent the pendulum from doing injury to anyone who might be in its path, a guard was placed on the base of the machine so as to restrict the pendulum to an arc of 25 degrees from its vertical position, v/hen an excessive frictional load is applied.

The feeding apparatus used to regulate the sup- ply of oil to the test bearing is that of the gravity

.1 0

« al

r'l

-ilitd a i

iiiiJ f i'> ■

en fl« q f ofoff

• A A'

'f f*

•xl

%o Hi.

rt fra:*

1 ol

type. The oil, which is containerl in a tank mounted on a platform scale, which in turn rests on a movahle suspended shelf, is supplied to the bearing-s through flexible rubber tubing and branches of pipinp leading to the brasses. Inserted in this piping are two small stop cooks which are used to control the rate of flow. ?rom the brasses the oil is distributed over the Jour- nal through grooves r\m diagonally across the babbitt face from the inlet holes, thus giving even and equal distribution. These channels are carefully gaged for an even flow to prevent dry spots or streaks appearing on the journal accompanied by sudden greatly increased friction. The grooves which are 1/8 inch wide and I/I6 inch deep carried the oil on the top surface length- wise of the bearing so that the moving surface passing the grooves or chamfers was bathed in, and coaterl with oil. In the lower bearing the grooves were cut so as to wipe off the oil, which has squeezed over toward

the end of the bearing, and thus prevented as little as possible working out. That which did pass out was drained down through holes in the plunger, nut, and plug to a receptacle where it was caught.

After enough oil had gathered in the reservoir, the oil was passed through a "White Star" filter where

jjf:

S-^'f •'^

J 8\I

1" St rr

.J

O TO

e.

the ant i-lubrl Gating natter was taken out so as to render it fit to "be used again. This reuse of the oil did not change the characteristics to any appre- ciahle amount as far as could he Judged from the re- sults ohtained. For, data acquired on different runs fulfilling the sane conditions excepting that in one case new oil was used and in the other, oil several times reclaimed, showed practically no discrepancies. According to tests made on a liartin's oil testing machine of oil used over and over again and recovered under proper conditions, the characteristics showed no variation hetween the values before and after use outside the limits of possihle ohservation. But use slightly increases the density, although it does not materially change the viscosity.

Before the machine was used for the tests, it was entirely taken apart so thrt it could he thorough- ly cleaner" v/ith gasoline in order to remove dirt, grit, and traces of other oils previously used.

While the machine was in this condition the vari- otiS parts were placed on a platform scale and the en- tire mass weighed.

Although the machine had been used, the moment had never heen actually determined. To ascertain

. ■ 'itiij r . ■ ■ --1

•. 0+ notff^t-^B-f : •/(* "0

^ ■ <J.;! i.' ■■ • , " . ■ Q ■."•■i.;;k

. • inoJ ' ■ rl:

isloanqetoaiJi! oc ^j i^oij ii:iT:(i i^jmim ^Jb^miaiosi ^^o-iJ:*

"'3.vori3 uoi.taitoToxr.':'>,.'lo o: , iiftnir

oliios r 'q a no ■ -^ro

f

this the pencl^^l^^n, with "brasses removed, was sup- ported in a horizontal position on a three-cornered file as follows: with the pendulum hanp-ing- verti- cally a line was drawn on the bearing- box at the intersection of the horizontal plane thi'u the Jour- nal axis ; the pendulum was then svamp into a hor- izontal position and the line of contact with the file was made to coincide with this line Just found. A strut upon a platform scale supported the free end of the pendulum, which was then leveled by means of wedges and a spirit level. The weight indicated by the platform scale was recorded as well as the dis- tance between the centers of support. Of course from the total weight, the weight of the strut and wedges had to be deducted. The accepted moment was taken as the average of the three sets of readings, the product of thr, net weight and the distance between the sup- ports being the only calculation needed in each case. The omission of the brasses had no effect on the re- sult because both of them weighed the same and occu- pied similar positions on either side of the aTis so that even if they had been in place, each would have had a neutralizing effect on the other.

The determination of this moment is very impor- tant for upon its accuracy hinges the value of the .c

ti! XO<i ^nltm^d flkrf-^ no sr:

O

no ^o«1:l« '. ••■asitf tjn'.t ^o nc •rfT

e.

