possible section, and the provision of separate nut and tool
FIG. 68. Nut carriage
'^. -L^UU LC1.J. llCL^l^. JL J.*JT WU. J.
Section of the precision lathe shown in Fig
FIG. 69. Tool carriage
Fig. 67.
carriages connected by massive rods in direct tension only,
whereby the lead screw and the screw in process of being cut are
placed in line with one another. As originally made the ma-
chine was provided with an independent driving head and an
equalizing driver of which portions are shown in Fig. 67, whereby
96 METHODS OF MACHINE SHOP WORK
the pull of the belt on the machine was eliminated. As now
used, however, the lathe is driven by hand. Its only use is to
finish screws which, except for the correcting cut, are made on
other lathes.
The correcting bar is plainly seen in Fig. 67 while Fig. 68,
which is a section through the nut carriage, shows the same
bar at a and in addition the tubular bed. Fig. 69 is a section
of the tool carriage and shows the support of the screw being
cut and the fact that the cutting tool is inverted from the usual
position, this arrangement being adopted because believed to
be more conducive to the avoidance of chatter. Screws
have been cut with this lathe of which the errors are less than
the ten thousandth of an inch in twelve inches of length. It
represents the culmination of the solution of the fascinating
precision screw problem which has engaged the attention of
many mechanics from Maudsley down, including Whitworth.
CHAPTER IV
THE MEASUREMENT OF ERRORS
Instruments for measuring errors embodying the multiplying lever
Uses of these instruments The dial gage and its uses The measure-
ment of errors with extemporized apparatus.
In a large number of cases of measurement the thing measured
is the error from truth. This error may be the error from
truth of size, in which case the process of measurement is one
of comparison with a standard, the measuring instrument giv-
ing the difference between the standard and the piece which
is compared with it. In other cases the error measured is one
of position or adjustment, in which, for example, the degree
of parallelism or of squareness of one piece with another is
determined. For both these classes of measurement instru-
ments embodying the multiplying lever possess advantages over
the micrometer because less time is required for their use and
the personal equation is eliminated.
APPLICATIONS OF THE MULTIPLYING LEVER
An application of the multiplying lever to the comparison
of parts with a standard is shown in Figs. 70 and 71 from
the Canadian Ingersoll-Rand Company. The standard with
which the comparison is made may be a gage or, as in this case,
a sample piece preserved as a standard. The illustrations
show an indicating beam caliper made by the Syracuse
Twist Drill Company in the act of gaging a rock drill slide valve,
Fig. 70 showing the instrument as it actually appears and Fig.
71 with a protecting cover removed in order to show the con-
struction. The stationary head of the instrument carries a
multiplying lever which plays over a graduated scale at the
top, the graduations reading to thousandths. The short end
of the lever is connected with the measuring finger, which has a
7 97
98
METHODS OF MACHINE SHOP WORK
FIG. 70. With cover plate in position.
FIG. 71. With cover plate removed.
Indicating beam caliper.
THE MEASUREMENT OF ERRORS 99
slight endwise movement. The sliding head carries a second
finger which is threaded and fitted with a knurled head. The
standard an end-measure rod or a sample piece preserved
as a standard, as in the present instance is placed in position
and the finger upon the moving head is adjusted until the
multiplying lever stands at the zero of the scale, this zero be-
ing at the center of the scale in order to read errors in both
directions. This being done, the parts as made are placed in
position and their correctness or their errors are determined
at once.
This is the cheapest method known to the author of intro-
ducing the limit gage system. The limits are tabulated and the
inspector has only to compare them with the readings of the
instrument in order to determine if the parts are within the
limits. The instrument has received far less application than
it deserves and this is the more strange because of the wide
application of the multiplying lever to other uses. It will be
observed that the instrument is in no sense a measuring instru-
ment. It is a comparator, that is to say, it compares parts and
determines their differences but it does not determine their
absolute sizes.
THE TOOL-MAKER'S INDICATOR
The widest application of the multiplying lever is to the tool-
maker's indicator, one of which, by Koch & Son, is shown in
Fig. 72, of which the lower view shows the sliding cover removed
and the multiplying levers exposed. Such indicators appear
in a great variety of forms as they are frequently home made.
