SSPC-PA 2

June 1, 1996

SSPC: The Society for Protective Coatings

PAINT APPLICATION STANDARD NO. 2
Measurement of Dry Coating Thickness with Magnetic Gages

1. Scope pressure on the coated surface during the entire measuring

operation. Electronic circuitry is used to convert a reference
1.1 GENERAL: This standard describes the procedures to signal into coating thickness. (See 8.1.)
measure the thickness of a dry film of a nonmagnetic coating
applied on a magnetic substrate using commercially available 2.3 USE OF STANDARD: This document contains the
magnetic gages. These procedures are intended to supplement following:
manufacturers’ instructions for the manual operation of the
• Calibration, verification and measurement procedures
gages and are not intended to replace them.
(Section 4).
1.2 The procedures for calibration and measurement are • Required number of measurements for conformance to a
described for two types of gages: pull-off gages (Type 1) and thickness specification (Section 5).
constant pressure probe gages (Type 2).
• Notes on gage principles and various factors affecting
1.3 The standard defines a procedure to determine if the thickness measurement (Section 8).
film thickness over an extended area conforms to the minimum
• A numerical example of thickness measurement over an
and the maximum levels specified.
extended area (Appendix 1).

2. Description and Use • A numerical example of verification of the calibration of

Type 2 gages using plastic shims (Appendix 2).
2.1 DEFINITIONS:

2.1.1 Gage Reading: A single reading at one point.

3. Reference Standards
3.1 The documents and standards referenced in this stan-
2.1.2 Spot Measurement: The average of at least three
dard are listed in Section 3.4 and form a part of this standard.
gage readings made within a 1.5 inch (4 cm) diameter circle.
3.2 The latest issue, revision or amendment of the refer-
2.2 DESCRIPTION OF GAGES:
enced documents in effect on the date of invitation to bid shall
2.2.1 Gage Types: The gage type is determined by the govern unless otherwise specified.
specific magnetic properties employed in measuring the thick-
3.3 If there is a conflict between the requirements of any
ness and is not determined by the mode of data readout, i.e.
of the cited documents and this standard, the requirements of this
digital or analog. This standard does not cover gages that
standard shall prevail.
measure the effect of eddy currents produced in the substrate.
3.4 NATIONAL INSTITUTE OF STANDARDS AND TECH-
2.2.2 Type 1—Pull-Off Gages: In pull-off gages, a per-
NOLOGY (NIST) STANDARD REFERENCE MATERIALS. (See
manent magnet is brought into direct contact with the coated
Section 8.15.)
surface and a calibrated scale measures the force necessary to
pull the magnet from the surface. Less force is required to
remove the magnet from a thick coating. The scale is nonlinear. 4. Calibration, Verification and Measurement
Procedures
• Type 1A - A magnet is attached to one end of a pivoting
balance arm. This assembly is connected to a calibrated 4.1 GENERAL
helical spring. By rotating a dial, the spring increases the
4.1.1 Access to Bare Substrate: All gages are affected to
force on the magnet and pulls it from the surface. The
some degree by substrate conditions such as roughness, shape,
Type 1A gages are commonly called “banana” gages.
thickness and composition. To correct for this effect, access to
• Type 1B - A magnet is mounted directly or indirectly to a the uncoated substrate is recommended. Another option is to
coil spring. The spring acts perpendicularly to the surface use separate uncoated reference panels with similar rough-
to pull off the magnet. The Type 1B gages are commonly ness, shape, thickness and composition. (See Sections 8.3 to
called “pencil” gages. 8.9.) These would be used as the bare substrate in the proce-
dures of Sections 4.2 and 4.3. Reference panels shall be of
2.2.3 Type 2—Constant Pressure Probe Gages: A con-
sufficient size to preclude edge effects. (See Section 8.7.)
stant pressure probe gage uses a probe which exerts a constant
Measurements on the bare substrate can be taken before

