—

The Measurement of Thickness

George Keinath^
The numerous methods for the measurement of thickness in laboratory or shop are
treated in groups according to physical operating principles: Mechanical
seven
weight/dimension relationships, acoustics, vibration, displacement with various con-
versions; chemical — —
stripping, spectrochemical analysis; electrical dielectric break-
—
down, resistance, electrochemical, capacitance, thermoelectricity; magnetic attractive
—
1

'
force, reluctance, saturation, inductance, eddy currents; optical microscopy (also elec-
—
1

tron microscopy K interferometry, diffraction, shadow; X-ray absorption, diffraction,

—
1

I
backscatter, spectrometry; radioactive radiation absorption, backscatter, tracers.
Ranges, accuracies, advantages, and limitations are discussed. A bibliography of refer-
ences, a limited list of suppliers, and a detailed index of the gages, methods, applications,
and trade names covered are appended.
I

5 I The measurement of thickness is the meas- the minimum. The low limit is 0.01 microinch,
2
i urement of the distance from a point on one as found in very thin coatings on glass. The
I bounding surface of a material body, through high limit may be one or more inches, as in
j

3
r the body, to the opposite bounding surface. If rolled or cast metals. This discussion will be
1 f
the surfaces of a body are parallel, as the two primarily concerned with the measurement of
1 i sides of a sheet of metal or the inner and outer the thickness of walls, sheets, and coatings of
I
surfaces of a pipe, the thickness is obviously the all kinds. Thickness of wires and threads is
I

I
I distance between the bounding surfaces meas- also included, although “diameter” is the term
r ured along a normal. However, if the surfaces more generally used if the sample has a circular
8 ! are not parallel the direction in which the thick- or an elliptical cross section. Distance measure-
* : ness is to be measured must be defined. Some- ment (as in spark gaps) is omitted, though the
times the measuring instrument itself deter- “feeler gages” for such measurements are called
j

jg
mines the direction along which the thickness thickness gages.
18
I
1 ismeasured. The methods of thickness measurements may
19 Further complications in measuring thickness be destructive or nondestructive, contacting or
1“ i are introduced by the lack of definition of the noncontacting, with relation to the sample. It is
l bounding surfaces of a body. For example, the very difficult to establish clear definitions of
thickness of a textile material is rather indefi- these terms for methods of measurement. For
i nite. Here, also, the measuring instrument im- instance, the use of a needle to penetrate an
;
portantly influences the “measured” thickness. insulating coat on a metallic base is a destruc-
I
Thickness measurements as above described tive method at the point of measurement. How-
deal with the measurement of a distance be- ever, the manufacturer of the gage may consider
^
tween two points. However, for some purposes his method nondestructive, because the tiny hole
the measurement of the average thickness of a in a paint coating is practically invisible and
body is of more significance. unimportant in the use of the product. In the
I This survey is intended to assemble for con- simplest problems the object of measurement is
I venient reference as many of the methods of a small specimen accessible from both sides.
! dealing with the problems of thickness measure- Difficulties arise when the object is accessible

jj ;
ment as possible. It is not limited to direct from only one side. Such an object might be a
measurement of thickness, but gives sufficient moving sheet coming from a steel mill red hot,
I background information on the types and prin- vibrating violently, covered with water or
ciples of operation of the devices to reduce the grease, and made almost invisible by a cloud of
I
need to seek the basic information elsewhere. steam. In such a case the use of noncontacting
I The survey includes information on the meas- thickness meters would be necessary.
i urement of physical parameters involved in the Units of thickness: In this compilation, thick-
practical measurement of thickness, such as ness has usually been expressed in “mils,” i.e.,
{ displacement, for example. It also deals with thousandths of an inch, so that values of thick-
some more general aspects of the measurement, ness will be easy to compare. It should be noted,
such as dynamic response. The extent to which however, that although this practice of stating
these are treated is determined in general, thicknesses of flat metal sheets and plates and
by the principles and makeup of existing diameters of wires in mils is increasing, the
instruments. specification of such sizes by numbers in one of
Considering only industrial problems, the several gage systems is still widely prevalent.
maximum thickness is about 10 million times In these systems, a given number is used to
designate different thicknesses depending on
1 Larchmont, N. Y. the material under consideration, e.g., steel.