The present instrument is unusual in that it is fitted with
compound levers, thereby making it extremely compact.
These levers are enclosed in a steel box which has at each end
a projecting finger which engages with the short end of the
lever system. The finger at the left slides endwise while the
one at the right is a bell crank, its external movement being
vertical in the position shown. Great flexibility of adjustment
is thus made possible, including internal readings by entering
the bell crank end into holes to be indicated.
The applications of this instrument are almost endless.
100 METHODS OF MACHINE SHOP WORK
Perhaps the most common is to the centering of work in the
lathe, as illustrated in Fig. 73, in which the vibration of the
lever as the lathe revolves shows the amount of untruth of the
piece of work. Fig. 74 shows the adjustment of a milling
machine vise at right angles with the cutter arbor to which
latter the indicator is clamped by a series of collars and a bind-
ing nut. By traversing the work table the indications of the
instrument indicate any lack of truth of the vise jaws. Fig.
75 shows the instrument used for transferring a measurement
to an otherwise inaccessible place. The height from the sur-
FIG. 72. Tool-maker's indicator.
%
face plate is taken in the indicator from the height gage, when
the indicator is lifted over the jig body into which the measur-
ing finger enters. The position of a point within the jig may
then be compared with the reading of the height gage. Fig.
76 shows the application of an instrument of different pattern
to the adjustment of a swivel ed angle plate. The small angle
plate clamped to the swiveled plate being known to be accu-
rate, it is obvious that by sliding the indicator and its base
about the horizontal surface plate, the parallelism of the small
plate and the perpendicularity of the swiveled plate with the
large plate are determined.
THE MEASUREMENT OF ERRORS
101
102
METHODS OF MACHINE SHOP WORK
a,
-d
i.
to
THE MEASUREMENT OF ERRORS
103
MEASURING ACCURACY OF POSITION
Applications of the multiplying lever to the determination
of accuracy of position are shown in Figs. 77 and 78, from the
Cadillac automobile works. In Fig. 77 the parallelism of the
crank pin and piston pin bearings of a connecting rod is being
tested. The crank pin bearing is clamped upon a true arbor
a which is mounted in suitable supports, the piston pin end
having inserted within it a similar true arbor b. A slide
may be reciprocated a short distance by the hand lever d and
FIG. 79. Testing accuracy of spacing of worm wheel teeth.
a swiveled lever e carried by the slide c be thus brought into
contact with the arbor b. A multiplying lever / plays over a
graduated scale at its left-hand end and thus shows any de-
parture from parallelism of the two arbors.
In Fig. 78 the squareness of the piston with the crank-pin hole
is similarly tested after the parts have been assembled. The
crank-pin bearing is clamped upon the pin a and the slide at
the right is adjusted until the swiveled lever b makes contact
with the piston. Two multiplying levers c, d, of which the lat-
ter plays over a graduated scale at its further end, show any
lack of truth, suitably magnified.
104
METHODS OF MACHINE SHOP WORK
Silver.
.Spider
Figs. 79 and 80, from the Cincinnati Milling Machine Co.,
show an application of the multiplying-lever
principle to the determination of the accuracy
of the spacing of the teeth of the worm wheel
of a milling machine dividing head. The
multiplying lever is forked at its lower end,
as shown more clearly in Fig. 80, and at its
upper end it carries a spider-web line which
is read against a line upon a silver disc at-
tached to the frame. Going back to Fig. 79,
the arbor upon which the worm gear to be
tested is mounted carries on its rear end an
accurately divided circle by which the worm
wheel may be turned through the spaces cor-
responding with the intended spacing between
the teeth. Were this spacing correct, it is
clear that the spider-web line and the line
upon the silver disc would agree for every
tooth and that the errors will be shown by a
lack of such agreement. The degree of pre-
cision of the equipment is sufficiently indi-
cated by the fact that both graduated circle and hair lines are
read by microscopes.
FIG. 80. Indicator
of worm wheel test-
ing apparatus.