1
SSPC-PA 2
June 1, 1996

the coating is applied or by masking off small representative over which the gage is accurate. All Type 1 gages have
areas during painting. If the coating has already been applied nonlinear scales and any adjusting feature is linear in nature.
to the entire surface, small areas of coating may be removed and Therefore, only a given segment of the scale can be accurate
later patched. Do not allow the removal process to alter the after adjustment.
condition of the substrate. Paint strippers should be used in
4.2.4 Measure the bare substrate at a number of spots to
order to retain the profile.
obtain a representative average value. This average value is the
4.1.2 Spot Measurement: Repeated gage readings, even base metal reading (BMR). Note the gage is not to be calibrated
at points close together, may differ due to small surface irregu- on the bare substrate.
larities of the coating and the substrate. Therefore, a minimum
4.2.5 Measure the dry coating at the number of spots
of three (3) gage readings shall be made for each spot measure-
specified in Section 5.
ment of either the substrate or the coating. For each new gage
reading, move the probe to a new location within the 1.5 inch (4 4.2.6 Subtract the base metal reading from the gage read-
cm) diameter circle defining the spot. Discard any unusually ing to obtain the thickness of the coating.
high or low gage reading that cannot be repeated consistently.
Take the average of the acceptable gage readings as the spot 4.3 CALIBRATION, VERIFICATION AND MEASUREMENT:
measurement. TYPE 2—CONSTANT PRESSURE PROBE GAGES

4.2 CALIBRATION, VERIFICATION AND MEASUREMENT: 4.3.1 Different manufacturers of Type 2 (constant pressure
TYPE 1—PULL-OFF GAGES probe) gages follow different methods of calibration or adjustment.
Calibrate the gage according to manufacturer’s instructions.
4.2.1 For Type 1 gages, use test blocks bearing calibrated
non-magnetic coatings that are traceable to a suitable national 4.3.2 With a properly calibrated gage, measure the dry
standard. (See Section 8.15.) The standards must be large coating as specified. (See Section 4.1.2.)
enough to exceed the critical mass of steel needed to satisfy the
4.3.3 Verify the calibration of the gage at the beginning and
magnetic field of the Type 1 (pull-off) magnets. Shims of plastic
the end of each work shift with one or more of the standards.
or of non magnetic metals which are acceptable for calibration
(See Appendix 2). During the work shift, if the gage is dropped
of Type 2 (constant pressure probe) gages should not be used
or suspected of giving erroneous readings, its calibration should
for calibration of the Type 1 gages. (See Section 8.1.1.) If the
be rechecked. If deemed appropriate by the contracting parties,
manufacturer’s instructions are in conflict with this standard by
initial agreement can be reached on the details and frequency of
allowing the use of plastic or other non-magnetic shims for the
verification or calibration. Record the calibration data and the
calibration of a Type 1 gage, the contracting parties must both
method used to verify the calibration. If the gage is found to be
be notified of this fact and agree on a method of calibration.
out of calibration at the end of the works shift, all measurements
If proprietary thickness standards are to be used, agree-
made since the last calibration are suspect.
ment between contracting parties should be reached prior to
starting the job.
5. Required Number of Measurements for
4.2.2 Using the Type 1 (pull-off) gage, measure the thick- Conformance to a Thickness Specification
ness of a series of calibration standards covering the expected
range of coating thickness. To guard against measuring with an 5.1 NUMBER OF MEASUREMENTS: Make five (5) sepa-
inaccurate gage, recheck the gage at the beginning and the end rate spot measurements (average of the gage readings, see
of each work shift with one or more of the standards. During the Section 4.1.2) spaced randomly over each 10 m2 (100 ft 2) area
work shift, if the gage is dropped or suspected of giving errone- to be measured. If the contracting parties agree, more than five
ous readings, its calibration should be rechecked. If deemed (5) spot measurements may be taken in a given area. (See
appropriate by the contracting parties, initial agreement can be Section 5.3.) The five spot measurements shall be made for
reached on the details and frequency of verification or calibra- each 10 m 2 (100 ft2) of area as follows:
tion. Record the calibration data and the method used to verify 5.1.1 For structures not exceeding 30 m2 (300 ft2) in area,
the calibration. If the gage is found to be out of calibration at the each 10 m 2 (100 ft2) area shall be measured.
end of the work shift, all measurements made since the last
calibration are suspect. 5.1.2 For structures not exceeding 100 m2 (1,000 ft 2) in
area, three 10 m 2 (100 ft2) areas shall be randomly selected and
4.2.3 When the gage no longer agrees with the standard, measured.
check the probe for cleanliness. If dirty, clean as described in
Section 8.5.1. If the gage still does not agree with the standard, 5.1.3 For structures exceeding 100 m2 (1,000 ft2) in area,
the gage is in need of repair or replacement. Some gages can the first 100 m2 (1,000 ft2) shall be measured as stated in Section
be adjusted to read accurately in a given range. Adjust the gage 5.1.2 and for each additional 100 m2 (1,000 ft2) of area or
to read correctly on a given standard. Then check the gage on increment thereof, one 10 m2 (100 ft2) area shall be randomly
standards of higher and lower thicknesses to establish the range selected and measured.