1
brass, or aluminum. Since 1941 the American References to the literature and to patents are
Standards Association has issued Standards for indicated by numbers, in brackets and paren-
Preferred Thicknesses for uncoated thin flat theses, respectively, referring to the biblio-
metals under one-half inch, covering in the graphic listings on pages 69 and 75.
latest issue the range from 4 to 236 mils in 40 In writing this mono^aph the author has
steps by values progressing in accordance with received a variety of assistance from the staff
the 40- series of preferred numbers, each step of the Office of Basic Instrumentation. This
being 3 percent higher than the preceding one. assistance is gratefully acknowledged. Deserv-
The 1941 issue has only 20 steps, with a differ- ing particular mention are the extensive edi-
ence of 6 percent from step to step. torial review and the contributions to the sec-
Conversions among a number of units in tions on application of interferometry, radio-
which thickness is expressed are presented in graphic, and nuclear radiation methods.
the accompanying table on preceding page IV.
1.00.

Methods of Measurement
Mechanical Methods is refired just enough to obtain a fire polish on
the oblique section of the enamel. After cooling,
In this section are assembled methods in
the specimen is immersed in 10-percent citric
which the first measurement action is mechani-
acid and dried. Colored wax pencil is applied to
cal. The mechanical displacement, which is a
the whole surface, which is then rubbed with a
function of the thickness, may in turn be meas-
cloth. The colored wax is readily removed from
ured by optical, pneumatic, electric, or other
the acid-resistant section, which will have re-
means, generally with the purpose of increasing
tained its fire polish. This section is now clearly
the accuracy.
bounded by the line aas, and the colored wax
Weight [147; 190, p. 7].^ If a material
1.01. deposit. From measurements made along the
is homogeneous and of known density, a sample parallel lines andn of the thickness of the total
of controlled size can be weighed, and its thick- coating and lengths along the cut, the thickness
ness calculated from its weight. It is possible of the acid-resistant coating is computed. The
with sheet materials to cut samples of a speci- method is accurate to about 0.5 mil in a total
each measurement and thus facilitate
fied size at acid-resistant thickness of 2.5 to 8 mils.
the determination of thickness. Wires are meas-
ured in standard lengths. Balances that indicate
thickness directly are available or can be cali-
brated according to need.
1.02. Volume [168]. The thickness of non-
porous insulation of wires may be determined
in the following way A
length of wire of about
:

8 inches is cut off and put in a narrow cylin-

drical measuring vessel, filled with alcohol. The
rise of the level gives the total volume. After
stripping and dissolving the insulation, the
measurement repeated. From the volume
is
difference the average thickness may be calcu-
lated. This method has been used successfully Figure 1. Sketch of specimen showing cross section and

for several years by the Physikalisch-Technische oblique section of enamel applied in three coats over iron.

Reichsanstalt in Berlin, for wires from 0.75- to

lO-mm^ section (AWG No. 18 to No. 7) with an 1.032. Case Depth (Hardness Traverse) [10,
accuracy of about 2 percent. 157, 172, 175]. Effective case depth in steel is
1.04. as a distance measured perpendicularly
defined
1.03. Oblique Cut.
from the surface of a hardened case to a point
1.031. Vitreous Enamel [111] (fig. 1). A of hardness equivalent to Rockwell “C” 50.
method has been developed to measure the thick-
ness of acid-resistant enamel coatings on iron Total case depth is defined as a distance meas-
plumbing fixtures (fig. 1). From a section of ured perpendicularly from the surface down to
suitable size, along a new-cut edge, the enamel the point where carbon enrichment ceases. The
is ground off in a plane at an angle of approxi-
specimen is cut, ground, and polished in either
mately 5 degrees to the surface, exposing an steps, taper, or perpendicular section. The ex-
oblique section of the enamel coating and part posed surface is tested for hardness until the
of the underlying metal. The lines (aag, aidj, effective case depth is reached.
a 2 d 2 and a^d.^) shown in figure 1 are marked Chord Cut [35] (fig. 2). Mesle’s chord
,

