Designation: D7438 − 13

Standard Practice for

Field Calibration and Application of Hand-Held Moisture
Meters1
This standard is issued under the fixed designation D7438; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.

1. Scope In-Line Dry Lumber Moisture Meters

1.1 This practice applies to the measurement of moisture 2.2 Other ASTM Sources:
content of solid wood, including solid wood products contain- ASTM Standards on Precision and Bias for Various
ing additives, that is, chemicals or adhesives, by hand-held Applications, 1992
moisture meters under conditions of end-use.
1.1.1 This practice includes calibration, use, and interpreta- 3. Terminology
tion of meters for conditions that relate to wood product
3.1 Definitions of Terms Specific to This Standard:
characteristics, such as nonuniform grain and growth ring
orientation, and to end-use process conditions, such as mois- 3.1.1 conductance meters—conductance meters are those
ture gradients. that measure predominantly ionic conductance between points
1.1.2 Meters employing differing technologies may not of applied voltage, usually direct current. Direct-current con-
provide equivalent readings under the same conditions. When ductance meters have been commonly referred to as
this practice has been applied, it is assumed that the referenced “resistance-type” meters. Most commercial conductance me-
meter is acceptable unless otherwise specified. Meters shall ters are high-input impedance (about 1012 Ω), wide-range (104
have been calibrated by Test Methods D4444. to 1012 Ω) ohmmeters. Their scales are calibrated to read
directly in moisture content (oven-dry mass basis) for a
1.2 This standard does not purport to address all of the particular calibration species and at a specific reference tem-
safety concerns, if any, associated with its use. It is the perature.
responsibility of the user of this standard to establish appro-
priate safety and health practices and determine the applica- 3.1.2 capacitive-admittance meters—capacitive-admittance
bility of regulatory limitations prior to use. meters transmit electromagnetic wave energy into the wood to
detect the influence of moisture in the wood on these waves as
2. Referenced Documents an estimate of moisture content. Wave energy is most often in
the radio frequency range; hand-held meters commonly are
2.1 ASTM Standards:2
placed directly on the wood surface.
D2915 Practice for Sampling and Data-Analysis for Struc-
tural Wood and Wood-Based Products
D4442 Test Methods for Direct Moisture Content Measure- 4. Significance and Use
ment of Wood and Wood-Base Materials 4.1 Hand-held meters provide a rapid means of sampling
D4444 Test Method for Laboratory Standardization and moisture content of wood-based materials during and after
Calibration of Hand-Held Moisture Meters processing to maintain quality assurance and compliance with
D4933 Guide for Moisture Conditioning of Wood and standards. However, these measurements are inferential; that
Wood-Based Materials is, electrical parameters are measured and compared against a
D6782 Test Methods for Standardization and Calibration of calibration to obtain an indirect measure of moisture content.
The electrical measurements are influenced by actual moisture
content, a number of other wood variables, environmental
1
This practice is under the jurisdiction of ASTM Committee D07 on Wood and conditions, geometry of the measuring probe circuitry, and
is the direct responsibility of Subcommittee D07.01 on Fundamental Test Methods design of the meter. The maximum accuracy can only be
and Properties.
Current edition approved April 1, 2013. Published May 2013. Originally obtained by an awareness of the effect of each parameter on the
approved in 2008. Last previous edition approved in 2008 as D7438–08. DOI: meter output and correction of readings as specified by these
10.1520/D7438-13.
2
test methods. Appendix X1 is a commentary that provides
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
explanation of the mandatory sections and discussion of
contact ASTM Customer Service at [email protected]. For Annual Book of ASTM
Standards volume information, refer to the standard’s Document Summary page on historical practices. Appendix X2 addresses the influence of
the ASTM website. process and wood variables.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States