2
 SSPC-PA 2
June 1, 1996

5.1.4 If the dry film thickness for any 10 m2 (100 ft2) area 8. Notes
(see Sections 5.1.2 and 5.1.3) is not in compliance with the
requirements of Sections 5.2.1 and 5.2.2, then additional mea- Notes are not a requirement of this standard.
surements must be made to isolate the non-conforming area.
8.1 PRINCIPLES OF THE MAGNETIC GAGE: Each of
5.2 SPECIFYING THICKNESS: Both a maximum and a these gages can sense and indicate only the distance between
minimum thickness should be specified for the coating. If a the magnetic surface of the steel and the small rounded tip of the
maximum thickness value is not explicitly specified, the speci- magnet or probe that rests on the top surface of the coating. This
fied thickness shall be the minimum. measured distance, from the top surface of the coating, must be
corrected for the thickness of any extraneous films or other
5.2.1 Minimum Thickness: The average of the spot interfering conditions on the surface of the steel. Such correction
measurements for each 10 m2 (100 ft2) area shall not be less is described in Section 4.2 for Type 1 gages and manufacturer’s
than the specified minimum thickness. No single spot measure- instructions for Type 2 gages.
ment in any 10 m2 (100 ft2) area shall be less than 80% of the
specified minimum thickness. Any gage reading may under-run 8.1.1 Type 1 (pull-off) gages use a calibrated spring mecha-
by a greater amount. If the average of the spot measurements nism to measure the force needed to pull a small permanent
for a given 10 m2 (100 ft2) area meets or exceeds the specified magnet from the surface of the coated steel. The magnetic force
minimum thickness, but one or more spot measurements is less holding the magnet to the surface varies inversely as a non-
than 80% of the specified minimum thickness, additional mea- linear function of the distance between magnet and steel, i.e.,
surements may be made to define the non-conforming area. the thickness of the dry coating (plus any other films present).
(See Appendix 1.) The Type 1A “banana” gages use a helical spring to pull a
small permanent magnet from the surface. Internal balancing
5.2.2 Maximum Thickness: The average of the spot mechanisms in most banana gages compensate for horizontal,
measurements for each 10 m2 (100 ft2) area shall not be more vertical and overhead positions so that there is no need to
than the specified maximum thickness. No single spot measure- recalibrate when changing orientation.
ment in any 10 m2 (100 ft2) area shall be more than 120% of the In a Type 1B “pencil” gage, a calibrated coil spring mea-
specified maximum thickness. Any gage reading may over-run sures the force necessary to pull the permanent magnet from the
by a greater amount. If the average of the spot measurements surface. Because of gravitational effects, these gages must be
for a given 10 m2 (100 ft2) area meets or falls below the specified recalibrated when the orientation of the surface changes; e.g.,
maximum thickness , but one or more spot measurements is a gage calibrated on a horizontal surface will not be accurate
more than 120% of the specified maximum thickness, additional when measuring a vertical surface. Some gages have three
measurements may be made to define the non-conforming separate indicators which compensate for horizontal, vertical
area. Manufacturers’ literature may be consulted to determine and overhead positions. Type 1B gages are generally not as
if higher maximum thickness readings are allowable under precise as Type 1A gages.
specific circumstances. Normally, Type 1 gages are not adjusted or reset for each
new series of measurements.
5.3 Other size areas or number of spot measurements may
Shims of sheet plastic or of non-magnetic metals, which are
be specified in the procurement documents as appropriate for
permissible for calibrating Type 2, (constant pressure probe)
the size and shape of the structure to be measured.
gages, should not be used for calibration of Type 1 gages. Such
shims are usually fairly rigid and curved and do not lie perfectly
6. Accuracy
flat, even on a smooth steel test surface. Near the pull-off point
6.1 To qualify under this standard, a gage must have an of the calibration measurements with any Type 1 gage, the shim
accuracy at least within ±10%. For thicknesses less than 25 µm frequently springs back from the steel surface, raising the
(1 mil), the gage must have an accuracy at least within ±2.5 µm magnet too soon and causing erroneous calibration readings.
(0.1 mil).
8.1.2 Type 2 (constant pressure probe) gages operate on
two different magnetic principles. Some Type 2 gages use a
7. Disclaimer permanent magnet. When the magnet is brought near steel, the
7.1 While every precaution is taken to insure that all magnetic flux density within the magnet is increased. By
information furnished in SSPC standards and specifications is measuring this change in flux density, which varies inversely to
as accurate, complete and useful as possible, SSPC cannot the distance between the magnet and the steel substrate, the
assume responsibility nor incur any obligation resulting from the coating thickness can be determined. Hall elements and mag-
use of any materials, coatings or methods specified therein, or net resistance elements are the most common ways to measure
of the specification or standard itself. magnetic flux density. However, the response of these ele-
ments is temperature dependent, so temperature compensation
is required.
Other Type 2 gages operate on the principle of electromagnetic