with a ceramic underglaze pencil. The specimen method for measuring the thickness of coatings
depends on barely cutting through the coating
- Figures in brackets indicate the literature references on page 69. on a plane surface, with a grinding wheel of
2
 ; ;

IS

ff

is

]•

c-

-
0

Figure 3. Roller thickness gage.

Figure 2. Chord method to determine thickness of

coating on a spherical surface.

it known radius fed perpendicularly on a curved

;

to surface, by a chordal cut. In either case the

a thickness can be computed from the chord and
ID
^
the radius of the segment removed. The ref er-
e-K ence cited [35] gives the formula for calculat-
lyf! ing thickness for a number of different inter-
ax
secting shapes: Plane surface and grinding
he wheel convex spherical surface and flat file
convex spherical surface and grinding wheel;
,js i
: and concave spherical surface and small-radius
he t'+ grinding wheel. The curvature of spherical sur-
faces is determined with a spherometer. The
ij measurement is difficult when coating and base
j
have about the same color, in which case the cut
w i surface is treated with chemicals that cause a
color change.
I
I With a grinding wheel radius of 3 to 4 inches Figure 4. Roller thickness gage.
si the least thickness measurable is 0.05 mil, the
chord length being about 40 mils. The accuracy The standard instrument with a wheel 2 in.
is about 5 percent for nickel or composite coat-
I in diameter has a measuring range of 0 to 4
on steel plates, and about 10 percent for mils. Each scale division on the eccentric wheel
/j
nickel coatings on copper or brass plates. For
I represents 0.2 mil and readings to the nearest
very thin coatings of chromium (about 0.02 0.1 mil are possible. The following instrument
I mil) used for decorating purposes, a very small
I ranges are also available 0 to 0.4 mil, 0 to 1 mil,
:

grinding wheel is recommended to produce a 0 to 2 mils, 0 to 12 mils, 10 to 30 mils, and 20 to

I
f| narrow but well-defined cut which can be meas- 60 mils; and the scale graduations, 0.02, 0.05,
ured with a microscope.
0.1, 0.5, 1.0, and 2.0 mils.
Dif 1.05. Manual Roller Gages.
Manufacturer: Henry A. Gardner Labora-
- 1.051. Euverard Gage for Wet Films (figs.
tory, Bethesda, Md.
gage, developed for the measure- 1.052. Euverard Gage for Dry Films (figs.
isffi ment of wet film thicknesses, is essentially a An adaptation of the wet-film gage de-
5, 6).
lj| graduated eccentric wheel supported by two scribed above makes it suitable for measuring
otl concentric wheels (fig. 4), all machined to an
the thickness of coatings, especially paint coat-
0,1 accuracy of 0.008 mil. When rolled over
a plane ings, on a nonconducting base like wood or
surface, there is a clearance ranging from 0 to
s.)'
glass. The same wheel is used, but the film is
4 mils (fig. 3) between the inner eccentric
lol!
wheel removed from the base along two suitably
and the surface upon which the outer two con-
lej
spaced strips, to permit the outside concentric
centric wheels rest. When the gage is placed
;t|,
on wheels of the gage to rest on the uncoated base.
-a wet surface in the position of
the greatest By the use of a suitably adjusted mirror or lens,
ifw clearance, and rolled until the clearance is elim-
the degree of clearance between the measuring
' mated, the eccentric wheel
will at some inter- surface of the gage and the film coating can be
contact
7 film. The measurement and “pick up” the wet seen with the aid of a diffused-light source di-
rj
is usually performed in rected toward the gage. As long as there is a
'!
perpendicular directions and the average gap between the measuring surface of the eccen-
of the readings taken as the
if! true wet-film tric wheel and the coating, it is possible to ob-
thickness.
serve a brightly illuminated area which has the

'L
 Figure 5. Roller gage to measure thickness of dry
coatings.