THE DIAL GAGE AND ITS
APPLICATIONS
A modification of the multiplying-
lever principle which has many appli-
cations is found in the dial gage of
the B. C. Ames Company, shown in
Fig. 81. In this instrument the
movement of the measuring finger is
shown by the turning of the index
on the dial, the two being connected
by multiplying internal mechanism
not shown. In the instrument
shown the readings are to thou-
sandths which are numbered from o to 50 in each direction, as
FIG. 81. Dial test indicator.
THE MEASUREMENT OF ERRORS
105
FIGS. 82 and 83. Testing the squareness of planer housings.
106 METHODS OF MACHINE SHOP WORK
FIGS. 84 and 85. Testing the squareness of a milling machine knee.
THE MEASUREMENT OF ERRORS
107
errors are as apt to lie in one direction as the other. The dial
may be turned to bring the zero under the index wherever it
may happen to lie at the first reading. This adjustment,
combined with the increased range of the instrument, makes
it more convenient for many purposes than the lever con-
struction previously shown.
Figs. 82 and 83, from the American Tool Works Company,
show the instrument used in combination with a square. By
moving the indicator and the block to which it is attached
FIG. 86. Testing the parallelism of a radial drilling machine arm and base and
the squareness of the spindle with the base.
vertically, it is obvious that the degree of squareness of the
planer housings with the V's of the bed in both directions is
quickly determined. Similarly Figs. 84 and 85, from the
Cincinnati Milling Machine Company, show applications to
milling-machine construction. In Fig. 84 the squareness of
the knee with the main frame as seen in plan, and in Fig. 85
as seen in elevation, is determined. The errors found are
removed by scraping.
108 METHODS OF MACHINE SHOP WORK
FIG. 87. Testing the alignment of a milling machine spindle.
FIG. 88. Testing the alignment of the spindle of a milling machine dividing head.
FIG. 89. Testing the alignment of milling machine centers.
THE MEASUREMENT OF ERRORS 109
Fig. 86, also from the American Tool Works Company, shows
applications that are typical of many. The indicator is attached
to an arm seen endwise but extending radially from an arbor
inserted in a radial drilling machine spindle. By traversing
the head along the arm and taking readings at various points
indicated by the three circles at the front of the base, the error
in the parallelism of arm and base is determined. Similarly by
revolving the spindle and taking readings at the four points
indicated by the circles, the squareness of the spindle with the
base is determined.
Fig. 87, from the Cincinnati Milling Machine Company,
shows an application to the testing of the alignment of a milling-
machine spindle and work table. The test arbor, which is
inserted in the taper hole of the machine spindle, being known
to be true, it is only necessary to revolve the spindle in order
to show any lack of truth of the hole since such lack of truth
will cause the arbor to vibrate and this vibration will appear
in the movements of the indicator pointer. Similarly by mov-
ing the indicator and stand to the inner end of the arbor, any
lack of parallelism of arbor and work table will be shown, and,
again, by traversing the work table on the knee, lack of par-
allelism between the arbor and the knee will appear.
Fig. 88, from the same source as the preceding illustra-
tion, shows a similar application to the taper hole of the work
spindle of a milling-machine dividing head. After the truth
of the hole has been proven, the head may be adjusted on its
swivel until the readings show the arbor to be parallel with the
base, when the zero of the graduated arc for reading the
angle of elevation may be located, or, if already located,
its truth may be proven. Still another application appears in
Fig. 89 in which the base of the stand for the indicator has
a tongue which drops into the T-slot of the main base. The
head and tail stocks have similar tongues and, the arbor be-
ing known to be true, traverse of the indicator in the slot
will show the truth of the alignment of the live and dead
centers.
These illustrations indicate the degree of precision that
enters into the construction of modern machine tools.
110
METHODS OF MACHINE SHOP WORK
TESTS WITH EXTEMPORIZED APPARATUS
A large number of entirely satisfactory tests may be made
with extemporized apparatus and an ordinary micrometer
caliper by measuring from the rear end of the barrel. Figs.
90-92 show such tests of the accuracy of a lathe. Putting a
well-centered arbor in the lathe and mounting a micrometer
upon it, as in Fig. 90, and taking readings against the face
plate at the ends of the vertical and horizontal diameters, it
FIG. 90.
FIG. 91. FIG. 92.
Extemporized tests with micrometer calipers.
is clear that the squareness of the plate and with it the align-
ment of the live spindle with the line of centers may be de-
termined. Again by mounting the calipers as shown in Fig.
