\
and with this value of a no force is needed to " slack" the
key. For well-finished keys, /= o. io f when the surfaces are
not lubricated more than is commonly the effect of handling,
and of may be taken above 10, i.e., a taper of about one to six ;
more usual values are I : 50 to I : 100 for keys fitted to gibs, and
half these values for cotters, or keys without gibs.
28. The Distribution of Pressure on surfaces subject to
wear by the friction of motion depends greatly upon their form
and on the character of that motion. Plane surfaces, if rigid
and subject to the wear of straight-sliding parts, of which they
form the bearing surfaces, if originally well fitted and of homo-
geneous material, and if kept in good order, exhibit uniform
intensity of pressure throughout, when the resultant pressure
passes through the centre of figure, and sustain uniformly vary-
ing pressure when the resultant is outside that centre. In the
latter case, the mean pressure may generally be assumed as a
uniformly distributed pressure in calculations. Inequality of
pressure leads, first, to unequal wear, then to exaggerated
variation of intensity of pressure, and finally to " cutting," or
abrasion, and destruction of the wearing parts. The maximum
permissible intensity of pressure is generally the less as the
speed of rubbing is the greater, and is usually but a small
fraction of that representing the " elastic limit " of the metal
resisting it.
Plane surfaces subject to wear under a motion of rotation,
even where the pressure is at first uniformly distributed, are
apt ultimately to take such form that the pressure is of vary-
ing intensity. The method of variation will be dependent
upon the form, and the fitting of the journal to its bearing.
As an example, a disk rotating about its centre will usually
wear differently at the periphery and toward the centre, and
thus ultimately is caused such a distribution of pressures as
will throw the greater part of the load upon the central part of
FRICTION AND LOST WORK.
the disk. The tendency is usually to effect such a distribu-
tion of pressures as will finally give permanence of form.
Curved surfaces may thus take pressure in many ways ; but
it probably rarely occurs in practice that the pressure is of per-
fectly uniform intensity. A number of cases will be considered
in the succeeding articles. The most important case is the fol-
lowing :
A cylindrical or spherical journal, if perfectly fitted, when
unloaded will, with its bearing, take such a form under load
that the intensity of pressure on the bearing surface will vary
as the cosine of the angle made by a radius passing through
the given rjoint in that surface with that radius with which the
resultant pressure coincides. Thus :
In the figure, let ACB be
the trace of the bearing surface
of a perfectly fitted unloaded
journal. When the load comes
upon it, the journal will sink a
minute distance, OO f , CC , into
the bearing, slightly compress-
ing the metal, and taking the
-- new position A'C'B'. As the
maximum intensity of pressure
in any well-proportioned jour-
nal is usually but a small frac-
tion of that which would pro-
duce a compression exceeding
FIG. o. DISTRIBUTION OF PRESSURE. ., t ,. ,. .. , ,,
the elastic limit of the metal,
and as within that limit the resistance is directly proportional
to the compression, every part of the surface, as E, will be sub-
ject to pressure of intensity proportional to the displacement,
El, of that point in the bearing. Thus the pressure at B re-
mains, as at first, zero, and contact simply is preserved ; at E
the pressure is proportional to El, and at C to CC. But the
vertical displacement, -CC, BB' , EJ, is at all points the same,
and the compression, El, at any point, E, being very small, is
measured by the product of that constant quantity into the
cosine of the angle, COE = 6, between the radius, OE, pass-
THEORY OF FRICTION. 39
ing through that point, and the line of the resultant bearing
pressure, OC.
The sum of all vertical components of these normal pres-
sures, each of which latter is measured by the product of a
constant into cos 6, is equal to the total load, W. Hence,
taking the intensity of pressure at any point, E, as represented
by/, and the constant as /,, the pressure on any element, ds, is
pds, assuming the length of the element unity, and this is
equal to/, cos dds. The vertical component, w, is
w = p cos Ods = /, cos 9 dds ;
and the total load and the value of /, are
But COS 6 = / a_~a ' ' an ^ & = - Q = & SCC # * tnen
~ = r ' sin ~~ l = I>
and the pressure on unity of area, at any point, E, is propor-
tional to cos 0, and is
WcosO
when r f is the radius of the journal.
It is evident that a similar demonstration applies to the
case of the sphere. The amount of compression is determined
by the magnitude of the modulus of elasticity of the softer
metal of journal or bearing, and by the intensity of pressure.
Thus, for a maximum pressure of looo pounds per square inch
4O FRICTION AND LOST WORK.
(703 kgs. per sq. cm.), a pressure often attained with steel crank-
pins, and with a modulus of elasticity of the bronze bearing of
12,000,000 (843,600 kgs. per sq, cm.), the maximum compres-
sion would be bat yyj-tnrth tne thickness of the " brass," or, for
journals of small size, about 0.00004 inch (o.oooi cm.). This
distribution of pressure remains constant so long as the maxi-
mum pressure is less than that producing wear.
In all cases which are to be here considered, Wis the resul-
tant pressure on the bearing surface. It is found by combin-
ing the weight of the parts carried by the journal with the
effort acting upon the journal, directly or indirectly, and pro-
ducing or tending to produce motion. The distribution of
pressure under light loads and at high speeds is sometimes
determined by the action of the lubricant, as illustrated in ex-
periments with the " oil-bath." This treatment is exact for
cylindrical shell-bearings in rigid frames, approximate only for
other cases. This investigation exhibits plainly the desirability
of securing the greatest possible rigidity of frames carrying
bearings.
29. The Friction of "Journals," as a source of lost work,
is of great importance to the engineer. A journal is a surface
of revolution, turning, loaded with a pressure due the weight
of the shaft and its load, within another surface of revolution,
called the " bearing," which should be of the same form, and
which should perfectly fit the journal without pinching.
These surfaces are almost invariably cylindrical ; but they are
sometimes conical, sometimes conoidal or ellipsoidal, and rarely
of other related forms. Axle or shaft journals, gudgeons, and
trunnions are the familiar forms of this element of mechanism.
A journal in thoroughly good order will fit the bearing
throughout the arc of intended contact : it is the custom with
many experienced engineers, however, to " free" the bearing at
the sides, leaving the two surfaces in contact only for about
one half the total depth of the bearing-piece, i.e., over an arc of
contact of 120. Journals also frequently wear loose, and thus
concentrate the load upon a limited area. Bearings are also
sometimes bored out a very little larger than their journals,
with a similar result. The theory of such cases is as follows:
THEORY OF FRICTION.
(i) A loosely fitting journal, ABC, when at rest, will lie
at the lowest point in its bearing; but, when moving will roll
up the side until it be-
gins to slide ; it then
retains this position so
long as the coefficient
of friction is unchanged,
and rises and falls as the
coefficient increases and
diminishes, continually
finding new positions of
equilibrium.
At any one instant
there are three forces in
..... , . , FIG. 10. LOOSE BEARING.
equilibrium : the weight,
W, on the journal ; the reaction, N, of the bearing ; and the
force of friction, holding the journal at the line of bearing on
the inclined surface : this latter force is F = fN. The angle,
FDE at, between the tangent to the common surface of COIK
tact and the horizontal is evidently that of an inclined plane
on which the mass would slide with uniform velocity, and
hence tan a =f= tan cp. These forces being in equilibrium,
they may be represented by the " triangle of forces," DNB.
Then, since the forces A^and .Fare at right angles,
= N* +
N =
W
I+/' 1
fW
(0
(2)
Wtzn
= = Wsmg>; . (3)
Vi tan
and the motion of the journal carries it around, in the direction
opposite to that motion, through the angle of kinetic friction,
= > ' -, Wn ; . (6)
when n is the number of revolutions made in the unit of time.
(2) A perfectly-fitted bearing may be made by careful work-
manship and fitting, while unloaded, when constructed ; or it
may be obtained by the wearing of the journal down into its
bearing. In the first case, the pressure on the bearing gradu-
ally increases, as has been seen, from o at the diametral line to
a maximum at the bottom, this pressure being at every point
proportional to the elastic, radial, displacement of the surface
where pressed. In the latter case the bearing wears until the
sum of the vertical components of all such elementary pres-
sures which sum is equal to the load is so adjusted as to
check the wear, and this may give a distribution of pressures
in any manner intermediate between the preceding case and
one in which the pressure is uniform through the supporting
" box," the latter value of the intensity of pressure being a limit
which may be closely approached, or even actually attained.
For the first of these cases, the pressure on any elementary
portion of the arc of the bearing, d9, is
(i)
in which N f is the normal pressure on an elementary area,
Ir 4 d0 t which has the length of this journal, /, and the breadth
THEORY OF FRICTION. 43
r,d9,p being the intensity of pressure at that part of the arc
considered. The sum of all the vertical components of these
normal pressures is equal to the load W. Then
W=f ** fir t cos OttO.
9 = -
But the intensity of the pressure,/, will be zero at i = - in-
creasing as cosine 6 to a maximum,/,, at 8 = o; therefore,
since/ = P l cos 0,
; .... (2)
-
p = 0.64
max = 0.64 -'
The intensity of the force of friction at any element is
- Hy cos(9 M
# = 0.64 ^ -; ...... (7)
and, at = o, (fp) max. = 0.64-7-^ ......... (8)
44 FRICTION AND LOST WORK.
The total pressure on the bearing is
cos 6 JO; . . (9)
0.64^2 sin -;
The total force of friction is
fJ>W=i.2 7 fW} ...... (10)
and the work wasted is
...... (11)
in which s is the distance traversed by the rubbing surface.
Otherwise the moment of friction is
M = Pfr, = 1.27 fWr^ . . . . (12)
and the energy lost is, per unit of time,
U=Ma \.27fWari = 2.^/nrJVit. . . (13)
Hence, in a bearing thus fitted, if the unloaded journal
is an absolutely perfect fit, the total friction is 1.27 times as
great as with a loosely fitted journal.
(3) A bearing in which the journal is so grasped as to give
uniform pressure throughout, produces a loss of power which
is also easily calculated thus :
The intensity of pressure is at all points constant, and may
be represented by /,. The vertical component is / a cos 6 ;
7HEORY OF FRICTION. 45
and the total weight, W, sustained by the journal is equal to
the sum of all vertical components. The pressure on any
element \sp^r l dO ; its vertical component is pjr^ cos 6d8, and
the total load is
/"'=+;
W r =/A,y__. cosfc; .... (I)
3
= 2pjr v ; (2)
W
Then the total pressure on the surface of the journal or of
the bearing is the product of this intensity of pressure into its
area, or
(4)
The total force of friction is
Ff= i.tffW. . . . . . . * .. (5)
The moment of friction is
M =Pf ri = i.vfWr^ . . . . (6)
and the work of friction is, per unit of time,
U^Ma^afPr l = i^7aflVr l \ .... (7)
= /* >,; ' . . -. .. (8)
i.e., it is 1.57 times as great as in the loosely-fitted journal, and
20 per cent, greater than in the last case.
4 FRICTION AND LOST WORK.
The first of the three cases just considered is often met
with, new journals being often purposely or carelessly bored to
make a loose fit, and old journals often wearing loose. The
second case arises when the journal is made an exact fit, when
new and unloaded ; and the last occurs when it has been
running smoothly and without jar, and has thus gradually
worked down into the bearing and has worn all portions of its
surface to a small but usually appreciable extent ; such a.
journal is always found to be in excellent condition. The
usual case in practice lies between these. The last case may
be also met with in those rare cases in which a new journal has
been fitted tightly into its bearing, and yet oftener where, as
sometimes happens, the heating of the " brass" causes it to
grasp the journal, closing over it so tightly as to cause as great
heating on the sides as on the bottom. The Author has some-
times met with such action in his own experience, even with
very large journals and bearings.
It is seen from the theory just developed that, while in any
journal the total pressure and the total resistance at the sur-
face of the journal are the same for any given load, whatever
the size of journal, the moment of friction increases with the
diameter of the journal, and the work lost varies in the same
ratio. It will be also noted that, since the liability of a journal
to heat varies, directly as the intensity of pressure and as the
amount of work done, and inversely as the area across which
this heat can be discharged, the diameter of a journal does
not within certain limits affect this phenomenon. This will be
better shown in another chapter. The bearing should evidently
be so proportioned that serious lateral pressures shall not be
produced when in operation.
With a flooded journal, as where the oil-bath is used, the
pressure is probably nearly always a maximum at the meridian
line, becoming zero at the edges of the brass. The second
case is therefore correct here.
(4) The quantity of heat produced by the friction of the
journal, in the several cases above treated, is obtained by divid-
ing the work of friction by the mechanical equivalent of heat.
Calling this J, and its reciprocal A, we have for the loose bear-
THEORY OF FRICTION.
47
ing, Case I, //"representing the heat produced in the minute or
the second, whichever may be the unit of time.
= ^=At7=2A
^ sin
o
THEORY OF FRICTION. 51
in which c = 0.0007 to 0.0009 for marine engine crank-pins, or
c = 0.0004 for locomotives, and e = 0.05 to 0.06 and e = 0.08.
Journals carrying uninterrupted loads require longer pins.
The pressure on journals is very generally reckoned, as
above, by reference to the projected area.
A Line of Shafting consists of a succession of iron or
steel shafts, or axles, connected end to end by " couplings,"
and carrying often a set of pulleys or of gearing, by which the
power transmitted to and through the line is distributed to the
driving shafts of various machines. This is called "line-shaft-
ing," to distinguish it from the "countershafts" and other
shafting of special machines.
Line-shafting is carried by a succession of bearings placed
40 to 60 diameters of the shafting apart usually, and the
journals are generally made three or four diameters in length.
These journals sustain the weight of the shafting, pulleys, and
belting, and the resultant pull of the belts, and are thus sub-
ject to considerable friction and consequent waste of power.
Since the power applied is all received at the end, it is evident
that the size of the shafting may be economically reduced, as
this power is distributed to the machinery driven in passing
from the receiving to the farther end.
Were this variation to be made by a gradual reduction of
diameter, and were the power all transmitted to the farther
end, the economical method of proportioning would involve
the measurement of the friction, and the determination of such
a size as would be the minimum required safely to transmit
the effort demanded to overcome the friction beyond the given
point, and to deliver the needed power.
Resistance to torsion varies as the cube of the diameter of
the shaft. Calling the diameter d, the moment safely applica-
ble to the shaft is
when A is a coefficient correct for the given case, and varying
with the material and the magnitude of the factor of safety,
which latter quantity ranges all the way from 6 to 30 in com-
mon practice.
52 FRICTION AND LOST WORK.
If the weight of the material of which the shafting is com-
pqs.ed be called >, the weight of a unit of length is
...... (2)
and its friction, nearly
,/V = Q.7854/wrf. ..... (3)
The moment of friction is
d
f "w r '- = 0.392; 'fwd* ; (4)
and the " exhaustive length," as it is called by Rankine, which
would be just sufficient to take up the whole applied moment,
by its friction, is
(5)
Then the maximum resistance of the shafting is Ad* ; the
moment of friction per unit of length is Bd* = o.3927fwd 3 ;
and the moment demanded to turn a tapering line of shafting
proportioned for minimum loss of power is
Ad* = o.3927fwLd* = M -(- 0.3927/0^ C
'dx, ..... .... (6)
when x is measured from the end farthest from that at which
the effort is applied. Taking x = o at the nearer end,
(7)
THEORY OF FRICTION. 53
being the useful moment transmitted. Then calling d=y,
(8)
where - & = 7;(i -*->);. . (3)
, ....... (4)
r
H %=--=#,
= rJt-
(5)
The tnean tension on the belt and its ratio to F are
I)'
Calling = 2?r, in which is the number of turns or the
part of a turn which the band or cord makes around the cylin-
der, and reducing for common logarithms, calling the modulus
4fi
T= eS* =
since
68 FRICTION AND LOST WORK.
and
T
common log -^ = 2.7288/0 ;
= T t (i io a -7 288 / i); (6)
P - 2.729 fn
*
For the quantity 2.7288/11 may also be substituted 0.00758
/(9 when is expressed in degrees, and 0.434294/0, if in cir-
cular measure, common logarithms being used in both cases.
The moment of friction is
M^Fr^rM-Tj; ...... (i)
and the work done in the unit of time is, as a maximum,
l -T,); .... (2)
or per revolution,
The values of M and of U may be less than the above, but
cannot be greater.
(3) In a " strap-brake" the band or strap is sometimes in-
tended to slip, the tensions being just sufficient to control the
load. In this case the value of f is that of the coefficient of
friction for motion. Here motion occurs between strap and
pulley, and heat is produced to the amount of
lf=2Axr l (r i -T,). ...... (4)
THEORY OF FRICTION. 69
The work and the moment on a slipping-strap are always
maxima ; if not slipping, the moment may be anything less, as
where the brake sustains at rest a small load.
The total friction-force is seen, both for the belt and the
brake, to be independent of the size of the cylinder upon which
it is coiled, and to depend solely upon the angular extent of
the circumference embraced or upon the numbers of turns
taken by the band, the ratio of tensions becoming rapidly
greater as the strap is wound on ; thus, iff = 0.333, as taken
by Weisbach, we have
EXTENT OF WINDING. T
* i
Q n_ n e& T t
90 = 2 = i revolutions lo - 22 * 1.788
180 = TT = " io-4S48 2.85
360 = 27f = I " I0'9096 8.I2I
720 = 47T = 2 " I0 x - 8 '9 a 65.95
1440 = STT = 4 " io3- 6 s84 4349
2880 = l6ff = 8 " I07' 2 768 I8,9I4,8(X)
3650 = IO7T = IO " IO9-9 I,247,38O,OOO
The total amount of work lost by friction in any case is, as
has been seen (7^ 7* a ) S, when the space, S, traversed by the
effort, 7*!, is given.
(3) The Friction of a Cord or Belt passing over the edge
of a rigid body is determined by the amount of the change of
direction taking place at the angle supporting it, by the value
of the coefficient of friction, and by the magnitude of the two
forces acting on either side the edge. If the edge is sharp, the
cord may be stretched with such force as to cut it, and the
resistance then becomes greatly increased ; but if the edge is
smoothly rounded, and the cord perfectly flexible and unin-
jured, the case is that of the friction of a cord on a cylinder of
very small radius, on which an arc is enwrapped by the cord
equal to the angle included between the two parts of the cord
or belt. The resistance due to friction has been seen to be in-
dependent of the radius of curvature of the arc, and it is evi-
dent that the case is precisely that already considered.
70 FRICTION AND LOST WORK.
Henee th& friction is
I) ; . . . (i)
when Wts the load and n and 6 are the measures of the angle
in parts of a circumference and in degrees, respectively.
The value of the pulling force is then
.... (2)
Ah approximate expression for the resistance of friction for
small angles is obtained by taking it as
/3
= ( 2 +f)W sin-, nearly ...... (3)
Where several edges are met, as in the " rendering" of a
chain over a barrel of polygonal section, the faces of the poly-
gon being equal in length to the links, the total friction may
be calculated by introducing the sum of the angles, 6, into the
first of the above forms, (i), or by raising the last (3) to a
T
power, n, equal i& the numbef of attgles, the ratio of-, 1 thus
* *
increasing in a geometrical ratio :
..... (4)
The work wasted Is
FS= WS (io*-s* - i). ...... ( 5 )
The useful work is WS and the total work PS.
THEORY OF FRICTION.
The following table gives the ratios of P : W for arcs less
than 300. For larger arcs see the preceding table.
VALUES OF - FOR BELTS AND CORDS.
Angle 0.
Value
5 Of/.
Degrees,
Circular
Measure, 9.
Parts of
Circumf., n.
0.2
0.3
0.4
o.S
30
0.52
O.08
.11
Values of
17
Ti -f- T a .
'23
M
45
0.79
0.13
17
.27
37
1.48
60
1.05
O.I?
.23
37
53
1.69
75
I-3I
0.21
30
.48
.69
1.92
90
1-57
0.25
.40
.60
-87
2.19
120
2.09
0.33
52
.88
2.3'i
2.85
150
2.62
O.42
.69
2.19
2.85
3-70
i8o
3-14
0.50
.88
2-57
3-51
4.81
210
3-67
0.58
2.08!
3-00
4-33
6.25
240
4.19
0.67
2.31
3-51
5-34
8.12
270
4.71
o 75
*-57
4.11
6.59
10.55
300
5-24
0.83
2.85
4.81
8.12
13.70-
32. The Friction of the Wedge, and of the Screw,
which is essentially a wedge, and both of which are illustrations
of the inclined plane, has already been given in principle.
(i) Applying 1 these principles to the case of the wedge
(Fig. 17), we have the weight, or
force driving the wedge, equilibrated
by the two lateral pressures and the
frictional resistance to slipping on
the sides ; and, a being the angle of
the wedge,
= 2/> (sin f+/ cos f)..(,)
When the wedge is forced back
by the lateral pressures,
FIG. 17. WEDGE.
7- FRICTION AND LOST WORK.
For other cases, simple and obvious modifications of the
theory of the inclined plane already given will suffice.
(2) For the screw, which is to be considered an inclined
plane wrapped around a cylinder, the pitch of the screw meas-
ures the height, the circumference is the length of base, and
the length of thread of screw per revolution is the length of
the inclined plane. We may take a (Fig. 1 8) for the angle
at the point of the wedge or inclined plane, r the radius,/ the
pitch of the screw or the height of the inclined plane, P the
force applied at the end of the lever-arm r, W the load, and N
the reaction at R normal to the plane. Then, resolving parallel
and perpendicular to the plane, we have
W
FIG. 18. SCREW.
Pcos a
P sin a N -\- Wcos a = o ;
and hence, for limiting values,
P _ sin a zpfcos a
W ~ cos a-.f sin a
The limits of value of the effort
required at the end of a lever, or
wrench, of the length r' ', is evi-
dently
p = P =
' ~
r' ' cos a f sin a
(2)
The values of P and P may be any values between the limit-
ing values thus derived.
The case of the weight being raised by an active effort, P, is
seen to be similar to that in which W acts to produce motion
and P resists ; the expression for the one being identical with
that for the other, with the sign of f changed. The value of
Pis thus a maximum when an active and a minimum when a
resisting force.
Friction-Couplings consist of a solid and a hollow cone, each
THEORY OF FRICTION. 73
on the end of a shaft, and so fitted that they may be forced
into contact, the one within the other, in such manner as to
make a firm connection when desired. The lever-arm is, as
has been seen ( 30),
r
-
and the intensity of pressure is
W
A (sin %a -f/cos %ot) '
when W is the total effort, A the area of common surface of
contact, and a the angle of the cone. Then the resistance due
to friction is
= fpA = -
W W
max. = ; Pmax. = /<4 = ;
and the limit becomes *
F max. = fAp max. = JK
For the plane disk, -Fmax. fW.
33. The Friction of Gearing is partly due to sliding of
the teeth upon each other, and partly to resistance to rolling.
That part of the work lost by sliding is measured thus :
Let a and ft be the angles made by the directions of motion of
the two teeth engaged with the normal to their surfaces at the
line of contact, and let P be the intensity of the normal pres-
sure. Then the resistance to sliding will be
R=fP.
See Weisbach, vol. iii.
?4 FRICTION AND LOST WOKK.
The work done against this friction will be, if s is their rela-
tive motion,
U=&=fPs=fP(v l tana + Vttan/y)t, . . (2)
when v l andz> 2 measure the absolute velocities of the two teeth.
Where several teeth are engaged,
(3)
The loss of work and energy by friction of the teeth of
gearing may be also measured thus :
Let the angular velocities of two teeth in contact be a f , a" ,
and call the distance of the line of contact from the pitch-point
of either tooth, s'. Then the relative velocity of rubbing is
v' = (a! -f- a")s', and the work expended in friction is
U = fPv't=fPs f (a' + a"}t ..... (4)
The loss due to rolling resistances is usually so small that it
may be neglected ; but the method of calculation is given in
Art. 25.
In Screw Gearing, in which a screw or "worm" revolving
in the plane of the gear drives the latter by engaging tooth
after tooth as they come around, the loss of work is mainly due
to sliding friction, and is often considerable. Here the resist-
ance is, at the surface of the tooth,
R=fP- ........ (5)
The work lost is
7+?\ ... (6)
in which r is the radius of the Worm and / the pitch, while n
is the number of revolutions made in the given time.
When 6 is the inclination of the worm-thread with the axis
of the worm, the total resistance is
tan
THEORY OF FRICTION. 75
in v/hich P is the effort at the pitch-line tending to turn the
worm, and R is the resistance at the same point, but on the
surface of the wheel, and in the plane of its rotation.
When we makey= tan q>,
P = R f cot (6 -
0.4
0.40
0.008
0.4
0.40
0.008
^
1
b - 8
Q 1.6
6^44
0.032
0.128
0.8
1.6
1.14
3-22
0.053 D
0.064^5
ar
n
i
3.2
25.75
0.5II
3-2
g.io
0.180
P
0.4
0.80
0.008
0.4
0.80
0.008
t'o.8
3.22
0.032
0.8
2.28
0.053 fi
^
JS
j 1.6
12.88
0.128
1.6
6-43
0.064 a
i
* I
3-2
5I.5I
0.511
3.2
18.20
o.iSo
3
FRICTION AND LOST WORK.
TARRED ROPE.
No. of Threads.
Weight per Foot.
Ibs.
Value of A.
Value of b.
6
0.02
0.15
0.008
15
0.05
0-77
O.O2O
30
I.OI
2-53
0.040
Weisbach's coefficients are :
For tarred rope,
British.
K = 3.3,1 ;
a = 0.22 ;
For untarred rope,
For wire rope,
For tarred wire rope and
hempen core,
a = 0.064.5 ;
= i .08;
# = 0.094 ;
= 1.21 ;
a .= 0.027 ;
Metric.
= LSI
a = 0.006.
K = 0.086 ;
a = 0.00164.
K = 0.49 ;
a 0.0024.
^=0.57;
$ = 0.0007.
The resistance of belts to flexure maybe calculated by
means of the simple formulas just given, and is .expressed in
terms of the tensions thus :
The resistance due to flexure is, according to Reuleaux,
aAP
But the pull, /> is
(0
when the whole circuit of the belt about both pulleys is taken,
and when r v r a , are their radii.
THEORY OF FRICTION. 79
The work lost is then
. . . (2)
a may be taken as already given.
36. The Friction of a Pulley or " Tackle" is due to two
distinct phenomena : the frictian of the pulley or " sheave"
on its axis, i.e., the pin fixed in the "block," and the rigidity
of the rope wound over the sheave. The first of these two
resistances is that of the cylindrical journal.
The load being W, the added resistances due these two
causes, reduced to a common line of resistance with W, being
F-{- 5, the total load becomes, for a single block,
(i)
The work done usefully will be Wk, where h is the distance
traversed by the load, and the total work will be
Ph = (w+ F+ s)h. . . Y *. . (2)'
The methods, of .calculating the magnitude of these several
forms of resistance have been already given.
37. The Friction of a System of Pulleys is the sum of the
frictions of all the elements of the system ; but as the load
transmitted from pulley to pulley or sheave to sheave between
the weight and the " hauling part" is continually augmented
by added frictional resistances, the relation of the one quantity
to the other must be determined by ascertaining the relations
of these quantities for each.
If the ratio
P W+F+S
W W
for a single pulley be known, and if this ratio be determined
~p
for each pulley of the whole system, then the ratio, =, for the
8O FRICTION AND LOST WORK.
system is obtained by the continued multiplication of these
values of C, and is
C = C t . C t . C 9 . C 4 , etc (2)
P f _
The final value of = is then known, P being the value
W
which exceeds the value of P, in a similar but frictionless
system, in the proportion in which C exceeds unity. The rela-
tion of the effort, P, required to raise any given weight, W, in
any frictionless system of pulleys may be experimentally de-
termined from the relation of velocities of the hauling and the
lifting parts. Thus, if these velocities are V and F,
since, friction aside, the power or energy exerted and absorbed
is the same at both ends of the system and
P V= WV. ....... (4)
Then, friction being considered,
The relations between the effort exerted and the resistance
overcome in systems of tackles are given in all treatises on
mechanics.
38. "Rolling Friction," or more correctly, resistance to
rolling, is a consequence of the irregularities of form and the
roughness of the surfaces of bodies rolling, the one over the
other. Its laws are not as yet definitely established, in conse-
quence of the uncertainty which exists in experiment as to how
much of this resistance is due to roughness of surface, how
THEORY OF FRICTION. 8 1
much to original and permanent irregularity of form, and how
much to distortion under the load. The first of these quanti-
ties evidently varies inversely as radius : the second similarly,
and the third as a function of the hardness and elasticity of
the material of which the two bodies are composed. The
total resistance, if the distortion does not exceed the elastic
limit, is proportional to the load carried at the line or band of
contact. In all actual cases the line of contact of two surfaces
originally tangent and unloaded becomes a band, of which the
width increases with the magnitude of the load and with the
softness of the material.
"Friction-Wheels" are often used to reduce the loss of
energy at a journal, when the load is small, its direction con-
stant, and the angular velocity small. In such case the jour-
nal or " gudgeon" is supported on the periphery of two
" friction-wheels," which are themselves supported on journals
turning with an angular velocity less than that of the supported
shaft, as the diameter of the journal is less than that of the
friction-wheels. A single wheel is sometimes used, in which
case the work lost by friction is reduced in the proportion
,-'
when U r, are the work done and the radius of the journal as
ordinarily mounted, and 7 2 is the work done against friction
when the friction-wheel is introduced ; r^ is the radius of the
friction-wheel.
When two supporting wheels are used,
* cos 2
in which a is the angle at the main journal-centre, subtended
by the two friction-wheel centres. ,
82 FRICTION AND LOST WORK.
39. The Laws of the Friction of Rolling are as simply
expressed as are those of sliding friction. It is customary to
take this resistance as proportional directly to the load and in-
versely as the radius of the rolling cylinder or wheel. Experi-
ment shows, however, that, with wheels capable of yielding
somewhat under load, the square root of radius should be
taken in the formula for rolling resistance.
The magnitude of the force of the friction of rolling is,
therefore, at the axis, in the first case,
W
*=/-. ........ 0)
in which f is the coefficient for the friction of rolling; W is
the load on the line of contact ; and r is the radius of the roll-
ing cylinder or wheel. Here the effort is taken at the axis of
the rolling body ; acting at the circumference of the roller or
wheel, as where straight-lined surfaces have relative motion on
interposed rollers, the force of friction becomes
The first of these two cases is illustrated in ordinary vehicles,
the second where a heavy mass on rollers has the hauling rope
or chain attached to the mass itself. In the latter case, two
frietional resistances are met at top and at bottom of the
roller. The moment of resistance is
M=Fr=fW.
The moment, of friction is evidently thus measurable by
the product of the load into an arm the value of which may
be determined by experiment, and the resistance is thus plainly
of the nature of a couple resisting rotation. This moment,
multiplied by the relative angular velocity of the two surfaces,
gives the work of rotation. The value of the arm as given by
THEORY OF FRICTIOX. 83
Coulomb and Tredgold are from f= 0.002 foot with iron to
f= 0.006 for hard wood; the load being multiplied by this
arm the moment of resistance is obtained.
The work of rolling is evidently measured by
Ma=U=Fs=Wfs,. . . . .- . (3)
i
in which s is the space through which the carnage is drawn.
The total work is this amount increased by the work of axle-
friction, and that of raising the body against gravity in passing
over the road.
Friction Gearing is sometimes used. It is made without
teeth, the periphery of the wheel being sometimes plain, some-
times grooved, on the one shaft, and made of wedge-shaped
section on the other, the one wheel driving the other by fric-
tion. In such cases the adhesion is usually found greater than
is due to ordinary friction-coefficients.
In this case the work done against rolling resistance is
measured by
(4)
where a is the relative angular velocity, b a constant depend-
ing on the conditions which affect rolling friction, and which
will be given later ; and P is the total pressure with which the
two wheels are held together. It is evident that the pressure,
P, must exceed the driving effort, P, in the proportion
(5)
or the surfaces will slip and the pair will refuse to drive.
With grooved wheels the pressure applied to hold them to-
gether may be reduced as the grooves are made with smaller
angles. The value of /is, in this case, taken as that of the co-
efficient for rest ; /= 0.15 as a minimum ; ^ == 7.
84 FRICTION AND LOST WORK.
40. The Draught of Vehicles, a case which illustrates the
first of the two methods of application of the impelling force,
for rolling friction is a matter demanding careful investigation.
Morin and later investigators disagree in their statements of
its laws. The former, who made very extended experiments,
states these laws as follows :
(1) On hard surfaces, as paved and macadamized roads, the
resistance is directly proportional to the weight of vehicle and
load, inversely proportional to the diameter of wheel, and in-
dependent of the breadth of wheel-tire. It increases with
velocity.
(2) On soft ground the resistance increases inversely as the
breadth of tire. It does not sensibly vary with velocity. Morin
concludes, also, that the line of draught should be horizontal.
Dupuit, working with carriages on macadamized roads,
found the resistance to vary nearly inversely as the square
root of the diameter of wheel, and directly as the load on the
wheel. He found the resistance on pavement to be increased
at high speeds by the concussions incident to rapid movement.
Clark obtains a somewhat less simple law, which he expresses
thus:
(i)
The work of hauling is then
..... (2)
This formula is deduced from the experiments of Macneil
on " metalled" roads.* The values of the constants for the
several formulas expressing these variously stated laws are, in
British measures, #=30; # = 4; =19 pounds per ton, v
being given in miles per hour ; these figures are derived from
Macneil's experiments.f
The resistance of all vehicles on common roads and streets
* Clark's Manual, .p. 964.
f Parnell on Roads, p. 464.
THEORY OF FRICTION. 85
is principally resistance to rolling, their axle-friction being
usually comparatively small. The work of hauling is, then,
U=Fs=fWs=fWvt ...... (3)
Railway trains are subject to the same laws as are carriages
on hard roads, although some elements of resistance here
enter which are absent in the latter case. Their wheels are
fastened rigidly to the axles, which rotate with them and
compel both wheels on the same axle to revolve with precisely
the same angular velocity. In turning curves, or where, as is
not infrequently the case, the wheels differ in size, this arrange-
ment gives rise to an increased resistance, which is sometimes
very considerable. This increase of resistance cannot occur
when the wheels are loose on the axle, as on other vehicles.
Another source of increased resistance is the friction of the
flanges of the wheels rubbing laterally against the rails.
A principal resistance of trains at ordinary speeds is, how-
ever, as with other vehicles, that of rolling friction. The re-
sistance of railway trains is commonly reckoned, in British
measure, in pounds of resistance per ton of weight of train.
Clark makes this resistance vary as a constant plus a term
which varies as the square of the velocity, thus:
(4)
the values of the constants in which are given by Clark as
a = 6 to a= 8 ; b = y^-j- to b = ^fa, the first set applying to
whole trains, the second to train exclusive of engine.
The work of hauling is then
Rs = (a + fa?)s = avt + bv*t.
On the best roads the resistance is often one half that given
above.
41. The Friction of Earth causes the retention of the
form of elevations, or the preservation of embankments when
soil is thrown up above the general level. The slope de-
86 FRICTION AND LOST WORK.
pends usually upon the internal friction of the mass ; and the
steepness of a bank of earth cannot permanently exceed the
minimum angle of repose of the material of which it is com-
posed under the most unfavorable conditions, as when soaked
by rains or floods.*
The resistance to displacement by sliding along any given
plane, in such a mass, is equal to the normal pressure exerted
between the parts of the mass on either side of that plane,
multiplied by the coefficient of friction, i.e., the tangent of the
angle of repose of the material. Thus,
F = p n tan ?>, (I)
where F is the resistance per unit of area, and p n is the inten-
sity of pressure normal to the assumed plane.
In order that no part of a detached mass shall slide, it is
thus necessary that the angle with the horizontal made by the
plane along which least resistance to motion is offered shall be
less than cp.
It is shown by Rankine, in the theory of the " Ellipse of
Stress," f that the relation of maximum and minimum pres-
sures must be such that
and
.
/, i sirup 9 * '
and hence that the ratio of their difference to their sum at any
given point must not be greater than the sine of the angle of
repose.
It is also shown J that the intensity of pressure in a direc-
tion parallel to the surface must be
cos |/(cos 3 cos a cp)
-- Z-T ^77 -5 --
cos -f- |/(cos cos a
,
A, = wx cos 6 -- Z-T ^77 -5 -- r*T - (4)
a
* Rankine "On the Stability of Loose Earth," Phil. Trans., 1856-7.
f Applied Mechanics, 112. f Ibid., 195-7.
THEORY OF FKICTION. 87
when w is the heaviness of the soil, x the depth of the point
of application, and 6 the angle of surface slope.
The intensity of vertical pressure at the same point upon a
plane parallel to the surface is obviously
p x = wx cos ...... . ..... (5)
When the surface has assumed a permanent slope at the angle
of repose, 6 =
p v wx- ........... (8)
i + sin
2 aw
u = H \ / -. : \ / 7 ; . . . (
Y i + sm (p y 3 w'
16)
which equations apply when the overturning moment is a mini-
mum.
Where jar or shake produces a displacement by settlement
of the earth behind a retaining wall, the maximum possible
pressure may be encountered, and we shall have
2 ,,!-(- sin cp
= r WH i - sin y ; ^
t = ffl + sV /l^. (I8)
I sin 9 y 3 zc;
THEORY OF FRICTION. 89
It is usually the safer course to assume these latter condi-
tions, and to give structures receiving such lateral pressures
the greatly enlarged dimensions and stability thus indicated.