1
 D7438 − 13
4.1.1 This practice provides for calibration and application mixtures of heartwood and sapwood. Further, these process
of wood products that contain commercial characteristics and variables may change or invalidate some of the calibration
that reflect the manufacturing environment. results obtained under laboratory conditions in Test Methods
4.2 Most uses of hand-held moisture meters employ correla- D4444.
tive (predictive) relationships between the meter reading and 6.2 Methods—The principles and procedures of calibration
wood areas or volumes that exceed that of the direct meter in Test Methods D4444 shall be applied to the degree possible
measurement (for example, larger specimens, pieces of lumber, and relevant to develop a meaningful relationship between
lots). The field calibration section of this practice anticipates meter readings and actual moisture content (MC).
the potential need for this type of sampling. These correlative 6.2.1 All field calibrations shall be referenced to direct MC
uses are examined in Appendix X3. measurements (Test Methods D4442).
6.2.2 Field calibration shall be carried out with meters that
5. Standardization
have been laboratory standardized and calibrated for appropri-
5.1 General—Standardization shall be performed to estab- ate wood variables, such as species and temperature using Test
lish the integrity of the meter and electrode under the field Methods D4444, and subsequently field standardized.
conditions of use. The meter circuit shall be tested by applying
the reference material in accordance with manufacturer’s 6.3 Field Variables—The calibration may be based on
recommendations, noting the corresponding meter response end-use environmental and product and process conditions that
value, and comparing with the manufacturer’s data. Standard- are more restricted than those evaluated by Test Methods
ization shall be done before calibration. If alternate electrodes D4444. In addition, the process conditions may produce
are to be used with a meter, standardization shall be done for all interactions that must be considered in the calibration.
electrode types and alternate assemblies. 6.3.1 Special care must be taken to minimize errors caused
5.1.1 Initially, standardization should be performed before by the influence of unintended wood variables, such as density
each period of use. The time interval may be extended if and temperature (uncorrected) on readings. Specimen size for
experience shows that the particular meter is stable for a longer field testing may be selected to represent the appropriate
time under equivalent use conditions. geometry of the target sample. Field meter readings are
5.1.2 Standardization procedures in the field will be affected conditional upon both the prior standardization and calibration
by the standardization performance of the meter during evalu- process, the influence of wood variables in the field test, and
ation under Test Methods D4444. The report of section 5.2.3 of application information supplied by the meter manufacturer.
Test Methods D4444 provides this information. 6.4 Calibration Steps—The field calibration shall be con-
5.2 The standardization shall be carried out with the ducted on specimens and in conditions that are representative
instrument, including electrodes, at the temperature of the of the process and are carefully documented. See Appendix X2
anticipated application. This shall include the range of antici- for discussion of process variables and wood characteristics.
pated conditions; the reference material shall maintain its 6.4.1 Sample Selection—The number of wood specimens
essential characteristics over this range. The sensitivity of this used for the calibration shall be selected following the concepts
standardization to temperature of the meter shall be part of the of Practice D2915, considering the variables to be represented
evaluation. and the desired precision of the calibration. For example, if the
5.2.1 If the environmental conditions change during the sample is to represent grain patterns, moisture gradients, etc.
usage period beyond those evaluated in the initial found in a lumber grade, these variables shall be considered in
standardization, the standardization shall be repeated. setting sampling criteria. (See also Test Methods D4444.)
5.2.2 If the manufacturer recommends an area, a method, or 6.4.2 Sample Preparation—While the sample may be in-
a standard specimen for standardization that does not reflect the tended to include process variables such as moisture gradients,
entire direct measurement area of the meter, this shall be noted temperature, etc., the measurement and subsequent preserva-
as the manufacturer’s recommendation. tion of these variables prior to and during meter measurement
5.2.3 Field standardization may be difficult to carry out shall be considered part of the sampling process. See Test
under some ambient field conditions and with the electrodes to Methods D4444 for discussion of other relevant issues.
be used. One example is the use in monitoring in-kiln perfor- 6.4.3 Testing—Field calibration shall be based on the rela-
mance. If the measurement conditions are difficult to reproduce tionship of the meter readings to Test Methods D4442 moisture
or are transient (for example, in a hot dry kiln), then it shall be measurement values. Because process conditions may be
understood that the validity of the meter readings are depen- transient (for example, temperature and moisture gradients, or
dent upon the laboratory standardization and manufacturer’s both), calibration that reflects these variables requires special
recommendations. treatment of specimens (such as subdividing specimens) or
additional equipment (such as temperature probes). Care shall
6. Calibration be taken to not distort the original specimen condition with
6.1 General—Under processing conditions, laboratory cali- these additional steps.
bration procedures maybe impractical, particularly because of 6.4.4 Determination of Corrections—To establish a correc-
moisture and temperature gradients, nonstandard temperatures, tion that reflects the influence of the measured variables, the
unverified species within commercial species groups, non principles of Test Methods D4444, section 6.2.4, shall be
straight-grain wood, and common production variables such as followed.