3
SSPC-PA 2
June 1, 1996

induction. A coil containing a soft iron rod is energized with an 8.8 PROXIMITY TO OTHER MASS OF STEEL: The older
AC current thereby producing a changing magnetic field at the two-pole Type 2 gages with permanent magnets are sensitive to
probe. As with a permanent magnet, the magnetic flux density the presence of another mass of steel close to the body of the
within the rod increases when the probe is brought near the steel gage. This effect may extend as much as three inches (7.6 cm)
substrate. This change is easy to detect by using a second coil. from an inside angle.
The output of the second coil is related to coating thickness and
8.9 CURVATURE OF STEEL SURFACE: Magnetic gage
this relationship can be determined experimentally.
readings may be affected by surface curvature. If the curvature
8.2 REPEATABILITY: Magnetic gages are necessarily is appreciable, valid measurements may still be obtained by
sensitive to very small irregularities of the coating surface or of calibrating or adjusting the gage on a similarly curved surface.
the steel surface directly below the probe center. Repeated
8.10 TILT OF PROBE: All of the magnets or probes must
gage readings on a rough surface, even at points very close
be held perpendicular to the coated surface to produce valid
together, frequently differ considerably, particularly for thin films
measurements.
over a rough surface with a high profile.
8.11 OTHER MAGNETIC FIELDS: Strong magnetic fields,
8.3 ZERO SETTING: Type 1 magnetic gages should not
such as those from welding equipment or nearby power lines,
be adjusted or set at the scale zero (0) with the gage applied to
may interfere with operation of the gages. Also, residual
either a rough or a smooth uncoated steel surface.
magnetism in the steel substrate may affect gage readings.
8.4 ROUGHNESS OF THE STEEL SURFACE: If the steel With fixed probe two-pole gages in such cases, it is recom-
surface is smooth and even, its surface plane is the effective mended that the readings before and after reversing the pole
magnetic surface. If the steel is roughened, as by blast cleaning, positions be averaged. Other gages may require demagnetiza-
the “apparent” or effective magnetic surface that the gage tion of the steel.
senses is an imaginary plane located between the peaks and
8.12 EXTREMES OF TEMPERATURE: Most of the mag-
valleys of the surface profile. With a correctly calibrated and
netic gages operate satisfactorily at 4°C and 49°C (40°F and
adjusted Type 2 gage, the reading obtained indicates the coat-
120°F). Some gages function well at much higher temperatures.
ing thickness above this imaginary plane. (See Section 4.3.) If
However, if such temperature extremes are met in the field, the
a Type 1 gage is used, the coating thickness is obtained by
gage might well be checked with at least one reference standard
subtracting the base metal reading. (See Section 4.2.)
after both the standard and the gage are brought to the same
8.5 DIRTY, TACKY OR SOFT FILMS: The surface of the ambient temperature. Most electronic gages compensate for
coating and the probe of the gage must be free from dust, grease temperature differences among the gage, probe and surface.
and other foreign matter in order to obtain close contact of the
8.13 VIBRATION: The accuracy of the Type 1 (pull-off)
probe with the coating and also to avoid adhesion of the magnet.
gages is affected by traffic, machinery, concussions, etc. When
The accuracy of the measurement will be affected if the coating
these gages are set up for calibration or measurement of coating
is tacky or excessively soft. Tacky coating films also cause
films, there should be no apparent vibration.
unwanted adhesion of the magnet. Unusually soft films may be
dented by the pressure of the probe. Soft or tacky films can 8.14 VARIATION IN THICKNESS - 80% OF MINIMUM/
sometimes be measured satisfactorily by putting a shim on the 120% OF MAXIMUM: In any measurement there is a certain
film, measuring total thickness of coating plus shim and sub- level of uncertainty. Two inspectors using the same gage will not
tracting shim thickness. necessarily record the exact same number for a given spot
measurement using the same 4 cm (1.5 in) diameter circle. To
8.5.1 Ordinary dirt and grease can be removed from a
allow for this natural fluctuation, an individual spot measure-
probe by wiping with a soft cloth. Magnetic particles adhering to
ment is permitted to be below the specified minimum thickness
the probe can be removed using an adhesive backed tape. Any
as long as other spots in the 10 m2 (100 ft2) area are high enough
adhesive residue left on the probe must then be removed.
to make the average thickness meet or exceed the specified
8.6 ALLOY STEEL SUBSTRATES: Differences among minimum thickness. Similar reasoning applies to maximum
most mild low-carbon steels will not significantly effect magnetic thickness. The 80% of specified minimum and 120% of speci-
gage readings. For higher alloy steels, the gage response fied maximum allow for the tolerance of the gage and calibration
should be checked. In any event, the gage should be recalibrated standards and for variations in the substrate.
on the same steel over which the coating has been applied.
8.15 Polished metal calibration standards are manufac-
8.7 PROXIMITY TO EDGES: Magnetic gages are sensi- tured by the National Institute of Standards and Technology
tive to geometrical discontinuities of the steel, such as holes, (NIST). The chrome plated panels are flat smooth steel 2.86 x
corners or edges. The sensitivity to edge effects and 2.86 cm (1.125 x 1.125 in) in size. Examples of some NIST
discontinuities varies from gage to gage. Measurements closer standards are:
than 2.5 cm (1 in) from the discontinuity may not be valid unless
the gage is calibrated specifically for that location.