42. The Pressures on Retaining Walls which sustain
level embankments are due to the resultant of the pressure
produced by a fluid mass of equal depth and density, and the
resistance to motion produced in such a mass by the friction
of its particles. The magnitude of the intensity of this re-
sultant pressure may be obtained from the expressions given
in the preceding article, or the following treatment may be
adopted :
Three cases may arise :
(1) The mass may be perfectly fluid.
(2) The mass may be semi-fluid or semi-solid, and friction
may act to reduce the pressure tending to cause the mass to
slide or to overturn.
(3) The mass may be of the kind last described, and its
internal friction may act to intensify the pressure upon the
back of the wall.
The wall, when yielding, may either slide or overturn. It
usually gives way by " bulging" on the face, and finally crumbles
down : it thus often overturns ; it rarely slides on the bed of its
foundation.
The First Case is illustrated by masonry dams and by re-
taining-walls subject to the pressure of wet quicksand or of
other soil capable of free flow.
In this case
y, (O
in which w is the weight of the unit of volume of the mass.
It is a maximum at the bottom, where/ max. = wH.
The total pressure on the unit length of a vertical wall is
the mean pressure, from top to bottom, multiplied by the
height H\ i.e.,
H (2)
90 FRICTION AND LOST WORK.
This is the pressure tending to cause the wall to slide. If
the friction of the wall on its bed is less, i.e., if
the wall will fail. If
F = fW>P, ......... (3)
the condition of stability in this respect is complied with, and
the wall will stand. For security, we should have
F=afW. .......... (4)
The point of application of this sliding effort, P, is deter-
mined by ascertaining the mean lever-arm of all the elemen-
tary efforts tending to overthrow the wall. Thus, the moment
of any elementary force, pdy, about the base, calling y the
depth from that point to the bottom, and taking unity of
length, is
.......... (5)
and the total moment is
. . . . (6)
This quantity being less, or greater, than the moment of
resistance of the wall, i.e.,
(7)
t being the thickness of the wall, the wall will stand or fall
accordingly.
THEORY OF FRICTION. 9!
Adopting for the factor of safety, a, any desired value, the
equation becomes
\Wt = \awIT; t = \ a -^; ... (8)
which gives the required thickness of wall.
The point of application of the resultant pressure on the
wall, measured from the bottom, is evidently to be found by
dividing Mby P\ i.e.,
/=?=** (9)
The " Centre of Pressure" is the point of application of the
resultant force, P, and is that point at which, if a force equal
and opposite to P be applied, it would produce an equilibrium
of efforts and of moments. Its position is measured from the
surface, as above, and the depth of the centre of pressure is
equal to the quotient of the moment of inertia of the surface
divided by its statical moment, which latter is equal to its area
multiplied by the depth of its centre of gravity.
The total pressure on the surface is thus equal to the
weight of a column of the fluid having that surface as a base,
and a height equal to the depth of the centre of gravity of this
area below the surface of the fluid.
The Second Case is met with when a mass of earth piled
against a wall, or an embankment sustained by a retaining-
wall, settles against the back of the wall without jar or other
action tending to increase pressure. In this case the pressure
is less than that produced by a fluid mass of equal density,
and is the less as the friction and adhesion of the soil are
greater. The friction and adhesion attaining a certain limit,
the soil stands without support ; or, passing this limit, it may
even require the exertion of a force to throw down a vertical
face.
To determine the pressure on the back of a vertical wall,
under the assumed conditions, we may use the equations
already given, or let the angle PEG = q> represent the angle
9 2
FRICTION AND LOST WORK.
of repose, or the angle at which the soil will lie undisturbed
by gravity. Assume a plane, BE y along which motion may
take place should the wall yield ; let its angle with the hori-
zontal be called 9, and let its angle with BP be ft.
As the angle ft increases from zero to 90 q>, the ten-
dency to slide increases from zero to a maximum ; but the
weight of the mass sliding, CBE, decreases from a maximum
FIG. 19. RETAINING WALL.
to zero. The pressure on the back of the wall is thus zero for
either ft = 90 q> y or ft = o, and is a maximum at an inter-
mediate value of ft.
Let WbQ the weight of the mass sliding, CBE, and P the
reaction of the wall, or its equal quantity, the pressure on the
wall. An equilibrium evidently exists between these two
forces, the pressure, P' y on the surface BE, and the force of
friction. Resolving perpendicularly and parallel to that sur-
face, since CBE = 90 8,
WcosO + Psm 8-P' =o; . . . . (i)
Wsin V -Pcose-/P'=o; .... (2)
W(sm 8 -/cos 8) _ P(cos 8 +/sin 0) = o;
and P= W
THEORY OF FRICTION. 93
sin /cos
cos0+/sin a
sin 6 cos tan cp
sin a tan ),
tan' i(y> -
Values of the functions of (p are given in Chapter VI.
The Third Case is illustrated by retaining-walls on which
the pressure is intensified by jar or change of volume due to
alternate freezing and thawing, the action of friction tending
to retain the maximum pressure, and by foundations.
Foundations, whether of structures or of machinery, resting
upon soil, depend for their permanence and stability upon the
friction of the particles composing it. The pressure upon the
bed of the foundation causes a tendency in the earth below to
slide laterally, and thus to permit the foundation and superin-
cumbent structure to descend. The liability to slide is zero
where the material is rigid, and becomes greater as the friction
and cohesion of the soil decrease ; until, in freely-flowing soils,
like quicksand and mud, the sole supporting pressure is that
due the hydrostatic head measured from the surface to the
given level, and is proportional to the density of the material.
The maximum horizontal pressure resisting this sliding is,
since the direction of friction-resistance is here reversed, and
we have -(-/"sin ^ in place of /"sin *u sin (pi
and the total weight which can be sustained is
/I + sin V , N
, = /' max.= awh I ! ^-} ; . . . (14)
A * *I sin (pi ^ ^'
96 FRICTION AND LOST WORK.
and the area and the total weight should be
sm
'
sin cp
/T 4-
= aAWh
the weight of building, if uniformly distributed, exceeding the
weight of soil displaced by its underground masonry in the
proportion
4- sin
r =
sn (p
Thus, for/= tan These may change places at times.
( Lard Oil )
(5) Rape-seed Oil.
(6) Other Seed Oils {Cotton-seed.
( Linseed.
(7) Castor Oil.
( Cod.
(8) Fish Oils -| Menhaden.
( Porgy.
(9) Whale Oil.
(10) Mineral Oils.
(11) Rosin Oil.
I 10 FRICTION AND LOS7" WORK.
The Best Lubricants are in general the following, for usual
conditions met with in practice :
Under low temperatures, as in rock-drills driven by com-
pressed air light mineral lubricating oils.
Under very great pressures with slow speed graphite,
soapstone, and other solid lubricants.
Under heavy pressure with slow speed the above, and
lard, tallow, and other greases.
Heavy pressures and high speed sperm-oil, castor-oil,
heavy mineral oils.
Light pressures and high speed sperm, refined petro-
leums, olive, rape, cotton-seed.
Ordinary machinery lard-oil, tallow-oil, heavy mineral oils,
and the heavier vegetable oils.
Steam cylinders heavy mineral oils, lard, tallow.
Watches and other delicate mechanism clarified 'sperm,
neat's-foot, porpoise, olive, and light mineral lubricating oils.
For mixture with mineral oils, sperm is best ; lard is much
used ; cotton-seed and olive are good.
Many different conditions must, therefore, be studied, and
the behavior of the lubricant determined with reference to
each before it can be known, with any degree of certainty, what
is its real value for any specified purpose, and it is equally
evident that the conditions under which the behavior of an
oil or other lubricating material is to be determined should
always be those approximating with the greatest possible ex-
actness to the conditions proposed in its actual use. An exact
theory of the commercial value of lubricants will be developed
in a later chapter.
52. Lubricants, as already seen, are sometimes solid, but
usually liquid ; and of the liquid unguents there are many
varieties in the market, which differ in their viscosity and
cohesiveness as widely as they do in nearly every other quality,
and range from the most limpid watch-oils to those " heavy
bodied " and densest of all the oils castor-oil and rosin-oil.
We have semi-solid lubricants, of which tallow, soap, cocoa-
nut oil, and wax are illustrations ; and still others are perfectly
hard and solid, as graphite and soapstone.
THE LUBRICAXTS. Ill
The engineer also uses what are known as " anti-friction
metals," one of the oldest and best known of which is the so-
called " Babbitt-metal." These are permanently fixed in the
bearings in the form of linings, and their peculiar use is to
present to the journal, instead of the hard, unyielding, and
resistant surface of the metal itself, a material which more
readily and perfectly adapts itself to the form of the journal
which it supports.
Lead has been introduced by Mr. Hopkins to act thus tem-
porarily, gradually, as it wears, letting the journal down to a
good bearing on the brass of the boxes.
Some anti-friction metals are used without lubricants, and
are therefore themselves as truly lubricants as are plumbago
and similar solid materials which are usually finely ground and
interposed between rubbing surfaces.
In some cases no lubrication will suffice to keep a journal
from heating, or even " cutting :" in such an event the
" brasses" are sometimes made hollow, and a stream of water
is made to circulate through them, thus effectually keeping
them cool.
In the " Palier-glissant" of Girard and the " Water-bearings"
of Shaw, the journal is supported upon a cushion of water
which is forced into a space in the journal beneath it by a
pump, and at such a pressure that the journal is perfectly
" water-borne," and revolves on the liquid cushion. Shaw has
applied this plan successfully in supporting vertical shafts.
The Oils are the most generally applied fluid lubricants;
the most common are the better known and cheaper kinds of
animal, fish, vegetable, and mineral oils : of these, sperm stands
admittedly at the head of the list ; lard, neat's-foot, whale, tallow,
seal, and horse oils are all largely used either alone or mixed.
The vegetable oils in use are olive, which is by far most gen-
erally used in other countries ; cotton-seed oil in the United
States, palm, rape-seed, oleine, colza, poppy, pea-nut, rosin,
cocoa-nut, and castor oils* are all more or less employed in
* Linseed-oil is a good reducer of friction, but dries and "gums" too
rapidly to permit its use as a lubricant.
112 FRICTION AND LOST WORK.
lubrication. Of the fish-oils, porpoise, cod, and menhaden*
oils, are most used. The mineral oils are of two classes:
the shale-oils, obtained from certain shales; and the well-
petroleums, which come from extensive oil-lakes, situated
usually far beneath the surface of the earth, and which are
principally obtained from oil-wells in Pennsylvania and other
of the United States. Glycerine is sometimes used as a
lubricant for light pressures.
Of these oils, sperm excels nearly all others in its power of
reducing friction, and generally excels them in endurance.
Rape-seed is in some districts now displacing olive-oil as a
lubricant ; but the mineral oils, pure or mixed, are rapidly
taking the leading place in all markets.f
53. The Semi-fluid Lubricants, or Soft Greases, are
usually of animal origin. The term grease is usually restricted
to those soft fats which permeate the tissues filling the cavities
of the animal system, especially about the loins and among
the intestines, and which are solid or nearly so at all tempera-
tures not greatly exceeding that of the living animal. They
usually liquefy at about this temperature, some of them be-
coming fluid at even lower temperatures than the normal.
Ignited, they burn freely, with a clear light, but with a smoky
flame.
The greases are composed of stearine, margarine, and oleine,
in variable proportions, and are the more fluid as the latter
constituent is present in larger proportion. They are partially
soluble in alcohol, and freely so in ether, in essential oils, and in
other oily compositions. When fresh they are white or light
yellow in color, and when old and altered chemically or by
mixture, often become darkened. They are always liable to
alteration, becoming rancid on exposure to air and sunlight.
This occurs by the development of the fatty acids, and this
change, which is readily detected by their odor and taste,
renders them injurious to the machinery on which they are
* The whale is not a fish, but an animal classed among the mammals.
f Portions of this chapter and of other parts of this work are from " Friction
and Lubrication," lectures by the author, published by the Railroad Gazette
Publication Co., New York, 1879.
THE LUBRICANTS. 113
used, and especially where heated, as in the cylinders of
steam-engines.
Tallow, which may be taken as the best-known example of
this class of lubricating materials, is the fat of domestic animals,
removed from the membrane in which it is secreted usually by
melting. Its quality and properties vary somewhat with the
animal, and with its age and other characteristics. It is solid
at common temperatures, white or nearly white, slightly
odorous, and readily saponifiable. The best is obtained from
mature animals, and usually, according to Chateau and other
authorities, from males of the domestic animals. The greater
part of the tallow of commerce is beef tallow and mutton
tallow.
The greases are sometimes used in the natural state, and
often mixed with other classes of lubricant.
Vaseline, and other similar preparations of mineral origin,
are to be classed with the greases, as are a number of vege-
table waxes and butters, as the so-called cocoa-nut oil. These
are rarely used in the lubrication of mechanism, however,
although the former class occasionally and the latter more
frequently are introduced into mixtures.
Vaseline and the other mineral greases are obtained by the
distillation of petroleum at low temperature in vacuo. The
vegetable greases are usually natural products.
54. For Hard Greases, as for use on railways, mixtures
of tallow and palm-oil with water rendered alkaline with soda
are often used. Two parts paraffine, one of lard, and three
of lime-water is a good grease for heavy, slow-moving jour-
nals.
A mixture of eight parts of bayberry-wax with one of
graphite is very good also, and is said by a U. S. Ordnance
Board to be the best-known preparation for rifle-bullets.
Grease is usually employed in lubricating axle-journals in
Great Britain, and is generally of palm-oil. The following are
said to be good compositions* for that climate :
*W R. Browne, Railroad Gazette, August 9, 1875
1 14 FRICTION AND LOST WORK.
RAILROAD AXLE GREASE.
For Summer. For Winter.
Tallow 504 Ibs. 420 Ibs.
Palm Oil 280 " 280 "
Sperm Oil 22 " 35 "
Caustic Soda 120 " 126 "
Water i,3?o " 1,524 "
On German railroads the following composition is used
Parts.
Tallow 24 . 60
Palm Oil 9.80
Rape-seed Oil i.io
Soda , 5 . 20
Water 59. 30
IOO.OO
The following is Austrian :
Tallow. Olive Oil. Old Grease.
For Winter 100 20 13
For Spring and Autumn TOO 10 10
For Summer 100 i 10
Tallow and " black-lead," or plumbago, " white-lead " and
oil, and mixtures containing sulphur are often used as semi-
fluid lubricants.
There exists a decided tendency to displace the more fluid
by the less fluid lubricants, to use tallow in place of the oils,
and to adopt manufactured hard greases where the more free
flowing materials have been formerly generally employed. The
change leads almost always, if not invariably, to loss of power
by increased friction a loss which is seldom noted while
saving in cost of lubricant by reduction of quantity used. In
many cases this is not economy, and a careful determination
and balancing of gains and losses is advisable before a final
choice is made.
The greases have advantages over the oils other than mere
reduction of cost of lubricating material. The cost of the
time demanded for the supply of the lubricant is usually less
with the greases ; the drip is less, and the injury by soiling
THE LUBRICANTS. 11$
floors and goods is correspondingly reduced ; danger of fire is
also less, and the journals will usually work more uniformly
cool. The greater the consistency of the lubricant, other
things being equal, the greater its endurance and economy.
The number of these greases in use is very great, and their
differences of value are sufficient to make their careful selection
by test a matter of serious importance. The method of appli- 4
cation is even a more important matter than the kind of
lubricant, or the conditions affecting it.
55. The Solid Lubricants are sometimes found to work
well when no fluid will answer at all. Some of them sustain
immense pressures without injury. Those in general use are
certain metallic compositions, mixtures of metallic with non-
metallic elements graphite, sulphur, soapstone, asbestos, lamp-
black, and white-lead (carbonate of lead). In some cases they
are permanently and solidly fixed, and sometimes are applied
at intervals between the rubbing surfaces, as are the oils.
Plumbago, or Graphite, and Soapstone are lubricants. The
former is a solid form of carbon, supposed to be the ultimate
product of the destructive distillation of the vegetable matter
of the forests of the carboniferous or, usually, earlier periods.
It is often distinctly crystalline, has a specific gravity of 1.8,
and is moderately hard. Very pure graphite, containing 99
per cent, carbon, is found at Ticonderoga, N. Y. ; in Cumber-
land, Great Britain ; and in the island of Ceylon. Crude and
impure graphite occurs in many other localities Very fine
graphite also comes from Siberia, supplying the demand for
the best grades of pencils. It is principally used for crucibles
and in pencils, but is an excellent lubricating material for heavy
work, and is also often found very useful for light machinery ;
it is used for silk-looms making delicate fabrics which would be
destroyed by oil. Its value as a lubricant is sometimes greatly
impaired by impurities, and especially if they are earthy and
gritty. Freedom from such impurities is essential to the suc-
cessful use of plumbago, either alone or mixed with other uii-
guents.
Graphite was mentioned by Rennie in 1829: he states that
"in all cases where plumbago was used it lessened friction."
H6 FRICTION AND LOST WORK.
General Morin, experimenting with it later, concluded that
it could be used to advantage where heavy pressures were to
be sustained. The author has found graphite, and graphite
mixed with certain oils, well adapted for use under both light
and heavy pressures. It is especially valuable to prevent abra-
sion and " cutting," under very heavy loads and at low veloci-
ties. Plumbago is used generally by interposition, although
often forming, as just stated, an ingredient in the composition of
mixed oils and of anti-friction and " anti-attrition" compounds
of the first class. It should always be absolutely pure and free
from grit, and should be ground to the condition of a flaky pow-
der.
Mr. T. Shaw found it superior to oil for the tables of heavy
planers.
Soapstone is a hydrated silicate of magnesia, known also
as talc and as steatite. It is very widely distributed. It is soft,
easily cut by the knife, and has an unctuous quality, to which
it owes its name. For use as a lubricant, it must be free from
gritty impurities, and can be then employed like graphite. It
is extensively used in the manufacture of " packing" for the
piston-rods and valve-stems of steam-machinery.
Some engineers express a preference for soapstone powder
as a lubricant for the axles of machines. For this purpose
it is first reduced to a very fine powder, then washed to remove
all gritty particles, then steeped for a short period in dilute
muriatic acid, in which it is stirred until all particles of iron
which it contains are dissolved. The powder is then washed
in pure water again to remove all traces of acid, after which it
is dried, and forms the purified steatite powder used for lubri-
cation. It is not generally used alone, but is mixed with oils
and fats, in the proportion of about 35 per cent, of the powder
added to paraffine, rape, or other oil ; the powder may be mixed
with any of the soapy compounds employed in the lubrication
of heavy machinery. These solid lubricants are both used in
making up packing for steam-engines, etc.
Plumbago and soapstone are both used, mixed with soap, on
heavy work, and especially on surfaces of woodworking against
either iron or wood.
THE LUBRICANTS. 1 1/
Asbestos is a silicate of lime and magnesia, having some
resemblance to soapstone in its physical properties, but dis-
tinguished by its structure, occurring in, often, long silky
fibres. It is spun into threads and ropes, and woven into
cloth, and even felted, and made into paper. It is used for
piston-rod packing and if well purified is excellent for this
purpose.
Sulphur, " White Lead" and some other solids are used
generally mixed with oils ; but they are not important mem-
bers of the class of substances here considered.
Woods, as lignum-vitae, beech, hickory, oak, maple, elm,
canewood, snakewood, are sometimes used as bearing surfaces,
and are almost always kept cool and prevented from wearing
seriously by flooding them with water. The best of these
woods are, like lignum-vitae, hard and tough in structure ; they
are usually obtained from the tropics.
56. The " Animal Oils" are usually derived from the fats
of the mammiferous animals, including the whales and their*
relatives ; but they are sometimes obtained from fish, as from
the " menhaden" or " moss-bunker." The .principal of these
oils are sperm and whale oils, lard and neat's-foot oils. Tallow-
oil is also used to some extent. They are generally obtained
by melting them out from the animal tissues in which they
are originally found, and by passing them through various
purifying processes. All have characteristic and persistent
odors, which are often, as in the case of the fish-oils, disagree-
ably powerful, and which are even perceived in the soaps made
from them. The liquid animal oils are principally derived
from the sperm and the " right" whales.
57. Sperm Oil, or spermaceti-oil, is the best known, and
for general purposes the most excellent, of all the lubricants.
It contains, according to Brande : carbon, 78 ; hydrogen, 1 1.8 ;
oxygen, 10.2. It is found in a large cavity in the head of the
sperm-whale, mingled with the solid fat, spermaceti, from
which it is separated by crystallization and pressure, without
heating. It is saponifiable with potash, but with difficulty,
and is one of the most permanent and most valuable of all the
oils. Its specific gravity ranges from 0.880 to 0.896, averaging
Il8 FRICTION AND LOST WORK.
about 0.885. Crude "head-oil" from the cask runs about
O.88. It is the lightest of all the lubricants. Sperm-oil is of
light-orange color in large masses, lighter in small quantities,
transparent, has a slight fishy odor, and precipitates needle-
like crystals of spermaceti at 47 F. (8.3 C.). It is solidified
by nitric acid.
Used as a lubricant, it is liable to sudden fluctuations of its
coefficient of friction in consequence of its changes of density
and fluidity, as the spermaceti contained in it alters with vary-
ing temperature. In lubricating quality, for light work, as for
spindles, it is only excelled by the very finest of the refined
mineral oils, and excels nearly all other oils under heavy
pressures, although often closely approached by fine petro-
leums. Exposed to the air it absorbs oxygen, becomes gradu-
ally " gummed " or resinous, and loses quality seriously. At
140 F. (60 C.) it gains two or three per cent, in weight in
twelve hours. It has a " flashing point" at about 500 F.
(260 C.).
Whale Oil is obtained from the " blubber" of the whale by
removing it from the animal in great strips, which are then
heated to melt the oil out from the tissues enclosing it. All
the whales, including not only the varieties classed with the
sperm and the right whale, but also the blackfish and their
relatives, the dolphins, furnish this " train-oil." Three varieties
of oil the so-called white, yellow, and black are brought into
the market, and are mixed to form the oil of commerce. Com-
mon whale-oil is brownish yellow, transparent, disagreeably
odorous, limpid at ordinary temperatures, solidifying at the
freezing-point, and precipitating at times more or less sperma-
ceti. Its density is about 0.93 at 70 F. (21 C.). It is much
used in making crude soaps and for illuminating purposes,
usually mixed with vegetable oils, and little used for lubrica-
tion.
58. Lard Oil is the most extensively used of all the animal
oils, and is an excellent lubricant, although inferior to sperm-
oil. It is obtained from the fats of the hog. It is exported
from the United States to Europe in large quantities for the
purpose of adulterating olive-oil. It is itself often adulterated
THE LUBR1CAXTS. 1 19
with cotton-seed oil, which latter is also used as a salad-oil, but
sold, however, as olive-oil. Ail three oils are good lubricants.
Lard from which the oil is expressed yields 62 per cent, of its
weight, the specific gravity approximating 0.925. It saponifies
readily, congeals at the freezing-point of water, and " flashes"
under fire-test at about 500 F. (260 C.). If sperm-oil be
rated at unity as a lubricant under ordinary conditions, lard-
oil will stand at 0.75 to 0.95. This oil is twice as viscous as
sperm. Exposed to air it absorbs oxygen with far less rapidity
than sperm-oil.
59. Neat's-foot Oil is one of the best of lubricants, and
has extensive use in the arts. It is obtained by boiling the
feet, and often other parts, of cattle, and skimming off the oil
which rises to the surface of the water. It has a very slight
straw-yellow color, which darkens with age ; it is odorless when
fresh, has a pleasant taste, is limpid at all common tempera-
tures, but congeals at about the freezing-point of water.
Its density at 60 F. (15.5 C.) is 0.916. It is very frequently
adulterated with other less expensive oils. When allowed to
stand for any length of time, it often deposits white flakes of
solid fats. Its low temperature of congelation makes it a
very useful oil for out-of-door machinery. It resembles lard-
oil in general appearance and qualities.
Tallow Oil is made from the tallow of beeves by pressure,
and has very similar qualities to the preceding. The tallow is
melted, the stearine separated by slow cooling and straining,
followed by pressing. The oil is a good lubricant, but is
principally used in fine soap-making.
60. Fish Oils, so called, include the whale-oils already
described, and the oil of the menhaden and other fishes.
Seal Oil is also often classed, even more improperly than
the whale-oils, with the fish-oils. It is not a common oil
in our markets, and is rarely used for lubrication, although a
good unguent.
Porpoise Oil is used as a watch-oil, for which purpose its
limpidity and stability of composition well fit it. It resembles
the best whale-oils. The " porpoise-oil " of the market is very
generally made, not from the porpoise, but from the jaws and
120 FRICTION AND LOST WORK,
the " melons" of the blackfish. It does not congeal at the
zero of the Fahrenheit scale ( 1 8 C.). It is refined by
straining cold. Rusty iron placed in the bottle with the oil
keeps it free from acid. It is very expensive. " Grampus"
oil is even better than porpoise or blackfish oil. Dolphin Oil,
Cod-liver Oil, Dugong or Sea-calf Oil, and the oils of the
herring, the sardine, and other fish, have still less use in the
mechanic arts.
Menhaden Oil has been used by the author for the pres-
ervation of steam-boilers out of use for long periods of time,
with very satisfactory results. It forms an impermeable and
almost unchangeable greasy varnish, which protects the iron
from oxidation very thoroughly.
All these oils, like the animal oils, generally dissolve to
a certain extent in alcohol. They are usually extracted by
maceration and pressure.
61. The Vegetable Oils are obtained from the seeds, and
occasionally from the fleshy part of the fruit, of plants. In
one case, that of the earth-almond, the oil is found in the
woody tissue of the root. These oils are usually limpid, but
sometimes are so hard as to be properly classed as greases.
The oils are expressed by grinding the seeds, adding water,
and finally treating the emulsion of water, oil, and albuminous
matter to separate the oil.
The vegetable oils are divided into two classes, the fixed
or non-drying, and the drying oils. The former are permanent
liquids, like the animal oils; the latter are subject to a process
of oxidation which causes their gumming, and the formation
of a resin which is useful as a kind of varnish, and as a vehicle
for holding colors in painting. The drying-oils, among which
the best known are linseed, castor, hemp-seed, walnut, and
poppy oils, are of little value for purposes of lubrication.
Castor-oil, when fresh, is a moderately good lubricant for
heavy pressures, although the fixed oils are vastly better for
common use. It changes much more slowly than linseed-oil.
The non-drying oils, of which the principal are olive, cotton-
seed, almond, rape-seed, cocoa-nut, pea-nut or ground-nut,
and colza, are all good lubricants. Of these the first named
THE LUBRICANTS. 121
is by far the best known ; although cotton-seed, pea-nut, and
colza oils are also extensively used.
The gain in oxygen and the loss of the hydrocarbons in
eighteen months, by the process of " drying," is thus shown
by analyses made by Cloe'z :
LINSEED OIL.
Fresh. Aerated.
Original weight. Per cent. Total weight. Difference.
C 77-57 67.55 72.299 - 5.271
H 11.33 9.88 10.574 o-756
O ii. 10 22.57 24.157 +13-057
CASTOR OIL.
C 74-361 72.125 74-058 - 0.303
H 11.402 11.108 11.405 0.003
14.237 16.767 17.217 -f- 2980
62. Olive Oil is obtained from the fruit of the Olea Europea,
one of the jasmines, which grows throughout Southern Europe
and Northern Africa, and in other semi-tropical countries.
The total production of the world is vastly less than the nomi-
nal sale, the commercial oil being adulterated to an enormous
extent. It is extensively used as a table-oil, as well as for
illuminating and lubricating purposes. The finer grades of
fruit are harvested by hand-picking, and reserved for the
manufacture of table-oils. The larger varieties of olive furnish
the less excellent grades of oil which are used in the arts.
Each part of the fruit, the outer skin, the pulp, the enclosed
seed or nut, supplies an oil of peculiar quality ; but they are
rarely separated. The oil from the pulp being comparatively
free from stearine, remains fluid at lower temperature than
that from the other portions of the olive, and is sometimes ex-
tracted separately as a watch or a clock oil.
In making olive-oil, the fruit is usually first stored about
two weeks in bins, and allowed to ferment slightly, in order
that the softened cells may yield their oil the more easily and
completely. The fruit is then crushed in an " edge-roller
mill," and the oil removed by exposing the pulp so produced
122 FRICTION AND LOST WORK.
to heavy pressure while enclosed in bags and under a screw-
press. The expressed oil runs into tanks of water, and is then
separated by skimming. The "virgin-oil" is that which first
comes off or often that which drains, unpressed, from the
crushed paste at the roller-mill. That which is afterward ob-
tained is called " ordinary oil." An inferior quality is obtained
afterward from the mixture of water and paste, which is left
to settle in a large reservoir called "Tenfer" and this oil is
therefore called " /mile d'enfer;" it is used for a cheap lamp-
oil.
Good olive-oil is limpid, unctuous, sometimes colorless,
but usually golden yellow or greenish yellow in color, trans-
parent, and if fresh very slightly odorous. Its taste is sweet
and fruity, and pleasant to the palate of many persons ; but it
becomes disagreeable and is unpleasantly odorous when it be-
comes rancid with age. Its density varies, according to Saus-
sure, from about 0.92 at the freezing-point to 0.86 at the boil-
ing-point of water. It congeals at a low temperature, deposit-
ing flakes of stearine as it approaches the freezing-point.
Heated, it begins to change to a darker color at about 248 F.
(120 C.), and fumes at 356 F. (180 C), without decomposing,
however, as a mass ; it must be heated to a considerably higher
temperature before breaking up.
All the burning and lubricating varieties of olive-oil arc
obtained after removing the virgin-oil and finer grades of ordi-
nary oil. They are allowed to remain stored, and are kept warm
in tanks for some months to precipitate all foreign substances:
they are thus easily and rapidly clarified in summer, less
rapidly and perfectly in winter. Good olive-oil is the best
vegetable lubricant. Exposed to air, it shows symptoms of
rancidity in a single day. It is much more viscous than sperm,
and less so than neat's-foot oil. The best olive-oil is, for some
purposes, equal to sperm ; and it is even claimed to be superior.
Good olive-oil is one of the most perfectly non-drying of all
the oils ; it experiences no other change with long exposure to
the air than an increase of viscosity, only slightly observable,
according to Cloez, after a year and a half ; it is then increased
in weight 3f per cent.
THE LUBRICANTS. 123
63. Cotton-Seed Oil has been produced since about 1856,
in large quantities, in the United States, from the seed of the
common cotton-plant as removed from the " boll " by the
" gins." It is obtained by crushing the seed and expressing
the oil, very much in the same way as other seed-oils. It is, in
large quantity, of a dark reddish yellow, and of a rather deep-
yellow color in smaller masses. It has a pleasant taste, is to
some extent a slightly drying oil, and is used in adulterating
non-drying lubricating oils, in making soaps, and for illumina-
tion.
This oil is nearly as permanent as olive-oil ; Cloe'z exposed
it to the air for a year and a half without observing other
change than a slight loss of fluidity.
The crude oil may be refined by Botch's method by stirring
several hours, with three per cent, of its volume of caustic-pot-
ash lye, of 45 B., or with six per cent, soda solution of 25 to
30 B., for an hour, at the boiling-point of the lye. Yellow,
clear oil, of density 0.926, separates from a brown soap-stock,
and is decanted. Forty gallons of oil are made from a ton of
seed : this is about one half the oil contained in the seed, which
averages about 25 per cent, oil, by weight.
64. Rape-seed Oil is expressed from the seeds of the sev-
eral kinds of brassica, of which Brassica napus and B. rap is a conically-shaped vessel, small
enough usually to be carried conveniently in one hand, which
FIG. 28. CRANK-PII* LUBRICATOR.
has a flexible and elastic bottom ; while at the upper and
smaller end of the eone a tapering tube is screwed which has
a very small orifice at its extremity. This little instrument
being filled or partly filled v/ith oil, held between the middle
fingers and inverted, the pressure of the thumb on the bottom
causes the oil to spurt from the point of the tube in a fine jet,
which is directed to the point at which the oil is needed.
85. " Oil-Pumps" are sometimes used where the bearing
to be lubricated is either peculiarly important, as the steps of
vertical shafts or the " thrust-bearing" of a steamship, or where
ft is unusually liable to heat. In such cases a reservoir of consid-
erable volume is placed in a convenient location and nearly filled
with oil, a pum$ connected by its suction-pipe with this reser-
voir, and by a force-pipe with the bearing, is kept in operation
METHODS OF APPLYING LUBRICANTS. 149
by connection with the mechanism to be oiled, and an ample
and continuous supply is thus secured. Even this arrangement
is liable to failure, and to cause the very accident that it is
intended to prevent if the oil used is not absolutely free from
foreign material, if the connections are not all well made, if
the valves of the pump leak or fail to seat properly, or if the
pump-plunger is not kept well packed.
86. Water-Bearings have been adopted in some cases, as
by Shaw and by Giffard, the "palters glissants' of the French
engineers, in which the weight of a revolving shaft is taken
by a cushion of water, or sometimes of oil, and in which the
journal does not bear upon metal at all, except as it may be
necessary to steady it. The journal enters a bearing so con-
structed that the liquid can be forced between the two adja-
cent surfaces in such quantity and under such pressure that
the journal is supported by and turned upon the fluid cushion
so formed. The excess of the liquid which flows out at the
end of the bearing returns to the reservoir below, and is again
circulated by the pump. Journals thus arranged have been
known to work many months without appreciable wear, and
even without discoloration of the liquid.
87. Unlubricated Bearings, cooled usually by the flow of
water across them, are sometimes found preferable to any
other device for sustaining parts having relative motion unde?
pressure. Thus the " stern-bearings" of screw-steamers are
almost invariably fitted up in this manner. The screw-shaft
of iron or steel is encased in brass and turns within a long,
hollow, cylindrical sheath, which is fitted with narrow strips of
lignum-vitae, separated by longitudinal spaces forming water-
channels. No lubrication is employed, and the bearing is kept
cool by the flow of water between the strips of lignum-vitae.
Such bearings wear well in clear water, but cut away rapidly
in shallow water over sandy bottom. The lignum-vitae if kept
cool will sustain enormous pressures, and will wear in such
situations better than metal.
88. Bearing Surfaces are of bronze or other alloys, of
cast-iron or other metal, or of wood, according to location, in-
tensity of pressure, velocity of rubbing, and nature of the
ISO FRICTION AND LOST WORK.
material of the journal. Ordnance bronze wears well under
heavy pressures and at high speeds if not subjected to intense
localized pressures by the springing or misfitting of parts ;
cast-iron has an advantage, if used under moderate pressures
and in ample extent of surface, in its porosity and absorptive
power and the persistence with which oil and grease adhere to
it ; wrought-iron and steel sustain heavy loads, if free from sur-
face defects; ''mild steel "is peculiarly valuable for journals,
and hard steel ground to shape and well bedded in its bearing
will safely carry pressures of enormous intensity ; wood is only
used in special cases. Too high a polish on the harder surfaces
is objectionable where thin oils and heavy pressures are
adopted, as the lubricant is difficult to feed between the
metals in contact, or to keep there while in operation.
It is nearly always advisable to make the bearing of the
softer metal, since its renewal is a matter of less difficulty and
expense than that of the journal, and since the journal must
usually have great strength. A hard bearing cuts the softer
journal, and gives rise often to serious expense. It is from
this consideration that bearings are often " babbitted "or lined
with the soft white alloys.
The fitting of the surfaces in contact is as important a
matter as the selection of the material of which they are com-
posed. The theory of friction is based upon the assumption
that all parts are accurately made to correct dimensions, and
exactly fitted ; and the conclusions derived are therefore in-
validated by any departure from such assumed conditions.
Precision and stability of form stiffness of all loaded parts
are essential elements of successful working. Stability of form
is dependent upon extent of surface exposed to wear: if this
area is ample, so that the two rubbing parts nowhere and at
no time come into unrelieved metallic contact, no appreciable
wear will occur, and their forms will be permanent.
Surfaces of similar area and form, even when well fitted, if
of different materials will wear very differently. Thus the
following table shows the comparative wear of axle-bearings.
Thoroughly pure bronzes, like those fluxed with phosphorus,
METHODS OF APPLYING LUBRICANTS.
were reported as wearing very much less than ordinary com-
positions.
BEARING.
COMPOSITION.
Cost
per
100 IDS.*
Miles
run
per Ib.
Wear
per .
TOO miles
for four
bearings.
Cop-
per.
Tin.
Anti-
mony.
83
82
3
5
17
18
90
85
7
10
$28 60
28 68
32 85
32 27
13 04
28 68
25,489
27,918
22,075
24,857
22,921
2,576
200grs.f
252 *'
366 "
28 4
308 "
274 "
White-metal
. i ii
Lead Composition: lead,
84* antimony 16
Gun-metal on brake- cars . .