.Q

^o.

tests, since it enters as an important factor into every calculation. >vith a knovvledg-e of the weight of the heavy hall and a neasurenent of the distance moved from the original position a nev/ moment may he calculatef' hy addinf or suhtractinp the product of the weight and distance, depending upon whether the hall ia lowered or raised.

The moment of the machine having been deter- mined, it was assembled into running order and, with the pendulum in a true vertical position, the pointer for indicating the deflection was set at zero on the graduated arc. So with everything calibrated, assem- bled, and with only the weight of the machine on the upper brass, power was thrown on, as well as a liberal feed of oil started.

After +he machine hcd been running for several hours, conditions of temperature and speed were found to be constant so that the next change to be made was to regulate the flow of oil to a minimum. This was done by m.aking several preliminary runs of fifteen minutes each until the desired feed was obtained, which was that of nearly approechinp intermittent oiling. This method of regulation was necessary because of the fact that the valves and variance of the head had not

'As •• + ' f '

' n i^ f •^

Sll fV

r.+ '^.-f" f } T f>.r\ ',:

,»«,

pif^ n"* .•'•sTr.'fn

on fl' Tt

a a

iU

rw

rlmr

"been calibrated as it was found impractica'ble due to the wide change of rates of flow caused "by the different "bearinf pressures that v/ere to be used. Had this "been attempted, the reg'ulation would have to have been calibrated for each bearing pressure,

FiTien the desired regulation of feed was obtained, the first run of the series of friction tests was started by taking readings of the following:

Time.

Weight of oil and reservoir on scale.

Deflection of pendulun from normal position.

H.P.M. of shaft (kept constant)

Bearing temperature.

Room tem.perature. These readings were recorded at intervals of ten minutes throughout the run, the duration of which was usually one hour. In a few cases where all conditions remained constant, and the rate of flow was sufficiently large so that small errors in scale readings might be neglected, runs one half hour in length were made, readings in such instances being taken eyerj five minutes.

At the expiration of this run the flow was in- creased until a noticeable decrease of deflection

,'Br - -<•

, »!Ji -

,t a«t»

JR fJ: .- ^^rj-J

'*««»r

iu:~ •" U^vJ*iO'>i r yi.

fror the previous dcYiation was had. Then the ma- chine was allowed to acqxiire constant hearinp tem- perature, after which the second set of readinps at ten minute intervals were taken. To get ade- quate data for each "bearing pressure at least six of these runs had to he performed.

By deflecting the pendulum and using a wrench on the head of the bolt at the hottom, the pressure on the brasses was increased gradually until the pointer indicated a pressure of 1,000 lbs. This increase of bearing pressure had to be accompanied with a delay of readings because the bearing tem- perature for this frictional load was also increased. So during this intermission the flow was decreased to one as near the minimtun as possible. 7ifhen con- ditions were finally fulfilled another set of six runs of an hour duration were made.

This method of procedure was carried out up to a bearing pressure of 5,000 lbs. at which load it was found impossible to obtain reliable data. Then too at this pressure it was found impossible to get minimum rates of flow because of driving difficul- ties encountered such as slipping and throwing off of the belt. This latter difficulty necessitated

*:ro

d-: no

ZOlJi^i

1.7^.

the complete throwing off of 'bearing' pressure In order to set the journal in notion again.

The preceding series of readings of five dif- ferejit "bearing pressures was again repeated for another sample of oil which not onlj- differed from the first in hody and density "but in viscosity. Al- though "both were mineral oils they were of different crudes, the first sample heing from the eastern wells, and the latter from the southern fields. The physi- cal properties of these oils were ohtained hy the ^^se of the proper instruments. The viscosimeter used was the standard Tagliabue's instrument, the principle of whidi is shown in the accompanying sketch. The oil was raises"' to a temperature of 150°F. , and the time for 50 c.c. to flow through the nozzle was re- corded. These results, of course, have no value except comparatively, to obtain the relative vis- cosities of the two oils. The flash and fire points were obtained hy means of a flash test machine of design similar to the one herein shown. The oil in the cup is heated and a spark from an indxiction coil passed across the surface at frequent intervals. The temperatixre at which the vapor rising from the sur- face is seen to flash is known as the flash point.

i^-fa

^.O

, ITi^ a. t f-Uft'^ hfa^, I M I ;- -.i -">

iBb

,-»«^

M.