91, taking a reading with the tail spindle drawn in, then loosen-
ing the tail stock upon the bed, running out the tail spindle
and repeating the reading, the horizontal alignment of the
tail spindle with the line of centers may be determined. Again,
mounting the caliper as in Fig. 92 and repeating the operations
just described, the vertical alignment of the tail spindle may be
determined.
CHAPTER V
GAGES
Relation of stiffness and sensitiveness of gages In large gages stiffness
must be sacrificed to lightness Expedients used under these conditions
Defects of snap gages Explanation of the popularity of common calipers
Limit gages Improved construction of snap gages Causes which
restrict the use of gages The Johansson combination gages, their
principles and properties Uses of these gages Screw thread gages
Independent measurements of the various elements of screw threads
Measuring the errors of pitch of long screws.
THE FUNCTION OF STIFFNESS IN GAGES
Referring again to the Brown and Sharpe plugs, Fig. 52, an
important lesson may be learned by comparing them by means
of a pair of common calipers. Using the gage shown with them
the difference between them may be detected by any one, but
using common calipers, an unskilled person will find this detec-
tion impossible. Using calipers a tool maker would detect the
difference with reasonable certainty but the fact remains that
while with the gage this detection is easy, with the calipers it is
difficult.
This difference between the two instruments is due to the
increased stiffness of the snap gage as compared with the
calipers. It is clear that when either gage or calipers is passed
over the larger plug the instrument must spring to accommodate
the increased size. Because of the stiffness of the snap gage,
the increased effort required to push it over the larger plug is
sufficient to be felt by the hand while, because of the flexi-
bility of the calipers, this increased effort is so small that, to all
but the highly skilled, it is imperceptible. This shows at once
the function of stiffness in gages which become more sensitive
as they are made stiffer with, however, a limitation which
grows out of the fact that increased stiffness is necessarily
accompanied by increased weight and this increased weight, if
111
112
METHODS OF MACHINE SHOP WORK
GAGES 113
carried too far, dulls the sense of touch. Were the snap gage
of Fig. 52 to weigh ten pounds, for example, the increased
effort necessary to force it over the larger plug would be lost
in the weight of the gage and the hand would not feel it.
This consideration has immediate application to large gages.
Were the weights of large gages made proportional to their sizes
in order to maintain the stiffness, the effect would be to destroy
the very object of this construction by reason of its effect in
dulling the sense of touch and, moreover, such gages would be
clumsy and unwieldy. It is, consequently, impracticable to
make large gages of the same relative stiffness as small ones and
this compels us, when using large gages, to resort to expedients.
Because of their comparative flexibility large gages are subject
to distortion from the effect of their own weight and, if satis-
factory measurements are to be obtained, it is necessary to find
means by which this distortion may be neutralized.
Figs. 93 and 94 show such an expedient used at the works of
the Westinghouse Machine Company. The piece of work to
be gaged shown beyond the operator and in the lathe is a
thirty-nine inch crank shaft, in front of which is a micrometer
caliper of suitable size. In order to combine lightness with
stiffness as far as possible, the frame of the caliper is made of
aluminum alloy. At its upper end it carries a micrometer
head which has a range of adjustment of one inch. At its
lower end is an adjustable anvil screw having a range of sev-
eral inches in order to give the instrument a corresponding
range and thus reduce the number of instruments required for
a given total range.
Fig. 93 shows the instrument in process of adjustment. Sup-
ported by a suitable stand is an end measure rod of steel which
is protected from the temperature of the hand by a casing of
wood. This rod is of a length equal to the number of whole
inches desired the fraction being obtained from the micro-
meter head. At the lower end of the supporting stand is a
stirrup supported in springs of a strength sufficient to carry
the weight of the micrometer. When adjusting the instru-
ment the micrometer head is first set to zero and then, with the
instrument resting in the stirrup in the position of Fig. 93, the
114 METHODS OF MACHINE SHOP WORK
anvil screw is adjusted until the micrometer screw makes con-
tact with the top of the rod. The fractional part of an inch
desired, if any, is then obtained by turning back the micro-
meter when the adjustment is complete.