82
18
In many cases the excessive wear of a bearing is due to a
misfit. The Hopkins bearing is a bronze bearing lined with a
thin layer of lead, which, when new and unfitted, can accom-
modate itself to the distorted journal and permit gradual wear
to a correct fit without danger of injury, such as occurs often
with the common hard, unlined " brass." In the Defreest
bearing a thin bronze bearing-piece is sustained by a strong
iron backing-piece, and between them is a sheet-lead filling.
Journals should be fitted without the use of emery or other
gritty grinding material, which may adhere to its surface and
thus produce injury.
Bearing Surfaces of Wood are, under the conditions already
described as favorable to their use, exceedingly durable, and
will carry enormous loads without abrasion. Thus lignum-
vitae will sustain pressures exceeding 1000 Ibs. per square inch
(70 kgs. per sq. cm.), where brass becomes rapidly abraded and
destroyed under but little more than one fourth of that load,
and will run continuously under 4000 Ibs. (281 kgs. per sq.
cm.) when bronze sets fast instantly. Camwood has been sub-
jected to pressures exceeding 8000 Ibs. per square inch (562
kgs. per sq. cm.), and has worked without injury ; snakewood
carries about as heavy a load as lignum-vitae.
* Including melting expenses, loss, etc. These figures are constantly
varying.
f Seven thousand grains per pound.
!52 FRICTION AXD LOST WORK.
The bearing surfaces of watch-work are often made of
ruby, agate, and other fine-grained and hard stones, and of
gems.
A comparison made by the author between surfaces of
gun-bronze, of " Babbitt"-metal, and of other soft, white alloys,
all working on steel, proved all to have substantially the same
friction. In other words, the coefficient of friction was deter-
mined by the nature of the unguent and not by that of the
rubbing surfaces, when the latter are in good order. The soft
metals, however, heated more than the bronze, running at
temperatures somewhat higher with equally free or even freer
feed. To retain the temperature at 135 F. (57 C.), in some
cases one half more oil- over 300 grammes, as against 200
was needed on the white metal than on the bronze. This
probably does not, however, necessarily indicate a serious de-
fect, but simply deficient conductivity. Lined journals may
be expected to run normally warmer than unlined bronze of
good quality. The following are the results of experiment
with a " Babbitts-metal, which was compared with bronze and
a second white alloy:
Bronzes. White Metal.
No. i. No. 3.
Mean Temperature, Fahr 133 152 137
Mean Coefficient of Friction o.oio 0.013 o.oio
Oil used per hour, ounces 7 17 12
These differences prove ordinary lubricated surfaces to
have contact, since they give differences in the values of f
where none could exist were the friction fluid-friction solely.
CHAPTER V.
THE INSPECTION AND TEST OF LUBRICANTS.
89. Systematic Methods of Examination of Lubricants
are always necessarily adopted by large consumers of lubri-
cants. The opportunity for adulteration is so great, and a mis-
take in purchasing is so liable to result in serious accidents and
large expenses for repair, or for wasted driving-power, that
very considerable expenditure of time and money is often jus-
tified in the endeavor to secure reliable determinations of the
quality of the unguent which it may be proposed to use.
These methods of test are often physical, sometimes chemical ;
and very frequently they consist of direct methods of deter-
mination of the value of the oil in reducing friction, and of
its durability under wear and under the conditions of every-
day work.
Of these tests the simplest is the measurement of the den-
sity of an oil; any variation from that of known pure oils of the
same nominal grade being evidence of adulteration or of prob-
ably low quality. The method to be described as u oleography"
is another physical test, and the so-called " fire-tests" are other
illustrations of this class. The chemical tests are usually pro-
cesses of qualitative analysis, and the last-mentioned systems
of test are generally practised by the use of " testing-machines,"
forms of which will be described later.
The density of the oils is always less than that of water,
and varies from 0.875, that of sperm-oil, to 0.99, that of
the heaviest rosin-oils. The gravity of the oil, except per-
haps in the case of sperm, is not a definite gauge either of the
nature or of the quality of a lubricating oil, as mixtures may
be made of any desirable density. There is also no direct re-
lation between their lubricating property and their density.
154 FRICTION AND LOST WORK.
The determination of density is therefore an aid simply, and
not a real test of quality
The Color of an oil is a noteworthy characteristic of a pure
oil, but is so readily imitated and so frequently purely the
result of accident, that it cannot be assumed to be a reliable
guide in selecting lubricants. The best oils are, however, usu-
ally either colorless or very slightly yellow : a few are brown-
ish or brownish reel, and olive-oil has a slightly greenish tint.
The Odor of oils is due in the case of the animal oils to
the presence of a volatile compound, generally acid, as butyric,
valeric, or other fatty acid, and in the hydrocarbons to vola-
tile vapors, as naphtha. The vegetable oils are often distin-
guished by odors peculiar to the plants from which they are
obtained.
The Fluidity of the oil is not only very different in different
cases, but is very variable with change of temperature. It is
quite independent of density.
90. The Detection of Adulteration is the principal ob-
ject of the tests of unguents. The most valuable of the oils,
as sperm and olive oils, are rarely found in the market per-
fectly free from adulteration. The former is adulterated with
blackfish and other cheaper oils, the latter with cotton and
other seed oils; and even the cheaper oils, as lard, are often
mixed with cotton-seed and various inexpensive but not al-
ways seriously objectionable oils. The lubricating oils in most
general use are now almost invariably mixed oils ; and the
greases are as universally made up by mixture, the character-
istic odor of the cheaper fats being concealed by that of oil of
almonds or other fragrant substance. It is evidently import-
ant that the engineer should be able to determine when an oil
or grease is pure, and to detect the nature and determine the
extent of adulterations if it should prove to be impure.
The modern methods of testing oils are directed to the
determination of a number of independent facts. These ob-
jects are :
(1) Their identification and the detection of adulteration.
(2) The measurement of density.
(3) The determination of their viscosity.
INSPECTION AND TEST OF LUBRICANTS. 155
(4) The detection of tendency to gum.
(5) The determination of temperatures of decomposition,
vaporization, and ignition.
(6) The detection of acidity.
(7) The measurement of the coefficient of friction.
(8) The determination of their endurance, and their power .
of keeping the surfaces cool.
It is sometimes sufficient for the user of an oil to identify
it and to be able to detect adulterations. Sperm and lard oils,
for example, are standard lubricants ; and if the consumer or
dealer can assure himself that the oil which he has in hand is
pure sperm or pure lard, that is often enough, since long expe-
rience may have taught him that this oil and no other is likely
to fully answer his purpose.
The tests for identification are chemical and physical. The
chemist can sometimes, by applying " reagents" which have
peculiar effect on an oil, determine whether that oil is sperm
or lard, or other, and detect adulterations. This is in some
cases quite easy to do and tolerably certain, since there are
usually very few oils of which the cost would be low enough
to permit their use as adulterants. For example, the chemist
would look for cotton seed oil, perhaps, in his tests of so-called
pure lard oil, since that, in the present condition of the mar-
ket, is about as likely to be used as an adulterant of lard as
any other oil. The chemical methods of test would rarely be
used, except by an expert chemist; and it is enough to describe
a few of the best known.
91. Chemical Methods of Test have been proposed in
great variety.
Animal and vegetable oils are distinguished by the fact that
chlorine turns animal oils brown and vegetable oils white. ^
Some special tests are quite reliable for certain adulterations,
and chemists have devoted much time to their discovery and
to perfecting methods.
The alkalies saponify fats and oils, and the soaps so made are ^
compared in the detection of adulterations. Potash gives soft,
and soda hard, soaps.
The strong acids destroy the fats, altering them in very
\$6 F.RICTI&M AX0 LOST WOKK.
much the same manner as does the application of heat ; and
their action is accompanied by the development of heat, the
amount of which is an indication of the nature of the oil. The
reactions of sulphuric and of nitric acids have been very thor-
oughly studied. Chlorine and iodine have also been much used
in this work. The action of the oil on metajs, as on copper or
brass, is indicative of the presence of acid in the oil ; the
amount of this action, as evidenced by alteration of color, is a
gauge of the quantity of acid present. Acid is not found ia
pure mineral oils. Sperm and neat's-foot oils, and tallow, are
very often acid either from chemical alteration or from the in,-
troduction of foreign compounds having acid reactions.
Professors Crace-Calvert, Cailletet, Chateau, Wurtz, and.
many other chemists have systematically studied the reactions
of oils with various chemicals, with a view to their identifica-
tion and the detection of adulteration.
When, without any previous knowledge of the nature of any
substance, it is proposed to discover all its constituent parts,
and to furnish a proof that, besides the elements exhibited by
analysis, it does not contain others, it is necessary to proceed
with a method, and to follow strictly a systematic plan. Meth-
ods of analysis may be numerous and of various kinds, but they
are founded upon the same principles and all present the same
character. In fact, in all methods of analysis certain reactions
are made use of, which enable us to divide all bodies, or all
those under consideration, into classes that are perfectly defined.
Such characteristics are always made use of that each of these
sections shall comprise, as nearly as possible, equal numbers of
bodies which exhibit in the same degree the reactions which
have served to establish the group. By another set of charac-
teristics, new divisions and subdivisions are established in each
of these classes. Proceeding in this way, a certain number o,f
substances are eliminated, with which we need no longer occupy
ourselves ; and after some tests, usually but few in number, we
acquire the knowledge that the elements of the composition
submitted to analysis belong to such or such section or class, or
to one of the divisions or subdivisions.
It is only after having arrived at this result that we seek to
INSPECTION AND TEST OF LUBRICANTS. 1 57
determine by a special method the body considered, by making
use of specific characteristics and particular reactions.*
92. Chateau's Methodsf are among those which by gene-
ral reactions form such classifications as facilitate the determi-
nation of the nature of the oil, and consequently allow its
purity to be judged.
These general reactions are
(1) The use of bisulphide of calcium, giving a soap which
remains colored or loses its color.
(2) The colors given with the sirupy chloride of zinc.
(3) The colors produced by ordinary sulphuric acid.
(4) The colors produced by forming bichloride of tin.
(5) The colors given, both cold and warm, with sirupy phos-
phoric acid.
(6) The colors given by the pernitrate of mercury employed
alone or together with sulphuric acid.
These general reactions are rendered complete by the use
of several other reagents, potassa, ammonia, nitric acid, etc.,
the use of which will be stated in the monography of the fats.
Finally, the nature of the oil will be ascertained with certainty
by testing for special characteristics and particular reactions.
The tests may be made in a large watch-glass placed on a white
paper, on a glass plate; also on white paper, or in a small white
porcelain capsule. In practice the watch-glass has been prfc*
ferred.
93. Preparation and Use of the Reagents.
Bisulphide of Calcium. This is easily prepared by boiling
a mixture of flowers of sulphur xvith chalk and water. After
boiling a half-hour it is filtered. That which has been prepared
several days is to be preferred.
Chloride of Zinc (sirupy). This reagent is prepared b> v
saturating pure hydrochloric acid with oxide of zinc and
evaporating to dryness. A sirupy aqueous solution is made of
the product.
* Precis d'Analyse Chemique Qualitative. MM. Gerhardt et Chancel,
f Guide Pratique de la Connaissance, et de 1'Exploitation des Corps Gras
Industrielles. Theodore Chateau Paris, 1864.
158 FRICTION AND LOST WORK.
Sulphuric Acid (commercial and colorless). This acid is used
in the proportion of 3 or 4 drops to 10 or 15 drops of oil.
Bichloride of Tin (fuming). This reagent is obtained from
dealers in chemicals. It is also called the "fuming liquor of
Libavius."
Phosphoric Acid (sirupy). A strongly concentrated solution
resulting from the action of nitric acid upon phosphorus, or
else a sirupy solution of phosphoric acid prepared in advance,
or, better still, bought of the druggist.
Pernitrate of Mercury. -- This is prepared by dissolving
mercury in an excess of pure nitric acid. The use of this re-
agent is twofold : 1st, in the observations of color produced by
the salt alone ; 2d, in observations of the colors produced by
sulphuric acid when poured over the oily mass after the action
of the salt of mercury.
Potassa. Concentrated solution of caustic potassa. Chateau
uses alcoholic potassa.
Ammonia. That of commerce colorless.
Nitric Acid (pure). Commercial.
All these reagents are employed by pouring a few drops
(four or five) on the oil, which is placed in a watch-glass, cover-
ing about half its surface.
With the concrete oils, the fats, tallows, and waxes, four or
five drops of the reagent are used with a piece of the fat of
the size of a pea.
94. The Reactions of Oils when they are subjected under
similar conditions to the general reagents already indicated are
given in the following tables by Chateau.
To facilitate and guide investigation, the oils are divided
into mineral oils, the drying and non-drying vegetable oils, and
animal oils.
INSPECTION AND TEST OF LUBRICANTS.
59
oq
xoi
>ic
S8
5
o
^
s
t
1 111= II
.1 rit i*
. c w ^ ^ . C
~ 4; w wU'T't)-;;
SB .> .>.>^'.> |
- cre5 6 So^S
| o u
55
160
FRICTION AND LOST WORK.
z u
ll
If
s!
sS
IS
"x
u .
!i
il
la
?E
?
l
si
. . w
5:= a." fe&fi
"5 S= ^^ w >
2
!f?!~
i?
CQ u
1^
n
R!
11
.2 a
' ."g
/ ^ x w
.
fa-i s
i ^ i
w ^ a
1^1
162
FRICTION AND LOST WORK.
*o
i
l
fa
o
^ o
o Y
w
,'aa
!i
15
1"
O
sT
o
-3
u
v,-
jm
--
O O CJOiCUUCQU
**-< OrtairtSiiajO
OO c^ UOiCUUt/jMU
Q
U
X
.> ^-.>
6^0
.c o
I
Ii
O"
o
O
*-; jy
i
O O O uoiu
S = c S C
5 i c= c o
>3a^S
Ul ^ ' rt
o
|^ 5|*f
|| |
^t 5 !
Itlll
llt
INSPECTION AND TEST OF LUBRICANTS.
163
P 3
Si
2 s
fc, V-
rt
"
13
s u
^^
S&
p
iiiiii
O <*
U
f?>iJ555 Sj
^cfi'- . ga*!>:
> IKS 16
6 5 I 81
z^>
-
164 FRICTION AND LOST WORK.
95. The Tables of Reactions are referred to after first
observing the indications furnished by organoleptic methods ;
the odor, taste, color, and consistency are characteristics that
often assist in determining the method of adulteration. Seve-
ral cases may be presented in the analysis of oils.
(1) Having a commercial oil the name of which is unknown
(without label or label effaced, for example), to ascertain what
it is.
(2) Knowing to what class an oil belongs, but not knowing
its name, to ascertain it. For example, knowing of an oil that
it is a drying, fixed, or animal oil.
(3) The name of an oil being certainly known, to ascertain
whether it is pure or not.
These are three questions that the chemist, the purifier, or
even the consumer, may at any time be called upon to decide
particularly the last.
^ First Case. Knowing nothing of the oil, to ascertain its
name.
First try the bisulphide calcium as directed in the instruc-
tions for preparing reagents. Suppose, for example, the oil
gives a golden-yellow emulsion which retains its color. The
oil tested may be linseed, nut, olive (fine or crude), sweet-al-
mond, colza, rape-seed, sesame, camline, cotton, sheep-foot, tal-
low, or sperm. If in the reaction it does not produce effer-
vescence and evolution of sulphuretted hydrogen, it cannot be
tallow-oil. That is eliminated.
Try next a current of chlorine for a quarter of an hour. If
it produces no black coloration, it is not sperm-oil.
Try chloride of zinc. This reagent gives a green, greenish,
or bluish-green coloration ; the table gives the linseeds of
India, Bayonne, and North Europe, colza, camline, sweet-al-
mond, refined olive, and the other grades of olive, cod, and ray
oils.
The oil tested cannot be the lower grades of olive-oil, cod-
liver, or ray-liver: bisulphide of calcium would have identified
them. On the other hand, it is not rape-seed, sesame, cotton,
English linsieed, or sheep-foot, as the chloride of zinc would
have detected them. We are thus limited to the linseeds of
INSPECTION AND TEST OF LUBRICANTS. 165
India, North Europe, and Bayonne, colza, camline, sweet-al-
mond, and the higher qualities of olive oil.
Try sulphuric acid. Assume it gives, for example, a dark
reddish-brown and " dragon's-blood " color. Consulting the
tables, it is seen that such effect indicates the linseed-oil of
different countries, and a series of fixed and animal oils which
had been eliminated by the preceding tests.
The oil is, therefore, linseed-oil, and it only remains to
determine its origin.
Thus, without using the remaining tables, the name of the
oil supposed to be offered for test is determined. By trying
the reactions given by the other reagents indicated, the nature
of the oil can be still more precisely ascertained. It is evident
that another order of operations might have been followed,
but it is best to commence with the bisulphide of calcium.
This reagent divides the oils into two great groups ; and we
next proceed, using first simple then the more complicated
tests.
Second Case. Having given, for example, a fixed oil, to
ascertain its name.
Try bisulphide of calcium. This reagent may give, for
example, a golden-yellow emulsion, which retains its color.
The oil can be neither olive of low quality, pea-nut, nor beech.
It is useless to try chlorine here.
Pass on to chloride of zinc. We may obtain, for example, a
greenish or bluish-green color; the oil cannot be a poor quality
of olive-oil, sesame, rape-seed, or cotton-seed. There remain
colza, olive, camline, or sweet-almond.
Test with sulphuric acid. This reagent gives, say, a red-
dish-yellow color. This eliminates colza and illuminating olive-
oils, leaving camline, sweet almond, and fine olive.
Apply the fuming bichloride of tin. Perhaps a light
brownish red may appear instantaneously, and with it a thick
mass of faint or straw-yellow color. The first reaction elimin-
ates sweet-almond and best olive ; the second confirms the
first.
The oil must then be camline. Special reactions given in
the monography of this oil will clearly identify it.
1 66 FRICTION AND LOST WORK.
The most unfavorable example has been selected to illus-
trate fully the use of these agents. Had a soap been obtained
which did not retain its color, it would have limited the
further investigation to only four oils. In such cases the labor
is vastly reduced. A similar process would determine the
name of any animal oil.
The bisulphide of calcium effects a primary division three
oils on one side and eight on the other. If the characteristics
developed indicate one of the eight, the use of chloride will
eliminate the fish-oils, leaving it to be decided whether it is
neat's-foot or horse-foot oil.
Third Case. To ascertain the purity of any oil indicated.
As an oil is only adulterated with oils less costly, it is
usually not difficult to decide upon a limited range of possible
adulteration. It is also evident that an oil can only be adul-
terated with a similar oil of inferior quality, or one possessing
very similar properties. Thus an edible oil could not be adul-
terated with an oil of strong odor, as olive with fish, etc. It
is true that a difference of price does not invariably limit adul-
teration, as the price varies in different seasons, and sometimes,
even, from day to day. Thus colza is at one time quite costly,
while linseed is cheap, and vice versa. The adulteration of
colza with linseed is therefore quite probable, it is practised to
a great extent, but in other seasons the contrary is the case.
Suppose the purity of edible poppy-oil is to be tested ?
After having noted the organoleptic indications, test with the
bisulphide of calcium. Suppose a soap obtained which retains
its color? All the oils giving a soap which loses color are thus
eliminated. Without further test, an examination of the
tables will show that the three animal oils, sheep-foot, oleic
acid, and sperm, are also easily eliminated, these oils having
characteristic odor and taste. The linseed-oils also have odor,
and are not edible. The adulteration could not be with fine
olive-oil, for it is too costly. Illuminating olive-oil has a
characteristic taste and odor, which throws that out. Cotton-
seed oil, by its color and taste, and the oil of sweet almonds,
by its price, are thrown out of the question. There remain
nut, colza, rape-seed, sesame, camline, and poppy.
INSPECTION AND TEST OF LUBRICANTS. 1 67
Try the chloride of zinc. Suppose a white or slightly yel-
lowish mass be obtained? This reaction eliminates colza, rape-
seed, and camline, leaving nut, sesame, and poppy.
Next use sulphuric acid, which may give a reddish-yellow
color. As the nut-oil does not give this reaction, there remain
sesame and poppy.
Trying the fuming bichloride of tin, it gives a faint-yellow
coloration and a straw-yellow solidified mass. We still find
these reactions to indicate sesame and poppy oils. It then
becomes certain that the poppy oil is adulterated with sesame.
Try phosphoric acid. This gives, perhaps, a faint yellow
orange yellow. The detection is complete, since poppy-oil
should give a white emulsion. Lastly, try Behrens* reagent,
which will determine the presence of the oil of sesame with
certainty.
These methods apply equally well to the greases as to the
oils.
The reactions of the oils have been studied by many chem-
ists, among whom are to be especially mentioned, besides
Chateau, Calvert, Prescott, Gerhardt and Chancel.*
These reactions, for greater convenience, have been col-
lected into a single large table for the author by Mr. L. S.
Randolph, which table is here given.
* Prescott's Organic Analysis. Precis d'Analyse Chemique Qualitative.
MM. Gerhardt et Chancel
FRICTION AtfJ) LOST WORK.
TABLE II. PHYSICAL AND CHEMICAL PROPERTIES OF OILS AND
COLOR REACTIONS.
[Compiled from CHATEAU, CALVERT, PKESCOTT, and other authors.]
KIND OF OIL.
S. G.
Con-
cealing
Point.
Natural Color.
Odor.
Taste.
Drying
Quali-
ties.
Calcium
Bisulphide.
Almond
0.918
0.920
0.925
0.930
0.963
0.914
-20 C.
-18* C.
-18 C.
Below
J4 F.
-i 5 C.
- 6 6 C.
Clear straw-yel-
low; limpid.
Yellowish.
Clear golden yel-
low.
Clear yellow to
red brown; acid
reaction.
Sirupy; colorless.
Limpid; clear yel-
lowish.
None.
Nearly in-
odorous.
Peculiar.
Fishy.
Nauseating.
Bland
sweetish.
Mild.
Peculiar.
Fishy.
Mild; acrid
after-taste.
Fixed.
Fixed.
Fixed.
Animal.
Drying.
Fixed.
Animal.
Drying.
Animal
Drying.
Animal.
Fixed.
Fixed.
Fixed.
Fixed.
Drying.
Fixed.
Fixed.
Animal.
Animal.
Animal.
Animal.
Animal.
Drying.
Animal.
Fixed.
Golden yellow ;
permanent.
Golden yellow ;
not permanent.
Golden yellow ;
permanent.
Golden yellow ;
not permanent.
Golden yellow ;
not permanent.
Golden yellow ;
permanent.
Golden yellow ;
not permanent.
Golden yellow ;
not permanent.
Dark gray ; ef-
fervesces, giv-
ing off H 3 S.
Permanent.
Not permanent.
Permanent.
Not permanent.
Not permanent.
Not permanent.
Beech-nut
Camline
Castor
Colza
Fish
Hemp-seed .
0.926
0.915
6934
0.916
0.917
0.917
0.963
0.924
0914
0.921
0.875
-25 C.
10 C.
to o C.
-27 C.
Below
0C.
5* C. to
2C.
i 4 *c.
+ 4 C.
" 3 C.
-18 C.
- 6C.
oC.
Greenish when
fresh, after-
wards brownish
yellow.
Colorless, or
nearly so.
Gold yellow to
brownish.
Yellowish.
Greenish or yel-
lowish; thick-
flowing.
Greenish yellow.
Golden yellow,
passing to
brown.
Made hot it is yel-
low; almost
colorless.
Limpid; straw
yellow.
Clear yellowish.
Yellow.
Limpid; orange
yellow.
Unpleasant.
Slight odor
of lard.
Strong.
None.
Slight pleas-
ant or none.
Insipid.
Lard
Linseed,
Strong.
Bland.
Mild
sweetish.
Neat's-foot
Olive
^Refined).
Olive
(Ordinary salad)
Olive
Very odorous.
Almost odor-
less.
Slightly pleas-
ant odor.
Disagreeable.
Mild.
Fishy.
(Huile d'enfer)
Pea-nut
I*oppy-seed
Rape-seed
Slightly
pleasant
taste.
Disagree-
able.
Mild.
Permanent.
Permanent.
Sperm
Seal
Not permanent.
Tallow, Mutton
Tallow, Beef...
Tallow, Veal...
Walnut
0.925
0.925
0.925
37 C.
37 C.
Melts
betw'n
fingers.
-i8C.
oC.
i.C.
Hard white.
Hard white.
Soft white.
Slightly greenish
or yellowish;
thick.
Brownish.
Yellow or brown;
yellow to color-
less.
Decays
rapidly.
Decays
rapidly.
Nearly odor-
less.
Disagreeable.
Mild nutty.
Disagree-
able.
Mild.
Permanent.
Not permanent.
Permanent.
Whale
Cotton-seed
INSPECTION AND TEST OF LUBRICANTS.
169
TABLE II. PHYSICAL AND CHEMICAL PROPERTIES OF OILS AND
COLOR REACTIONS Continued.
KIND OF OIL.
Chloride of
Zinc.
Sulphuric
Acid.
Fuming
Bichloride
of Tin.
Thickened
Mass, from
SnCl a .
Cold
Phosphoric
Acid.
Almond
White mass,
slightly yellow
or uo color.
Flesh rose.
Yellowish green
to bluish green.
Greenish yellow
to bluish yellow.
Yellowish rose.
Greenish yellow
to bluish yellow.
Yellow to
brown.
Yellow.
(eddish brown.
Reddish yellow.
Violet red,
crimson violet,
hen darkbrown
bright yellow,
then reddish
yellow.
Green veins or
greenish color.
brownish black.
Green veins or
green color.
R.ed brown.
Dark brown;
brownish red.
Yellow, then
orange yellow.
Yellow.
Yellow.
Yellow, then
reddish yellow
Dark brown to
reddish brown
Bripht yellow,
then orange
yellow.
Green veins or
greenish color
Yellow to
reddish yellow
Brownish red.
Dark brown.
Yellowish.
Pale yellow;
when stirred a
reddish yellow
Canary yellow
slightly orange
Reddish brown
Brownish red.
Reddish brown
No color.
Reddish yel-
low.
Jro \vnish yel-
low to reddish
brown.
Green to green-
ish blue.
No color to
golden yellow.
Green to green-
ish blue.
Deep reddish
brown.
Green.
Reddish.
Bluish green.
Reddish yel-
low.
Yellow (?)
Yellow.
Reddish yel-
low.
Distinct brown
Reddish yel-
low.
Greenish.
Faint yellow.
Purplish; red-
dish brown.
Brownish.
Canary yellow;
H,S0 4 deep-
ens the tint.
Deep yellow.
Canary yellow;
H,S0 4 deep-
ens the tint.
Reddish yel-
low.
Orange yellow
Orange yellow
Canary yellow.
Reddish yellow.
r aint yellow.
discolored.
White.
Discolored.
Reddish yellow.
White.
Greenish.
Reddish yellow.
Dark green.
31ear yellow.
Straw-yellow.
Yellow.
jreenish.
Greenish.
Greenish.
Straw-yellow.
White.
White.
Straw-yellow
and orange
yellow.
Straw-yellow.
Distinct brown
red.
No color.
Beech-nut
Cod-liver
Castor
7 aint yellow.
Yellow to brown.
Deep sepia.
Dark green.
Does not thicken;
brown red.
Brownish.
Orange yellow.
Colza
Fish
Lard
Reddish-yellow
emulsion.
Greenish yellow;
bluish yellow.
White mass,
slightly yellow
or no coior.
Greenish yellow
to bluish green.
Greenish yellow
to bluish green.
Greenish yellow
to bluish green.
Yellow to
brown.
White mass or
no color.
Yellow to
brown.
No color or
white mass.
No color or
white mass.
Reddish brown.
No color.
No color.
Unseed
Xeat's-foot
Olire
(Refined).
Olive
Orange yellow.
Reddish yellow.
Brownish red.
Yellow.
Dirty green.
Yellow.
Orange yellow.
(Ordinary salad)
Olive
(Huile d'enferj
Pea-nut
Poppy-seed
Rape-seed
Sesame
Sperm
Seal
Tallow, Mutton.
Tallow, Beef . .
Tallow Veal
Stringy yellow
mass.
Walnut
White mass,
slightly yellow
or no color.
Yellowish brown
Dark brown.
White.
Straw yellow,
then orange
yellow.
Golden yellow.
Whale
Clear mahogony.
Yellowish brown
Cotton-seed
FRICTION AND LOST WORK.
TABLE II. PHYSICAL AND CHEMICAL PROPERTIES OF OILS AND
COLOR REACTIONS Continued.
KIND OF OIL.
Hot
Phosphori
Acid.
Pernitrate
of
Mercury.
Addition of
Sulphuric
Acid.
Potash.
Ammonia.
Almond
Beech-nut
Camline
Cod-liver
Faint
yellow.
Faint
yellow.
Faint
yellow.
Dirty green
Bright
yellow.
Brown.
Blackish.
Green or
greenish.
Golden
yellow;
effervesces
Bright
yellow.
Bright
yellow.
Reddish
yellow.
Gray
Gray
Grayish
white.
No colora-
tion.
Straw
yellow.
Straw
yellow.
White
emulsion
Greenish.
Golden
yellow.
Greenish
after
stirring.
No color.
Greenish.
Reddish
yellow.
Golden
yellow.
Greenish
yellow.
Light
chocolate.
Light reddish
brown.
Reddish brown
then chocolate
Dark brown.
Canary yellow
golden yellow
at first.
Dirty flesh
color.
Brownish
black.
Dark reddish-
brown.
Violent effer-
vescence; choc
olate brown.
Reddish brown
Reddish
yellow.
Raw sienna.
Reddish
yellow.
Reddish
yehow.
Chocolate.
Dark brown.
Brownish gray.
Orange-yellow ;
green veins.
Jght brown
and black,
irownish
black.
Slight choco-
late.
Vhite precipi-
tate ; brown-
ish violet.
Vhite precipi-
tate; sienna
passing to
sepia.
Sudden effer-
vescence.
Dark chocolate
brown.
Light
chocolate.
Greasy yellow soap
Thick white emul-
sion.
Greasy yellow soap.
White emulsion, when
hot.
Castor
Colza
Flocculent white
soap.
White emulsion.
Fish .
Hemp-seed
Lard
Greenish-yellow soap,
very thick.
White soap, very thick;
gelatinous when
heated.
Clear golden-yellow
emulsion.
Thick white emulsion.
Reddish yellow
soap; very thick.
Pale yellow emul-
sion.
Difficult to saponify
Linseed
Neat's-foot
Olive
(Refined).
Olive
(Ordinary salad)
Olive
Thick yellowish-
white soap.
3 ale-yellow soap,
like a precipitate
5 ale-yellowish
soap.
Greasy emulsion;
not homogeneous.
Deep yellow, homo-
geneous soap.
Very thick, gelatinous
soap; very white.
I!lear-yellow sonp, be-
coming yellowish
white.
Thick soap; slightly
yellowish.
Yellow emulsion.
Deep-yellow emulsion,
becoming homogene-
ous and pale clear
yellow.
(Huile d'enfer)
Pea-nut
Poppy-seed
Gray
Jo color.
Jrown.
Faint
yellow,
iright
yellow.
Blackish.
Greenish
yellow.
Greenish
yellow.
Faint
yellow.
Slightly
yellowish.
Straw-
yellow.
White.
\o color.
Reddish
yellow.
Pale rose.
losy when
cold, dis-
appears
when hot.
>Jo color at
first, after-
wards flesh-
color,
^o color.
r aint
yellow.
r aint
yellow.
Rape-seed ...
Sesame
Sperm
Yellow emulsion,
slightly reddish,
leddish-yellow
soap.
'ale-yellow emulsion.
Thick reddish-yellow
soap.
Seal
Tallow, Mutton.
Tallow, Beef. , . .
Tallow, Veal....
Walnut
Bright
yellow,
{.eddish
brown.
r aint
yellow.
Whale
Orange emulsion,
changing to thick
soap.
Homogeneous red-
dish-yellow soap,
with green veins.
Yellow emulsion, pass-
ing to yellow-white.
Cotton-seed
INSPECTION AND TEST OF LUBRICANTS.
To detect acid, dissolve a small piece of sodium carbonate
in an equal volume of water, and introduce the solution with
the oil to be tested into a flask, and agitate thoroughly. The
quantity of precipitate will be a gauge of the amount of acid
present.
The application of the senses of taste and smell, in the test-
ing of lubricants, to be satisfactorily useful demands great
familiarity with, and experience in the use of oils, and can be
practised with satisfactory results, usually, only by experts.
Some oils, however, are so characteristic in taste and odor
that a novice may readily recognize them. It is always best
to compare the suspected oil with a sample of known purity.
The characteristic odor of an oil can be brought out more
strongly by warming it. The taste, odor, and " feel " of the oil
are sometimes considerably modified by the locality whence it
is obtained, by the season during which it is prepared, and by
the method of manufacture.
METHODS IN DETAIL are given as follows by M. A. Re-
mont :*
Qualitative Analysis should be preceded by an examination
of the organoleptic properties of the oil, the manner in which
it behaves under the influence of heat, and of its specific grav-
ity. If the specific gravity of the sample is below 0.900, it con-
tains a mineral oil; if from 0.900 to 0.975, it may contain the
most complex mixtures; but if it is above 0.975, it is certainly
an oil of resin.
Begin by treating the sample with carbon-disulphide, freshly
prepared, which gives a clear solution with all oils. If oleic
acid or a fatty oil has been mixed with alkali to raise its spe-
cific gravity by the formation of soap, there will be a precipi-
tate. In such case the liquid is filtered, and the residue washed
with carbon disulphide. It may be shown to be soap by its
solubility in water, its alkalinity, and the turbidity more or less
marked, which is caused by an acid poured into the solution.
The filtrate is next freed from the carbon-disulphide by
distillation : I c.c. of the residue is mixed with 4 c.c. of alco-
* Bulletin de la Socitie' chimiqite de Paris. Chemical News, 1880.
1 72 FRICTION AND LOST WORK.
hol at 85. If solution takes place, fatty acids are pres'ent,
pure or mixed, and an excess of alcohol is gradually added
If after having poured in 50 c.c. the liquid is limpid or very
slightly cloudy, which cloudiness disappears on adding a drop
of hydrochloric acid, the sample consists of oleic acid, pure or
mixed with resin. If the specific gravity does not exceed
0.905 at 15, the sample is pure oleic acid. If the specific grav-
ity is higher, it contains resin. By way of confirmation it may
be examined with the polariscope, either alone or dissolved in
carbon-disulphide ; and if there is a deviation the presence of
a resinous mixture is indicated.
If persistent cloudiness is observed in the alcoholic solution
the fatty acids contain an oil sparingly soluble in this solvent,
and in greater quantity as the cloud appears earlier. This
process renders it possible to detect 2 or 3 per cent, of mineral
oil, of resin, or fatty oil in the oleic acid known as oleine. The
turbidity produced in the alcoholic liquid resolves itself after a
time into little oily drops, which line the sides of the vessel
and which can by jarring be made to fall to the bottom of the
tube. The volume of this residue shows approximately the
proportion of insoluble matter.
In the usual case 4 parts of alcohol do not completely dis-
solve I part of oil. A larger quantity of the latter is then
taken and agitated with an equal volume of alcohol. After
settling, the alcoholic solution is decanted, and evaporated in a
capsule. The nature and the quantity of the residue give a
clew to the nature of the mixture.
Next submit the oil to the action of caustic soda, employ-
ing the method of M. Dalican for the analysis of tallows. In
a capsule of porcelain, or preferably of enamelled cast-iron,
there are weighed about 20 grammes of oil, and heated to 100
to 1 10. There is then poured in a mixture of 15 c.c. soda-lye
at 36 B., and 10 c.c. of alcohol; the mixture is stirred and
heated until the alcohol and the greater part of the water have
disappeared. Then 150 c.c. of distilled water are added, and
the boiling is kept up for half an hour, when three cases may
occur :
(i) The oil under the influence of the alkali is merely
INSPECTION AND TEST OF LUBRICAN'IS. 173
emulsified, and on the addition of water it separates distinctly ;
this indicates either a mineral oil, a resin-oil, or a mixture of
the two. The aqueous solution is decanted off, and is mixed
with sulphuric acid. If there is no precipitation, or if but
slight cloudiness is produced, the sample is a pure mineral oil.
If there is a considerable precipitate which collects in brown
viscid drops, giving off a strong odor of resin, and soluble in
an excess of alcohol, we have a resin-oil, pure or mixed. The
oil is examined with the polariscope, and if it acts upon polar-
ized light this is a confirmation of the presence of resin-oil. If
the specific gravity^is below 0.960, there is some mineral oil
present. A test may be made by distillation if one of the oils
is not in too small proportion. The distillation should be
fractional as far as possible, and conducted slowly. As the
resin-oils boil, as a rule, at lower points than the heavy mineral
oils, it follows that, in place of having specific gravities which
increase with the boiling-points, as happens with the heavy
mineral oils or pure resin-oils, there are observed with their
mixtures very abrupt transitions. The sample ought to be
tested with tannic chloride, and if the violet coloration is not
very distinct, the same reagent should be applied to the first
products of distillation, since the colorable product contained
in the resin-oils is there chiefly met with.