4
 SSPC-PA 2
June 1, 1996

Certified Coating Thickness Calibration Standards Suppose this structure is 30 m2 (300 ft2) in area. Mentally
Nonmagnetic Coating on Steel divide the surface into three equal parts, each being about 10
m2 (100 ft2).
SRM 1358 Set of 3 80, 225, 1000 µm (3, 9, 40 mil)
SRM 1359 Set of 4 48, 140, 505, 800 µm (2, 5.5, 20, 31 mil) Part A - 10 m2 (100 ft2)
SRM 1362a Set of 4 40, 80, 140, 205 µm (1.6, 3, 5.5, 8 mil) Part B - 10 m2 (100 ft2)
SRM 1331a to 1339a Single standards from 3 µm (0.1 mil) Part C - 10 m2 (100 ft2)
to 62 µm (2.4 mil)
First, measure the coating thickness on Part A. This
8.16 CORRECTING LOW OR HIGH THICKNESS: The involves at least 15 readings of the thickness gage. (See
contracting parties should agree upon the method of correcting Figure A1.) Assume the specification calls for 64 µm (2.5 mils)
film thicknesses that are above the maximum or below the minimum thickness. The average thickness for area A is then
minimum specification. This method may be specified in the the average of the five spot measurements made on area A,
procurement documents, may follow manufacturer’s instruc- namely 66 µm (2.6 mils).
tions or may be a compromise reached after the non-conforming
area is discovered. Spot 1 64 µm 2.5 mils
Spot 2 76 3.0
APPENDIX 1—Numerical Example of Average Spot 3 53 2.1
Spot 4 76 3.0
Thickness Measurement
Spot 5 58 2.3
The following numerical example is presented as an illus- Avg. 66 µm 2.6 mils
tration of Section 5. (Reference JPCL, Vol 4, No 5, May 1987)