2
 D7438 − 13
6.5 Report—Useful application of field test calibration is electrode measurement, in-exact moisture gradients, and other
conditional upon the relevance of the test sample. specimen variables. Where these non-uniformities and uncer-
Consequently, accurate reporting of the wood and process tainties cannot be measured or corrected, their presence shall
variables (see 6.3 and 6.4) is critical. The report shall follow be noted in the report and quantified where possible.
the practice of Test Methods D4444, section 6.2.5.
6.5.1 Field samples often contain uncertainties with respect
to exact species or species mixtures, temperature at the point of

APPENDIXES

(Nonmandatory Information)

X1. COMMENTARY

INTRODUCTION

The purpose of this appendix is to supply auxiliary information on the basis for and practice of this
practice. It is organized with paragraphs that correspond by section number to those in the mandatory
text; text paragraphs needing no explanation are not listed. This concept permits changes at any time
in order to keep the practice current and to improve its usefulness.

This is a practice standard; thus, it describes and standardizes, to the degree possible, the calibration
and measurement practices that occur outside the environment of the testing laboratory.

TABLE X1.1
Section Comments
1.1 The principal concepts of this practice, as first incorporated in Practice D2016 in 1965 and then in subsequent editions of this practice, addressed
only meters based on the change of wood conductance or dielectric properties with moisture content. Specific electrode configurations were
anticipated, based on early commercial use. Meters were classified as “resistance-type” and “dielectric-type”; no provisions were made other elec-
trode configurations or measurement technologies. Meters are now classified as “conductance” rather than “resistance-type,” and “capacitive-
admittance” rather than “dielectric-type” to better reflect current understanding of the underlying physics of their function. The current practice
makes no distinction between meter measurement technologies for standardization and calibration requirements. Provision for unique characteris-
tics of measurement technologies is accommodated in Appendix X1 – Appendix X3. The use of “field” to describe calibrations and measurement
issues denotes conditions that cannot be controlled as in a laboratory, yet the conditions are very commonly the environment in which the meters
are used.

1.1.1 This practice targets use outside of the laboratory where controlled conditions are not usually possible. In addition, most commercial wood prod-
ucts are not “clear” and straight-grain and are heterogenous in other characteristics. Sampling is necessarily tied to commercial product descrip-
tions.

1.1.2 Requiring calibration under Test Methods D4444 ensures prior technical evaluation of a meter, with an associated report describing performance
under controlled conditions. Although the intended use may not adhere to these same conditions, the performance in the laboratory establishes
the minimum performance criteria for field use as well as reference points on sensitivity to variables such as species, density, temperature, etc.

3.1 This practice is designed to apply to meters using technologies other than the two technologies included in the section. Conductance and
capacitive-admittance meters are included because they provide the generic descriptions of principal, current commercial meters. Individual char-
acteristics of commercial meters are not intended to be covered in these generic descriptions. As other meter technologies are developed, more
generic descriptions should be added to this section.

4.1.1 Much of the content of this practice was incorporated in previous versions and drafts of Test Methods D4444, and some earlier in Practice D2016.
The mixing of test methods and practices in one standard is not desirable; thus, this practice attempts to capture the critical elements of the many
and varied commercial applications of hand-held meters while Test Methods D4444 concentrates on the base-line laboratory test methods. Spe-
cific issues of meter technology in use are covered in more detail in Appendix X2.

4.2 Correlative methods of data analysis are critical to many meter uses; however, they vary widely and are difficult to characterize as true calibration
of a meter. Consequently, Appendix X3 addresses both calibration of a meter and predictive uses as separate topics of standard practice.

5.2.3 The standardization of a meter under severe environmental conditions can be a serious operating issue. If the manufacturer’s recommendations
(for example, standard specimens or standard methods for standardization) cannot be followed, the correct operation of the meter is in question.

6.1 The essence of this practice is that many uses of meters demand performance on products and under conditions that are not covered by labora-
tory calibration (Test Methods D4444). Field calibration is the only recourse. It is emphasized that field-type calibrations often will apply very nar-
rowly to the conditions of that calibration and do not extend to other uses or conditions.

6.2.2 Although the calibrations carried out in the field may be the only ones of commercial interest because they reflect the actual operating conditions,
the meter must have the basis of laboratory calibration to ensure satisfactory operation under stable conditions and knowledge of response to
variables important in the field.

3
 D7438 − 13
TABLE X1.1 Continued
Section Comments
6.3 Common wood variables encountered when conducting a calibration on commercial samples include non-uniform grain and growth ring
orientation, moisture and density gradients within the measurement zone, mineral streaks, and “wet wood” (bacterially infected pockets). In many
cases, the size and shape of the wood specimens is different from that used in laboratory calibration because the goal of the field calibration may
be to calibrate the meter assembly for a particular application in which influence of specimen size and shape is an important element. These
factors pose a challenge to field calibration because they produce results conditional upon all the incorporated factors; the resultant calibration
may not extend to other uses. The results also have a component of sample “error” that may not be clearly identified (see Appendix X3). A
common example of a field calibration with a narrow scope is a calibration of a meter used with deep members (for example, 200 mm) dried in a
commercial kiln. A gradient is expected and accommodated in the calibration; however, this calibration may not be valid if the depth of the
specimens in the next kiln charge is different from the original or if the kiln schedule is changed to accommodate the change in size.