(2) There is formed by the action of caustic soda a paste-like
mass of soap, which on treatment with water and boiling for
some time gives a clear liquid. It is diluted with cold water
and then supersaturated with acid. The fatty acids liberated
collect on the surface after decantation of the water, and if
exposed to cold crystallize. A small portion is melted in a
tube at a low temperature, and 4 parts of alcohol at 85 are
added first, and later an excess. Here two cases are possible:
A. If no precipitation takes place it is because the fatty
acids are pure, which shows that the oil examined is a pure
fatty oil, or, which rarely happens, mixed with resin. The
specific gravity of the fatty acids may here give good indica-
tions, but it cannot be taken at ordinary temperatures, at
which fatty acids are solid. They must be melted, and the
specific gravity taken at a definite temperature. M. Baudouin
174 FRICTION AND LOST WORK.
has given a table of the specific gravities of the fatty acids of
certain oils taken at 30 C. Except for linseed-oil, which
marks 0.910, the fatty oils have specific gravities ranging from
0.892 to 0.900. To reduce the specific gravities of the fatty
oils examined to the temperature of 30, deduct from the
density found, calculated on the litre, as many times 0.64
gramme as there are degrees below, or, if the temperature is
higher, to add to the density found as many times 0.64
gramme as there are degrees above. If the specific gravity
indicates that the neutral oil contains resin, an attempt may
be made to separate it, in part at least, rapidly by agitating
5 or 6 c.c. of the original oil with an equal volume of alcohol,
decanting after settling, and evaporating in a capsule. There
is thus obtained a solid or semi-fluid residue in case of resin.
Further examination is then made with the polariscope.
B. The fatty acids derived from the decomposition of the
soap give a precipitate if treated with an excess of alcohol. If
it is not, redissolve by I gramme of hydrochloric acid, and if
after some time it is resolved into oily drops, it is mineral oil
or resin-oil. A fatty oil containing 10 to 15 per cent, of one of
these oils is completely saponified, and yields with boiling
water, not an emulsion, but a soap completely soluble. The
turbidity should yield oily drops, for there are certain fatty
acids those, among others, of the oil of the ground-nut or
pea-nut (arachis) which are soluble in a small proportion of
alcohol at 85, but an excess of alcohol precipitates a sparingly
soluble portion of arachidic acid in small flocks. These flocks
may be collected on a filter, and examined as to their com-
plete solubility in alkalies. If their melting-point is near 73
they may be attributed to pea-nut oil.
(3) Or, lastly, the oil on treatment with soda may give a
paste more or less firm, which, if placed in boiling water for
half an hour, allows oily drops to rise to the surface, which are
due to a mineral oil or a resin-oil. After settling for some
minutes, a part of the supernatant liquid is decanted and
mixed with an excess of a saturated solution of common salt.
There is produced a precipitate of soap, which is filtered off
on cooling. The filtrate is supersaturated with an acid. If
INSPECTION AND TEST OF LUBRICANTS. 1/5
there is produced a slight turbidity, and if the liquid, which
was almost colorless when alkaline, gives off an odor of fatty
matters, we have a neutral oil mixed with a non-saponifiable
oil. If, on the contrary, the solution was highly colored after
filtration, and gives, when acidified, a flocculent precipitate of
a resinous odor, the sample is a mixture containing resin. In
these two cases the components of the mixture may be recog-
nized by means of the operations indicated above.
Quantitative Analysis. If it is desired to know the elements
attacked by alkalies, the following method is to be followed :
If the sample has yielded bodies insoluble in carbon-disulphide,
it is separated, and the operation is confined to the residue of
the distillation. Let it be assumed that the composition of
the residue is as complex as possible, containing fatty oils,
mineral oils, resin-oils, and solid resin.
The mixture is saponified. Into a flask closed by a stopper,
through which passes a long tube, are introduced 20 grammes
of the oil, and a mixture of 15 c.c. of soda at 36 B., and 15
c.c. alcohol at 90 to 95 per cent. The flask is then set upon
the water-bath for half an hour, and is often shaken. At the
end of this time the whole is poured into a funnel fitted with
a tap and previously warmed, and which is left in a stove at
50 to 60 until a complete separation of the non-saponifiable
oil from the alkaline liquid has taken place. The latter is then
decanted into a porcelain capsule, and in its stead is poured
15 c.c. of boiling water, which has served to rinse the flask. It
is shaken well so as to wash the non-saponifiable matter, and
decanted anew after settling. Finally it is washed a third
time with boiling water. The oil in the funnel is received in
capsule and weighed. What adheres to the sides is washed
with a little ether, and the solution is received in another cap-
sule, which is exposed to the air till the ether has principally
disappeared. It is then gently heated to expel the rest, and is
weighed.
The alkaline liquid is kept boiling for some time to expel
the alcohol, and after cooling it is mixed with an equal volume
of a saturated solution of common salt freed from magnesia
by being boiled for a few moments with caustic soda and then
1/6 FRICTION AND LOST WORK.
filtered. In this manner the soap is precipitated in firm clots,
carrying with it the last portion of non-saponifiable matter.
The saline solution after settling is decanted by means of a
pipette, and neutralized with an acid. If a notable turbidity
is produced which collects in flocks, it is due to the presence
of resin. The flocks are collected, dried, and weighed. The
clots of soap are thrown upon a filter, washed twice with salt
water, the last traces of which are removed by pressing the
mass between sheets of blotting-paper. The soap is then
placed in a glass beaker, moistened with about 100 c.c. of car-
bon -disulphide recently rectified, stoppered, gently shaken at
intervals, so as not to break the clots, three or four times, and
left to settle. After an hour or two the carbon-disulphide,
which is colored yellow by the dissolved oil, separates in the
lower part of the beaker. It is decanted by means of a pipette,
and in its place is added a fresh portion of the solvent. It is
shaken, left to settle, decanted, and so on, till the carbon-
sulphide runs off almost colorless. The whole is then thrown
upon a filter and washed for the last time. A portion of this
last washing, if evaporated upon a watch-glass, should leave
an insignificant residue.
The soap on the filter is exposed to the air till the carbon-
disulphide with which it is saturated has escaped. As for the
carbon-disulphide solution, it is distilled gently on the water-
bath. The last portions of the solvent are expelled by blowing
air into the flask while placed in boiling water. When cold it
is weighed.
The last portion of the non-saponifiable matter thus ob-
tained should have the same appearance as the first portion.
If it is less fluid it still contains a portion of soap. In this
case it is again taken up in carbon-disulphide, at a gentle heat,
in presence of a few drops of water, to hydrate the soap,
which without this addition would again be partially dissolved.
It is then filtered, and the washed soap is added to the princi-
pal mass.
The non-saponifiable oil may consist of mineral oil, resin-
oil, or a mixture of both. The means of detection have been
given, but a satisfactory process for their separation is needed.
INSPECTION AND TEST OF LUBRICANTS. \TJ
The soap insoluble in carbon-sulphide, which lies on the
filter, contains resin and fatty acids combined with soda.
The separation of these substances presents many difficul-
ties. Several methods have been published, but none of them
gives satisfactory results. That of M. Jean consists in exhaust-
ing the barium-soap with ether, which should dissolve the
resinate and leave the soaps of the fatty acids untouched. It
is difficult to avoid the partial solution of the barium-oleate.
Substituting for the ether boiling alcohol at 85 per cent., it
dissolves much less of the oleate, but still takes up too much.
As far as possible the soap is separated from the filter and
placed in a capsule. The filter is put back in the funnel and
filled with boiling water, The solution is effected slowly, and
it filters by degrees; it is received in the capsule where the
detached portion has been already placed.
The solution of soap after cooling is mixed with caustic
soda until precipitation ceases, and is left to settle. All the
soap of the fatty acids is deposited, carrying down with it the
chief portion of the resinate, a part of which, however, remains
in solution and colors the liquid strongly. The whole is
filtered, the filtrate accurately neutralized with sulphuric acid;
the flocks of resin deposited are received upon a filter, which is
weighed anew after washing in water and drying at a low tem-
perature. The soap is redissolved in a little lukewarm water
and an excess of barium-chloride is poured into the solution
with agitation. The clots of barytic soap are drained in a
filter-pump, replaced in the capsule in which the precipitation
has been effected, and thoroughly dried in the water-bath or
the stove. The mass is then powdered, and treated with 50
or 60 c.c. of alcohol at 85 per cent., which is kept near the boil-
ing-point, working it up with a pestle. It is left to settle for
a few moments, and the supernatant alcoholic liquid is then
decanted into a vial. 20 to 25 c.c. of alcohol are again poured
upon the residue, let boil, decanted after settling, and so on
till a portion of the alkali which has been used leaves on eva-
poration scarcely any residue, which happens generally after
120 c.c. of alcohol have been used.
The alcoholic liquids are mixed and distilled till there re*
1/8 FRICTION AND LOST WOKK.
mains only about 50 c.c. Hydrochloric acid is aaJed to decom-
pose the resinate, and the resin, set at liberty, floats in the
liquids. On cooling, it collects in a solid mass at the bottom
of the vessel. It is thrown into a capsule, melted under water,
and weighed after desiccation on the water-bath.
The residue insoluble in alcohol is treated in a similar
manner to obtain the fatty acids.
Olive-oil is sometimes tested for purity by simply applying
heat.
This test is very simple, and can be performed by any one
possessing a good chemical thermometer. About a teaspoon-
ful of oil is put in a test-tube, and a thermometer suspended
in the oil, which is now to be heated to 250 C. (472 F.).
For a comparison, a second test-tube of pure oil may be treated
in like manner. Pure olive-oil, when heated, grows rather
lighter in color, but most other oils, like cotton-seed, pea-nut
oil, etc., grow darker. The latter, also, evolve a penetrating
and disagreeable odor, but olive-oil has a pleasant smell not
unlike strawberries. This test, devised by Merz, is considered
worthy of a trial.
When mixed with cotton-seed oil, the following method is
proposed by Dr. Nickels :*
Pure olive, or " Gallipoli," oil, as examined by a Browning
" direct vision" or pocket spectroscope, presents a deep shadow-
ing, or cutting-out, of the blue and violet ray, with a fine,
almost indistinct, line in the green, and a strong deep band in
the red.
Refined cotton-seed oil similarly examined presents exactly
the same appearance, but as regards the blue and violet ray
only, the green and red being continuous.
If we take as a standard a given stratum of pure olive or
Gallipoli oil in a test-tube, and a similar stratum or thickness
of the standard oil in admixture with cotton-seed, there is no
discernible difference as regards the shadowing in the blue and
violet ray, but an almost entire fading out of the delicate line
in the green, and a considerable diminution in the depth and
* Chemical News.
INSPECTION AND TEST OF LUBRICANTS. 179
intensity of the strong band in the red, consequent upon
" dilution" or " thinning down." With 50 per cent, in admix-
ture, the loss in intensity is considerable; with 25 per cent,
the variation is marked and discernible.
A suspected sample compared with and differing thus from
the standard, and in the absence of any direct chemical evi-
dence as to the nature of the oil in admixture, might fairly
fall within the range of strong presumptive evidence pointing
towards " cotton-seed " oil as the probable dilutant.
Pure olive-oil is exceedingly difficult to secure with certain-
ty when purchasing in large quantity, as it is often greatly
adulterated at the point of production. It is usually very diffi-
cult to distinguish the several vegetable oils in any mixture of
them.
96. Alterations of Composition occur in the animal and
vegetable oils, with exposure to air and light and with advanc-
ing age, which may sometimes cause some uncertainty in the
chemical work already described. These changes are usually
in the direction of those modifications which lead to the pro-
duction of resins. The oils become darker, more viscous, less
susceptible to the action of reagents, and, if time be allowed,
finally become " gummed," and completely altered into resins
of various degrees of solidity. Such changes are so plainly
observable, however, that no special tests are necessary to in-
dicate their commencement or their progress. The mineral
oils are not subject to such alterations to any serious extent,
unless very long exposed to the action of oxygen and of light,
in which case the absorption of the gas and its conversion into
ozone, with some loss of lubricating power and greater reduc-
tion of its value as an illuminant, become matters of some im-
portance.
97. The Action of Oils on Metals is sometimes important.
Copper and lead, and other soluble metals, are sometimes found
in oils; and Dr. Stevenson McAdam found that the second of
the two metals above named may go into solution to such an
extent as to injure the quality of the oil as an illuminant very
seriously. In such cases the metal is usually absorbed by the
oil from the metallic walls of the vessels in which it is stored.
ISO FRICTION AND LOST WORK
Dr. McAdam found this to occur to such an extent as to clog
up the wick and ultimately diminish its capillary attraction so
much that the light was extinguished. The wicks when
charred left a fine net-work of lead. The action of the oil on
tin, copper, and iron was slight, and its illuminating properties
were not much diminished. Zinc, however, was quickly at-
tacked, and the oil was as seriously affected as by lead. While
the vessels for the retention of paraffine-oil may be safely con-
structed of or be lined with tin, copper, or iron, it would
evidently be preferable to use tanks lined with enamel for
storing the oil.
Detection of Copper and Lead. To detect the presence of
copper, mix a small portion of the oil with twice its weight of
nitric acid in a test-tube, and shake well ; then, separating the
acid from the oil, add ammonia to the former: if copper is
present, the reaction will give a blue color by the formation of
an ammoniacal solution of that metal.
To detect lead, add to a portion of the oil, contained in a
test-tube, a small quantity of sulphuric acid, of carbonate of
soda, or of caustic soda : if lead is present the solution will be-
come white, and will yield a precipitate of similar color. To in-
sure certainty, add to the solution caustic soda until the acid,
if used, is neutralized, or add acid, if soda has been used, and
a few drops of sulphur-solution, the presence of lead will be in-
dicated by a dark-brown precipitate. With bichromate of potas-
sium or the iodide of potassium, a yellow precipitate is found.
Dr. Watson concludes,* in regard to this action
(1) That of the oils used, viz., linseed, olive, colza, almond,
seal, sperm, castor, neat's-foot, sesame, and paraffine, the samples
of paraffine and castor oils had the least action, and that sperm
and seal oils were next in order of inaction.
(2) That the appearances of the paraffine and the copper
were not changed after 77 days' exposure.
(3) That different oils produce compounds with copper vary-
ing in color, or in depth of color, and consequently rendering
* Paper read in the Chemical Section of the British Association, Plymouth
Meeting, 1879.
INSPECTION AND TEST OF LUBRICANTS. l8l
comparative determinations of their action on that metal from
mere observations of their appearances impossible.
He later * experimented further, with the following results,
noted, after one day's exposure, with iron :
(1) Neafs-foot. Considerable brown irregular deposit on
metal. The oil slightly more brown than when first exposed.
(2) Colza. A slight brown substance suspended in the oil,
which is now of a reddish-brown color. A few irregular
markings on the metal.
(3) Sperm. A slight brown deposit, with irregular mark-
ings on the metal. Oil of a dark-brown color.
(4) Lard. Reddish brown, with slight brown deposit on
metal.
(5) Olive. Clear and bleached by exposure to the light
and air. The appearance of metal same as when first im-
mersed.
(6) Seal. A few irregular markings on metal. The oil
free from deposit, but of a bright clear red color.
(7) Linseed. Bright deep yellow. No deposit or marks on
metal.
(8) Almond. Metal bright. Oil bleached and free from
deposit.
(9) Castor. Oil considerably more colored (brown) than
when first exposed. Metal bright.
(10) Paraffine. Oil bright yellow, and contains a little
brown deposit. The upper surface of the metal on being
removed is found to have a resinous deposit on it.
The tendency of an oil to act on metals varies with the
proportion of free acid and kind of oil, and also with the
nature of the metal. Nearly all fatty oils act more rapidly on
copper than on iron. The following table shows results ob-
tained by Watson with iron exposed to the action of oils for
twenty-four hours and with copper after ten days' exposure :
* Swansea Meeting, British Association, 1880.
182
FRICTION AND LOST WORK.
ACTION OF OILS ON METALS.
OILS.
Iron dissolved
in 24 days.
Copper dissolved
in 10 days.
.0040 gn
.0048
.0800
.0250
.0050
.0875
.0062
.0045
.0050
.0460
lin.
. 1030 grain.
Castor .
.0170 grain.
Lard
.Sooogra
.1100
.2200
.0015 '
.0485
.0030
n.
Olive
Seal
There is evidently no relation between the action of an oil
on copper and the action of the same oil on iron : in several
instances, those oils which act largely on iron act slightly on
copper, while those which act largely on copper act little on
iron. The total amount of action of the same oil (with the
exception of paraffine and probably other mineral oils) is greater
on copper than on iron.
98. Impurities in Mineral Oils consist, usually, of the
gritty and earthy substances which rise in the well with the
oil, and of the " still-bottom" impurities which are produced in
the process of refining. The presence of the latter in other
oils is the best possible evidence of the admixture of the min-
eral oils. They may be detected by dropping a little of the
suspected oil on white blotting-paper, which absorbs the oil,
leaving the impurities visible as black specks on its surface.
The abnormally low temperature at which the oil vaporizes in
contact with these particles is also a means of detecting their
presence. The presence of mineral oils in other oils may
sometimes be readily detected by holding a bottle of the oil to
be examined up to the light, and shaking it well, when the
appearance of fluorescence in the bubbles of air so formed is
an unmistakable sign of the presence of petroleum.
The following method of estimating the proportions of
mineral and other oils in the common mixtures is given by
Mr. C. C. Hall,* as based on a method suggested by Sir Wil-
liam Thomson and Mr. A. H. Allen.
* Trans. Am. Inst. Mining Engineers, 1882.
INSPECTION AND TEST OF LUBRICANTS. 183
Four to five grains of the oil under examination are weighed
out into a porcelain capsule of 75 c.c. capacity. Thirty c.c. of
a ten-per-cent solution of potassium-hydrate are added, and
the capsule, covered with a watch-glass, is placed in a water-
bath heated to about 93 C. The mixture of oil and alkali
should be stirred frequently, and after three quarters of an
hour it is boiled with stirring, to secure complete saponifica-
tion of all vegetable or animal oil. After boiling some time,
a thick scum of soap forms on the surface ; a little bicarbonate
of soda is then added to convert the excess of caustic alkali
into carbonate. When the contents of the capsule have be-
come pasty, an equal bulk of fine clean sand is stirred in,
which makes the soap granular, and facilitates the removal of
the last traces of alcohol. The capsule is heated for two hours
more on the water-bath. After cooling, the contents are trans-
ferred to a short-necked funnel, having a thin plug of asbes-
tos, and washed with petroleum-ether, or other light petro-
leum-spirit. The ether dissolves out the mineral oil from the
soap, and is collected in a quarter-litre flask having a short
neck. Care must be taken to effect a complete removal of the
oil. This can be tested by letting a drop of the ether, as it
comes through, fall on a piece of tissue-paper. If no greasy
stain is left after the ether evaporates, the solution may be
considered complete.
Most of the ether is removed from the oil by distillation,
and can be saved. The heat of the water-bath is sufficient to
boil it, and the fumes may be condensed by passing them into
a condenser. The oil is now transferred to a weighed 5O-c.c.
flask, which has a hole blown in its side; and dry, warm air is
forced into the flask through its neck in order to remove the
last traces of the ether. The flask should not be heated above
the point where it can be borne in the hand : if this precaution
is heeded, there is no danger that any of the oil will be volati-
lized. The passage of the air should be continued until the
flask and oil are constant in weight.
Sperm-oil cannot be separated from mineral oil by this
method, owing to the impossibility of completely saponify-
ing it.
1 84
FRICTION AND LOST WORK.
To determine the proportion of earthy matter in the
gummy masses sometimes found in steam-engines in which
organic oils and steam carrying dirty water from the boilers
have come in contact:
Weigh out any convenient amount of the deposit; wash
well with benzine until it ceases losing weight and all oily
matter is removed ; dry, and weigh again.
The proportion of mineral matter usually ranges from 85
to 95 per cent.
99. The Density of Oil is the-first of its physical charac-
teristics noted by the inspector in the attempt to determine
its character. It is, perhaps, the
simplest and easiest method of iden-
tifying a standard oil, although by
no means a certain one. This may
be done by carefully weighing an
exactly measured volume of the lu-
bricant, and comparing its weight
with the standard volume of a stand-
ard substance, or by the use of the
"densimeter," or oleometer. This
little instrument, generally known
as the hydrometer, takes its specific
name from the application for which
it has been designed ; as, for example,
lactometer when used to determine
the density of milk, and alcoholome-
ter when used to measure that of al-
cohol. It consists (Fig. 29) of a glass
or metal cylinder, usually of an inch
(2.4 cm.) or less diameter, and sev-
eral diameters in length, carrying at
the lower end a bulb loaded with
shot, or mercury, or other heavy
substance, and on the upper end a cylindrical stem graduated
in such a manner as may be best suited to the work for which
it is intended. A cylindrical tank or jar, with attached ther-
mometer, is nearly filled with the liquid to be examined.
FIG. 29. OLEOMETER AND
JAR.
INSPECTION AND TEST OF LUBRICANTS. 1 8$
Placing the instrument in the liquid, it floats upright, with
the loaded end downward, and sinks to such a depth that the
figure on the stem reads the density or the specific gravity
(the terms are not precisely synonymous) of the liquid.
The liquid must usually be tested at standard temperature,
say, 60 F. (15 C.), as its density is considerably affected
by heat or cold. The hydrometer has a thermometer attached
to the lower end. This is intended to assist in making cor-
rections for a temperature above or below 60. When
the thermometer indicates a temperature above 60, which is
shown by the figure on the right side, the corresponding num-
ber opposite must be added to the indications on the scale
above. If the thermometer stands below 60, the correspond-
ing number opposite must be deducted.
100. Specific Gravities and Baume's Scale, often used in
this work, are not proportional, the latter scale being conven-
tional. The specific gravity of a substance is proportional to
its density, and is the ratio of the weight of a given volume of
the substance to that of an equal volume of water, both being
usually taken at the temperature of maximum density of the
latter. For liquids lighter than water, , T. 7 = specific
130+ Baum
gravity, and - - 130 = B, the reading of Baume".
XT* S *
As illustrating the use of the instrument, assume it to be
used for obtaining the gravity of an oil sperm, for example :
finding it to be 0.8750, or 30 Baum, it would be at once
concluded to be impure ; because sperm should give about
0.8810 or 0.8815, corresponding to 29 B. Oils often differ
considerably in density, although nominally the same.
The following table gives the specific gravities and Baume"s
"degrees" for liquids heavier than water, as obtained by
various authorities.* It is evident that the determination
of the specific gravity, or the use of a carefully standardized
Baume" scale, only can give satisfactory figures.
* Chandler and Wiechmann.
1 86
FRICTION AND LOST WORK.
BAUME'S SCALE AND SPECIFIC GRAVITIES.
"0 .
c/i ^.
ro
c'oo
I
1
*?
c *
s *
^ - .
2-27
.261
I.26l
-1632
1.2624
263
1.200
1.263
.2608
.262
1.2605
.2612
1.256
1.245
! . .
2044
2 75
1-275
2743
1.2735
.274
1 -273
1.274
2719
269
1.27,6
2724
1.267
1.256
29 ''2
^86
1.286
2857
1.2849
7.285
1.284
1.285
.2831
.285
1.2828
.2838
1.278
i 267
-. ,
.298
1.298
2973
1.2964
.296
1.296
1.297
.2946
.293
1-2943
2954
1.289
i 277
34
32^7
309
1.309
.3091
1.3081
3 08
T -37
1.308
.3063
309
i 359
.3072
i 300
1.988
.>
-321
I. 3 2I
32"
1.3201
.320
* 3>5
1.320
.3181
317
i 3177
.3190
1.312
i 299
3'
3401
334
1-334
-3333
1 3323
332
1.329
.3302
334
1.3298
33"
1-324
1.310
3 7
3592
346
1.346
3458
1-3447
345
1.339
1-345
3425
342
1.3421
3434
J-337
i 321
38
3725
359
'359
3585
r -3574
358
r -359
1-357
3551
359
I-3546
3559
1-349
J -333
3)
3 86l
372
1-372
37H
371
J -37 2
I -37
3679
368
1.3674
3686
1.361
x -345
40
3999
384
1.384
.3846
I-3834
384
I -375
1-383
.3809
1.3804
3815
1-375
1-357
41 .
414!
398
1.398
3981
1.3968
397
1-399
J -397
3942
395
1 -3937
3947
1388
1.369
42 .
4285
412
1.412
4118
1.4104
.410
1-413
1.410
4077
413
1.4072
4082
1.401
1-381
4 ;
4433
426
1.426
4267
1.4244
424
1.427
1.424
4215
422
i 4210
4219
1.414
L395
44
4583
44
1.440
4400
1.4386
.438
i 441
1.438
4356
441
1-4350
4359
1.428
1.407
45
4735
454
i 454
4545
M53
453
1-455
M53
.4500
451
1-4493
45*
1.442
1.420
46
4893
470
1.470
4694
i 4678
.468
1.466
1.468
.4646
470
i 4640
4645
1.456
47-
5053
485
1.485
48 15
1.4829
483
1.482
1-483
4795
480
1.4789
4792
i 470
''448
48.
5217
5 01
1.501
5000
1.4983
.498
1.500
1.498
4949
500
1.4941
4942
1.485
1.462
49-
5384
516
1.516
5158
1.5140
514
i 515
5104
I -597
5096
1.500
1.476
5)..
5555
532
1-532
5319
1.5301
530
1 S3 2
i-53o
5263
531
I-5255
5253
1-515
1.490
51
573
549
5484
I-5465
.546
1 55
1-540
5425
541
1.5417
5413
1.531
1-505
52..
5909
566
1.566
1.5632
563
1.566
1-563
5591
562
I-S583
5576
1.546
1.520
53-
6092
583
1583
5824
1.5802
.580
1.586
1.580
5760
*73
1-5752
5742
1.562
1-535
54
6279
60 1
1.601
6000
I-5978
598
i 603
1-597
5934
594
I -59 2 5
59 12
1578
'55 1
6471
618
1.618
6179
1.6157
616
1.618
1.6.5
6m
616
1.6101
6086
1.596
1-567
16. .
6667
638
I - 6 37
6363
1.6340
634
1.639
1-634
6292
627
1.6282
2.6264
1.615
1-583
57-
68(8
1.659
1.656
1.6527
653
..660
1.652
6477
650
1.6467
1.6446
1.634
i. 600
NOTE. Where the modulus was not given, it was calculated by the formula n
py.
p-
in which
n = modulus, P = specific gravity, d Baume H*orree (). 66 was taken for d whenever the correspond-
ing specific gravity appeared.
INSPECTION AND TEST OF LUBRICANTS.
IS 7
The next table gives a similar comparison for liquids lighter
than water with, also, the pounds weight per gallon. In metric
measure the specific gravity also measures the weight of the
litre in kilogrammes.
SPECIFIC GRAVITIES AND DENSITIES, PER BAUME.
DENSITY.
Lbs. in one
Gallon.
DENSITY.
Lbs. in one
Gallon.
B.
S. G.
B.
S. G.
IO
I. 0000
8-33
44
. 8045 6 . 70
II
.9929
8.27
45
.8000 6.65
12
.9859
8.21
46
-7954
6.63
13
.9790
8.16
47
.7909
6.59
14
.9722
8.10
48
.7865
6.55
15
.9655
8.00
49
.7821
6.52
16
.9589
7-99
50
7777
6.48
17
.9523
7-93
5i
7734
6-45
18
9459
7.88
52
.7692
6.41
19
9395
7.83
53
.7650
6-37
20
9333
7.78
54
.7608
6.34
21
.9271
7.72
55
.7567
6.31
22
.9210
7.67
56
.7526
6.27
23
.9150
7.62
57
.7486
6.24
24
.9090
7-57
58
.7446
6.21
25
.9032
7-53
59
.7407
6.18
26
.8974
7.48
60
.7368
6-15
2?
.8917
7-43
61
.7329
6.12
28
.8860
7-38
62
.7290
6.09
29
.8805
7-34
63
.7253
6.05
30
.8750
7.29
64
.7216
6.02
31
.8695
7-24
65
.7179
5-99
S 2
.8641
7.20
66
.7142
5-95
33
.8588
7-15
67
.7106
5.92
34
.8536
7.11
68
.7070
5-89
35
.8484
7.07
69
.7035
5-86
36
.8433
7-03
70
.7000
5.83
37
8383
6.98
75
.6829 5.70
38
8333
6.94
80
.6666 5.55
39
.8284
6.90
85
.6511
5-42
40
8235
6.86
90
.6363
5-30
4i
.8187
6.82
95
.6222
5-i8
42
.8139
6.78
IOO
.6087
5-01
43
.8092
6-74
101. Densities of Commercial Oils are often determined
by the more accurate method of determining specific gravity
by weighing on the chemist's balance. A standard tempera-
ture is usually adopted, and all results reduced to stand-
ard by first determining the coefficient of expansion, which
for pure olive-oil has been determined by Mr. C. M. Still-
188 FRICTION AND LOST WORK.
well to be 0.00063 for i Centigrade, or 0.00035 P er degree
Fahrenheit.
Mr. Stillwell's determinations are given in the following
table :
SPECIFIC GRAVITY OF ANIMAL AND VEGETABLE OILS.
15 C.
COEFF. OF EXP. = .00063 FOR 1 C. 59 F.
= .00035 FOR 1 F.
Sperm, bleached, winter 8813
" natural, winter 8815
Elaine 901 1
Red, saponified 9016
Palm 9046
Tallow 9137
Neat's-foot 9142
Rape-seed, white, winter 9144
Olive, light greenish yellow 9144
Olive, dark green 9145
Pea-nut 9154
Olive, virgin, very light yellow. 9163
Rape-seed, dark yellow 9168
Olive, virgin, dark clear yellow 9169
Lard, winter 9175
S .a. elephant , 9 r 99
Tanners' (cod) 9205
Cotton-seed, raw 9224
Cotton-seed, refined, yellow 9230
Salad (cotton-seed) 9231
Labrador (cod) 9237
Poppy 9245
Seal, natural , 9246
Cocoa-nut 9250
Whale, natural, winter 9254
Whale, bleached, winter 9258
Cod-liver, pure 9270
Seal, racked 9286
Cotton-seed, white, winter 9288
Straits (cod) 9290
Menhaden, dark 9292
Linseed, raw 9299
Bank (cod) 9320
Menhaden, light 9325
Porgy 9332
Linseed, boiled 9411
Castor, pure cold-pressed 9667
Rosin, third run 9887
INSPECTION AND TEST OF LUBRICANTS. 189
The mineral oils are usually lighter than those of animal or
vegetable origin.
The following are the densities of some of the compounds
found in petroleums:
MINERAL OILS, 60 F., 15 C.
S. G. B.
Rhigoline 6220 95
Benzine 6510 85
Naphtha .7000 70
7500 57
Illuminating Oil 8000 45
Lubricating Oil (heaviest) 8860 26
Paraffine Wax 8900 27
The " sperm "-oils of the market vary considerably in den-
sity, partly in consequence of natural differences due to differ-
ences in age, size, health, and condition of the sperm-whale
which may have supplied all or part of the oil, and partly be-
cause of variations in the character and extent of the adultera-
tion. Professor Ordway found " spindle-oils" to vary in den-
sity from 0.840 to 0.92, averaging 0.880. Ten so-called sperm-
oils varied from 0.880 to 0.896, averaging 0.884. Oils from
newly arrived cargoes ranged from 0.877 to 0.888. Lard-oils
average 0.917, ranging from 0.914 to 0.920. Neat's-foot oil
gives an average of 0912, ranging from 0.910 to 0.920 for a
sample known to be pure. The addition of refined, odorless,
heavy mineral oils to other lubricants is a usual cause of in-
crease of density ; this is particularly the case with lard-oil.
The common method of making these determinations is by
the use of the " looo-grain bottle," or other such apparatus.
In using the various areometers as oleometers, large jars and
densimeters having slender, finely graduated stems should be
employed, their scales reading to o.ooi. This can be done by
constructing the instrument as an oleometer purely, thus being
able to distribute a small range of density over an extended
scale. Special oleometers are sometimes made for the mineral
oils, and others for the organic oils.
102. The Viscosity of Oil is generally closely related to
its density, but is not proportional to specific gravity, and is
190
FRICTION AND LOST WOKK.
occasionally found to decrease with increase of density. The
relative viscosity of oils may be determined with some degree
of accuracy by simply filling a pipette with the oils to be com-
pared, one after another, and
permitting them to flow out
through a small opening, noting
the time required to discharge
equal quantities. A very com-
plete apparatus for this pur-
pose is that exhibited in Fig. 30,
a form adopted by Mr. J. V.
Wilson.
In the figure, A is a glass
tube about I in. diameter, grad-
uated from I to 100, to contain
about 100 cubic centimetres of
oil. BB is a glass jacket, about
3 in. diameter, filled with water
as shown; C a thermometer, in-
dicating temperature of water
in jacket ; D a small brass cock
for withdrawing water from
jacket ; E a glass flask for generating steam to heat water in
jacket ; F a glass pipe connecting the steam flask E with jacket
B, delivering at bottom of jacket ; G is a small cock for per-
mitting escape of steam in order to regulate quantity sent into
jacket ; H a spirit-lamp on a stand ; J a glass beaker to contain
oil, and KK cast-iron stand, with adjustable arms, for carrying
the apparatus.
The following table gives the time required, by each of seve-
ral oils, to flow through the orifice of the above-described ap-
paratus, and the temperature observed in the same oils when
used on a journal 3 in. (7.2 cm.) diameter, making 1500 revolu-
tions per minute, the average being noted for an hour and a
half. It is seen that, as a rule, the more viscous the oil the
more heat developed by friction. The stearine found in tal-
low-oil may cause the apparent discrepancy noted there.
FIG. 30. VISCOSITY OF OILS.
INSPECTION AND TEST OF LUBRICANTS.
VISCOSITY OF OILS.
NAME OF MATERIAL.
S. G. at
60 F.,
15 C.'
RATE OF FLOW.
Temperature
Developed
by Test.
6oF.,
15" C.
120 F.,
49 C.
180 F..
82 C.
Water
1000
960
990
FAHR.
CEN.
Castor Oil
132
41
I 5 8
155
70
63
Rosin Oil
Solid
143
112
108
96
92
47
45
30
41
37
40
41
38
37
30
26
25
29
30
28
28
25
141
61
Rape Oil
916
916
915
880
905
875
I 4 8
146
143
133
121
117
64
63
62
56
49
47
Lard Oil
Olive Oil
Sperm Oil .
Mineral Oil
M
It is sometimes customary to make the viscosity of oils a
standard test of quality. In such cases it is usual to compare
the oils so tested with some well-known oil, as rapeseed, as a
standard of value. In these cases the size of the containing
vessel, of the nozzle and its orifice, the head producing flow,
the material of which they are made, the temperature, and
other conditions should be carefully specified and made as
nearly constant as possible. The specific gravity of the oil
should be ascertained and stated.
It has been proposed to adopt a standard " viscosimeter" *
of dimensions as follows:
A glass cylinder, 22 in. (55.9 cm.) long, ij in. (3.18 cm.)
diameter, has a brass lower head in. (0.318 cm.) thick. An
orifice is bored in the centre -fa in. (0.794 cm.) in diameter,
with bevelled edges chamfered back in. (1.27 cm.), thus pro-
ducing a sharp-edged orifice. A line marking the i8-in. (45.72
cm.) level is cut with several finer lines above and below, -J in.
(0.318 cm.) apart, ranging from 16 to 21 in. (40.64 to 53.34 cm.)
above the orifice. The standard temperature is usually 60 F.
(15.5 C.). A total flow of as nearly 100 c.c. (6.103 cu - m O * s
secured by adjusting the supply so that the head shall be as
nearly as possible equal to 18 in. (45.72 cm.) of water, deter-
* Chemical News, 1884. W. P. Mason.
I9 2 FRICTION AND LOST WORK.
mining this head by calculation from the specific gravity of
the oil.
Note the time required to discharge the 100 c.c. (6.103 cu -
in.), and divide this time by that required where water under
a head of 18 in. (45.72 cm.) is used. This ratio is the measure
of the viscosity.
Large consumers of oil sometimes purchase on the basis of
this kind of test solely. It is regarded as quite as satisfactory
and reliable as any single physical or chemical test known, and
as second only to the best testing-machine methods.
The less the viscosity, consistently with the use of the oil
under the maximum pressures to be anticipated, the less is,
usually, the friction. The best lubricant, as a rule, is that hav-
ing least viscosity combined with greatest adhesiveness. Vege-
table oils are more viscous than animal, and animal more so
than- mineral oils. The fluidity of an oil is thus to a large ex-
tent a measure of its value.
The close relation between the viscosity and the friction-
reducing power of the oils is well shown in Fig. 31, which
graphically exhibits this relation as determined by Mr. C. N.
Waite.* The curves show the relation between the viscosity
and lubricating power of lard and of light paraffine oil ; the full
lines represent the readings on the machine, at different tem-
peratures, multiplied by a constant, and the dotted lines the
viscosity of the oil. The curves are approximately correct.
The true curves are probably smooth, and their form mathe-
matically determinable.
The relation of viscosities of oils at ordinary temperatures
is not a measure of their relative standing in this respect at
higher temperatures, as in steam-cylinders. Oils of great vis-
cosity at low temperatures are often very limpid when heated.