15.

16.

The heating is continued until the oil finally takes fire and "burns, the temperature at which this occurs heinf known as the fire point. In ohtainiii>;the chill point a small quantity of the oil is placed in the hottom of a test tube which is cooled hy an ice-salt mixture, until a ther- mometer inserted into the tube picks up lumps of the congealed oil on the end. The thermometer is held at an angle of 45° and the temperature at which the oil is seen to drop off is the chill point.

The theory of the machine is as follows: -Let

R =» length of moment arm in inches

F = net weight on scale in lbs.

r ■» radius of journal in inches.

b ■• width of projected area of brass.

a = total projected area.

1 = length of projected area of brass.

W = weight of pendulum com.plete.

P OS total pressure on journal.

p = pressure per sn. in. of projected area.

T = tension on spring (read from graduation)

e = angle of deflection.

f = coefficient of friction

• X-

i»

e>.r «■ a

0* SB fl

A' .•'.f'

f ' \ 7 *'^ ^ ia /-. I

17.

Q = total friction.

N = revolutions per minute. From the construction of the machine it is readily seen that the pressure on the Journal is made up of efjual pressures caused by the action of the spring on the upper and lov/er brasses, and of the pressure due to the weight of the pendulum, which acts only on the upper brass. Since in the machine both brasses are loaded, the projected area is.

a = 2bl But the total pressure is,

P = £T + W Then the pressure per square inch is,

p - P/2bl = ^f^. Since the moment of friction is equal to the external moment of forces acting.

When © = 90°

FR = Qr and is a maximiun.

I'tTien at any other angina the moment arm = Rsin.e; anc! the moment,

FR sin.© = Qr

1. • -u ^ FR sin. © From which, Q «= .

osLor

,ei e*i

'\Jii m a

^'in ff.

.8i

'T«

:«.r-.

18.

By definition, the coefficient of friction equals, f - Q/P = gR^sin.tt

But P = 2T + W

mv ^ 4» FH sin. » Therefore f — -.

(ilT -♦- i;)r

FR It will he readily seen that jT- is constant for

each individual machine, and the value is knovm as

the "constant of the machine".

Thus,

f — Q sin.e

~ P

Since the principle is similar to that of the prony "brake, the horse power loss is,

H.P. = 2-JI NFR sin.Q^ ^ .00001586NCr 8in.« 12 X 33000

't(, +

• .i:lit 0

„ I-

0

.a

19.

SAMPLE 1. Physical Characteristics.

Color Reddish hrown

Baiun6 (Degrees) 30

Specific Gravity .8750

Pounds Per Gallon 7.29

Flash Point -- 417°?

Fire " 4640F

Chill " — 32°F

Viscosity (referred to distilled water at 150°Fh-1.77

■.*■•■■ -■:; fir*^fnArrn

_. . - " iiliiU

fCO.

ei.

24.

B5.

26.

Q7.

ee

g.9

31

SAHIPLE 2 Physical Characteristics.

Color Blueish Yellow

BauinS (Degrees) 20

Specific Gravity ^9333

Pounds per Gallon 7.78

Plash Point 365°F

Fire " 3900F

Chill " 12°F

Viscosity (referred to distilleci water at 150°F) - 1.89

- -- " 9tt'i

- " IllrfJ

c5?.

■34

35.

3<?

3fc

39.

40.

41.

CALCULATIONS FOR DETEHJ!IMTIOU OF MOMENT.

0

Determination No. 1 S 3

Distance "between Supports (inches) — 27.12-31.12-34,31

Weight of Scales (lbs) 85.2 -75.3 -69.7

Weig-ht of Strut (Ihs) — 16.5 -16.5 -16.5

Net Weig-ht (Ihs) 68.7 -58.8 -53.2

Moment (inch Ihs ) 1861 1831 1826

Average Moment = 1840 in. lbs.

Constant of J:/[achine = — g -gg = 820. where 2-1/4" is

the radius of the journal.