Figure 7. Pfund film-thickness gage.

Figure 6. Roller gage to measure thiekness of dry

coatings.

width of the measuring surface and the height

of the space between this surface and the coat-
ing. The gage is slowly turned until the illum-
inated area disappears, at which point the read-
ing is taken, in two directions if desired. A
magnifying lens is provided for better observa-
tion of the luminous area. The scale can be read
from the mirror without moving the operator’s
eye from the viewing position (figs. 5, 6). It is
claimed that this method, which applies no
pressure to the film, gives the true thickness
more accurately than other methods which re-
quire decoating.
Manufacturer: Henry A. Gardner Labora-
tory, Bethesda, Md.
1.06.Sphere Penetration Gage (figs. 7, 8, 9).
The Pfund paint-film thickness gage is used for
the determination of the thickness of wet paint
or varnish films (fig. 7). A convex lens, L,
whose lower surface has a radius of curvature
of 250 mm, is mounted in a short tube, Ti, which
slides freely in the outer tube, T 2 The compres-
.

sion springs, S, keep the convex surface out of

contact with the paint film until pressure is
applied to the top of T,. The instrument is
rested on the painted surface and the lens is
forced down as far as it will go. Upon releasing
the pressure and removing the gage, a circular
spot is left on the lens as well as on the painted
surface. The diameter of the spot on the lens is
measured to 0.1 mm
with a steel scale and then
converted to thickness by using tables. The
accuracy has been found to be within 0.04 mil
for films 1.6 to 2.6 mils thick.
Manufacturer Koehler Instrument Co., Ja-
:

maica 33, N. Y.
1.10. Mechanical Gages, Direct Reading [78;
190, p. 6-8] Direct-reading mechanical gages,
.

examples of which are discussed below, are the

most universally used thickness meters, espe-
cially for making measurements on small pieces.
Their use is prescribed in many of the ASTM
specifications and standards. They give the Figure 9. Paint thickness as a function of the paint-spot
thickness at the spot where the measurement is diameters for the Pfund gage.

4
 I

Figure 10. Slide caliper with vernier attachment.

made. For large objects one must determine

j
the average thickness by making a number of
'

measurements at a number of evenly distributed

spots and calculating the arithmetic mean. The
thickness of very fine paper, such as that used
for capacitors, is often measured by stacking an
assembly of 10 sheets and using a parallel-plate
gage. With paper about 0.25 mil thick, using
dial gages or screw micrometers, the thickness
^
!
of one sheet might be measured as 0.34 mil,
'
whereas the average thickness based on 10
stacked sheets might be only 0.24 mil. The dif-
ference in values between measurements on a
single sheet and measurements on stacked sheets
depends on the “packing fraction” (the extent
to which the thicker areas of one sheet coincide
with the thinner areas of another sheet) The .

stacked thickness values are preferred as a prac-

tical matter because they tend to be more nearly
uniform.
1.11. Slide Calipers [27] (fig. 10). Slide cali-
_ pers are among the simplest devices for thick-
J ness measurement. The reading accuracy is
I
about 0.1 mm
(4 mils) for direct reading, be-
M cause the smallest discernible distance is about
ji 0.07 mmfor the average human eye [27]. With
^
a vernier attachment the accuracy can be much
increased, with 0.018 mm as the possible limit.
In view of the thickness of the graduation lines
a practically feasible value is 0.050 mm (2
mils). By using a magnifying lens, a vernier
- with 50 divisions, and a line thickness of not
more than 0.07 to 0.1 mm, a least reading of
> 0.02 mmmay be obtained. Figure 11.
1.12. Joint Calipers (fig. 11). There are three a» Firm-joint caliper ; b, firm-joint caliper with screw adjustment;
,
c, spring caliper.
I types of joint calipers (fig. 11) which are cus-
,
tomarily used as comparators to transfer a fixed nut in the frame, and the spindle, which is
'
dimension to a steel scale for measurement: the unthreaded portion of the screw, moves up
1 (a) Firm-joint calipers with legs that hold or down relative to the anvil. The spindle moves
.. their position by friction, giving the advantage through a definite distance depending on the
of very quick adjustment. pitch of the screw. As a rule, there are 40
(b) Screw-adjusted firm-joint calipers, simi- threads to the inch hence one turn of the thim-
.! lar to (a), but with fine adjustment. ble moves the spindle ’/4o
;