Tallow and castor oils are more viscous than sperm when cool,
but they become very much more fluid when heated, as in
steam-cylinders.
103. Gumming, or Drying, is a method of alteration of
oils usually caused, as already stated, by the absorption of oxy-
* Proceedings N. E. Cotton Manufacturers' Association, No. 28, 1880.
INSPECTION AND TEST OF LUBRICANTS. 193
gen and the gradual conversion of the oil into resin. It goes
on rapidly with the " drying"-oils, slowly with the fixed ani-
mal and vegetable oils, and is not observed in any important
2W
is;
220
2:0
200
180
170
iec
150
140
130
\
\\
\
\
\
\
. \
\\
\
\
\\
\
\
^
\
^
N
L
\
no
100
90
70
eo
SO
40
r
\\
\
x
\
^
>c
\
\.
\
^
fe
\
%
^X
X
^
^
^
X
^
!>s.
^
- *
=^.
)0 J S 80 J 90 100 11
0= 12(
FIG. 31. VISCOSITY AND LUBRICATION.
degree in the mineral oils. This gradual increase of viscosity
and tendency to final conversion into the solid form is one of
the phenomena noted by the inspector in his examination of
IQ4 FRICTION AND LOST WORK.
lubricants. The methods of determination of the character
of the lubricant in this respect, as practised by various observ-
ers, differ greatly. The most satisfactory method is probably
that in which the lubricant-testing machine is employed : this
method, as conducted by the Author, is simply to test the oil
as received ; then to expose the journal, still wet with oil, to
the action of the air, but keeping it protected from dust, one
day or more, according to the kind of oil, and then to again
test its friction-reducing power. This process will be fully
described later (Arts. 132, 136).
104. Nasmyth's Apparatus for observing the viscosity and
gumming of oils is very simple. The observer places a drop
at the top of an inclined plane, and notes the time required
for it to run down the plane. Of oils which do not gum, the
least viscous reach the bottom first. Drying and gumming
oils are retarded in proportion to the rate of drying or of gum-
ming. Nasmyth used a plate of iron 4 inches wide by 6 feet
long, on the upper surface of which six equal-sized grooves are
planed. This plate is placed in an inclined position say, I
inch in 6 feet.
The mode of testing is as follows : Assume that six varie-
ties of oil are to be tested, to determine which of them will
for the longest time retain its fluidity when in contact with
iron and exposed to the action of air; pour out simultaneously,
at the upper end of each inclined groove, an equal quantity of
each of the oils under examination. This is very conveniently
done by the use of a row of small brass tubes. The six oils
then make a fair and even start on the race down-hill : some
are ahead the first day, and others are still ahead the second
and third day ; but on the fourth or fifth day the bad oils be-
gin to fall behind by gradual coagulation, while the good oil
holds on its course : at the end of eight or ten days there is no
doubt left as to which is the best. Linseed-oil, which makes
capital progress the first day, is, in the case given, set fast af-
ter having travelled 1 8 inches, while second-quality sperm over-
reaches first-quality sperm by 14 inches in nine days, having
traversed in that time 5 feet 8 inches. The following table
shows the state of the oils after a nine days' run :
INSPECTION AND TEST OF LUBRICANTS. 195
VISCOSITY OF OILS.*
DESCRIPTION OF OIL.
First Second
Day. Day.
Third Fourth
Day. Day.
Fifth
Day.
Sixth
Day.
Sev'th
Day.
Eighth Ninth
Day. | Day.
Best Sperm Oil
Common Sperm Oil.
Gallipoli Oil
Lard 0.1
Rape Oil
Linseed Oil
ft. m.
ft. in. ft. in.
4 5% 4 6
4 6% 4 "
o % o *
j!2i
ft. in.
'>%
iM
ft. in.
4 6
5 4
ft. in.
4 6}i
5 6*j
i 9
ft. in.
stat.
5 T%
i
= 0.7854 P~ r nearly; . . . (2)
the work of friction is
and the heat produced becomes
H -j = o.ooi Pv-, nearly. . (4)
120. The Friction of Fluids and of Semi-Fluids, such
as gases, liquids, resins, and in some cases earth, follow laws
varying greatly from those governing the friction of solids, and
these laws have been already stated in Chapter II. The
friction of liquids and of gases has been experimentally studied
by many distinguished investigators. These researches con-
firm the principles embodied in the mathematical analysis of
the case. The friction of any fluid is found to be independent
of the pressure, as first shown by Coulomb, who measured the
friction of a rotating disk submerged in water, applying vary-
226 FRICTION AND LOST WORK.
ing pressures to the surface of the mass, and by many later
observers who find the frictional losses of head of fluids tra-
versing pipes, under different pressures, to be the same at the
same velocities.
The law that the resistance is, with velocity constant,
directly proportional to the area of surface is almost axio-
matic; it is fully confirmed by experiment. It is found, how-
ever, that where a body moves in a large mass of fluid, the
friction of the leading portions of the surface of the moving
body causes some motion of the adjacent fluid in its own
direction, thus reducing the relative velocity, the velocity of
rubbing, from forward aft, and correspondingly reducing the
resistance of large bodies, as those of long ships.
Low velocities are found to give variations from the law
assumed in the theory, while high velocities more closely ac-
cord with that law. This variation is only important for
velocities considerably less than one foot (0.31 in.) per second.
The smoothness or roughness of surfaces exposed to fluid-
friction has been found to considerably affect this resistance.
For all velocities usually met with in engineering, the ex-
pression
R=fAV*=f'DA^,U = fA V = f'-DA^
given in Chapter II., may be adopted, where R and 7 measure
the resistance and the work of friction, A is the area of rub-
bing surface, D the density, V velocity of relative flow.
121. The Flow of Gases is subject to modification by
changes consequent upon variation of temperature due to fric-
tion, and problems relating to such flow are therefore compli-
cated with calculations of the effect of heat ; but where no heat
is- lost by conduction there is no loss of head by friction, ex-
cept such slight losses as are due to the imperfectly fluid
character of known gases.
The loss of head may be taken as the same as for liquids,
and the method of flow is similar. Unwin obtains for air
7=0.005(1 + 3.6^)
EXPERIMENTS ON FRICTION TESTING MACHINES. 22/
when d is expressed in inches, and the velocity is 400 feet per
second or more, the data being obtained from experiments by
M. Arson. Experiments at the St. Gothard tunnel give, for
probably rougher surfaces,
/= 0.0028
122. The Friction of Liquids, as affecting the work of
the engineer, is always a cause of lost work by resisting the
relative motion of the liquid and some solid which is driven
through it, as when a ship moves across the ocean, or which
constitutes a channel along which the liquid is impelled.
Fluid-friction occurring between the touching surfaces of a
solid and a liquid is proportional, according to accepted authori-
ties, to the area of surface of contact and to the density of the
fluid, and is found, as already stated, to be nearly as the square
of the velocity of their relative motion ; i.e.,
in which F is the measure of the resistance when f is the co-
efficient of fluid-friction, D = the density of the fluid, A = the
area of surface of contact, V = the velocity of flow, and g =
the measure of gravity = 32.2 feet per second, while h is the
F a
head due the velocity, and equal to
o
For iron pipes, according to Eytelwein,
, 0.00144
/ =0.0056 -f
or, according to Weisbach,
7=0.0036 +
and for average value, f = 0.0064.
228 FRICTION AND LOST WORK.
The mean velocity of a stream of water, according to Prony,
s
10.25 + F
where v is the mean and F the maximum velocity of the
stream as measured at the middle thread of its surface ; the
difference between v and Fis due to friction.
In flowing streams, according to Eytelwein,
or, according to Weisbach,
, 0.00023
/= 0.00741 H
and an average value is f= 0.0076,* The value is somewhat
variable.
The method of variation of this friction depends both on
the nature of the fluid and on the character of the surrounding
solid surfaces. Froude found in salt water, and with surfaces
of small area coated with tallow or with shellac varnish, that
the resistance to the motion of ships, which in well-formed
vessels is principally frictional, varies as F 1 ' 83 ; surfaces coated
with tinfoil gave F oc F*-*. With surfaces of considerable
area, the character of surface seemed comparatively unimpor-
tant.
The total loss of head, in any case of friction of water in
orifices or pipes, may be taken as a loss of head equal to
O*
~
* Rankine, Applied Mechanics, 638.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 2 29
in which
F = 0.054 for an orifice in a thin plate ;
F= 0.505 for an entrance into a pipe from a reservoir;
F= 0.505 +0.3 cos i + 0.23 cos* / for a mouthpiece mak-
ing the angle i with the side of the reservoir.
Q is the quantity of water flowing and A the area of sec-
tion of the channel.
Where the ajutage has the form of the contracted vein, its
cross-section at a distance radius from the side of the reservoir
is of a diameter equal to 0.7854 the diameter at the side ; in
this case the value of F becomes practically zero.
Ib
In pipes and conduits, ^ = /-~J>
in which expression /"has the value already assigned ; /,', and
A are, respectively, the length, breadth, and area of cross-sec-
tion of the stream.
Substituting for -r, its value, the reciprocal of the hydraulic
mean depth, = -^ , we may write F = / .
Friction is somewhat increased by bends and " knees" in
pipes; and from Weisbach's experiments are deduced, for
smooth bends,
in which i is the angle through which the pipe is bent, r is the
radius of the curve, and d is the diameter of the pipe ; for
knees, i.e., rectangular or abrupt changes of direction, we find
F= 0.95 sin* i + 2 sin 4 ~ .
The values of f and f in the expressions for fluid-friction
vary with circumstances. The values obtained by Froude and
230 FRICTION AND LOST WORK.
other experimenters accord well with the following, as given
for f and f in the simpler of the expressions given at the
opening of Article 120:
f. /'
Painted iron (Unwin) ........................ 0.00489 0.00473
Smooth, painted wood (Beaufoy) ............. 0.00350 0.00339
Iron ships (Rankine) ......................... 0.00362 0.00351
Varnished surface (Froude) ................... 0.00258 o 00250
Fine sand (Froude) .......................... 0.00418 0.00405
Coarse" " .......................... 0.00503 0.00488
The resistance of ships is often expressed by the formula
of Rankine,
C '
in which 5 is the area of "augmented surface" in square feet,
V the speed in knots per hour, and C a coefficient, which
ranges from 20,000 to 25,000 in full to fine vessels. The aug-
mented surface is measured by the product of length, mean
wetted girth, and a coefficient of augmentation obtained by
taking the sum of unity, four times the mean of the squares
of the sines of greatest obliquity of water-lines, and the mean
of their fourth powers.
Sudden enlargements and sharp bends often cause serious
losses of head and of pressure.
Notches discharge less than the quantity which should pass
if no such loss as is above described takes place. For a rectan-
gular notch, the volume discharged is
Q = \cbd
= 5
in which c is a coefficient usually not far from 0.6, b and d are
the breadth of notch and the depth of stream issuing through
it. If W is the width of the channel,
'=Q-57 + o-i- nearly.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 23!
123. The Friction of Earth has been the subject of many
experiments. The alteration in form and location of any mass
of earth by the action of gravity, as has been seen ( 41), is re-
sisted by both friction and adhesion. Where the latter occurs
to any considerable extent, as in clayey soils, a bank may even
overhang its base at a measurable angle. Where adhesion is
inappreciable, as in dry, sandy soil, the surface assumes a
uniform slope at an angle with the horizontal which is the
"Angle of Repose," the tangent of which measures the "Coef-
ficient of Friction." The latter is also the limit of declivity
assumed by any soil or earth in which, as is always liable to be
the case, adhesion is destroyed by moisture or other cause.
In calculations relating to the sustaining power of earth under
foundations or the pressure upon a retaining-wall, the angle
of repose, as obtained by direct experiment, must be known
to insure safety.
The angle of repose is in some cases liable to be reduced
to a very small value by the presence of water, as in flooded
quicksand or in saturated clayey earth. The least probable
value should in such cases be assumed. In some cases the soil
should be considered as a perfectly fluid mass of maximum
density, and its pressure calculated as if it were a liquid.
Calling cp the angle of repose, experiment gives the follow-
ing values of fWdR sec ot, on conical journals. . . . (5)
*W. R. Browne, Railroad Gazette, August 16, 1878.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 239
And
H.P. = 0.00003/^2;, on flat surfaces ; (6)
= o.oooooZfWdR, on cylindrical journals ; . . (7)
= o.*)fWdR, on cylindrical pivots ; . . . (8)
= o.oooooSflVdR sec a, on coned journals ; . . (9)
= o.ooooo^fWdR cosec a, on coned pivots. . (10)
Mr. D. K. Clark * takes the values of/, from various sources,
as averaging /= 0.07 and/ =0.043, for cases of ordinary and
of free lubrication respectively, and thus gets
[/= 0.0182 WdR, for ordinary oiling; . . . . (u)
= o.oi 12 WdR, for continuous oiling; . . . (12)
H.P. = 0.0000005 WdR, for ordinary oiling; . . . (13)
= 0.00000033 WdR, for continuous oiling; . (14)
the free supply giving a gain of 40 per cent. In these equa-
tions, W is the load in pounds, 5 the space in feet, R the revolu-
tions per minute, d the diameter in inches, a the angle of the
cone.
127. The Size of Journals has been seen (Chap. II., Art.
29) to be determined by the magnitude of the friction, only as
to its length. The diameter is made sufficient to insure safety
against springing and permanent distortion, and the length is
determined by the limit of intensity of pressure allowable;
while this limit is fixed, as will be seen more clearly hereafter,
by the speed of rubbing and the temperature of the surfaces
in contact. The usual maximum pressures, the pressure at
which the limit of safety against abrasion is approached, has
been given as 500 or 600 Ibs. per square inch (35 to 42 kgs.
per sq. cm.) for iron crank-pin journals, and as about double
these figures for steel. It is, however, variable with change of
speed, etc. The maximum pressure on timber, as on the
launching-ways of vessels, is below one tenth that for iron. All
bearing-surfaces should have sufficient area at least to reduce
the intensity of pressure below these figures, and should be
increased beyond this extent in the manner given below, with
* Manual, p. 763.
240 FRICTION AND LOST WORK.
increase of speed, or for journals subjected to uninterrupted
pressure.
The two surfaces usually differ the one being hard enough
to bear the maximum pressure without change of form, and
the other being less hard, in order that it may not abrade the
first. With such an arrangement, the surfaces, if properly
cared for, take a fine smooth, mirror-like polish, and give a
minimum frictional resistance. Cast-iron surfaces are usually
less satisfactory than good wrought-iron, although where the
areas can be made large, cast-iron bearings work very satisfac-
torily, and homogeneous and moderately hard steel is vastly
better for journals than iron. A pressure of 800 Ibs. to the
square inch (56 kgs. per sq. cm.) can rarely be attained on
wrought-iron at even low speeds, while 1200 Ibs. (85 kgs.
per sq. cm.) is not infrequently adopted on the steel crank-
pins of steamboat engines; but double this pressure has been
reached on locomotives, at the instant of taking steam. Seven
to nine thousand pounds pressure per inch is reached on the
slow-working and rarely moved pivots of swing-bridges. In
practice with heavy machinery, higher pressure than 600 and
IOOO Ibs. per inch (42 to 70 kgs. per sq. cm.) on iron and on
steel are rarely adopted, and in general practice we make the
pressure less as the speed is greater, since the amount of heat
developed is directly a measure of the amount of work done in
overcoming friction, and is proportional to the speed as well as
to the pressure. Reciprocating motion in journals compels the
adoption of greater length than continuous revolution. Slowly
moving journals are often but one diameter in length ; fast-
working journals are sometimes 6 and 8 diameters long. Under
steady pressure, this length must be greater than under inter-
mitted loads.
By watching the behavior of the journals of the engines of
naval steamers in 1862, the author determined the following
formula for the size of journals for such engines and for sta-
tionary steam-engines:*
/= 6_^L'
* Materials of Eng., vol. i.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 241
in which / is the length of the journal in inches, P the average
load in pounds, and V the velocity of rubbing in feet per
minute ; d is the diameter in inches. Rankine published, in
1865, the following as applicable to locomotive practice:
These are intended for iron journals; those of steel may
sometimes work well if of one half the length given by the
formulas.
The length being known, the mean pressure per square inch
admissible is within the limits above given,
60,000
/ = (Thurston).
Where journals are exposed to dust, as in locomotives, or to
unintermitted pressure, it is advisable to make them of greater
length than where they are fully protected. This difference is
observed in the two formulas just given. The best makers of
mill-shafting make the journals about four diameters long.
The expressions above given can only be taken as correct
for such cases as are familiar to the engineer as representing
good current practice. They are subject to great variation,
with variation of condition and kind of surface, temperature,
nature of the lubricant, etc., etc.
For rapidly revolving pivots, lower pressures and corre-
spondingly increased areas of surface must be usually adopted.
Fairbairn would restrict pressures, in this case, to less than
240 Ibs. per square inch (18 kgs. per sq. cm.), which he thinks
a critical pressure. Trautwine takes pressures 40 per cent.
lower for iron "steps," and 25 per cent, higher for steel
both to be used for general mill-work. Railway turntable-
pivots, and those of drawbridges, which turn exceedingly
slowly, sometimes work under pressures approaching the elastic
242
FRICTION AND LOST WORK,
limit of the metal. Chilled iron and hardened steel work well
if properly cared for, under loads of 6000 Ibs. per square inch
(422 kgs. per sq. cm.) when kept well lubricated.
In all these cases ordinary methods of oiling are assumed.
Where the oiling is intermittent, the pressure intermitted, the
speed of rubbing small, and the lubricant fluid, these limits
should never be exceeded ; if, on the other hand, the lubrica-
tion is very free, as with the oil-bath, the pressure intermitted
or reversed, as on crank pins, the speed of rotation of journal
high enough to force the lubricant between the surfaces, and
the latter at the same time of good "body," much higher
limiting pressures may be sometimes attained. A steady,
unintermitted pressure will not permit maximum intensity of
pressure to be maintained.
The experiments at the Brooklyn Navy Yard, made under
the direction of the Bureau of Steam Engineering, and under
these conditions, were reported to indicate the following limits
of pressure for a velocity of rubbing of about 200 feet (60 m.)
per minute, and a temperature of 116 F. (47 C.), the pressure
and speed being unintermitted.
PRESSURE.
OIL. Lbs. per Kgs. per
sq. in. sq. cm.
Summer Sperm Oil 86 6
Winter Sperm Oil 70 5
Winter Lard Oil 62 4.3
Tallow Oil 50 3.5
PRESSURE.
OIL. Lhs. per Kgs. per
sq. in. sq.cm.
Heavy Mineral Oil 73 51
Light Mineral Oil 65 4.5
Paraffine Oil 55 4
Mineral and Fish Oil.. .. 48 3.5
These figures are very much smaller than would be given
by either of the rules above given, which at 200 feet would be
from 200 to 300 Ibs. per square inch (14 to 21 kgs. persq. cm.).
In other words, the apparent factor of safety is here at least
2 or 3 for the best oils. The rules reduced to this basis would
read
15000
/ = jr-> nearly,
for sperm-oil. As previously given, however, they have been
adopted in the design of many steam-engines and other
machines, and have given satisfactory results. The adoption
EXPERIMENTS ON FRICTION TESTING-MACHINES. 243
of the latter will give good results for light machinery, but would
produce journals of impracticable size if used for heavy work.
The pressure at which the film of oil is displaced and the
friction becomes altered from liquid friction to mixed, or
" mediate," friction by contact of the metals, varies greatly
with different oils and at different speeds, and is not exactly
known for any one lubricant. These pressures are perhaps
not far different from those last given. Mr. C. N. Waite sup-
poses this point to be reached with a pressure of about 84 Ibs.
per square inch (6 kgs. per sq. cm. nearly) with neat's-foot oil,
one half this figure with lard, 70 Ibs. (5 kgs.) with sperm, and
deduces the conclusion that a light paraffine-oil is best for low
pressures and a heavy mineral oil for heavy loads. This point
varies, however, very greatly with velocity of rubbing, becom-
ing as a rule greater as the speed increases. It is also, as
already stated, very much greater where the pressure is inter-
mitted, as on crank-pins of steam-engines, and less with vibrat-
ing journals, as on the " beam-centres " of engines having
" working-beams."
128. Machines for Testing Lubricants are used in the
most important of all the tests to be applied to determine the
precise value of a lubricating material, and in that which most
completely and satisfactorily reveals that value, the machine
being specially constructed for the purpose.
In order to determine precisely what oils are adapted to any
special purpose, or to ascertain for what uses any oil is best
fitted, it is necessary to make an examination of the lubricant
while it is working under the specified conditions. That is to
say: The oil should be put upon a journal of the character of
that on which it is proposed to use it, and, subjecting it to the
pressure proposed, running it at the speed that the journal is
expected to attain ; its behavior will then show conclusively
its adaptability to such an application. While running, it is
necessary to measure the friction produced, and to determine
its coefficient, which, as we have seen, is its measure, as well
as to be able to note its durability and the rise in temperature
of the bearing. These qualities being determined and recorded,
all is known of the oil that is needed to determine its l
244 FRICTION AND LOST WORK.
ing power, and its value for the purpose intended. A number
of such machines have been invented, although but two or
three are in use.
One of the oldest is that of McNaught. It consists of two
disks. The upper one is loose ; the lower one is turned by a
pulley on its spindle. The oil is interposed between the disks,
and the friction causes a tendency on the part of the loose
disk to turn with the other. This tendency is resisted by a
pin on its upper side coming in contact with the short arm of
a bell-crank lever, the long horizontal arm of which carries a
weight which can be adjusted to measure the friction.
The oil to be tested is placed between these two disks. As
the lower one turns, the friction between them carries the
upper one with it, but its motion is restrained by a pin, which
comes in contact with another pin, in the end of the arm of
a T-lever. A movable weight slides on the arm, on which is a
scale to note its position. A counterweight is attached to
the opposite end of the lever, so as to afford the means of a
more delicate adjustment. It is evident that the resistance due
to the friction between the two disks may in this way be very
readily measured by the position of the weight.
Napier's machine consists of a wheel, of which the smooth,
wide rim is pressed by a brake-block, which is forced against it
with any desired amount of pressure by the action of weighted
levers. The effort of the wheel to carry the block around is
resisted by another weighted lever, and by it the friction is
measured, as in the later machine of Riehle.
The machine of Messrs. Ingham & Stapfer consists of a shaft
running in two bearings and carrying a third journal between
them. This latter has adjustable bearings, which are set up
to any desired pressure by weighted levers. A thermometer
in the top brass enables the heating of the bearing to be ob-
served. A later modification of this machine is seen in that of
Ashcroft. In this machine the friction cannot be measured;
but the durability of an oil and its effectiveness in keeping a
bearing cool can be observed. A somewhat similar but much
larger machine has been used at the Brooklyn Navy Yard
several years.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 245
The work done on the Ingham & Stapfer machine is some-
times plotted as in the accompanying diagram :
230 - -
-
1 | | i
S&:::::::::::::::::
ram
:|:::
X
-"200
4mmm
a::::::ji:::::: = j:
,. i... T* i ; i i | : ; , ' ;
--
' I ! ! T
^T- lio
|i
i:::|l::::::::::
_
[ ' i j -f
f-H- ?
r-rf 50
10000 15WW 80000 85000 30000 3WOO 40000 45000 60000 53UW WOW WUOO 70000
FIG. 36. OIL TEST. HEAT AND WORK.
The two dotted lines show the behavior of two different
samples of oil under test. The line of large dots shows an
excellent quality of prepared and purified sperm, that, starting
at a temperature of 67 F. (19. 5 C), has with 70,000 revolu-
tions only attained 176 (80 C.) ; while the other, an indifferent
mixed oil, attains 200 (93.3 C.) with only 19,000 revolutions.
By means of such a diagram a permanent record of all tests
can be kept for future guidance.
The value of the lubricant is assumed (improperly) in the
use of this machine to be determinable simply by observing
its durability and its effect upon the thermometer. In making
experiments of this kind, Mr. W. H. Bailey proposes that all
should begin at the same standard temperature say 60 F.
and should terminate at the same point, which he would make
200. He enters the data, as obtained, on a record-sheet thus
arranged :
NAME OP OIL.
Price.
Total Rev.
to 200 F.
Temp, of
Atmosphere.
Rev. per Degree.
In a test thus made to determine the gumming of oils,
Wheeldon obtained the following table : *
* Lecture by Mr. W. H. Bailey, Manchester, G. B.
246
FRICTION AND LOST WORK.
TESTS OF OIL ON BAILEY'S MACHINE.
Resistance to Oxidation. ( Wheeldon.)
Name.
Price.
Rev.
Temperature.
Elevation
of Temp.
Rev. per
Degree.
First day 1
No. i Ox.
5/6
1 3, OO^
From 80 to 200
120
108
Second day 2 . . . .
11,787
" 78 tO 200
122
97
First day 3
Sperm.
Q/O
16,044
From 65 to 200
135
no
Second day 4 . . . .
13,104
" 62 tO 200
138
95
First day 5
Mineral(?)
3/6
II 831
From 65 to 200
135
88
Second day "
1 First trial; new oil. 2 No fresh oil added. s First trial; new oil. 4 No
fresh oil added. 6 First trial. 6 Second trial; after standing 24 hours the bear-
ings were found glued to the test journal, and the machine refused to start.
The last of these trials could not have been made with an
oil of the kind indicated by the name given. Mineral oils do
not gum; this was undoubtedly a mixed oil of poor quality.
The Zeitschrift deutcJicr Ingtnieure, 1871, gives the follow-
ing:
OIL. Price per cwt. Rev. Relative Cost.
Refined Rape seed $1125 69975 100
Mineral 750 41.850 111.4
Impure Rape-seed 9 60 26392 225.9
Lieut. Metcalfe, of the Ordnance Corps, U. S. A., in experi-
ments made at the Frankford Arsenal* in 1873, on axle and
trunnion friction, has adopted .Rankine's method f of noting
the time required by a fly-wheel running loosely on a shaft to
lose a given quantity of energy while stopping under the
opposing efforts of its own inertia and the frictional resistance
of its lubricated bearing on the stationary axle. From this he
deduced the coefficient of friction thus:
The energy thus destroyed is
* Ordnance Notes No. LXXXIV. Washington, July 15, 1878.
f Machinery and Millwork, p. 397.
EXPERIMENTS ON FRICTION TESTING-MACHINES.
(W\
in which M is the mass ^ J of the wheel, k its radius of gyra-
o
tion, and a is the initial angular velocity.
The work of resistance by friction is U = U, and is meas-
ured by
and
.
U = 2Fnrn = - a
2
471 rn y
in which F is the effort of friction resisting motion, r the radius
of the shaft or journal, and n the total number of revolutions
made while stopping. The mean velocity a' is one half the ini-
tial velocity a. Then
=
where / is the time of retardation in seconds.
F F _ Afinn _ r n
^W = M = ''~rf r '' ?'
in which last expression C is a constant to be determined for
any wheel used.
In Metcalfe's experiments the pressure was about 100
Ibs. per square inch (7 kgs. per sq. cm.), and whale-oil gave
/= 0.015 to/= 0.016, sperm-oil, 0.088 ; castor-oil, 0.028 ; axle-
grease, 0.030.
The average revolutions were 53 per minute.
This affords a very convenient method of comparing the
values of lubricants used upon the wheels of vehicles ; the
wheel itself may be used as the storer and restorer of the energy
expended in friction.
129. The Ashcroft and Woodbury, the Wellington, the
Tower, and the Riehle Machines for testing oils are improve-
ments upon the earlier testing-machines. All embody provi-
sions for ascertaining the value of the coefficient of friction.
248 FRICTION AND LOST WORK.
The Ashcroft machine is a modified Ingham & Stapfer
instrument, as seen in Fig. 37.
It is operated in the same manner. The illustration shows
the test-arbor, weighted lever producing pressure, the ther-
mometer indicating changes of temperature, and a dial show-
ing the friction-resistance. The oils tested are compared by
noting the rise of temperature during test as already described,
FIG. 37. ASHCROFT OIL MACHINE.
the maximum allowed being taken usually at a little below the
'jjfcoiling-point of water.
Mr. Woodbury has improved the Nasmyth machine.*
The machine is shown in perspective in Fig. 38.
The lower disk is secured upon the top of an upright shaft,
its top being an annulus, ground to a true plane surface. Upon
this rests the upper disk, which is a hollow ring of hai 1 compo-
sition.
* Trans. Am. Soc. Mech. Engrs., vol. vi., November, 1884.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 249
A partition divides the interior of the hollow ring forming
the upper disk, and water can be introduced through the con-
FIG. 33. THE WOODBURY MACHINE.
necting tubes to control the temperature of the disks or to re-
tain the heat of friction. The sides and top of the upper disk
are surrounded by a case of hard rubber, and the space is filled
with eider-down.
250 FRICTION AND LOST WORK.
Ice-water is used to reduce the temperature of the disks to
nearly the freezing-point of water, and the friction is noted at
each degree of rise in temperature.
A tube of thin copper, closed at the bottom, reaches through
to the bottom of the disk, and a thermometer with its bulb
placed within this tube indicates the temperature of the fric-
tion-surface. A tube leading through the upper disk conducts
the lubricant under trial to a recess in the middle of the lower
disk. The upper end of this tube, being of glass, exhibits the
supply and rate of feeding of the oil. As the friction of a jour-
nal depends quite largely upon the method of lubrication, uni-
formity in the manner of supply is of the utmost importance.
The axes of the upper and lower spindle do not coincide,
but are on parallel lines about one eighth of an inch from each
other. This prevents the surfaces from wearing in rings, be-
cause the same points are not continuously brought in con-
tact 'with each other.
A counter records the number of revolutions made during
any given time.
The dynamometer on the right-hand side of the machine
consists of segments and pinions multiplying the deflection of
a steel bar, and indicating the stress necessary to produce such
deflection by the position of the hand on the dial. When the
machine is in operation the lower disk is revolved, and tends
to carry the upper disk around with it, by a force equal to the
friction due to the lubricant between the disks.
The frictional resistance is thus obtained : The reading on
the dynamometer indicates the force of a couple whose arm is
the length of the lever projecting from the upper disk, and
this couple is opposed by a couple of equal moment, of which
the dimensions of the frictional surface form the data for com-
puting the arm, and the frictional resistance of the lubricant is
the unknown quantity.
The coefficient of friction is deduced from the data of ob-
servation in the following manner: Let
W Weight on disks, Ibs.
r 2 = Outer radius of fractional contact, feet.
r Inner "
EXPERIMENTS ON FRICTION TESTING-MACHINES. 2$ I
r = Radius of any infinitesimal ring or band of the fric-
tional surface, feet.
N = Number of revolutions per minute.
F = Reading on dynamometer, Ibs.
L = Length of lever arm of upper disk, feet.
f = Coefficient of friction.
Suppose that the annular surfaces of the disk be divided
into an infinite number of elementary areas by equidistant
circles and radial lines, then will
Width of elementary band = dr ........ (i)
Angle between two successive
radial lines = dO ........ (2)
Length of arc between two radii = rdd ....... (3)
Elementary area = rdrdO ..... . (4)
Area of annulus = n(r* r^) ..... (5)
W
Pressure per unit of area = 73 57 ..... (6)
^v a ^i )
Wrdrdd
Pressure on elementary area = -r^-t:> .... \j)
fWrdrdO
Friction on elementary area = , a --- ^ ..... (8)
Moment of friction on elementary area
fWfdrdB
(9)
fW /*r a / s ff
Moment of friction on entire disk = -/-^- z I I r*drdd.(io)
n(r*r*}Jri J Q
27tfW
252 FRICTION AND LOST WORK.
2fW(r 3 r 3 )
Substituting the limits = r~^ ~^\' ""'' ( I2 )
"
2
Work of friction per minute - , a ^ r d . . (13)
,-.; 3v a ' i )
The work of the dynamometer = 2nLFN. ..... (14)
The friction equals the resistance ; hence
s-
= aF+W\ (16)
in which the constant coefficient may be easily determined by
each machine.
The work done by this machine will be referred to at some
length in the succeeding chapter.
In the construction of the Riehle machine, which is shown
in Fig. 39, the inventors have introduced a "balanced" weigh-
ing arrangement, and the combination, first used by the Au-
thor, of a device for indicating the coefficient of friction with
those for determining pressure and velocity of rubbing.
The counter-pulleys admit of running the journals at dif-
ferent speeds, and any pressure can be applied up to 2200
Ibs. (1000 kgs.). The thermometer and counter indicate the
heat of the journal during the different stages of the testing,
and the number of revolutions made by the journal. The
coefficient of friction can be accurately determined by observ-
ing the pressure and friction as indicated by the beam, in
connection with size of journal. The beams are graduated
like scale-beams, and balanced. One weighs the pressure pro-
duced by the wheel and screw on the journals, one is used as
a counterbalance, while the third measures the friction pro-
duced when the machine is in motion.
EXPERIMENTS ON FRICTION TESTING-MACHIXES. 253
130. Thurston's Lubricant-Testing Machine. The ma-
chine devised by the Author was, so far as he is aware, the
first in which it was made possible to obtain from indices on
the machine measures of the velocity of rubbing and speed
of revolution, the total pressure and the intensity of pressure
on the journal, the temperature and the friction, and easily to
determine the exact value of the coefficient of friction. The
Author, some time previous to the year 1872, found that the de-
FIG. 39. THE RIEHLE" MACHINE.
termination of the amount of frictional resistance had been sel-
dom attempted, but that the simple measurement of the heating
by means of machines of the Ingham & Stapfer class had been
relied upon alone, and that results obtained were of value only
by comparison. He therefore endeavored to devise a machine
which should not only exhibit the heating of a lubricated jour-
nal, under pressures and speeds variable at will, but one that
should also give at the same time and with accuracy the more
delicate but much more important measure of the friction.
It was desirable that the machine should give not only a
254
FRICTION AND LOST WORK.
measure of the resistance due to friction, but an exact meas-
ure of the relation which that resistance bears to the total
load on the journal ; in other words, it should give, directly
and precisely, the value of the " coefficient of friction/'
The construction of this machine is shown in Figs. 40
and 41, below.
At F is the journal on which the lubricating material is to
be placed for test. This journal is carried on the overhung
extremity of s^aft A, which is sustained by the journals BB ',
on a standard, D, mounted on a base-plate, . The shaft
FIG. 40. THURSTON'S MACHINE.
FIG. 41 THURSTON'S MACHINE.
is driven by a pulley, C, at any desired speed. A counter is
placed at the rear end of this shaft, to indicate the number
of revolutions. The shaft is usually driven at a fixed speed,
corresponding to a velocity of rubbing surfaces approximating
that of the journals on which it is proposed to use the oih
The testing-journal, F, is grasped by bearings of bronze, GG',
and with a pressure which is adjusted by the compression of a
helical spring, /. This spring is carefully set, and the total
pressure on the journal and the pressure per square inch are
both shown on the index-plate, N, by a pointer, M. Above
EXPERIMENTS ON FRICTION TESTIXG-MACHINES.
the journal is a thermometer, g, of which the bulb enters a
cavity in the top " brass," and which indicates the rise in tem-
perature as the test progresses.
The " brasses," thermometer, and spring are carried in a
pendulum, H, to which the ball, /, is fitted ; and the weights
are nicely adjusted, and, as nearly as may be, in such a man-
ner that the maximum friction of a dry but smooth bearing
shall just swing it out into the horizontal line. The stem,
KK'y of the screw, which compresses the spring, projects from
the lower end of the pendulum, and can be turned by a
wrench. A pointer, (9, traverses an arc, PP', and indicates
the angle assumed by the pendulum at any moment. This
angle is large, with great friction, and very small with good
lubricating materials. This arc is carefully laid off in such
divisions that, dividing the reading by the pressure shown
on the index, N, gives the corresponding coefficient of fric-
tion.
The figures on the arc are the measure of the actual resist-
ance of .friction on the surface of the journal. Dividing this
frictional resistance by the total load gives the value of the
coefficient. As there is no intermediate mechanism, this meas-
ure is obtained without possible error ; and, as the resisting
moment changes very rapidly at low angles, great precision of
measurement is obtained, as will be seen when the results of
experiment are given. The machine can also be arranged to
give readings of this coefficient directly.
The theory of the machine is as follows : Let
R = radius to centre of gravity of pendulum;
F= effort due to weight of arm ;
r radius of journal;
/ = length of journal ;
W= weight of pendulum complete ;
P '= total pressure on journal;
p = pressure per square inch of longitudinal section ;
T== tension on spring;
= angle between arm and a perpendicular through axis;
/"= coefficient of friction ;
Q = total friction.
256 FRICTION AND LOST WORK.
When is equal to 90,
FR=.Qr ......... (i)
And when any other angle,
PR sin = Qr ....... . . (2)
Solving equation (2) with respect to Q,
- ....... (3)
The coefficient of friction is
The pressure per square inch is
P
From this last equation the graduations on the right-hand side
of the index-plate are deduced.
From the equation
(6)
the numbers on the left-hand side are determined.
By substituting in equation (i) the value of Q, in terms of
the coefficient and total pressure, from (4) it becomes
....... (7)
Solving with respect to/, equation (7) becomes
FR
r ........ (8;
From the numerator of the second number of equation (8)
the graduations on the arc are deduced.