FIGURE A1
Part "A" of Structure (Area Approx. 10 m2 [100 ft2])
SSPC-PA 2
June 1, 1996

The average, 66 µm, exceeds the specified minimum of 64 The thickness of a plastic shim is typically accurate to within
µm and thus satisfies the specification. However it must be ±5%. After calibration according to manufacturer’s instructions,
decided if the lowest spot measurement, 53 µm, is within 80% the gage is probably accurate to within ±5% also. Therefore, for
of the specified minimum thickness. Eighty percent of 64 µm is the gage to be in agreement with the shim, the average thick-
51 µm (0.80 x 64 = 51). Although 53 µm is below the specified ness measured by the gage should be within ±10% of the shim’s
minimum, it is still within 80 percent of it, so the specification is thickness. If the average thickness measured on the 51 µm (2.0
satisfied. [The average, 2.6 mils, exceeds the specified mini- mil) shim is 56 µm (2.2 mils), the gage is in agreement with the
mum of 2.5 mils and thus satisfies the specification. However it shim because 56 is within ±10% of 51 (2.2 is within 10% of 2.0).
must be decided if the lowest spot measurement, 2.1 mils, is Similarly, if the average thickness of the 254 µm (10.0 mil)
within 80% of the specified minimum thickness. Eighty percent shim is measured to be 279 µm (11.0 mils), the gage calibration
of 2.5 mils is 2.0 mils (0.80 x 2.5 = 2.0). Although 2.1 mils is is verified because 279 is within ±10% of 254 (11.0 is within
below the specified minimum, it is still within 80 percent of it, so ±10% of 10.0).
the specification is satisfied.] In summary, if the average measurement of the 51 µm (2
There are individual gage readings of 38 µm at spot 5 and mil) shim is between 46 and 56 µm (1.8 and 2.2 mils), and if the
46 µm at spot 3, both of which are clearly less than 51 µm. This average measurement of the 254 µm (10.0 mil) shim is between
is allowed because only the average of the three readings (i.e., 229 and 279 µm (9.0 and 11.0 mils), the calibration is verified.
the spot measurement) must be greater than or equal to 51 µm. Check that another shim of intermediate thickness, for
[There are individual gage readings of 1.5 mils at spot 5 and 1.8 example 127 µm (5.0 mils), is also within ±10%. If the calibration
mils at spot 3, both of which are clearly less than 2.0 mils. This is verified on both the high and the low shims, it is almost always
is allowed because only the average of the three readings (i.e., verified on the intermediate value shim.
the spot measurement) must be greater than or equal to 2.0 NOTE: With some gages it may be more practical to adjust
mils.] the gage at the intermediate thickness (e.g., 127 µm shim) first
Since the structure used in this example is about 30 m2 (300 and then verify that the gage also reads the high and the low
ft 2), the procedure used to measure the film thickness of part A shims correctly.
must be applied to both part B and part C. The measured
thickness of part B must exceed the 64 µm (2.5 mils) specified
minimum as must the thickness of part C.
To monitor the thickness of this entire 30 m2 (300 ft 2)
structure, at least 45 individual gage readings must be taken,
from which 15 spot measurements are calculated. The five spot
measurements from each 10 m2 (100 ft2) part of the structure are
used to calculate the thickness of that part.

Appendix 2—Example of Verification of the

Calibration of Type 2 Gages Using Plastic
Shims
This example describes a method to check if a Type 2 gage
is properly calibrated.

Suppose the coating thickness is specified at 100 µm (4.0

mils). The Type 2 constant pressure probe gage being used has
been calibrated according to the manufacturer’s recommenda-
tion. Now its calibration over blast cleaned steel must be
verified. A test coupon which had been blast cleaned during the
time the structure was blasted and has a profile representative
of that under the coating is available. After selecting a 50 µm
(2.0 mil) and a 250 µm (10.0 mil) plastic shim, proceed to take
thickness readings of the shims lying on the bare blasted
surface.
Because of the randomized nature of a blast cleaned
surface, repeated readings may exhibit significant variation.
Thus, make at least 10 measurements on each shim and record
their averages. There is no need to keep track of each individual
reading. Many gages will compute the average for you.