0.025 in. 25 = ^
!:
(c) Spring calipers, the most widely used mils. The major divisions of 25 mils are marked
: type. on the hub and 25 subdivisions are marked on
,
1.13. Screw Micrometers [15, 20, 27, 118, the beveled portion of the thimble attached to
'
206] (fig. 12). the screw.
1.131. Micrometer Caliper [27]. The mi- Othe'' micrometers of greater precision are
s)»T crometer is essentially a calibrated screw. By also m.ade. It should be noted, though, with the
turning the thimble the screw revolves in a more finely divided screw and the more precise

5
 instrument, greater care must be taken in its 1.132. Screw Micrometer With Optical Indi-
use. Accuracy of measurement depends largely cation of Pressure [206] (figs. 13, 14). This
on the pressure exerted on the work and on the micrometer (figs. 13, 14), which utilizes optical
compressibility of the material. The more com- pressure indication, has been developed to pro-
pressible the material, the greater the error. vide the highest accuracy of screw micrometers
Many types of screw micrometers use a friction [206]. The micrometer screw, machined -with
clutch or ratchet for rotating the screw in order great precision, carries a 6-in. -diameter alumi-
to obtain constant and correct pressure. Gen- num wheel with 0.1-mil graduations about 100
erally the accuracy obtained is ±0.1 to ±0.5 mils apart, a magnification of 1,000. This per-
mils. mits readings of 0.01 mil. Comliined with this
One design of micrometer caliper with a non- mechanical-screw micrometer is an optical indi-
rotating sleeve reads by vernier down to 0.025 cator of pressure, which enables the operator to
mil. Optical magnification with a simple 5 or apply exactly the same pressure on the object
lOX hand lens is often used with such instru- for each measurement. This pressure indicator
ments. In the most elaborate designs [27] the is a combination of an optical flat, a chromium-
reading error is reduced to 1 /x (0.04 mil) and plated steel flat, and a red selenium-glass screen,
even 0.1 The total accuracy of measurement so arranged that the slightest upward pressure
is, of course, less than the reading accuracy. on the micrometer spindle causes interference
Procedure and permissible errors of microm- bands to move past a reference line. Each inter-
eter calipers with and without ratchets are ference band is a measuring unit of 0.01 mil.
discussed in ASTM standard D374. The bands are spaced about Vs in. apart. Any
measuring load from 0 to 2.5 lb can be applied.
Ratchet Screw When in proper adjustment, two or three bands
will show, from the left-hand edge of the steel
Ratchet Spring'
flat up to the cross-reference line, when the
Ratchet Plunger spindle is not touching the anvil. Each addi-
tional band that is moved up to this reference
Ratchet Thimble
line corresponds to an increase of measuring
Cap with Ratchet load of 6 oz. The load is selected to suit the
material under test. The ranges of this light-
Cap Only
wave screw micrometer are 0 to 1, 0 to 2, or 0 to
Tension Nut 3 in. The threshold sensitivity of measurement
Tension Adj. Sleeve

Retaining Spring

Screw Tension Spring

Thimble

Screw Nut

Hub
Lock Nut

Lock Nut Shoe

Spindle

Frame

Figures 13 and 14.Pressure indication on light-wave

Figure 12. Micrometer caliper. screw micrometer.