In applying the foregoing equations to the machine shown
EXPERIMENTS ON FRICTION TESTING-MACHINES. 2$?
in the engraving, the following numerical values may be given
to the respective symbols :
F= 2.5 Ibs.; R= loin.; r .625 in.; 1= 1.5 in.; 4/r = 3.75
sq. in.; ze/ = 61bs. Also, a compression of if inches of the
spring corresponds to a tension of 100 Ibs. ; hence, for each
pound's tension the spring will be compressed .01375 of an inch.
The graduations on the right-hand side of the scale are
obtained from equation (5) :
(4)
The first graduation will naturally be that value of p when
T is equal to o, which value is 1.6.
The speed of the machine, when the belt is upon the largest
pulley of the cone, C, should be that which will give at the
surface of the testing-journal the least speed of rubbing,
which is expected usually to be adopted.
The figures on the arc PP, traversed by the pointer O,
attached to the pendulum, are such that the quotient of tlie
reading on the arc PP, by the total pressure read from, the
front of the pendulum at MN, gives the "coefficient of fric-
tion," i.e., the proportion of that pressure which measures the
resistance due to friction.
A printed table furnished with each machine gives these
coefficients for a wide range of pressures and arc-readings.
To determine lubricating quality, remove the pendulum^
HH from the testing-journal GG 1 ', adjust the machine to run
at the desked pressure, by turning the screw-head K proj-ect-
ing from the lower end of the pendulum, until the index M
above shows the right pressure, and adjust it to run at the
required speed by placing the belt on the right pulley, C.
Next throw out the bearings, by means of the two little
cams on the head of the pendulum, H, in the small machine,
or by setting down the brass nut immediately under- the head
in the large machine ; then carefully slide the pendulum upon
the testing-journal, GG f , and at the same time see that no
scratching of journal or brasses takes place,
Oil the journal through the oil-cups or the oil-holes, set the
258 FRICTION AND LOST WORK.
machine in motion, running it a moment until the oil is well
distributed over the journal. Next stop the machine; loosen
the nut or the cams which confine the spring, and, when it is
fairly in contact and bearing on the lower brass with full pres-
sure, turn the cams or the brass nut fairly out of contact, so
that the spring may not be jammed by their shaking back
while working. Start the machine again and run until the
behavior of the oil is determined, keeping up a free feed
throughout the experiment.
At intervals of one or more minutes, as may prove most
satisfactory, observations and records are made of the tempera-
ture given by the thermometer, Q, and the reading indicated
on the arc P, of the machine, by the pointer O. When both
readings have ceased to vary, the experiment may be termi-
nated.
The pendulum is then removed, the pressure of the spring
being first relieved, and the journal and brasses are cleaned
with exceedingly great care; care is taken to have no particle
of lint on either surface, or any grease in the oil-cups or oil-
passages.
The journal may be cleansed, after each test, either with
alcohol, gasoline, or benzine. The effect of an oil is often felt
in successive tests, long after starting with a new lubricant.
A comparison of the results thus obtained with several oils
will show their relative values as reducers of friction.
Steam-cylinder lubricants are tested upon bearings heated
to a temperature corresponding to any desired steam-pressure.
When the maximum temperature has been attained the flame
is removed, and the behavior of the oil noted as the tempera-
ture falls to 212 F., which corresponds to atmospheric pres-
sure or to zero on the steam-gauge. Any effervescence or
excessive friction at the higher temperatures condemns the
lubricant. It is the custom to take the average of the coeffi-
cients of friction for temperatures ranging from 340 F. cor-
responding to a gauge-pressure of 104 Ibs. to 212 F.
In each case the results are recorded in tables on the blanks
(of which a copy is given on the next page) which are sent
with the machine, and which exhibit
EXPERIMENTS ON FRICTION TESTING-MACHINES. 259
K
IS
leg
-Hi
--
e>--c o^i
liiiilij
& ! f :
jo
oopouj
spunod
jo
spunod
'uouauj
snoimioA3-y
JO
spnnod
uonouj
sainuicn
260 FRICTION AND LOST WORK.
(i) The pressure and speed of rubbing at each trial. (2) The
observed temperatures. (3) The readings on the arc of the
machine. (4) The calculated coefficients of friction.
At the end of the trial the average and the minimum co-
efficients are entered, and the total distance rubbed over by the
bearing surfaces.
To determine the liability of the oil to gum, the bearings
are lubricated with a definite quantity of the oil, and the ma-
chine run a certain number of revolutions. The temperature
of the bearings and the friction at the end of this period are
noted. Both journal and brasses are then removed, placed
under a glass receiver, which excludes the dust yet permits the
entrance of air, and are left there for any desired length of
time, as for one day. At the end of that time the bearings
are replaced in the machine, and the latter is driven until the
temperature of the bearings is the same as at the previous
trial ; the friction is then again noted. Any increase of fric-
tion above that previously observed must be due to the gum-
ming of the lubricant. For the machine described, the stand-
ard quantity of the lubricant is 1 6 milligrammes, which is ample
to afford perfect lubrication of the bearing surfaces during the
trials. The number of revolutions at the first trial is often
5000; it may, however, vary considerably without affecting
the results, so long as it is too small to affect the wearing
qualities of the lubricant, as within this limit the friction
remains constant with a constant temperature. Changes in
temperature and friction always accompany each other; it is
for this reason that great care is taken to obtain the same
temperature of bearing at each trial.
To determine durability, proceed as in determining the fric-
tion, except that the lubricant should not be continuously sup-
plied, but should be fed to the bearing a small and definite
portion of time as a drop or two for each two inches length
of journal. Extreme care should be taken that each portion
actually reaches the journal and is not lost, either in the oil-
hole or by being wiped off the journal, and that the portions
applied arc exactly equal. When the friction, as shown by the
pointer O, has passed a minimum and begins to rise, the ma-
EXPERIMENTS ON FRICTION TESTING-MACHINES. 26 1
chine should be carefully watched, and should be stopped,
either at the instant that the friction has reached double the
minimum, or when the thermometer indicates 212 F. ; or
another portion of the lubricant should be then applied to the
journal.
This operation should be repeated until the duration of
each trial becomes nearly the same ; an average may then be
taken either of the time, of the number of revolutions, or of
the distance rubbed over by the bearing, which average will
measure the durability of that lubricant. Next carefully clean
the testing-journal, and proceed as before with the next oil to
be tested.
In making comparisons, always test the standard, as well as
the competing oils, on the same journal and under precisely the
same conditions.
It was formerly the custom to continue the trial until the
temperature of the bearing, as indicated by the thermometer, at-
tained a certain point, as 120 or 200 F., and to take the number
of revolutions of the journal or the number of feet traversed,
up to that point, as a measure of endurance. The real endur-
ance, however, of the lubricating material bears no definite
proportion to the range of temperature thus observed.
Another method is adopted by the boards of U. S. naval
engineers sometimes appointed to test oils at the navy-yards.
The quantity of oil required to keep down the temperature of
journal to a certain figure, as lio or 115 F. (44 to 46 C),
during a definite period, as one hour, five hours, or twenty-four
hours, is measured, and the endurance is taken as inversely
proportional to these amounts.
The Author considers the endurance of a lubricant to be
measured by the length of time that it will continue to cover
and lubricate the journal and prevent abrasion. When an oil
is placed upon a journal, and there subjected to wear without
renewal, it gradually assumes a pasty or gummy condition,
slowly losing its lubricating power, and finally either increases
friction to an objectionable extent, or oftener becomes so far
expended as to permit the two rubbing surfaces to come into
contact. It has been the custom of the author to run until
262 FRICTION AND LOST WORK.
this occurs, and then to take the length of the run as a meas-
ure of the endurance of the oil.
It is extremely difficult to obtain successive measures of
similar value even by this method ; but by taking an average
of several successive trials or many, if necessary the true
measure of the endurance of lubricants can be obtained with
any desired or necessary accuracy. This method involves
more risk of injury to the journal than the other, and some-
times considerable loss of time in bringing the rubbing surfaces
back into good condition again before going on to make other
tests. The determination of the real value of the lubricant is
usually of sufficient importance, however, to justify whatever
time, trouble, and expense may be thus incurred.
This machine did such good work as to encourage the Au-
thor to design one especially fitted for railroad work.
The journal of this machine is of standard size, 3^ inches
diameter and 7 inches long. The speed is intended to be
adjusted to velocities varying from that of a twenty-six-inch
engine-truck wheel at sixty miles an hour down to that of a
forty-two-inch wheel running fifteen miles an hour. The pres-
sures are adjustable from a minimum total pressure up to 400
Ibs. per square inch (28 kgs. per sq. cm.), or a load of nearly
10,000 Ibs. (4545 kgs.) on the journal.
Fig. 42 is a side elevation of the larger machine, with the
journal and pendulum in section, and Fig. 43 a front elevation.
It consists of a shaft, AB, which is driven by a cone-pulley, C,
the whole mounted on a cast-iron stand, D, terminating in a
forked end' at the top, with two bearings, E and F, in which
the shaft runs. The shaft projects beyond the journal F, and
the projecting part A is provided with a sleeve or bushing,
mm, the outside of which forms a journal on which the tests
of oil are made. A pendulum, AG, is suspended from this
journal with suitable bearings, aa, which work on the journal
mm ; the heavy weight, G, attached to the lower end, is now
omitted. It is evident that the friction on the journal mm
will have a tendency to move the pendulum in the direction of
the revolution of the shaft, and that the greater the friction on
the journal the farther will the pendulum swing. A scale or
EXPERIMENTS ON FRICTION TESTING MACHINES. 263
dial, HI, is attached to the stand, and the distance the pendu-
lum swings may be read off on this scale, which thus indicates
the coefficient of friction of the lubricant on the journal. In
order to get any desired pressure of the bearings on the jour-
nal, the pendulum is constructed as follows: A wrought-iron
pipe, J, which is represented in Fig. 42 by solid black shading,
FIG. 42. FIG. 43.
THURSTON'S "RAILROAD MACHINE."
is screwed into the head K, which embraces the journal and
holds the bearings aa in their place. In this pipe a loose piece,
b, is fitted which bears against the under journal-bearing a'.
Into the lower end of the pipe a piece, cc, is screwed with a
hole drilled in the centre through which a rod, J, passes, the
upper end of which is screwed into a cap, d\ between this cap
264 FRICTION AND LOS 7' WORK.
and the lower piece, cc, a spiral spring shown in section in Fig.
42 is placed.
The upper end of the rod has a cap, e, in which it turns and
which beats against the piece , which in turn bears against
the bearing a'. If the rod is turned with a wrench applied to
the square head at/, it is obvious that the cap d will be either
drawn down on the spiral spring, which will thus be compressed,
or it will be moved upward, and the spring will thus be released,
according to the direction in which the rod is turned. If the
spring is compressed, its lower end will bear against the under
cap and on the piece cc, by which the pressure will be trans-
mitted to the pipe/, and thence to the head K, and from that
on the upper journal-bearing a ; while at the same time the
upper end of the spring bears against the cap d, which, being
screwed on the rod/, transmits its pressure upward to the cap
e, and from that to the loose piece b, and from that to the up-
per journal-bearing a. It will thus be seen that any desired
pressure within the limits of the elasticity of the spiral spring
may be brought upon the journal and bearings by turning the
rod /. The piece b has a key, /, which passes through it and
the pipe/. This key bears against a nut, o, which is screwed
on the pipe, its object being to provide a ready means of re-
lieving the journal of pressure by simply turning the nut o
when it is desired to do so. An index, /', is attached to the
spiral spring so as to show the position of the latter.
A counterbalance is sometimes used to reduce the " mo-
ment" of the pendulum, when very fine readings are desired.
This modification necessitates a corresponding change of the
scale on the arc of the machine. (See Frontispiece.)
The " brasses" are cast hollow, and when desired a stream
of water is driven through them to keep the rubbing surfaces
cool and at uniform temperature. This plan was adopted
many years ago by Him, to secure uniformity and manageabil-
ity of temperatures. This provision insures great exactness of
determinations. Provision for lubrication by the oil-bath is
sometimes advisable for special work.
The oil is fed to the journal by means of oil-cups, LL, on
the top of the head K, and a thermometer, 7", is attached be-
EXPERIMENTS ON FRICTION TESTING-MACHINES. 26$
fi
0009V <
o:
0009V 13
O
OOOtl U-
O
O
ooou >
0000 V
0006
0008
OOOi
0009
0009
000*
iO NOIJ.VIA3Q
266 FRICTION AND LOST WORK.
tvveen the two cups, and from it the rise in temperature is ob-
served. A cord, s, is attached to the pendulum in some cases,
to prevent its being thrown beyond the intended limit.
The Pratt & Whitney Co., of Hartford, U. S., and Messrs.
W. H. Bailey & Co., of Salford, G. B., the builders of these
machines, have slightly modified some of their details, but have
retained all essential features as in the frontispiece.
131. Lux's Improvement on Thurston's machine consists
in the addition of an automatic recording apparatus. The
pendulum of the machine carries an arm, which raises and de-
presses a slide at the right, which slide carries a pencil. A cyl-
inder is mounted behind the pencil-slide, and is connected with
clockwork, by which it is made to revolve uniformly at any con-
venient rate. Paper wound on this cylinder is thus made to
move under the pencil at a constant rate, and the rise and fall
of the latter is proportional to the swing of the pendulum, and
varies with the friction at the journal. The paper is suitably
lined, in such manner that the diagram so made can be conve-
niently read, the abscissas of the curve measuring the times and
the vertical scale giving the friction. The pressure is adjusted
and the temperature readings taken as before.
The preceding figure exhibits the form of diagram obtained
during tests of oils in the manner just described.
132. Illustration of Method, Record, and Report
Results of Trials of an Oil marked X, and its comparison with
Standard Bleached Winter Sperm and Pure Lard Oils.
In illustration of the method frequently adopted by the
Author in making a tolerably complete investigation, we have
the following:
These oils were tested on a " lubricant-testing machine"
of the " 77" style, by the method already described. The
standard bleached winter sperm and a pure lard oil were tested
with the X oil on the same bearing and under precisely simi-
lar conditions. The following are records of data obtained
during these tests:
RECORD OF TESTS OF LUBRICANTS. WINTER-BLEACHED SPERM AND LARD OILS.
b
Laboratory Nos. 90 and 93 ; Original Marks, Standard Sperm, Penn. Lard ; Sources, New Bedford and P. R. R. ^
Tnv M ,ir a ,inn T n H^rmin. th. PowPr nf r^nrincr Fr.Vfinn - r^ffi.i. t f PVi,M _ Rea din* On Arc Pressure ^
,
{( UNIV
>\
ERIMENTS ON " FRICTION TESTING-MACHINES. 26
'lotal Pressure Friction.
2 Sperm, i, 2, Lard; Pressure on Journal, Ibs. per square inch, 50, 100 Sperm, jo, 100 Lard. Total pressure on Journal,
ard; Amount of oil used on Journal, continuous supply ; Average Coefficient of Friction, .0050, .0037 Sperm, .0100, .0062 Lar
lied by rubbing surface, per minute, 237.3, 235-i Sperm, 233.8, 229.2 Lard; Elevation of temperature, "max., 8, 10 Sperm, 9, 12 ]
SPEKM OIL No. i.
uo Suipea^i
to 10 to 10
d
8888
S ,,S
3M M C1
2 2 2
suopn{OA3^j
samara
UM
MU
MM
j.'SBSJo
UO SuipE3^J
M M M M
d
6
Q
K
M M M
i
H
M M
ajnicasdmax
^^
^888
*^ 8
mk
suoimioA3-g
sajnuim
8 to o >n
ro m
ItM
O 10 O to
N W CO ro
W M en ro
uonoijjj
uo Suipna^
10
?
to to to
IO to 10
m.
a, SS
am
O (f>tr> t**
O> O ON O
No. of Test, i,
400 Sperm, 200, 400 L
number of feet trave
^r
j-. 5
|-, B
ti
2 tog JO
268
FRICTION AND LOST WORK.
IS 8.
5 S3
o ^ I is
i H p" ^ <->
o w-c'o
t^ 6
2^2 = SW
<; rt u jj *>j^
3 = Pc^-j
S Q S"
= S ^til"
I a ll K'"
g .. -
PI
< I'S
UOIJDUJ
jo ;u3'p^903
ojy
UO 3UlpB3^
10
VO . -4- >OVO . lovo _ VO
"3 ^ ?
>o
00 Ov O t^OO 00 O O 00 O 2
'o 'o 'o
ajmTUSdOiaj,
| | -
o o o 0:2 Q o o
OuOOOuOOCO
1 11
S88.18S 8 ggj 8 S|S
.2 i2 .-
5 -3 -c
.W5 C/J s tO
suopnjoAa^i
0* 0* 0*
vo o ~~
CJ "2
VO t^. t-x
W N N
^ ^ ^
^ J 1
sainuioi
'aujji
t^ ON O N ** 10 t^.
vS'vS'S Sv8*K K. ^
UOHDUJ
JO 5U3IOIJJ303
ojy
UO SutpE3-JJ
O 1O
4- lO^O _ r<1 lOVO ^ fO 10
^ ^
10 1O 10
Ov O iovo t*xOO 00 ON O IT*VO t^
'o 'o
TJ T3
3jn}i3j3dai9j,
3 3
8 8 8-c 8 8 8-c 8 8
j
O
iJ Si
S 3
00 O^ O O O O O O o O O O
W(N>-IN(MWo r^oo ON O w fo
^-10 10 10 10 10 10VO VOVOVO
UOnOUjJ
jo }u3pyj9o3
ojy
to
iO>O t^oo Ov
uo SuipBa^j
'o
ajrvjBjaduiax
10 10 O oo 10 ro r>oo O O
VO vO t^ t^OO Ov O\ C- O O
1
^^s^sjf'Hlsgg*^
.a
suopn[OA3^
8^
sainuirn
'aoJij,
rt
8 MIO Sff8.5-8,
<
ti
I Htn S5?81?8, ^^^5-?
<
EXPERIMENTS ON FRICTION TESTING-MACHINES. 269
1
22
II
O 2
"0*0
Si
13
i*3
l*.-
ii *<
2 II
1
id
Sg
i If!
I --o
I ^
c "5 *?
.5 3
UOIIDUJ
JO 1031010303
ei
o"
UO 3uipB3^I
IO 10
S?2"8 8
'o "ft
1 1
^ |
1 1
suorjn[OA3^[
8 "8
egos SoJ
N
sainuiiu
M CO^lOOlOQMCO -
t^ t^ t^ t^OO C3O O\ O* O* ON
in 10 t-.
vo r^ t- t^
UOI13UJ
jo ;u3ioijpo3
uo Snipes^
IO
to O o co * 10 co
"c
"3
3jrUBJ3dai3j,
10 1000 ooOioOO30OfOioo
N W M m m * -^-^ - Tf o oo oo o C 'C;
10 IOIO 10 10
O ' M M fj O co
3JnjBJ3dui3X
,oloooJ
^ooooo
U5 OB
suorjn[OA3^
"8 "8
^ 8*
IO O*
s->inuiui
r?
i
t:
2/O
FRICTION AND LOST WORK.
ii
.2
^a
C.3
ill
ii =
i
a
xi
^w
hofc^
r^ O
2
2J-3
-
s
e^ b. |S
C ^ -*^ . vn
*S'S sS
5 >
M!
3S U
A Si8,
g ir>
S&
jo
888
uopouj
JO 1U3IDUJ303
N 0?
IO O >O >O
o 1 *O
u,
0000*0
CN N M O
S 1
c w co ro
3.
^Jrg
as g,^
8
3 H no
4J
<3
2 1
EXPERIMENTS ON FRICTION TESTING-MACHINES. 2? I
From the preceding logs of tests were deduced the follow-
ing results and conclusions :
AVERAGE COEFFICIENTS OF FRICTION.
I.ab.
No.
NAME OF OIL.
Pressure per Square Inch.
100
50
Average.
90
93
W B Sperm
0.0037
0.0168
0.0062
0.0050
0.0206
O.OIOO
0.00435
0.0187
O.OoSl
X
Lard
The relative values of these oils in reducing friction, taking
sperm-oil as a standard, and giving it a value of 100, will be
represented by the quotients obtained by dividing the coeffi-
cients for sperm by those for each of the other oils, and multi-
plying by 100.
The following table gives these quotients :
RELATIVE POWER OF REDUCING FRICTION.
NAME OF OIL.
Pressure per Square Inch.
100
5
Average.
W B Sperm .
IOO.O
IOO O
IOO O
X
22.
24.2
23.2
Lard
5Q 6
50.0
53-7
The speed was about 700 revolutions per minute (244.3 ft-*
74 m.), giving a speed of rubbing surface corresponding to
about 35 miles per hour for a 33-inch (79 cm.) wheel in railroad
service. Dividing the coefficients for the oils by the coefficient
for sperm and multiplying by IOO, we obtain the following
tabulated figures as the relative amount of power consumed in
using the respective oils.
A common standard pressure and speed for such tests is,
on some roads, 250 Ibs. per square inch, and a speed equivalent
to 15 miles per hour for the axle-journal, at a temperature of
100 F.
2/2
FRICTION AND LOST WORK.
RELATIVE POWER CONSUMED.
NAME OF OIL.
Pressure per Square Inch.
TOO
5
Average.
W B Sperm
100.
481.0
167.6
IOO.O
412.0
200.0
IOO.O
429.8
186.2
X
Lard
As regards friction, sperm excels, lard stands next, and X
next.
From the results of the tests of durability, we find the fol-
lowing:
DURABILITY.. OR WEARING POWER.
Revolutions. Ft. travelled.
W. B. Sperm 27,870 9.726.6
X (average) 26,380 9,206.6
Lard 24,500 8,550.5
Taking bleached winter-sperm oil as a standard, and assum-
ing its value to be 100, the values of the oils as regards dura-
bility will be represented by 100 times the quotient obtained
by dividing the number of revolutions or feet travelled of each
oil by the feet run by sperm. We thus obtain the following:
RELATIVE DURABILITY.
W. B. Sperm 100.0
X 94.6
Lard 87.9
The figures in this last table are measures of the lengths of
time that equal quantities of each oil would run, so that the
greater the figures of this table the more valuable the oil.
The value of an oil may be taken as greater in proportion as
the figures in the above table are greater, and as the figures in
the table headed " Relative Power of Reducing Friction" are
greater, so that combining the results given in both tables, the
relative values of the oils, sperm-oil being the standard and
taken at 100, may be represented by one one-hundredth the
product obtained by multiplying the figures in the last column
of the table headed " Relative Durability" by those in the last
column of the table headed " Relative Power of Reducing
Friction." The following are therefore the relative values.
EXPERIMENTS ON FRICTION TESTING-MACHINES. 273
RELATIVE VALUES OF THE OILS.
W. B. Sperm
X
Lard..
100. o
21.9
47.2
SECOND TEST.
A second test consisted in cutting a square hole in the lower
box and packing it with waste saturated with the oil to be
tested. The oil to be tested was spread on the journal and a
pressure of 100 Ibs. per square inch (43 kgs. per sq. cm.) was
applied ; the machine was then started and allowed to run
until the friction had increased to double the least amount
shown at any time during the test. Both the X and the lard
oils were tested by this method. In each case 743 milli-
grammes weight of waste was used as packing. The waste
was in each case thoroughly saturated with the oil and weighed
before and after the test. In the case of X, the waste absorbed
4.806 grms. and contained 2.229 g rms - at the end of the test,
so that the oil consumed was 2.577 grms. In the case of the
lard, 4 grms. were also absorbed by the waste ; 7.265 grms.
remained ; so that the useful consumption was 2.735 grms. X
ran 266,226 ft. = 54.2 miles per gramme consumed, with an
average coefficient of friction of 0.0318, and lard-oil ran
182,528.7 ft. = 34.5 miles per gramme consumed, with an
average coefficient of friction of 0.0244, the former excelling
the latter about sixty per cent.
THIRD TEST,
A third test was made upon the 4< R. R. Standard Machine, 1 '
and the following are the coefficients of friction obtained :
AVERAGE COEFFICIENTS OF FRICTION.
OIL.
Pressure per Square Inch and Total.
150,
2629
525
Average.
W. B Sperm..
0.008
0.024
O.OOg
0.0046
0.0(5
p.0059
0.0063
0.0195
X .
Lard
CHAPTER VII.
FRICTION OF LUBRICATED SURFACES-LAWS AND MODIFYING
CONDITIONS.
133. Variations of Friction of Lubricated Surfaces oc-
cur, as has been already stated, with every change of physical
condition of either the bearing and journal surfaces, or of
the lubricant applied to them.* A rough pair of surfaces ex-
hibits great resistance to relative motion, while this friction is
constantly reduced as they become smoother with wear ; but
under some conditions the smoothness and the nicety of fit
may be made too perfect, and the friction then increases again.
An oil which works well, and gives a comparatively low coef-
ficient under low pressures, may prove an inferior lubricant un-
der heavy loads, and the same unguent may be a good, a bad,
or an indifferent lubricant according to the temperature or the
speed of the rubbing surface to which it is applied. It is even
sometimes found to be the fact that, with some lubricants, and
especially with light mineral oils, the total frictional resistance
may be reduced, while nevertheless the bearing may show in-
creased wear, the increase of resistance due to the exceedingly
slow wear being compensated by the decrease in fluid resist-
ance.
The conditions which produce most serious differences in
ordinary work are the nature of the unguent, the pressure, and
the temperature. Velocity of rubbing determines a limit be-
yond which the intensity of pressure cannot be carried without
danger of heating; but the effect of its variation upon the
* Friction and Lubrication. New York, 1879.
FRICTION OF LUBRICATED SURFACES.
275
coefficient of friction is usually less considerable than is that of
either of the other conditions specified.
The lubricating value of oils is even affected by moisture.
It affects mineral oils very little, the moisture slightly increas-
ing their resistance in the bearing. They have little tendency
to absorb moisture from the atmosphere. Fatty oils are some-
what hygroscopic, and are quite sensibly affected by a trace of
moisture.
Exposure to air produces a tendency in organic lubricants
to acidify or to become resinous, the non-drying oils exhibiting
the one and the drying-oils the other method of change. The
purer the oil, as a rule, the less is the liability to change.
Hirn, experimenting on the oils named below, found that some
were rather better lubricants at the period of incipient rancid-
ity than when fresh. Cocoa-nut oil was 7 per cent, and rape
seed 3 per cent, better, while with other oils less difference is
observed.
Working the oils for a week together, using an oil-bath,
Hirn finds sperm-oil to alter least of all, very slowly increasing
in resistance ; neat's-foot next, then olive and rape-seed ; while
cocoa-nut oil depreciates most rapidly, and at a rapidly acceler-
ated rate.
" The time required to exhibit an acid reaction was as below :
OILS. Time, hours.
Sperm, first quality. ....... 36
" second quality 36-38
Lard 24
Neat's-foot 30
Olive, limpid 24-30
OILS. Time, hours.
Cocoa nut 4
Poppy 5
Rape seed, refined 12
crude 24
Sperm-oil was found to be the best lubricant in all these
experiments.
The method of supply should b' carefully looked to, and a
very free " feed," with a system of collection and reapplication
of the oil leaving the bearing, will be found to give by far the
greatest economy of power and cost. Experiments made for
the Institution of Mechanical Engineers, in which oiling by a
pad as in railway work, by a siphon lubricator or oil-cup, and
by a bath, which keeps the surfaces flooded with oil, gave the
2/6
FRICTION AND LOST WORK.
COEFFICIENTS OF FRICTION.
[Journal of Cast Iron ; Bearing, Bronze ; Velocity, 750/^(230 m.} per minute
Temperature, 70 F. (21 C.). Intermittent feed through oil-hole .]
NAME.
PRESSURES:
LBS. PER SQUARE INCH AND PER SQUARE CM.
8
0.56
16
I . 12
3 2
2.24
48
3-36
Avge.
Min.
Avge.
Min.
Avge.
Min.
Avge.
Min.
GROUP I.
.1720
2505
. 1920
.1866
.1986
.3296
.1979
.2386
2242
.1840
.1585
.1928
.1668
.2156
.2826
.1817
2 597
.1598
. 1910
.2125
.2765
*75
2375
2475
.2776
.2530
1875
1537
1833
2550
2330
.1272
.2607
.2072
1755
.2369
1747
!959
.1746
.1839
. 1716
.1259
1557
l6 37
1330
.1500
1583
1333
.1500
1833
::m
1500
i5<->o
333
1333
1333
I 577
.1666
*333
.2000
*333
- 1500
.1666
.2650
1333
.1916
1500
.2166
.1660
*333
.1500
!333
1500
.2165
. IIOO
.2000
J333
.1166
.2166
1333
1583
1500
!333
.1666
.1166
*333
1333
. 1627
.1410
.1600
1383
.1482
. 1902
.1916
1575
1621
1460
1378
.1650
1575
I7S7
.2041
1567
.1842
.1215
.1688
.1401
.2452
.1066
.1380
.1488
.1666
.1238
. 1604
1583
2333
.2067
.1729
1453
.1777
SECON
.1661
.1678
1250
.1483
.1770
1254
"75
1435
.0981
. 1006
.1685
.1083
. IOOO
*33
.09166
.0916
!25O
. 1166
.1000
. IOOO
.1083
.1083
. IOOO
.1250
.1250
1250
.1500
0833
1333
.1249
1500
.0916
.1125
.1250
. 1500
. IOOO
.1416
.1500
. 1500
.1500
. 1416
. IOOO
1333
D SERI
.1291
.1291
. IOOO
"33
.1250
. IOOO
.0916
.1166
.0833
.0833
.1083
.102
.0958
.1172
.1109
.I 3 l6
.0925
.1086
.1405
.1166
935
.1190
.0862
.1681
.1444
.1116
.1187
.1277
J 347
. 10052
.1166
. 1170
.1062
. 1026
.1016
.0970
. IOOO
.0861
.1277
.1250
1275
.1250
.1777
1343
ES OF
.1302
.1083
. IOOO
1333
.1095
.1198
.0902
.1000
.0983
.0895
.0982
.0833
.0875
.0916
.0874
.1086
.0750
. IOOO
. IOOO
.0916
.0750
.0916
.0791
. IOOO
1083
0584
.0833
.0833
0750
.0792
. IOOO
.0833
.0 79 lt
.0708
.0666
0833
.0917
.0791
.1125
.1166
.1250
.1250
1500
.1125
TESTS.
.0958
.0958
.0750
0833
.0666
.0791
.0750
0833
.0666
.0750
.0625
.1180
.08,3
.09907
.0881
.0951
.1444
0993
. 1005
.1138
.1166
0986
.0766
.0930
.0996
IOI 3
1063
.1305
.0962
.0833
. IIOO
.1028
.0794
.0944
.0805
.0880
. I22O
.0944
1277
. 1222
1555
.1770
I5OO
.2222
"55
.0811
.0777
.0986
.0758
"59
T 344
.0822
.0861
.0758
.0963
1050
.0750
.0944
.0777
.0722
. IOOO
.0705
.0750
1055
.0844
.0750
.0611
0555
.0694
.0666
.0722
.1111
.0609
.0550
.0800
.0844
t.o6ii
.0722
.0661
0833
. IOOO
.0944
.1277
.1222
.1444
.1770
.I5OO
.2222
.0888
.0750
.0666
.0666
.0666
.IOOO
.0611
0555
.0750
.0722
.0888
Bleached " "
" Winter Whale
Bleached " " ...
Winter Lard Oil
Extra Neat's-foot Oil
Tallow Oil
Refined Seal Oil
Bleached Winter Elephant Oil
GROUP II.
Olive Oil
Cotton seed Salad Oil
Palm Oil
Rape-seed Oil
Elaine Oil . .
Linseed Oil*
Pea-nut Oil
Refined Cotton-seed Oil .. . . .
Rosin Oil
Cocoa-nut Oil
Cold-Pressed Castor Oil
GROUP III.
Labrador Cod Oil
Tanner's Cod Oil
GROUP IV.$
Mineral Sperm Oil
Deod White Lubricating
Bleached Deod. Lubricating
Unbleached Deod. Lubricating
Paraffine
GROUP I.
Natural Winter Sperm ...
Bleached " "
Natural " Whale
Bleached " "
Winter Lard
Extra Neat's-foot
GROUP II.
Olive Oil
Refined Rape-seed (Yellow)
Winter-pressed Cotton-seed (White)
Winter-pressed Cotton-seed (White)
GROUP III.
Menhaden Oil . .
* Not a lubricant.
t Values somewhat uncertain.
j All mineral oils here described are of uncertain composition.
FRICTION OF LUBRICATED SURFACES.
277
following figures, showing an enormous advantage in the use
of the last method :
METHODS OF OILING (RAPE-SEED OIL).
Velocity of rubbing, 157 feet (46 m.) per minute.
Actual
Load.
Coefficient
Kilogs. per
sq. cm.
Ibs. per
sq. in.
of Friction.
Friction.
Oil Bath
18.5
263
0.00139
j
Siphon Lubricator. . . .
Pad under Journal. ...
17.7
I 9 I
252
2 7 2
0.00980
0.00900
7.06
6.48
The lowest of these values of the coefficient are below any
reached by the Author, or, up to their date, probably, ever
recorded.
134. Commercial Oils, under moderate pressures, vary
greatly in their power of reducing friction. The table of
values (p. 276) obtained by the Author by experiment, using
the testing-machine devised by him, exhibits the effect of varia-
tion of pressure in changing these values, as well as the differ-
ences in oils, all of which were supposed to be pure. These
values may probably be assumed as correct, and applicable in
the ordinary work of the designing engineer.
In this case the journal was of cast-iron, running in gun-
bronze bearings, and was in very good, but not in the very
best possible, condition. As will be seen, much better figures
may be obtained. The oils were here supplied intermittently,
but frequently, in the usual manner, and the results may be as-
sumed to be substantially the same as with continuous feed.
The first series were not all fresh ; the second set were fresh and
pure.
To show how these figures were obtained, the results are
given below in detail and in the usual tabular form, as obtained
o
by the Author by trial of a good sample of winter-bleached
sperm-oil. It should be remembered that precise agreement
between two tests of even the same oil, under nominally the
same conditions, never can occur except by a rare accident, as
the oil itself is never precisely alike throughout sperm-oil, for
example, varying in quality with its purity and age, and with
278
FRICTION- AND LOST WORK.
the age, sex, health, and habits of the fish from which it was
taken, etc., and the conditions of the journal and the other
circumstances affecting the trial can rarely if ever be dupli-
cated with absolute precision. These differences are not usu-
ally of practical importance, but the precaution is always taken
to compare each oil tested with a standard pure sperm, care-
fully preserved, to be tested immediately before or immedi-
ately after the test of the oil to be examined. The quantity
of oil here adopted was 332 milligrammes enough to flood
the journal at one application.
DETAILS OF TEST.
BEST WINTER-BLEACHED SPERM OIL.
FIRST TRIAL.
Amount used upon the journal 332 milligrammes.
Speed of rubbing surface 736 ft. (224 m.) per minute.
Pressure per square inch and per cm 8 Ibs. (0.56 kgs ).
Total pressure 3 o Ibs. (13.6 kgs.).
1
h
Temperature
of
Brasses.
Friction,
Ibs.
Coefficient
of
Friction.
|
H
Temperature
of
Brasses.
fl
Coefficient
of
Friction.
u
8
H
Temperature
of
Brasses.
Friction,
Ibs.
Coefficient
of
Friction.
At
Deg.
Fahr.
Degr.
Fahr.
Deg.
Fahr.
Start.
min.
75
39
160
3-5
79
190
5-5
i
85
8
4 1
160
3-5
81
190
5 5
3
120
7-5
43
163
3-5
83
T 95
5-5
5
145
55
45
161
3 5
85
T 95
5-5
7
1 60
5-5
47
163
3-5
87
200
5-75
9
170
6
49
165
3-5
89
200
5 75
ii
178
6
Si
^ 5
3 5
9 l
205
6-5
3
185
6
53
165
3-5
93
2IO
5
15
190
5
55
167
95
205
4-5
J 7
190
5
57
167
97
20.S
5
19
193
5
59
1 68
99
197
5
21
193
45
61
170
IOI
195
5
23
190
4
63
170
103
195
5-5
25
'jjS
3 1
65
170
J 05
197
5-75
27
1 80
3
67
170
107
200
6-5
29
*75
3l
mini-
69
i75
5-5
109
2OO
6-5
31
170
3[
mum.
7 1
J75
5-5
III
212
6-5
33
168
3
O.IO
73
180
5-5
"3
216
6.5
35
37
^ 5
165
3)
3-5
75
77
180
185
5 5
5 5
"5
"7
216
220
6-5
5-5
119
220
6-5
139
235
7
159
2 4
7
121
218
6-5
141
235
7
161
240
7
I2 3
220
6-5
H3
235
7
163
240
7
125
2 2O
9 5
MS
238
7
165
243
7
127
22O
9-5
H7
238
7
167
243
7
I2 9
230
75
149
238
7
169
243
7
131
2 3
7-5
J5 1
238
7
171
2 44
7
*33
232
7
153
240
7
173
245
7
135
234
65
155
240
7
175
250
7
Mean
137
234
7
I 57
240
7
177
260
7
0.1875
OF LUBRICATED SURFACES.
279
SECOND TRIAL.