This moment entered as a factor only in runs on Sample 1 for total pressures of 670 lbs. and 2670 lbs. For all other runs the moment was smaller due to a falsing of the ball, and necessitated calculation as follows:

Weight of Ball 177.5 lbs.

Distance moved 3-1/32 inches

Decrease in Moment 177.5(3.03) = 540in. lbs.

New Moment 1840 _ 540 = 1300 in. lbs.

New Constant of Machine '^^^^■- - ^"^"^

2.25

xi 6C

I

'•«id

(•(fX

SK (1

.r.o. OTd ^0

♦■ f - 1 -sr

rxa? "to j-rf-Je,

I =

44

SAMPLE CALGULATIOIJS Oil Sample 1. Total Pressure 2670 Its, Run Ko.l.

Data as obtained :-

Pounds fed per hour ,01

R.P.M. - 20£

Constants :-

Width of "brass *-^ fb) 3 inches

Length of trass (1) 8 inches

Constant of Ilachine — (C) - 820

Spring Tension (T) 1000 lbs.

Weight of -Machine (W) - 670 Ihs.

Total projected area (a) 2hl « 2(24) — 48 sq.in.

" pressure (P) (2T + W) = 2000 -j- 670 .. 2670

(lbs.

It was assumed in all rims that the total pres- sure was evenly distrihuted on both hearings, even though the weight of the m.achine rested wholly on the top hearing.

Then : -

/ , P 2670 _

Pressiire per sq. in (p) — -=— = — 56 lbs,

^ ^ ^ a 48

C s in d The coefficient of iYiction (f )

f»xti'sa»

HI

dO.V '^,

. .K

>

.f«fl Or.ni (v) /

.%fl -

, 4. 0 !

, ^/r^.r-, fi0^B(>t nnXA^MT " ■ ♦ * »> ♦■''■!->" 1 /♦ .-f-Mn,'4

- - — n( ■■!'. _- \ r .frrnaai*^

«

'I

45.

C sin 9 _ 820 sin 7.06 P ~ 2670

^ .0378 Horse Power Loss (H.P) — .00001586KCr eln « .00001586 (202) 820 (2. 25). 1230 = .725

B'

•j^axoc^ .

0.

Oil Sample 2. Total Pressure 670 l"bs. Run No. 1.

Data as obtained :-

Pounds fed per hour .01

Average deflection (e) 2.44°

R.P.M. 198

Constants : -

Spring Tension (T) — 0

Weight of JIachine -- (W) — 670 lbs.

Constant of I.Iachine- (C) — 577

Total Pressure -(P) 2T -|- W - 670 Ihs.

Pressure per sq. in. = 14 lbs.

Coefficient of Friction -ff) C sin e

P

C sin e _ 577 sin 2.44

p 6T0

s= .0568

Horse Power Loss (H.P)- = .00001586IICr sin e

.000015861ICr sin 9 = ,00001586(198)577(2.25) .0428

= .174

..■;<fj

10.

Ok

-; ^l^l^tt^Q •« :':^ail

• ) rtc ' •«v»^»irA

r QVc^ —

- ; p,;' or

-..■#fr

It)- m oD

' r /\r\r ~ « , > . i r '-inr •

*Vi. =

4

In view of the fact thet reliable results are obtained only through the uee of accurate instru- ments, greet care v/as teken with the scales and thermometers used, to obtain the proper correction factors. The knife edges of the scales were cleaned so as to render the apparatus more sensitive, and the scales were tested v;ith standard v/eights, both before and after the completion of the experiment. The thermometers were similarly tested against stand- ard instruments; and all the results herein tabu- lated are corrected for errors. Effort was made in every direction to make the results reliable.

In testing sample 1, it was found at first that considerable time was censured in allowing the machine to settle down to a state of constant con- ditions. Even after such a point had been reached the data as recorded on the riuining log showed con- siderable variation, especially at lov/ tates of feed. These discrepancies were probably due to the light body of the oil, which v/as unable to withstand the pressure when a very slight irregularity in the flow resulted from eddy currents or pockets, and allowed the film on the bearing to "streak". The higher rates of flo7/ overcame this difficulty to a great extent.