When gaging the shaft two light sling chains are passed
around it as shown in Fig. 94, which chains carry a spring sup-
ported stirrup identical with the one on the stand for the end
measure rod. The caliper is placed in the stirrup and the size
of the shaft is read from the micrometer. The object of the
whole arrangement will be seen to be to place the instrument in
the same position as regards gravity when adjusting it and
when measuring with it, by which expedient the deflection due
to its own weight is obviously nullified.
GAGES CONTRASTED WITH CALIPERS
Snap gages have the defects of their virtues. The stiffness
which gives rise to their extreme sensitiveness gives rise also to a
property which makes them, for many purposes, an unsatisfac-
tory substitute for spring calipers. Because of their flexibil-
ity, the calipers may be pushed over a piece of work before it
has reached the final size without disturbing the adjustment or
doing other damage and the skill of the workman connects the
pressure required to push the calipers over the work with the
amount of metal remaining to come off. This is an extremely
valuable property of the calipers. Because of its stiffness, the
snap gage cannot be pushed over the work until the work has
reached size. If it does not go over, it tells the workman that
there is more metal to come off but it gives no indication of how
much more, the result being an increased number of trial cuts.
By reason of their flexibility the calipers may be used for
work of various degrees of accuracy. For close work the size
is made such that the feel of the contact between calipers and
work is very light, while for coarser work the contact is heavier
the calipers going over the work in both cases. The stiff snap
gage will not go over the work at all until within very narrow
limits the work is as small as true gage size. To tell the
workman to work to the gage means, therefore, that in many
GAGES 115
cases he will make fits that are unnecessarily good and hence
unnecessarily expensive. On the other hand, to depart from
this by authorizing him to use his judgment regarding the de-
gree of correspondence between gage and work, is to lose that
very control of the sizes for which the gage system is adopted
and so practically abandon the system at the outset.
All this is epitomized in the expression that "gages give no
warning" as the desired size is approached, and, since they give
none, the size must be approached with extreme care and in
constant fear that a cut may be too heavy and so spoil the work.
For these reasons calipers hold, and always will hold, their
own for many kinds of work, especially those in which the work-
man adjusts the tool for each piece produced, ordinary lathe
work being a typical example.
LIMIT GAGES
These facts, combined with a recognition of the further fact
that exact duplication of sizes is an impossibility, has led to the
system of limit gages of which three forms are shown in Figs.
95-97. 1 Of these Figs. 95 and 96 are external and Fig. 97
internal gages. In all cases one end is larger than the other
by a predetermined amount as indicated by the figures stamped
on Figs. 96 and 97. In the use of these gages one end must go
over or within the work, as the case may be, and the other must
refuse to go over or within it. The work is thus always between
the sizes of the two ends and, by a proper determination of the
difference between the ends, any required grade of workmanship
may be established.
It is obvious that in use these gages, like all end measures, are
subject to wear and it is also obvious that this wear is confined
chiefly to the end that goes over or within the work. Conse-
quently it is customary, especially in the case of internal gages,
to make the end which enters the longer of the two, thus pro-
viding increased wearing surface and also showing at a glance
which is which. For purposes of distinction the two ends are
1 Limit gages were first advertised foi sale by the Brown and Sharpe Manu-
facturing Company in 1875, after about ten years prior use in their own works.
116
METHODS OF MACHINE SHOP WORK
distinguished as maximum or minimum and as go or not-go.
The terms maximum and minimum are not satisfactory because
the maximum external gage goes over the work while the maxi-
mum internal gage does not go in. The terms go gage and not-
go gage avoid this ambiguity and are to be preferred.
FIG. 95.
.249
ISZ'l
FIG. 96.
FIG. 97.
Limit gages.
Fig. 98 shows a modification of the snap gage designed to avoid
the necessity for renewal after wear has taken place. This
effect is accomplished by the combination of the center piece a
and measuring jaws b. The original size of the gage is deter-
GAGES
117
mined by the center piece while the effect of wear is confined
to the jaws. After the jaws have worn it is only necessary to
remove them, lap them flat again and then replace them in
order to restore the gage to its original size and with a trifling
degree of expense. This construction will be recognized as
FIG.
FIG. 99.
Modified snap and limit gages.