Amount used upon the journal.
Spsed of ruL.bhig' surface
Pressure per square inch
Total pressure
332 milligrammes.
.736 ft. (224 m.) per minute.
i61bs.
6oibs.
a
2
2
emperatu
of
Brasses.
Friction,
Ibs.
y
cj "C
|
emperatu
of
Brasses.
II
Coefficien
of
Friction.
i
H
emperatu
of
Brasses.
Is
.'oHIii-int
of
Friction.
H
H
H
At
Start.
Deg.
Fahr.
Deg.
Fahr.
Deg.
Fahr.
min.
TO
*7
275
"5
35
325
10.5
i
So
15
fo
19
290
"5
37
325
II
3
130
IO
21
35
11.5
39
327
11.5
5
170
10
6
23
315
11.5
4i
333
"5
7
200
9
8
25
320
11.5
43
337
12
9
215
9
a
27
320
9-5
45
342
IS
ii
239
9
8
29
320
10
47
352
12
T 3
250
8*
"5
31
318
n. 5
49
355
IO
15
262
9
ii
33
322
10.5
51
350
10
53
350
ii
*
59
335
10
65
318
10-5
55
35
IO
61
325
10
67
325
IO
a'vge
57
340
10
63
9-5
68
325
0.1776
THIRD TRIAL.
Amount used upon the journal.
Speed of rubbing surface
Pressure per square inch
Total pressure
332 milligrammes.
736 ft. (224 m.) per minute.
32 Ibs.
120 Ibs.
e
H
|
Temperatu
ot
Brasses.
jl
fc
a
-
cJ
i
H
Temperatu
of
Drasscs.
t"
O .
Coefficien
of
Friction.
V
B
H
Temperatu
of
Bruises.
Friction,
Ibs.
Coefficien
of
Friction.
At
Start.
80
o
5
2IO
11.5 |
o 096
ii
295
15
V
mini-
to
i
95
21
7
235
ii 5 )
mum.'
12*
320
25
av'ge
3
170
'5
9
260
o 1317
FOURTH TRIAL.
Amount used upon the journal.
Speed of rubbing surface
Pressure per square inch
Total pressure
332 milligrammes.
.736 ft. (224 m.) per minute.
48 Ibs.
...iSolbs.
|
Temperature
of
Brasses.
8 .
th
Coefficient
of
Friction.
8
H
Temperature
of
Brasses.
.
5a
Coefficient
of
Friction.
8
H
Temperature
of
Brasses.
i^
.5 i
1 1
At
Start.
80
100
20
3
5
180
235
16)
0.0833
mini-
mum.
7
8
9
285
320
345
30
to
40
av'ge
0.1104
2 SO
FRICTION AND LOST WORK.
135. The Relative Standing of Oils, such as are found
in the market, as determined by their power of reducing fric-
tion, and economizing work and energy, when used on ma-
chinery in which the pressures are low, is readily determined
by the study of the preceding table. The columns of mini-
mum values of the coefficient of friction may be taken to rep-
resent the values of the oils there named when lubrication is
continuous and free; and these values are those to be selected
for the purposes of such a comparison.
Comparing the oils tested at any one pressure, it is seen
at once that they differ greatly in their power of reducing
friction at whichever pressure they are compared. All give
lower coefficients as the pressure rises; but the differences are
great at all pressures. The following table exhibits the rela-
tive standing of the oils named at the several pressures re-
corded :
RELATIVE STANDING OF LUBRICANTS.
First Series.
ORDER.
PRESSURES.
[Lbs. per sq. in. and kgs. per sq. cm.]
8
0.56
16
I . 12
32
2.24
48
3-36
i
Crude Mineral Lubri-
cating-.
Nat. Summer Sperm
B. S. Whale.
B. W. Whale.
Refined Seal.
B. W. Elephant.
Olive.
Rape-seed.
Cocoa-nut.
Mineral Sperm.
Bl. Deod. JMin. Lub.
N. W. Sperm.
N. S. Whale.
Ex. Neat's-foot.
Tallow.
Pea-nut.
Lab. Cod.
Deod. W. Min. Lub.
Unbl. W. Min. Lub.
Natural Whale and
Cocoa-nut.
Nat. W. Sperm.
Ex. Neat's-foot.
Tallow.
Olive.
Menhaden.
Crude Lub.
N. S. Sperm.
Ref. Seal.
B. W. Elephant.
C. P. Castor.
Palm.
Labrador Cod.
C. P. Castor.
N. W. Whale.
Tallow.
Pea-nut.
Olive.
R. W. Elephant.
Cocoa-nut.
Labrador Cod.
FRICTION OF LUBRICATED SURFACES. 28 1
RELATIVE STANDING OF LUBRICANTS Continued.
ORDER.
PRESSURES.
[Lbs. per sq. in. and kgs. per sq. cm.]
8
0.56
16
1. 12
32
2.24
4 8
3-36
Cotton-seed.
B. W. Sperm.
Menhaden.
W. Lard.
Palm.
Ref. Cotton-seed.
N. W. Whale.
B. W. Whale.
N. S. Whale.
Ex. Neat's-foot.
Tallow.
Pea-nut.
Lab. Cod.
Deod. W. Min. Lub.
Unbl. W. Min. Lub.
Cotton-seed.
B. W. Sperm.
Menhaden.
W. Lard.
Palm.
Ref. Cotton-seed.
N. W. Whale.
C. P. Castor.
Elaine.
Paraffine.
Tanner's Cod.
Rosin.
W. Lard.
Ref. Cotton -seed.
N. W. Whale.
Cotton-seed.
Palm.
Rape-seed.
Lab. Cod.
B. W. Sperm.
B. W. Whale.
Pea-nut.
Paraffine.
Mineral Sperm.
Elaine.
Rosin.
Tanner's Cod.
Deod. W. Min. Lub.
Bl. W. Min. Lub.
Unbl. W. Min. Lub.
Cocoa-nut.
Mineral Sperm.
N. S. Sperm.
Rape-seed.
Elaine.
Cocoa-nut.
Tanner's Cod.
B. W. Sperm.
Ex. Neat's-foot.
Ref. Seal.
Menhaden.
B. W. Whale.
Winter Lard.
Olive.
Ref. Cotton-seed.
Cotton-seed.
N. S. Whale.
N. W. Sperm.
W. Lard.
Ref. Seal.
Menhaden.
B. W. Whale.
W. Lard.
Olive.
Ref. Cotton-seed.
Cotton-seed.
N. S. Whale.
Deod. W. Min.
Paraffine.
Bl. Deod. Min.
Unbl. Deod. Min.
Crude Min.
Palm.
Cotton-seed.
B. W. Whale.
N. S. Whale.
Rape-seed.
C. P. Castor.
N. W. Sperm.
W. Lard.
Ref. Seal.
Ref. Cotton-seed.
Tanner's Cod.
Tallow.
Rosin.
B. W. Sperm.
Mineral Sperm.
N. W. Whale.
Tanner's Cod.
Tallow.
Rosin.
B W. Sperm.
Mineral Sperm.
N. W. Whale.
Menhaden.
N. S. Sperm.
Ex. Neat's-foot.
Elaine.
HI. Deod. Min.
Deod. Min.
Unbl. Min.
Crude Min.
Paraffine.
6
8
12
13
*4
15
l6
18
ig
282
FRICTION AND LOS 'T WORK.
RELATIVE STANDING OF LUBRICANTS.
Second Series.
ORDER.
PRESSURES.
[Lbs. per sq. in. and kgs. per sq. cm.]
8
0.56
16
I. 12
32
2.24
1,,
W. P. Cotton-seed.
B. W. Whale.
Olive.
Menhaden.
Ex. Neat's-foot.
W. Lard.
Ref. Rape-seed.
N. W. Whale.
W. P. Cotton-seed.
N. W. Whale.
Ex. Neat's-foot.
Menhaden.
B. W. Whale.
Ref. Rape-seed.
W. Lard.
N. W. Sperm.
B. W. Sperm.
Menhaden.
W. Lard.
W.P Cotton-seed.
N. W. Whale.
Olive.
W.P Cotton-seed.
Ex. Neat's-foot.
B. W. Whale.
Ref. Rape-seed.
N. W. Sperm.
B. W. Sperm.
Ref. Rape-seed.
Olive.
N. W. Sperm.
B. W. Whale.
W. Lard.
W.P.Cotton-seed.
N. W. Sperm.
Menhaden.
Ex. Neat's-foot.
2
6
Studying these tables, a number of interesting facts are
revealed. It is seen that when under moderate pressures
whale-oil is better than sperm, while as pressures rise the
sperm gains in value, finally excelling whale. This difference
will be found still more marked under very heavy pressures.
The mineral oils fall at the end of the list under pressures ex-
ceeding the lowest here given, although standing well under
the minimum. As will be seen elsewhere, these light oils
make excellent spindle-oils, and are good lubricants for such
low pressures as are met with in the working of textiles.
They vary enormously in quality, however, and the Author has
met with refined petroleums which fully equal sperm under
the heaviest pressures. This has since been observed by other
investigators. Olive-oil stands well under all pressures here
reported on, as do the other vegetable oils generally. Castor-
oil is too viscous for general use, however. Tallow and neat's-
foot oils are better at the lower than at the higher of these
pressures ; the reverse is the case with palm and cotton-seed
oils.
It is to be remembered that the order of standing just deter-
mined is liable to be changed by a change of velocity or of
temperature, and by alteration of pressure outside the range
here given.
It was found by Mr. Woodbury that the best neat's-foot oil,
FRICTION OF LUBK1CATED SURFACES.
283
used as a spindle-oil, absorbed 3.2 times as much power as the
best refined light petroleums. The mixed oils are sometimes
best for heavy machinery; unmixed refined petroleum of low
density is probably best for light machinery. The following
are the figures obtained by test at low pressure, moderate
speed, and standard temperature, the conditions being as nearly
as possible those met with in spinning-frames.
COEFFICIENTS OF FRICTION FOR SPINDLE-OILS.
Order of
Value.
OIL.
Coefficient of
Friction at 100
Degrees, F.
9
12
10
4
5
i
14
3
ii
8
2
6
13
15
Refii
Lard
Blea
Unb
Blea
Seal
Neal
0.1187
0.1233
0.1208
O.III3
0.1132
0.0756
o 2181
0.1067
O.I2I7
o 1170
0.0956
o 1147
o 1141
0.1608
0.2427
j Lbs. per square inch.
Pressure : j K ilos per square cm.
0.56
IO
9
25
i-75
150
10.5
200
14
250
17-5
275
12.3
300
21
500
35 o
Sperm
0.12
0.08
0041
0.056
0.0090
0.0136
0.0120
0.0096
0.0127
0.0095
0.0086
O.OIIO
0.0081
0.0091
0.0090
O.OIOO
0.0046
0.0059
o 0033
0.0044
Lard . . .
West Virginia
FRICTION OF LUBRICATED SURFACES.
297
Pressure
j Lbs.
4 5
o.2S 0.35
0.18 0.17
The experiments of Mr. Woodbury* give the method of
variation of the figures for still lower pressures, thus :
per sq. inch I 2 3
cm... 0.07 0.14 0.21
Values of/. 0.38 0.27 0.22
These values of the coefficient of friction of motion were
obtained on new surfaces at a temperature of 100 F. (38 C),
and at a velocity of 600 feet per minute. The surfaces were
probably not quite equal to those just described, or the lubri-
cant may not have been equally good ; the figures are consid-
erably higher.
Here it is seen that the figures are as widely different from
accepted values at high pressures as at low, but that the differ-
ence is upon the other side. At those pressures, therefore,
which are most used in heavy machinery the resistance of
friction is vastly less than we have been led to suppose, while
the friction of very light machinery is very much greater. The
fact that the journals here used were of steel, instead of iron
in the first case, does not modify these conclusions. Steel,
cast iron, and wrought-iron all give very nearly the same
figures up to their limits of pressure, when well worn.
The next table exhibits the results of experiment up to
still higher pressures, and with other journals and bearings:
COEFFICIENTS OF FRICTION, OF MOTION, AND OF REST.
(.) Cast Iron Journal and Steel Boxes.
B
o
1
B. W. Sperm. 1 West Virginia.
Lard.
?
i
I
I
I
si
tarting/'
Instant of
pping/".
fj
tartiug/
[nstant of
Pping/'.
IB
y
tarting/
[nstant of
Pping/".
Temperature in all cases
less than 115 Fahrenheit. Ve-
locity of rubbing, 150 feet per
minute.
I
M
1
8-
o r^
8 8
88
O O HI
o o o
si
Q H
CO CO
88
11
t^ oo r>
88S
il
I s -
o
8-
i i
1 1
r^. co O
883
l
8-
88
r^ r^
co co
8 8
O M HI
in O O
Q O
O O
a.
s|
-
CO CO
\n in
8 8
co \O
88
CN in m
M &
&
u gj 3
^
co O
8 8
8 8
T in N
888
$
8*
l|
I?
in in r-
8 8 8
00
^C O
co co
o o o
CO CO CO
3
1
&3
8 8
fO co
8 8
CO CO CO
8 8 8
a
V
1
I-
I 1
8 8
8 8 8
o
1
8,-
Tj- C
88
M in
T co
8 8
CO CO CO
CO CO CO
8 8 8
&
8j
1
O
co m
O
O
o O O
V
00
O
m in
W C>4
if?
3
g
1
S3
*3
in r^
CN CO
s 1
in in
r^- i^
8. 8
vO m co
8 8 8
a
1
ft
8-
O O
in M
8 o
"1" Tf
CO CO CO
888
s
5 1 S
O O
co oo in
CO W M
888
i;
o<
i-
8 S>
S 1 8
\O M
o o
1 If
||
fr
i^-a |
CO N
M M
O Q O
M M
4I.S
C/l
Jfls.
S
SS .8
T^ O*
in rf
CO 00 N
JY
Mineral-oil f oc pp
The apparent law thus varies with the character of the
lubricant, with variation of pressure, although usually giving
values of friction varying as the square root of the velocity.
The work of the Author, exhibited in Figs. 48, 49, illus-
trates the peculiar variation of friction with velocity of rubbing,
through a wide range of speeds, pressures, and temperatures.
These curves, which were constructed for the Author by the
late Mr. W. G. Cartwright, indicate the existence of a definite
law of variation of the coefficient, for each definite set of con-
ditions, taken as unvarying in other respects. At low speeds
the coefficient decreases, in all cases, with great rapidity ; passes
a minimum, usually at between 100 and 200 feet (30 and 61 m.)
per minute, and then gradually increases again up to the high-
est speeds attained.
For sperm-oil, the increase at 100 Ibs. per square inch (7 kgs.
per cm.) is very uniform in these experiments, and is very
3*4
FRICTION AND LOST WORK.
COEFFICIENTS OF FRICTION
.005 .010 .OT5 .OSO .025 .030 .035 .040 .045 .050
SPEED IN FT.
PER MIN,
400
FIG. 49. VELOCITY AND FRICTION.
FRICTION OF LUBRICATED SURFACES. 315
nearly proportional to the increase of speed, but is most rapid
at the lowest temperatures noted. The latter is the fact also
at higher pressures ; but less difference is usually observed with
change of temperature.
Heavy petroleum, as shown in the last of these figures, ex.
hibits the same general behavior at 100 and at 150 Ibs. (7 and
10 kgs. per sq. cm.) ; while at the lowest pressure, 50 Ibs. per
square inch (3.5 kgs. per sq. cm.), the action of varying temper-
ature becomes exaggerated to such an extent as to become very
plainly observable. It is seen that with such lubrication as was
here obtained the best temperature for this pressure is the
highest as usual, while at 90 the coefficients steadily increase
from the lowest speeds.
These curves are all established by too limited a set of ob-
servations to permit definite formulation of results, and those
presented must be received and used with caution until more
work is done and these laws are more completely ascertained.
As confirming the general deduction that the higher speeds
met with in machinery give reduced coefficients, it may be stated
that Mr. Pearce, of Cyfartha, reports less indicated power re-
quired to drive an unloaded rolling-mill engine at high speeds
than at lo\v.
140. Rest and Motion, not only as already stated, give
coefficients of friction differing greatly in value ; but experiment
indicates that they follow entirely different laws. The varia-
tions of both coefficients will probably prove to be influenced
by every change of condition of surface or of method of lubri-
cation, or of operation. Figs. 50, 51, 52, exhibit graphically
the results of experiments made on the testing-machine of the
Author with a wide range of pressure, and the comparison of
these coefficients when using sperm, lard, and mineral oils.
The temperature was in each case 115 F. (46 C.).
Under the conditions of surfaces and of lubrication by oil-
cups here adopted, the speed of rubbing being 150 feet (46
m.) per minute, the sperm-oil (Fig. 50) exhibits a minimum co-
efficient at 400 to 500 Ibs. per square inch (28 to 35 kgs. per sq.
cm.), while the coefficient for rest rises very rapidly as pressures
increase toward 100 Ibs. (7 kgs.), less rapidly to 500 Ibs. (35 kgs.),
FRICTION AND LOST WORK.
PRESSURE IN LBS. PER SQ. INCH
100 200 300 400 500 600 700 800 900 1000
COEFFICIENTS
OF FRICTION -
:::: :: :: ? ^?_ -_:_= = -.
:: ::::: ::: ^
^ _ _ _
__
.17 ---
.16- -
.__?__
1C
__^ k ! ) < ?
p fc --
14 -f^Pr
:
13 /
II
12 -J-
r
I I 1 I I i I 1 I III
I
jl 1 __.
OF REST AND OF MO1
'ION. - -
1
10 i - - -
:::::::::::::::::^fflB^ffi
i
1
C9 4-
jr
08 t -
III
III
07 -f-
06
i
05 . . ._ .
.04
.03 :::
,02
oiMJ_}4JJJ_M I III
::|:::!' = :-i55
- 15O FT. PER MIN. - ^
r _
, ''Ml LJ.J I I I I I I
FIG. 50. FRICTION OF REST AND MOTION.
FRICTION OF LUBRICATED SURFACES.
317
.20
COEFFICIENTS
OF FRICTION
.19
PRESSURES IN LBS. PER SO.. INCH
100 200 300 400 500 600 700 800 900 1000
FIG. 51. REST AND MOTION.
318
FRICTION AND LOST WORK.
PRESSURE IN UBS. PER S.O.. fNC.H.
100 200 300 400 500 600 700 800 900
1000
COEFFICIENTS --
OF FRICTION
-IK
.18
4
--->H
--/- -
._.._. .4 . .
::^3?:::
Jy|_.
J6
>yj * -
i::::::::g::::::::::::::::
.15
t'i _
s\
^ *
,14 --,-**--
js+5
COEFFICIENTS OF FRIG
TION -
- OF REST AND OF M01
'ION. - -
/
i _
.n--* *--
i : MI i ,| |
JO--- -
TEMP. 115 U FAHR.
AQ .
np . __
Q7 .
06
.05
04 _. ..__
03 -i- .
02 lmi ImN
X)l -i.. -.
! \^\\\
oEB
." * "
FIG. 52. REST AND MOTION.
FRICTION OF LUBRICATED SURFACES. 319
more rapidly again to 750 Ibs. (52,5 kgs.) ; while the last obser-
vation at 1000 Ibs. per square inch (703 kgs. per sq. cm.) gave a
lower figure, which however may have been an accidental and
exceptional departure from the general law.
Lard-oil (Fig. 51) exhibited the same behavior when in mo-
tion, passing the minimum at the same pressure, and having then
a little higher value. The coefficient for rest also varies at the
start in exactly the same manner, rapidly increasing with rise
of pressure up to 100 Ibs. per square inch (7 kgs. persq. cm.) as
before ; but it then decreases with rising pressures, passing the
maximum at about 150 Ibs. (10 kgs.), and a minimum at 500
Ibs. (35 kgs.), and rising to a second maximum at highest pres-
sures.
The general character of the curve is the same as that for
sperm-oil, but with the terminal portion depressed.
Heavy lubricating petroleum behaved (Fig. 52) very much
like sperm-oil passing the minimum on the moving journal ; at
a somewhat higher figure (750 Ibs. ; 53 kgs.) it gives exactly the
same form of curve of coefficients for rest that was obtained
with sperm ; and the lines for the two oils are almost identical
in location. It is thus evident that these peculiar curves are
not obtained by a merely accidental set of conditions for either
oil.
In these experiments the minimum coefficients for motion
were for sperm 0.004, f r lard-oil 0.005, anc ^ f r mineral oil the
same as lard. At the same pressures the coefficients for quies-
cence were 0.15, o.io, 0.15 for the three oils. Lard-oil permits
starting most easily, but it loses its superiority as soon as
motion begins.
These relations of value probably differ, however, with every
change of speed and temperature as well as of pressure.
141. Temperature modifies Friction to a very important
degree, as is seen by examining the tables already given, and
especially by studying the following values, which were ob-
tained by heating the bearing by its own friction to a maxi-
mum 170 Fahr. (77 C), well within that liable to produce al-
terations of the oil, and then noting the friction at successive
decreasing temperatures while cooling. It should be remem-
320 FRICTION AND LOST WORK.
bered that no temperature-readings can be taken as more than
approximate.
FRICTION AND TEMPERATURE.
Steel Journals. Lubricant, Sperm Oil. Velocity, 30 feet per minute.
square inch. Temperature, Fahr. Coefficient of Friction: f.
200 150 0.0500
200 140 0.0250
200 130 0.0160
2OO I2O O.OIIO
200 110 0.0100
200 100 0.0075
200 95 0.0060
200 90 0.0506
150 no 0.0035
ioo no 0.0025
50 no 0.0035
4 no o 0500
200 90 o . 0040
150 90 0.0025
ioo 90 0.0025
50 90 0.0035
4 90 0.0400
The figures just given would indicate that the sperm-oil
used in this instance, and under these conditions, including
that of exceptionally low speed, works best at lowest tempera-
tures, and that a heating journal gives rapidly increasing fric-
tion and rapidly increasing danger. At usual temperatures
90 to 110 F. (32 to 43 C.) the best pressure seems to have
been from ioo to 150 Ibs. on the square inch. The study
of the last table is exceedingly interesting and instructive.
There are there given coefficients of friction for temperatures
from 90 to 150 F., for pressures up to 200 Ibs. per square
inch, and for velocities of rubbing up to 1200 feet per minute.
It has been seen that at the low speed of 30 feet (9 m.) per
minute, the coefficient increases rapidly with increase of tem-
perature, and that at 200 Ibs. pressure (14 kgs.), an increase of
50 F. (10 C.) may increase its value to nearly ten times the
minimum, the rate of increase rapidly rising as pressures are
greater.
FRICTION OF LUBRICATED SURFACES. 321
It is now found, at speeds of 100 feet (31 m.) per minute.
that the friction does not vary between 90 and 150 F. (32
and 66 C C.), at pressures below 50 Ibs. per square inch (3.5 kgs.
per sq. cm.) ; but that it rises nearly 300 per cent, at a pres-
sure of 200 Ibs. (14 kgs.), over 100 per cent, at 150 Ibs. (i I kgs.),
and 33 per cent, at 100 (/ kgs. per sq. cm.).
At speeds exceeding 100 feet (31 m.) per minute, heating
the journal within this range of temperature decreases the re-
sistance due to friction, rapidly at first ; then, slowly and
gradually, a temperature is approached at which increase takes
place and progresses at a rapidly accelerating rate. It is seen
that this change of law takes place at a temperature of 1 20
F. (49 C.), and upward ; at all higher speeds the decrease con-
tinues until temperatures are attained exceeding those usually
permitted in machinery and very commonly not far from 150
F. (66 C), and sometimes up to 180 F. (82 C), or probably
even higher. The Author has found the decrease at 1200 feet
(37 m.) per minute to continue up to 175 F. (79 C.), at which
the value, at 200 Ibs. (14 kgs.) pressure, was, in the cases deter-
mined, 0.0050. The limit of decrease is reached under loo
Ibs. (7 kgs.) pressure, at 150 F. (66 C.), when running at this
high speed.
At 200 Ibs. (14 kgs. per sq. cm.) pressure, the temperature
of minimum friction for conditions here illustrated seems to be,
in Fahrenheit degrees, about
On either side this point on the thermometric scale it may
be assumed, for a narrow range, to vary, as the temperature de-
parts from that point, directly or inversely, as the case may
be, as the temperature. The coefficient of minimum friction
is found usually nearly constant over quite a wide range of
emperature.
Again, studying in this most instructive of these tables the
method of variation with pressure at higher temperatures, we
find the effect of change of pressure to be much more marked
at the higher temperatures at low speeds ; and we note, as
when studying the effect of variations of friction with change
122 FRICTION AND LOST WORK.
of temperature at a standard pressure as affected by variation
of speed, that we here find a change of law for the higher
speeds.
At a velocity of 1200 feet (37 m.) per minute, the coefficient
remains practically uniform with varying pressure at 150 F.
(66 C.), while below that temperature the friction coefficient
diminishes with increasing pressure. At velocities of rubbing
of 250 to 500 feet (75 to 150 m.) per minute the temperature
of the constant coefficient is about 100 F. (38 C.) ; at 100 feet
(31 m.) this peculiar condition is seen at about 120 F. (49 C.),
when extreme pressures (4 to 200 Ibs., 0.28 to 14 kgs.) are
compared, but the value is seen to be a little over one half as
much at 50 and 1 50 Ibs. (3.5 and 1 1 kgs.), and to become a mini-
mum 0.0019 at 100 Ibs. (7 kgs.) pressure; a similar behavior
is noted at the lowest speed observed 30 feet (9 m.) at about
125 F. (52 C.), and the same fall to a minimum occurs at the
intermediate pressure. It would seem that at all times there
is a tendency to an acceleration of outflow from the journal,
with increase of fluidity due to increasing temperature,
which tends to cause an increase of friction, while the effort
of capillarity to resist this outflow seems effectively aided by
increasing the velocity of rubbing. A balance between these
opposite influences is seen to take place at the slowest speed
when the pressure is somewhere below 4 Ibs. per square inch
(0.28 kgs. per sq. cm.) ; this occurs at a speed of 100 feet (36 m.)
per minute at a pressure of 50 Ibs. (3.5 kgs.), at 250 feet (77 m.)
when the pressure becomes about 150 Ibs. (u kgs.) probably ;
it happens at a speed of 500 feet (155 m.) at somewhere about
the same point ; and at 1200 feet (37 m.) per minute the bene-
fit of increased speed is sufficient to produce this balance when
the pressure exceeds 200 Ibs. per square inch (14 kgs. per
sq. cm.).
142. The Law of Variation of Friction with Tempera-
ture is evidently not a simple and definite one.
Studying all the results obtained, as above, it becomes evi-
dent that every pressure demands a certain degree of viscosity
and capillarity in the lubricant to secure at the same time
thorough lubrication and minimum friction. The effect of
FRICTION OF LUBRICATED SURFACES.
323
2UO
DN
190
180
170
icn
\
\
N
\
150
140
130
^
\ N
\
^
X
\
Sj-.
^
^o n
120
110
100
90
BO
\
^s
\
"N
X^
Ss
\
>
>v
s^^
\
X
\
^
' : ^
^^
N
\.
^
.S o
> ..So
200 rev.
09 ft.
>er min.
(6 4 m.)
> .So
BS
O *> v m OVD U H
O M ^ O^ f>\O V M
> n'O
M a
v 18
14
7
258
205
100
22
18
8
252
123
0.0056
o 0098
0.0125
0.0057
0.007
0.0146
0.0063
0.0077
0.0152
0.0068
0.0082
O.Ol63
0.0132
o 0144
0.0087
O.OI7I
0.0178
(9*7, Pad under Journal.
NOMINAL
LOAD.
ACTUAL
LOAD.
Tempera-
ture.
COEFFICIENTS OF FRICTION, FOR SPEEDS AS BELOW.
2Ls
v> C/3
IM
g
en .
8$.
Is
> E-.
Ed a a
8 o .S o
ids s
^s.S
o 0099
0.0105
o v- < ^-
00^*0
N 0.^
KB?
O OJ U. O>
u^vo 4^ t^
o * v- 10
III!
111!
24
22
2T
19
18
14
IO
7
328
310
293
275
258
205
100
40
35
32
19
12
582
551
520
458
364
272
178
C. F.
32 90
28 82
24 76
25 77
26 78
28 82
23 74
24 75
0.0107
0.0099
0.0105
0.0102
0.0092
0.0097
0.0098
0.0099
0.0097
O.OII2
O.OIO5
o 009
0.0099
0.0095
0.0087
0.0096
0.0109
0.0088
0.0085
O.OIO2
0.0122
o o^ci 4
o 0073
0.0105
0.0133
0.0082
0.0085
0.0119
0.0144
0.0083
O.OI
0.0125
0.0154
0.0102
0.0105
The following table illustrates the variation of friction with
alteration of temperature through a limited range. The re-
sistance decreases enormously, in this case, with a moderate
rise in temperature, becoming but one third the maximum.
FRICTION AND TEMPERATURE.
Bath of Lard Oil. Load, TOO Ibs. per sq. in (7 kgs ptr sq. cm.}
COEFFICIENTS OF FRICTION, FOR SPEEDS AS BELOW.
Tempera-
ture.
loo rev.
105 ft.
per min.
(30 m.)
150 rev.
i57 ft-,
per min.
(18 m.)
200 rev
209 ft.
per min.
(6 1 m.)
250 rev.
262 ft.
per min.
(79 m.)
300 rev.
314 ft.
per min.
(95 m.)
350 rev.
366 ft.
per min.
(113 m.)
400 rev.
419 ft.
per min.
(128 in.)
450 rev.
47i ft.
per min.
(143 m.)
C. F.
49 120
43 "0
38 100
32 90
27 80
2T 7.
0.0024
0.0026
0.0029
o 0034
0.004
o 0048
0.0029
0.0032
0.0037
0.0043
O.OOS2
0.0063
0.0035
0.0039
0.0045
0.0052
0.0063
0.008
o 004
o 0044
o 0051
0.006
o 0073
o 0044
0.005
0.0058
o.oo6g
o . 0083
0.0047
o 0055
0.0065
0.0077
o 0093
0.0051
0.0059
o 0071
0.0085
0102
0.0054
0.0064
0.0077
0.0093
O.OII2
i6 5 Co
o 0059
0.0084
0.0103
o 0119
0.013
0.014
0.0148
o 0156
FRICTIOtf OF LUBRICATED SURFACES. 337
The oil-bath used in these experiments by Mr. Tower is
not in common use, and cannot always be adopted when de-
sired. The conditions are not, therefore, those of usual prac-
tice ; but they may be taken as representative of conditions
toward which practice should be made to approximate as
closely as possible. It is seen that the mixed friction, here
met with, approaches more nearly fluid friction than is
usual.
Other experiments, reported later by Mr. Tower, exhibited
fluid pressures between journal and bearing rising to 625
Ibs. per square inch (43 kgs. per sq. cm.), and varying in
very nearly the same ratio from the centre-line of the crown
" brass," either way to the edge. The journal was found to
be thus completely " oil-borne" at speeds as low as 20 revolu-
tions per minute. The coefficient of friction at the latter
speed was found to vary nearly inversely as the pressure, ex:-.
hibiting a minimum at maximum nominal pressure, 443 Ihjs.
square inch (31 kgs. per sq. cm.), as follows:
COEFFICIENTS OF FRICTION.
Journal 4 inches diameter, 6 inches long. Revolutions per- minute, go; 1
per minute (61 /.) speed of rubbing ; 90 F. (32 C.), Mineral Oil.
o. 00139
0.00168
0.00247
0-00440
The experiments just summarized were all made at the
high pressures usual in heavy machinery. The accompanying
table of coefficients obtained by Woodbury at light pressures,
and of which the graphical representation has already been
given (142), are very complete, and are valuable as eomple^
mentary of the work of other engineers on heavy work. Thp
same general laws are here exhibited, and these values, \vjth
those already given, furnish a valuable set of data.
NOMIN
A*. LOAD.
Lbs. per sq. in.
Kgs. per sq. cm.
443
31
333
23
211
-5
8 9
6
338
FRICTION AND LOST WORK.
FRICTION OF PARAFFINE OIL.
Velocity of rubbing, yxz feet per minute.
Flash 342 Fahrenheit.
Fire 410 "
Evaporation by exposure to 140 Fahr. for twelve hours. 0.02
Specific gravity 0.888
PRESSURE
IN LBS
PKK SQ. IN.
TEMPERATURES.
40
43
50
55
60
65
70
75
80
8 5
90
95
100
COEFFICIENT OF FRICTION.
I
0.5380
0.4760
0.4260
o.382o'o.343o'o.3O2o o 2680
2383 O.2I2o'o. 1900 O 1700
0.15000.1383
2
3
0.2990
0.2107
0-1853
0.1660
0.1487
o'i 33 3 i o'.i20o
0.1083
. 0983^0 08800.08000.0733
o!o675 ! o!oS
4
o. 1670
o 1465
o 1310
5- "75
o 10600.0960
0.0870
.0795 0.0725 0.0665 0.0605
0.05 so 0.449 5
5
0.1400
o. 1232
o. 1104
3.0966
o 0900 0.0816
0.0740
.0676 0.0620 0.0592 o 0520
0.0476
0.04 ;6
6
0.1217
o. 1067
0.0960
o 0870
0.078710 0717
0.0653
.0597 0.0550 0.0503 0.0465
o 0427
o 0390
7
o 1089
o 0949
0.0847
o 0774
o. 0706! 0.0643
0.0583
.0540 o 0497 '0.0460 o 0423
0.0388
0.0360
8 0.0978
0.0858
0.0775
o 0705
0.0642
o 0585
0.0540
.0498 o. 045^0. 0423 0.0390
0.0359
0.0335
9 jo 0900
0.0791
0.0715
o 0651
o 0593 0.0544
0.0500
.04600 0427 o.0395;o. 0567
0.0340
0.0316
10
ii
0.0836
0.0782
0.0732
o 0687
o 0666
0.0624
0.0606
0.0571
0.05540.0508
0.0524 0.0482
0.0468
0.0445
4 H
0411
0.0402 0.0372 0.034-$
0.03^4 0.0356 0.0330
0.0324
o 0311
0.0302
o 0289
12
0-0735
0.0648
0.0592
0.0542
0.0498
0.0458
0.0423
0.59
o 036510.03400.0315
0.0297
0.0277
J 3
0.0695
0.0615
0.0561
o 0515
0.0474
0.0437
0.0405
375
0.0349 0.0328 0.0306
0.0285
0.0266
14
0.0663 o 586
0.0533
3.0491
0.0451
0.0419
0.0389
.0361
0.0337 0.0317 0.0296
0.0263
0.0259
15
0.0633
0.0561
0.0513
0.0475
0.0435
0.0403
0.0375
0.0349
0.0525
0.0305 O.O28o
0.0268
0.0257
16
0.0608 0.0540
o 0494
3-0455
0.0420
0.0390
0.0363
0.0338
0.0316
O.O295
0.0278
0.0261
0.0244
17
0.0582
0.0520
0.0477
3.0441
o 0407
0.0378
0.0353
0.0328
0.6308
0.0289
0.0272
0.0255
o 0240
18
0.0564
o 0504
o 0462
3 0426
0.0396
0.0364
0.0342
0.0321
o 0501
0.0282
0.0264
0.0250
0.0235
19
0.0545 o 0487
0.0448
3.0414
o 0384
0.0358
0.0335
0.0314
0.0295
0.0278
o 0262
o 0245 0.0233
20
0.0528
0.0473
0.0435
o 0405
0.0375 jo. 0349
0.0327
0.0507
0.0289
O.O273
0.0257
0.0241
o 0227
21
0.0510
o . 0460
0.0424
3.0394
0.056410.0342
o 0320
o 0302
o 0284
o 0268
0.0252
o 0238
o 0224
22
0.0496
o 0450
0.0414
0.038410.0358
0-0334
0.0314
o 0296
o . 0280
0.0264
0.0248
O.O234 O.O22O
23
o 0483 0.04 59
0.0404
0.0574
0.0350
0.0527
0.0308
0.0290
0.0274
0.0258
0.0244
0.0230 0.0216
24
0.0471
0.0436
0.0396
0.0568
0.0342 o 0320
0.0302
0.0285
0.026910 0254
o 0241
O.O229
0.0213
25
0.0460
0.0418
0.0386
o 0360
o 0336(0 0314
o 02 )6
0.0279
0.0265 0.0250
0.0236
0.0226
O.O2IO
26
o 0448
o . 0408
0.0378
0.0552
0.0328:0.0308
0.0290
o 0274
0.0260 o 0246
0.0233
O.O22I
O.O2O8
27
0.0439 0.0400
0.0370
0.0346 o 0322
3.0502
0.0286
0.0270
0.0256 0.0243
0.03300.0218 0.0206
28
o . 04 30 o 0392
0.0364
o 0340 0.0318
0.0298
0.0282
0.0266
o. 02:52
0.0240
0.0228 0.0216 0.0204
29
0.0421
o 0386
0.0358
0.0334 0.0313
o 0294
o 0277
0.0263
0.0250
o 0237
0.0225
O.O2I3 O.O2OI
3
0.0413
0.0378
0.0352
0.0328
0.0307
o 0289
0.0273
0.0259
0.0246 0.0234
0.0222 0210
0.0199
0.0404
0.0371
0.0347
0.0323 0.0304
0.0284
0.0268
0.0255
0.0243
0.0231
O.O2I9;O.O2O8 O.OI97
32
0.0397
o . 0364
o 0339
o 0318
o 0298
0.0281
0.0265
o 0252
o 0240 0.0228
02IOJO 0205
0.0195
33
0.0390 0.0358
0.0335
0.0313 0.0294
0.0277
o . 0262
0.0249
0.0237
0.0226
0.0214 0.0203 O - OI 93
34
0.0382
o. 03=53
0.0330
0.0309
o . 0290
0.0274
0.0260
0.0246
0.0235
0.0224
O.O2I3 O.O2O2
O.Oigi
35
o 0376
0.0347
0.0325
0.0304
o 0286
0.0270
0.0256
o 0243
0.0231
. 0220
0210 0.0200
O.OigO
36
0.0370 0.0542
o 0320
0.0300 0.0283
0.0267
0.0254
o 0244
0.0230
O.O2I9
o 0208 0.0198
0.0188
37
o 0364
0.0336
0.0315
0.0297
0.0279
o . 0264
0.0251
0.0239
0.0228
O.O2I7
o.o2o6|o.oi96
0.0186
38
0.0358
0.0332
0.0312
0.0293 0.0276
o 0262
0.0248 0.0235
0.0226
0.0215
o . 0205
0.0195
o 0185
39
0-0353
0.0328
o . 0308
0.0290
0.0274
o 0258
0.0246 0.0234
0.0223
O.O2I3
o . 0203
O.OI93
o 0183
40
0.03490.0323 0.0303
0.0289
0.0271
0.0256
0.0243
0.0232
O.O22I
O . O2 1 1
O.O2OI O Oigi
0.0181
The fact that the coefficient of friction varies greatly with
change of pressure is here exhibited with no less certainty. It
is also seen that the method of variation varies somewhat with
different lubricants, in some cases varying very nearly in-
versely with the intensity of pressure, and the total frictional
resistance remaining nearly constant within wide limits of
alteration of pressure. It is here found, as in the experiments
FRICTION OF LUBRICATED SURFACES. 339
of the Author, that the increase of speed raises the pressure
per unit of area attainable, and that the speed giving minimum
friction rises with increasing pressure.