.08 «■''■ Offlf)

-jiOu tiiatHiiOo '-0 -! « c^ ixwoi? wSfiea o.t *>f(i.io art

;»»<! lijwf ti Joq « rfours -r^tl^is r .^ttottlh

46.

since the supply was sufficiently large to make slight variations negllgable; but the tine ele- ment entering into the settling down to constant conditions was not much reduced.

Sample 2, however, which was of heavier hody, though slightly lovrer in viscosity, proved less trouhlesome in every way; for it not only took less time to hfing ahout constant conditions, hut the readings showed practically no variations during the runs. Since it was not until the highest pres- sure was reached thf.t any irregularities were noticed it was concluded that the heavier body was respon- sible for this smoothness of running. In addition to the fact that the second specimen of oil showed better running qualities thail the first, figures obtained show it to be more economical as regards friction loss. Comparing the runs of the two oils for each pressure, it v/ill be seen that in every case the coefficient of friction for sample 2 is less than that of sample 1. The data on the horse power loss also brings this out very effectively, which can be seen at a glance by noting the curves. It may be notice^' on the coefficient curves of Sample 1 that the point vyhere the curve begins to

00 oi iiftob ^li;' of ztno ^ot\ it lo't ;^w tt«t» nl •nioe<»ltfrrott

'^ifjt' .i^ii ■d'i.JP^-V Oil '■ 'I

r ton I eo£i£8 : trf*

I'rtrif .. . , •.^! * *'r -A Tii-'i.T'-'f r ■n_^. t '"* ^flri«iAf>■■'i -in:; nfrrH

I't SAW ^6o«f taiv

• »J o* • fo

^.^. 2 •Iqmo lol noi •«ao

fiPTO r rifft no ntafi «!-f -ft fflnoT

. a f» fl»«a stf HAo i(oi;{w

"U j 9.' ' ' ■ ,;., I oq :■,.' ^L

49

flatten oixt is much further out than the correspond- ing point on the curves for Sample 2, That is, the point heyond which the increavse of rate of feed pro- duces no appreciahle saving is a much lower value in the case of Sample 2 than in the case of Sample 1 which is another point in favor of the second spe- cimen tested.

The data from Sample 1 proved to he very incon- sistent as the plotted points indicate; in almost every case for that specimen the points distrihuted themselves in a "shot-gun pattern" It was not so for Sample 2; for almost all of the points fell a- long a very well defined path. Thxis, in drawing the curves for the first sample, reference was made to the characteristic shape as foxmd in Sample 2. The horse power curves of Sample 1 shov; the absurdity pf these results much more clearly, for the horse poT/er loss at the 2670 lb. pressure appears greater than that at 4670 or 6670 lbs. pressure. This is considerably different from the other case in which the curve seems to rise by even steps as the pres- sure increases. In as much as the pressure on the bearing enters as a factor into the coefficient of friction, a wide range of figures resulted which

J s»rtt ul UBdf d e't'i.'i'. lo epao orft xtl

» ■ t) . .-.-.J.' ut*,' iJi.y

^' • ■' 'a<T «'«'■ - '^" a al aoTltSBHirf*

r, «,r+ tworr^j r *fMMo« "n firf»,r-r.r.-i ^ewocf snioxl 9i{^

;q .cfl OVdS ».-ft f» BRoI f#woc

•■a n»T(» •^^d' ftsl-r ot iMnen ^vni/o erf*

»i He lo •";> .coltoltti:

which rendered it izapractical to superpose the curves as was done in the case of the horse power, thus hiding the freakish nature of the results.

A study of the data aind curves herein contained irrespective of comparison of the oils, will show the advantage of the continuous system of lubri- cation over the restricted feed method. Such a system permits the flooding- of the hearing? at all times, thus reducing the friction to a miniriun, which saving may be traced back to the coal pile, and figured up in dollars an-^ cents. Then, too, the greater part of the oil v/hich is lost in the restricted feed method, may be reclaimed and used over and over again, \7hich is another distinct saving. Of course, small plants are not to be considered in recommending the installation of these flooded systems; but the large stations v/here the saving amounts to several hundred horse power find this method almost necessary for eco- nomical operation.

'p- ,T.rJo '■■■'J '^ '^oiitfl'^tttoo to •rtrtnoraet'iJt

i

■■■^

l4\

i?; *V

Ij.'lt! 'i *** '^'