The journals in the cases here cited were so arranged that
the pressure was unintermitted. It remains to be determined
how intermission of pressure modifies the laws affecting fric-
tion. It is only known, as yet, that it permits the use of
much higher pressures sometimes double those safely used in
the former case.
Some of the most important conclusions which have been
deduced from the later experiments described above were
anticipated by Mons. G. A. Him,* who found by experi-
ment, about 1855, tnat a lubricant gives least friction after
working some time; that friction is diminished by increase of
temperature ; that, under favorable conditions of lubrication,
friction increases in ordinary cases as velocity increases ; and
that the resistance is proportional to the square root of the
product of area and pressure ; i.e., the coefficient varies in-
versely as the square root of the pressure a conclusion later
confirmed by the Author.
144. Fluid Pressure and Friction are here controlling
conditions. The former evidently in some cases, as seen above,
more than mere capillarity, sustains the load, and holds the
two surfaces out of contact ; the latter produces the observed
resistance. The intensity of this pressure was found to be, in
experiments already cited, sometimes more than 200 Ibs.
per square inch (14 kgs. per sq. cm.) when the average load
on the journal was one half that amount. In cases such as
this, in which no oil-grooves are made in the bearing or in the
cap to which the oil-cup is attached, difficulty is often found
in securing a free feed of the oil. In nearly all cases the en-
gineer cuts small channels or " oil-grooves" from the oil-hole
across or diagonally, or in both directions, to the further por-
tions of the "brass," and thus succeeds in supplying them with
oil. Those 4< reservoir-boxes" in which the oil-bath is incor-
porated give the best adjustment of fluid-pressure.
* Introduction a la M6canique Industrielle; Poncelet.
340 FRICTION AND LQS7* WORK.
145, Conclusions.* Specified Qualities may, by the pro-
cesses here described, be secured by the identification by test
of a lubricant possessing such properties. If an unguent is
desired for heavy pressures, or an oil for very light work, or
for high or low speeds of rubbing under known pressures, the
methods of study of the available lubricants which have been
described will enable the engineer or the manufacturer to
select that which is best suited to the specified purpose. He
may go still further, and, by repeated mixing and test gradu-
ally improve the mixtures, may finally secure compounds
having the best possible qualities for the various proposed
applications. The Author has in this manner sometimes
produced lubricants for manufacturers which have been found
peculiarly well suited for special lines of trade.
Studying the facts here stated, and the data acquired by
many hundreds of other experiments, made on one or the other
of these last-described machines for testing lubricants, we may
recapitulate the facts and figures for ordinary use in machine-
design and in estimating losses of power by friction as follows :
(1) The great cause of variation with well-cared-for journals,
since they must work at ordinary temperatures, is alteration
of pressure and variation in methods of supply ; and it is seen;
that the higher pressures give the lowest percentages of loss
of power by friction.
(2) The value of the coefficient is greatly modified by the
state of the rubbing surfaces ; a single scratch has its effect in
wasting power. A good journal usually has its surface as
smooth and as absolutely uniform as a mirror. Every well-
kept journal acquires such a surface.
(3) For general purposes and for heavy work, as in the ex-
periments of the Author, and at considerable speeds, the value
of the coefficient varies nearly inversely as the square root
of the pressure, for pressures ranging from 50 to 500 Ibs. per
square inch.
(4) The coefficient for rest or starting may similarly be
* See Trans. Am. Inst. Mining Engineers, 1878; Journal Franklin Institute,
November, 1878.
FRICTION OF LUBRICATED SURFACES, 34-1
taken to vary nearly as the cube root of the pressure. For
closer estimates and other conditions, the tables just given can
be referred to directly.
(5) The coefficient for the instant of coming to rest, under
the special conditions here referred to, is nearly constant, and
may be taken at 0.03.
(6) The resistance due to friction varies with velocity, de-
creasing with increasing velocity rapidly at very low speeds, as
from i to 10 feet per second, and slowly as higher speeds are
reached, until the law changes and increase at ordinary tem-
peratures takes place, and at a low rate throughout the whole
range of usual velocities of rubbing met with in machinery.
Its amount and the law vary with method of lubrication,
however. With oil-bath lubrication the value of f usually
varies more nearly as the square root of the velocity.
(7) With pressure and velocity varying, we may take the
coefficient as varying as the fifth root of the velocity, divided
by the square root of the pressure for such work as is repre-
sented by the experiments of the Author.
(8) The effect of heating journals under conditions here
illustrated is, to increase the friction above 90 or 100 F., at a
speed as low as 30 to 100 feet per minute, while at higher
speeds and low pressures the opposite effect is produced, and
the coefficient often decreases more nearly as the square root
of the rise of temperature.
(9) The temperature of minimum friction, under the con-
ditions of the experiments here referred to, varies nearly as
the cube root of the velocity, for a pressure of about 200
Ibs. per square inch.
(10) The endurance of any lubricant should be determined
by actual wear upon a good journal under the pressures and
velocities proposed for its use.
The economy with which it can be used will be dependent
upon its natural method and rate of flow, and upon its capillary
qualities, as well as upon its intrinsic wearing power and the
method adopted in feeding it. Greases, therefore, are usually
more economical in cost than oils, even if having less wearing
capacity.
342 FRICTION AND LOST WORK.
(11) The only method of learning the true value of a lubri-
cant and its applicability in the arts is to place it under test,
determining its friction-reducing power, and its other valuable
qualities, not only at a standard pressure and velocity, and at
ordinary temperatures, but measuring its friction and endur-
ance as affected by changing temperatures, speeds, pressures,
and methods of application, throughout the whole range of
usual practice.
(12) The true value of an oil to the consumer is not pro-
portional simply to its friction reducing power and endurance,
under the conditions of his work ; but its value to him is
measured by the difference in value of power expended, when
using the different lubricants, less the difference in total cost
of oil or grease used; but for commercial purposes, no better
method of grading prices seems practicable than that which
makes their market value proportional to their endurance,
divided by their coefficients of friction.
The consumer will usually find it economical to use that
lubricant which is shown to be the best for his special case,
with little regard to price, and often finds real economy in
using the better material, gaining sufficient to repay excess in
the total cost very many times over.
(13) To secure maximum economy, the journal should be
subjected to a pressure the limit of which is determinable by
either Rankine's or Thurston's formula (Art. 127); the most
efficient materials should be chosen for the rubbing surfaces;
they should be reduced to the most perfect state of smoothness
and perfection in form and fit ; a lubricant should be chosen
which is best adapted for use under the precise conditions
assumed ; the lubricant should be supplied precisely as needed,
and by a method perfectly adapted to the special unguent
chosen. The real problem is often not what oil shall be used,
but how to secure most effective lubrication.
(14) The semi-fluid lubricants, when equally good reducers
of friction, are usually the most economical for heating jour-
nals, in consequence of their peculiar self-regulating flow, as
the rubbing parts warm or cool while working. They are
usually too viscous for economical use in ordinary work.
CHAPTER VIII.
THE FINANCE OF LOST WORK AND THE VALUATION OF
LUBRICANTS.
146. The Conditions affecting Values, both of the lost
work produced by friction and of the unguent used in reducing
its amount, have been already stated (Art. 5 1, Chap. III.) to in-
volve other and far more important considerations than the
market-price of the lubricant. The principles involved were
stated by the author in an earlier work ;* the treatment to be
here given is a more complete development of the subject.
Demand usually, if sufficient time is allowed for its operation,
brings prices into a correct relative order, but not necessarily
into a true proportion of values for any one specific applica-
tion. It is generally the fact that t4 the best is the cheapest"
to the consumer, and this rule is probably almost always appli-
cable in the purchase and use of lubricants. It is frequently
the fact that the consumer can better afford to use the highest-
priced article than to take those of lower value as a gift.
A very roughly approximate value by which to compare
the oils can be sometimes based on the assumption that they
will have a money-value proportionate to their durability and
to the inverse ratio of the value of the coefficient of friction.
Thus: Suppose two oils to run, one 10 minutes and the other
5, under a pressure of 100 Ibs. per square inch, and both at
the same speed, and suppose them to give on test for friction
the coefficients o.io and 0.06 respectively.
Their relative values might be taken at -J-J = I and = 0.833.
If the first is worth one dollar the second should be worth 83^
cents.
* Friction and Lubrication. R. H. Thurston, New York, Railroad Gazette
Pub. Co., 1879.
344- FRICTION AND LOST WORK,
In many cases, however, about the same quantity would
be applied by the oiler, whatever oil might be used, and their
values to the consumer would be taken in the inverse propor-
tion of the values of their coefficients of friction, i.e., as, in the
above case, 6 is to 10, thus making the value of the second
$i.66f, and showing that it would be better to use the latter
at anything less than this price than the first at one dollar.
Engineers have been accustomed to use these methods of
comparison in reporting 1 upon the values of lubricants simply
because they are generally considered to be correct by dealers
and users, and because there has been no better method sug-
gested of assigning an approximate figure for market price.
The real difference in values of any lubricants, to any user,
may, nevertheless, be determined in any given case when
the cost of power is exactly known, and when the quantity
of the several unguents required to do the same work has
been found, and their several coefficients of friction given.
The difference in actual value to the user, where any two
unguents are compared, is measured by the difference in the
costs of power and other expenses expended in driving the
machinery when lubricated first with the one and then with
the other of the two materials. As power is usually much
more expensive when developed in small, than when demanded
in large, amounts, the economy to be secured by adopting a
good lubricant is the greater as the magnitude of the work is
less. In large mills, and wherever work is done on a very large
scale, the cost per horse-power and per annum may be taken
roughly at about $50 a year, while for small powers this figure
is doubled or even trebled and quadrupled.
Every reduction of power to the extent of one horse-power,
by the introduction of an improved material or system of lubri-
cation, thus effects a saving of $50 to $100 a year; the differ-
ence between this amount and the extra cost of the new kind
of lubricant represents the annual profit made by the change.
Should it happen, as is sometimes the fact, that the better
unguent is also the cheaper, an additional profit is made which
is measured by that saving in cost.
In an ordinary small mill or in a machine-shop in \\ hich 100
THE FINANCE OF LOST WORK. 345
horse-power is used, a change in lubricant will often effect an
average saving of 5 horse-power and a consequent economy of,
probably, $500 a year. The total amount of oil used in such
a case might considerably exceed 100 gallons.
The consumer could in such a case better afford to pay $5,
or perhaps even more, per gallon for the good oil than accept
the less valuable lubricant as a gift.
In mills filled with light machinery, where the mean value
of the coefficient of friction is greater, and where a larger pro-
portion of the total power expended is used in overcoming the
friction of lubricated parts, a saving of 15 or 20 per cent, has
been made by the substitution of a good oil for a worse, i.e., a
gain of 75 to 100 horse power on 500, and of $3000 to $5000
per annum in power alone. In a case reported by Mr. Comly,*
a reduction of cost of oil on a single engine from 3.53 to 0.78
cents per hour was effected by the use of a slowly-flowing
grease instead of a freely-flowing oil. The cost of lubrication
of shafting was similarly reduced 44 per cent., but the loss by
increased friction was not noted. An instance is reported by
Mr. Woodburyf in which a gain of power of 33 per cent, was
effected by change of grease for a light oil, the loss in cost of
lubricant becoming comparatively unimportant ; in still another
instance the production of a mill was thus increased 5 per
cent., while also greatly reducing the lost work of friction.
This subject is of such importance, and has as yet received
so little attention, that it has been considered advisable to de-
vote a chapter to its development.
The differences in value of good oils, and the enormous
wastes of power, and of other costs, with unguents of poor
quality, are easily exhibited. Assuming the cost of a good oil
at $i per horse-power per annum, in any case, a variation of
one per cent, in the coefficient of friction produced by a change
of oil will produce a gain or loss of from 50 to 100 per cent,
of the total cost of oil used in the shop or mill, and of
other costs of power accordingly as the mean coefficient is
high, as in cotton and other mills filled with light mechanism,
* Trans. Am. Soc. M. E., 1884. f Ibid.
34-6 FRICTION AND LOST WORK.
or low as in the locomotive engine and other heavy machinery*.
The use of good instead of bad, or of an oil with low " cold-
test" in winter instead of one easily stiffened by low tempera-
ture, may enable an engine to haul two or three additional
cars in a train, or a mill to be driven easily and economically,
where otherwise it could not be driven, if at all, by an engine
of proper proportions except very wastefully.
The use of a poor quality of cylinder-oil will sometimes
cause losses by increased friction of engine, and even on loco-
motives by breakage of rods and rock-shafts, sufficient to com-
pensate many times over the gain in money cost of oil. Under
heavy pressures, also, the cost of wear and tear of journals and
bearings may become a serious item.
All lubricants should be purchased with careful regard to
their value, rather than by reference mainly to their price.
Their value is determined principally by their friction-reducing
power, and their reduction of wear of rubbing parts. Unguents
of low grade cause losses, direct and indirect, which are out of
all proportion to their low cost, and may invariably be expected
to produce such losses by waste of power, by injury to jour-
nals and bearings, and by destruction of valuable machinery,
to say nothing of the dangers of fire which often accompany
their introduction, that the user can generally better afford to
pay many times their value for the privilege of declining to
use them, than to submit to the enormous losses sure to follow
their application to his machinery. In every case the lubri-
cant should be carefully selected for the special use intended.
147. The Defects in the Usual Methods of valuation of
lost work and of lubricants are readily seen to arise from the
fact that they include simply a comparison of the market-
price of available kinds and qualities with their endurance and
friction-reducing power. It is usually assumed that, of two oils
having endurance and friction-coefficients in the inverse ratio
of their prices, the purchaser may take either with practically
equally good financial result. No comparison is usually made
of the relative costs of wasted power and of total expense for
oil. This system is obviously entirely wrong, as is every
method which does not take into account every item of profit
THE FINANCE OF LOST WORK. 347
and loss variable with change in quality and quantity of lubri-
cant, and which does not make up an account including all
these items. The real question is not whether the difference
in price of any two oils is justified by the difference in their
intrinsic qualities, but whether the profit or loss to be made by
the substitution of one for the other is compensated by the
total loss or gain in expense.
148. An Exact Method of valuation of lost work and of
lubricants must include a determination of the intrinsic quali-
ties of the latter, tjieir influence upon the magnitude of the
former, and of the money-value of every item of gain and loss
in the purchase of the lubricants, in the variation of the quan-
tity of power used, and in all incidental expenses, such as wear
and repairs, taxes, insurance, rents, availability of the property,
and many other items that may be usually determined in any
given case. An expression must be obtained for the total of
all these costs of wasted power and of lubricant for the actual
and for the proposed case, and a comparison of the amounts
so determined will indicate the magnitude of the gain or loss
to be produced by the proposed change.
149. The Theory of the Finance of Lost Work includes
a comparison of economy in the use of various lubricants, which
is evidently not that of the relative cost of operation with and
without lubricants, but of the relative total costs of working
with two or more available unguents. The costs include the
expense of the lubricant and of repairs, and the value of the
work wasted by friction in the several cases.
If the cost of the lubricant per unit of quantity is k, and
if the quantity used in the assumed time be q, the cost of the
lubricant is kq. If the amount of work lost by friction in the
given time be U, and if its total cost be k f per unit of work,
and for the assumed time, the expense chargeable to lost work
is k'U\ while the total expense due to friction of the apparatus
is, neglecting other expenses as unimportant,
...... (i)
But the work is
U = a/PS = afPVt, ..... (2)
34^ FRICTION AND LOST WORK.
the product of the coefficient of friction,/", the total load, P,
the mean velocity of rubbing, V, the time, /, and a constant,**,
dependent upon the relations assumed for space and time ;
hence,
..... (3)
For any given cases taken for comparison, the only vari-
ables in the second member of the above equation are q and/",
and, making ak'PS =&,
(4)
in which b is determinable for each case of comparison. That
lubricant which gives the least value of K is best. The true
value of a proposed oil will vary as
The above equations show that the value of the lubricant is
inversely as the quantity required, and, when the cost of un-
guent is small in comparison with the value of the lost work or
wasted power, its commercial value, which varies with the de-
crease effected in K, is directly as some function of its lubri-
cating power, i.e., nearly as the reciprocal of the coefficient of
friction. If the cost of oil is large, the comparison becomes
one of the expense for lubricants.
Two oils being compared, the costs of lost work are, re-
spectively,
and the saving effected by the substitution of a better lubri-
cant is
K,-K, = k tq> -k,/, (I - A)]. . . (12)
i l\
A higher cost causes loss, a lower is a gain ; this value of k
being that which the buyer can pay for the lubricant in place,
on the journal, without losing by the change.
It is obvious that b may be expressed in any units of cost
that may be convenient, as on railroads, in repairs, fuel, or other
material expended per train-mile. Thus on railroads the ex-
penses of hauling trains are measured by the costs of oil, re-
pairs, and of power per train-mile, and
(13)
* See Friction and Lubrication; also, Encyclopedia Britannica, art. "Lubri-
cants."
THE FINANCE OF LOST WORK. 351
in which q is the quantity of oil used and df is the cost of
power and attendant expenses per train-mile. This makes
the criterion
k-k = d
x * Q
Where, as may often occur, the reduction of friction is ac-
companied by increased expenses on account of wear of journals
and bearings, a third term must be introduced and the varia-
tion of the total thus obtained noted. For ordinary pressures,
in well-designed mechanism, the last item may probably be
neglected ; but in some cases, as in transportation on railway ,
it may become, and probably often is, a very serious item of
expense, and must be taken into account.
150. Data required in Applying the Theory, although
usually obtainable with satisfactory exactness in any given
case, are not sufficiently uniform to permit their statement in
figures for general use.
The total expense chargeable to lost work in machinery
consists of the following items:
(1) Cost of power produced, only to be wasted, including
all items of cost in the motive-power department.
(2) Expense incurred by " wear and tear" of the driven
machinery and its repair and replacement.
(3) Indirect, casual, and remote money-losses due to in-
efficiency caused by friction and by wear.
(4) Cost of lubricants and of their application.
The first item includes all running expenses of die motor,
including fuel and supplies, interest on invested capital, wages,
insurance, and taxes on the engine, boilers, and buildings
covering them. The second, which is a large item, includes
the replacement of worn bearings and journals, and parts in-
cluded in their depreciation, sometimes the latter involving
finally the whole machine. In fact this is the usual limit of
the life of the machine. The third item cannot be calculated,
since it includes accidents, but it may usually be covered, like
352 FRICTION AND LOST WORK.
other casualties, by a system of insurance. The fourth item is
the least important of all. It includes the purchase of the
lubricant, its transportation, and the expense of its application
and removal and of keeping the bearings clean. Although the
smallest of these expenses, this is most obvious to the con-
sumer, and is wrongly allowed to determine, usually, the selec-
tion of the unguent. A change of lubricant usually effects
enormous changes in the magnitudes of the first three items,
and comparatively insignificant alterations of cost in the last.
As the total resistance is composed partly of friction of fluids,
and partly of that of solids, some lubricants are found to give
reduced resistance, while nevertheless increasing wear inordi-
nately. In such cases, the lubricant is found to have too small
viscosity, and the decreased fluid resistance, although not com-
pensated by increase of solid friction, is more than counter-
balanced in the expense account by cost of increased wear.
151. The Units of Measurement to be adopted in the
commercial theory of lost work will be determined by circum-
stances. As a rule, the cost of power is measured in dollars or
cents per horse-power, or per foot-pound, per hour of working
time, which is usually about three thousand hours per annum.
The usual charge for the horse-power in New York City, for
example, in small amounts, is $100 per annum, equivalent to
$0.033 P er hour. The cost of wear and tear and of deprecia-
tion is very variable, but can be best estimated as a percentage
of the value of the machinery; 2\ per cent, for renewals and
something more for minor repairs is a common figure. All taxes
and insurances are reckoned by a similar method. The cost of
lubricants may be reckoned from the quantity used per hour.
All expenses being thus reduced to one measure money-
cost it becomes easy to solve any problem of this kind aris-
ing in practice when the requisite data are obtainable.
The costs are thus made to appear finally as two items
the one the cost of the lubricant, and the other that of the
wasted power which are regarded as independent variables,
although evidently dependent according to some law which
may possibly be sometimes easily expressed. The data re-
quired are often exceedingly difficult of determination, and
THE FINANCE OF LOST WORK. 353
approximate results only can be reached. This is especially
true of cost of wear and repairs.
152. The Values of Quantities entering the preceding
theory are often ascertainable : they are mainly costs of
power, of oils, and of depreciation. The cost of power will
vary according to amount, efficiency of engine, costs of wages,
fuel, and minor items, from $40 per annum, or $0.013 per
hour, to $200 per year, or $0.07 per hour, nearly : the higher
figures being for very small, and the lower costs for large and
economical condensing engines, with cheap fuel and labor.
The mean may be assumed as $60, or $0.02 per hour, for good
non-condensing, stationary engines of 100 to 200 horse-po'ver.
This annual expense is divided, in some cases noted by the
Author, thus :
Total. Coal and Oil. Wages. Minor Costs.
Small engines $200 $50 $100 $50
Medium " 60 25 25 10
Large " 40 20 10 10
In marine work, the cost of fuel often becomes a larger per-
centage of the total ; perhaps 60 to 80 per cent, may be con-
sidered a common allowance.
The power demanded for overcoming friction of engine
and shafting of mills may be taken at from 0.20 of the total on
heavy work, to 0.30 on light, the total power ranging from lo
to 20 horse-power, averaging 15, per 1000 spindles and "pre-
paration."
The cost of oils in the market has no direct relation to
their values as lubricants, and is not infrequently in the inverse
order, the best costing least, and the most expensive having
a comparatively low position as unguents for the specific pur-
pose considered. Taking them as they come, however, the
following may, for purposes of illustration, be assumed to be
fair relative values :
Sperm-oil, per gallon $i 10
Neat's foot oil, per gallon I oo
Lard oil, " " o 70
Tallow-oil, " " 070
Olive-oil, " " 090
354 FRICTION AND LOST WORK.
Cotton-seed oil, per gallon . o 50
Greases, per pound , o 25
Mineral oil, heavy and fine o 80
" fair 050
4 " light 040
4 " spindle, light . 030
44 natural W. Va 025 >
4 <4 kerosene o 10
The quantity used will vary greatly with its use and the
method of application. Cotton -mills use from 10 to 30
gallons per 10,000 Ibs. of cloth made, or about 10 gallons
per annum per horse-power, at a cost averaging $0.70 to $1.00
per gallon. A mill of 60,000 spindles, making 3,000,000
Ibs. of cloth per year, and demanding 1 200 horse -power,
uses about $2000 worth of oil. The cost of replacement of
wearing parts is small. Railway-engines use 0.005 to o.oi
gallon per " train-mile/* and 40 to 60 Ibs. of coal. Cylinder
oils are used in the proportion of from 2OO to 600 miles run
per gallon.
The ordinary passenger locomotive on New England rail-
roads averages an expenditure of between 60 and 70 Ibs, of coal
per mile, at a cost of not far from 15 cents; while an expense of
one half Cent per mile for oil and tallow is considered a good
showing. A run of 30 miles per ton of coal and of 100 miles
per gallon of oil is not an unusual figure on Western roads. The
cost of fuel is often about one third the total cost per mile ;
that of oil about two or three per cent of the total. Two or
three times as much oil is used under a passenger car as under
a freight car. The cost of repairs is enormously variable. It
has been found in some cases of good practice that a pound of
bearing and a pound of journal are worn away by, respectively,
twenty-five thousand and seventy-five thousand miles of travel.
But the cost of this form of depreciation alone is enormously
greater than the mere cost of material per; pound. Using a
black oil, the cost of wear has been found five times that of the
lubricant and twice that of power.
A large machine-shop is reported to have used one thou-
sand tons, of coal per annum for all purposes, including heating,
to demand. 120 horse-power from, its .engines, and to use 450
THE FINANCE OF LOST WORK. 355
gallons of oil, the cost being $6500 for coal and $250 for oil.
Another moderately large shop uses but 60 gallons of oil per
year, or about 0.02 gallons per hour of working time. The
cost of wear should be insignificant.
153. Illustrations of Application may be taken as below :
Calling the total value of the horse-power $100 per annum,
or $0.03 per hour, the value of b will be found as a function of,
k'afPS. The value of V will be
k' = ' 3
1,980,000*
if a is taken as unity, i.e., one hour, and
b 0.000,000,01 5 fPS.
Assume PS 4,000,000,000 a fair figure for an iron-work-
ing establishment wasting 100 horse-power in friction. Then
b $60 = 0.6 H. P. ; and if in equation (4)f= 0.05, k v = $0.50,
and q l = 0.02 gallon per hour,
KV = ,?, + i = 0.01 + 3.00 $3.01.
Assume , = $0.25 ; q^ = 0.03 ; /= 0.06 ; then
K, = k& + bf % = 0.0075 + 3.60 = $3.6of ;
K^ K^ = $0.60 nearly.
The cost of lost power is increased 20 per cent, and $0.60
per hour is lost by a saving of one quarter of a cent per hour
in cost of lubricant by the substitution of an oil giving a coeffi-
cient of 6 per cent., and demanding one half more oil for a
lubricant giving a mean coefficient of 5 per cent. The saving
in cost of oil is insignificant; the loss in cost of power is com-
paratively enormous ; although the difference in the coefficient
is but one per cent.
If by freer supply of the cheaper oil, as by the oil-bath, the
356 FRICTION AND LOST WORK.
value of /, can be reduced, as is not unlikely, to/j = O.O2, if
, = 0.40 and k = 0.25, we get
^T a = 0.10+ 1.20 = $1.30;
K t - K\ = $2.30; k,q, - k\q\ = $0.0925 ;
and the expenditure of nine cents per hour for additional oil
produces per hour a gain of $2.30, i.e., a profit of about 2500
per cent.
If one oil gives a mean coefficient of friction, /^ = 0.05 and
another /j = 0.06, using 0.02 gallon per hour of each, the real
value of the latter becomes (Eq. 7)
_ o.Q i +60(0.0$ -0.06) _
^ ~ ~~02~, O., December 23. 1884.
GENTLEMEN' : In reply to your request I will state that I have used your Capital (. yin.cier
Oil for the last year on a fo.npoutui engine, cylinders thirty and titty-six inches in diameter,
and it give entire satisfaction, and I can cheerfully recommend it as the best oil I have ever
used. Respectfully, J. RIGG, Chief Engineer Steamship Wo Co Ken.
December, 1884.
DEAR SIR: The Capitol Cylinder Oil I have used for the last three years, and have stcuied
better results from it than from any cylinder oil I have ever used.
W. H. SEEMAN, Chief Engineer Steamship A. Everett.
CLEVELAND, O., December 5. 1884.
I have used your Capitol Cylinder Oil for three vears, and in that lime I have used, or
rather tried to use, several other brands of oil, and never found any to come anywhere near
to the Capitol Cylinder Oil. I: keeps the cylinder and rings always clean and free from gum.
I have used it with a pressure of 60 to 140 Ibs. of steam and it m-vi-r failed to do its work with
me. It is the best cylinder oil manufactured. Respectfully.
J. B. MILLKK, Cliui i^ii^ineer Barge Business.
December, 1884.
I have used the Capitol Cylinder Oil for five years, and find it to be a splendid lubricant
on both compound and high-pressure engines. Respectfully,
W. S. SEMPLB, Engineer Steamer H. L. Worthington.
Eldorado Engine Oil.
Prof. Thurston's Report of Eldorado Engine Oil.
STHVENS' INSTITUTE OF TECHNOLOGY, HPBOKEN, N. J , Feb. 9. 1883.
A comparison of the results obtained from tests of ELDORADO with those obtained at
the same time from "Standard Laboratory Lard Oil " leads to the following conclusions:
With a free feed and a pressure of 100 Ibs. per square inch and a speed of 250 revolutions
of the test-journals, the minimum coefficient of friction was about six-tenths of one per cent, for
ELDORADO ; the minimum coefficient of friction for lard-oil was seventy-three one-hundred ths
of one per cent, (the average being eighty-two one-hundred ths of one per cent.)
Tke oil is therefore superior to lard oil for reducing friction ; reducing the friction b-
scrv^d f.ir lard-oil about twenty per cent.
When a weighed amount (eight milligrams, about one drop) of each oil was placed on the
test-j mrnal an.i the machine sinned and run. as in the case of a free feed, the number of revo-
liKi'.ns made by ihe test-journal before ihe oil ceases lo lubricate or wears oul, will give wh;it
is k:v>wn as our "endurance-test." The coefficient of friction is, of course, larger in tl.is
case than with a free feed or the " friction-test."
The record shows thai the lard-oil endured through lo.ooo revolutions, while ELDORADO
continued to lubricate up to ab f ratrd as 20 f>er cent superior in reducing friction, and 30 per cent
more enduring tkan f>ur t - lil. anil during the " free-feed " test we used less oil.
We find, on referring to the similar test m.ide last summer, that the results obtained then
are practically the same as now. R. H. I HUKSTON, Director.
ELDORADO ENGINE OIL. M \NUFACTURED BY
The American Lubricating- Oil Company, Cleveland, Ohio.
VI
THE BEST IS ALWAYS THE CHEAPEST."
For Nineteen Years
VACUUM OILS
Through superiority of process and care in manufacture, have led
all others, affording
Perfect Lubrication at Lot est Cost.
THEY 8 A VE MONEY,
SAVE POWER,
SAVE FUEL,
SAVE MACHINERY.
Send for new Pamphlet Circular with Autograph Testimonials.
VACUUM OIL CO., Rochester, N. Y.
51 Purchase St., Boston ; 96 Water St., New York ; 70 St, Peter St., Montreal,
THE DAVIS OIL COMPANY,
MANUFACTURERS OF
LARDIEENEATSFOOT OILS,
SOLE MANUFACTURERS AND PROPRIETORS OF
WELL-KNOWN BRAND,
"DAVIS' REFINED OILS."
Our Oils are especially adapted for Compound-
ing purposes and Export.
PURl'lY and QUALITY GUARANTEED.
Works : Brooklyn, N, Y, Office : 37* Water Street, New York,
CHAS. W. HAND, Manager.
Vll
D. A. STUART & COMPANY,
MANUFACTURERS OF
Oils $ Lubricating Compounds,
CHICAGO, ILL.
Send for Descriptive Price List.
Dealers who furnish Consumers with Lubricants are
invited to send for sample, etc., of
STUART'S CASTOR GREASE.
JAS. H. PLATT, Pres't. G. C. THORP, SecV.
ANDREW WASHBURN, Vice-Pres't. G. S. RICHARDS, Treas.
NEW YORK REFINING CO.,
141 Maiden Lane, New York.
FACTORY: Newtown Creek, Blissville, L /.
MANUFACTURERS OF
Pure Mineral Lubricating Oils 9
CARBOLINE CYLINDER OILS,
CARBOLINE MACHINERY OILS,
CARBOLINE ENGINE OILS,
PURE PARAFFINE OILS,
BLACK OILS 9 SIGNAL OILS,
STANDARD LUBRICANTS,
PURE MINERAL GREASES.
For all kinds of Machinery.
AXLE GREASE trnd CARRIAGE LUBRICANT.
Owners of Machinery wi 1 find it to their advantage to communicate directly
with us in the sehct'on and purchase of their Lubricants.
SPECIAL CONTRACTS WITH LARGE CONSUMERS.
Vlll
T ZE3I IE
PRATT & WHITNEY CO.,
HARTFORD, Conn., U. S. A.,
MANUFACTURERS OF
MACHINE TOOLS
For RAILWAY and GENERAL MACHINE SHOP SERVICE,
-AND-
Special Machinery.
MACHINERY & SPECIAL TOOLS
FOR
Armories, Sewing Machine and Agricultural
Implement Manufacturers,
ilOLL GROOVING MACHINES
For Fluting CJiilled Rolls for Flouring 31 ills.
U. S. STANDARD TAPS and DIES, REAMERS, GAUGES,
FIXTURES, and ALL TOOLS necessary for
INTERCHANGEABLE WORK.
ALSO, MANUFACTURERS OF PROF. R. H. THUPSTON'S
R. R. Standard Lubricant Testing Machine.
SEE FRONTISPIECE
Illustrated Catalogue and Pries Lists furnished on application.
IX
The Cleveland Refining Go.
Works and Office, Bessemer Ave. and C. & P. R. R., Cleveland, 0.
RKKINERS OR
PETROLEUM AND ITS PRODUCTS.
MANUFACTURERS OK
Lubricating Oils, Naphtha and Gasoline.
Water White Diamond Light Carbon Oil a Specialty.
RAILROADS WILL PLEASE NOTICE
That we make a Specialty of soliciting their Trade for either
Lubricating or Burning Oils.
THE NEW AND COMPLETE CATALOGUE
OF THE
[PUBLICATIONS
OF
JOHN WILEY & SONS
15 ASTOR PLACE, NEW YORK,
CONTAINING
Scientific and other Text-Books for Colleges, Industrial Schools and Theological
Seminaries, with many Valuable Practical Works for Architects,
Engineers, Mechanics, etc.,
Including Works on Agriculture, Assaying. Astronomy, Book-keeping. Chemistry,
Drawing and Painting. Electricity, Engineering. Metallurgy, Machinery,
Mechanics, Mineralogy, Seamanship, Steam-Engine,
Ventilation, etc., etc.,
And for Theological Sem'.naries. Hebrew and Greek Bibles. Testaments, Lexi-
cons, Grammars, Reading-Books and Concordances.
Also a full List tf their Editions of
JOHN RUSKIN'S WORKS.
Will be sent free by mail to any one orderinsr it.
X
OCEAN OIL Co. THE TIDE WATER PIPE Co., LIMITED. THE CHESTER OH. Co.
POLAR OlCOIPAIY
Cor. Broadway and Beaver Sts., NEW YORK.
205 La Salle St., CHICAGO, ILL.
MANUFACTURERS OF FINE GRADES OF
MACHINERY,
ENGINE, AND
CAR
OIXL.S
AND
REFINED WAX.
"V.SCO" MACH.NERY,
"REVOLUTION" ENGINE,
" POLAR VALVE,"
"ELECTRIC "^SIGNAL,
Correspondence solicited. Samples on application.
THE CHESTER OIL CO.,
S. E. Cor. Third and Walnut Sts., No. 205 La Salle Street,
PHILADELPHIA, PA. CHICAGO, ILL.
REFINERS OF PETROLEUM.
MANUFACTURERS OF
Illuminating Oils, Naphthas and
Lubricating Oils.
PHILADELPHIA AGENTS FOR THE POLAR OIL CO.
OCEAN OIL CO,,
NEW YORK.
THE TIDE WATER PIPE CO., Limited,
TITUSVILLE, PA., and NEW YORK.
POLAR OIL CO.,
NEW YORK.
14 DAY USE
RETURN TO DESK FROM WHICH BORROWED
LOAN DEPT.
This book is due on the last date stamped below, or
on the date to which renewed.
Renewed books are subject to immediate recall.
16 191)0
LD 21A-50m-9,'58
(6889slO)476B
General Library
University of California
Berkeley