6.5 Because field calibrations can have a very narrow scope, the report must clarify the variables evaluated, the variables not controlled, the “grade”
description, etc. These provide boundaries within which the results may apply.

X2. METER USE

X2.1 General accompanied with documentation on whether the corrections

X2.1.1 Measurement of moisture content in end-use appli- are for either the dial calibration species, a specific species, or
cations requires consideration of meter technology, wood for a species market group. Where appropriate correction data
characteristics and environmental influences. Sampling and are not available, calibrate the meter in accordance with
related analysis are additional essential elements. Appendix X2 Section 6.
incorporates historical observations on end-use applications. X2.2.3.2 Other Species Correction Considerations—There
X2.1.2 Meter Technology—Different meter technologies re- are numerous species-related effects that may result in different
quire differing operating procedures in end-use. Some operat- meter readings for the same actual moisture content. These
ing variables, however, have a generally common influence on include wood property variations related to site or genetics and,
most meter/electrode assemblies. The following sections in- for some species, differences in between the heartwood and
clude both these generally common influences and those that sapwood portions. Quantifying these effects to a precision
are technology specific. sufficient to justify a separate correction, such as for
temperature, may be difficult. Similarly, species market groups
X2.2 Common Operating Variables (such as Hem-Fir and Spruce-Pine-Fir) may contain species
X2.2.1 Among the more common operating variables that that cannot be visually separated at the point of moisture
can be addressed in a generic fashion are environmental factor measurement or where such separation is impractical. In field
of temperature, the wood variable of species, the sensitivity to measurements where these influences cannot be practically
chemicals, and the issues of sampling. separated, make some judgment in defining the population that
X2.2.2 Temperature Influence and Corrections: encompass these sources of variation and calibrate the meter in
X2.2.2.1 Temperature Effect on Meter—Meter circuits can accordance with Section 6.
be temperature sensitive, therefore, frequent zero or span X2.2.4 Chemical Additives and Adhesive Influence and
adjustments, or both, may be necessary during use. The Corrections:
manufacturer should indicate the optimum range of tempera- X2.2.4.1 Chemicals—Wood products which have been
ture for operation of the meter without loss of accuracy due to
treated with preservatives, fire retardants, or dimensional
temperature. It is recommended that whenever possible, the
stabilization agents may give abnormal readings (usually high).
meter be equilibrated with the measurement environment
Of these chemicals, creosote and pentachlorophenol solutions
before readings are taken. The intent is that in no case
appear to have insignificant effects.3 However, salt solutions
temperature or humidity alter the operating characteristics of a
meter (that has been equilibrated and adjusted) to the degree may cause abnormally high readings that should be considered
that the accuracy is impaired. qualitative or semiquantitative at best. Conductance meters
X2.2.2.2 Temperature Correction—Temperature corrections having insulated pins can be used to measure MC of materials
are obtainable from manufacturer’s data, published data, or that have been surface-treated with chemicals provided that
using built-in adjustments in the meter. Temperature correc- confirmation is made of the accuracy through direct MC
tions require special care to obtain the wood (not air) determination (for example, Test Methods D4442).
temperature, and may be unreliable to correct some species. A (1) Specific Treatments—CCA-C treatment4 has been re-
reference temperature of 2°C shall be standard for zero ported to be less conductive than salt treatments, reducing the
correction. Clearly indicate the reference temperature at some error of readings of treated southern pine to about 2 % MC in
point on the meter. Always make temperature correction before
species correction.
3
X2.2.3 Species Influence and Corrections: James, W. L., “Effects of Wood Preservatives on Electric Moisture Meter
Readings,” U.S. Forest Service Research Note, FPL-0.06, 1965.
X2.2.3.1 Species Correction—Only use manufacturer’s 4
Richards, M. J., “Effect of CCA-C Wood Preservative on Moisture Content
data developed in accordance with acceptable calibration Readings by Electronic-Type Moisture Meter,” Forest Products Journal, Vol 40, No.
procedures for the particular meter. The data should be 2, pp. 29–33, 1990.

4
 D7438 − 13
the range of 12 to 24 % MC. New preservative treatments were point, an approximation can be obtained for average MC of the
introduced in 2004, replacing some traditional treatments; no cross section. Using the same principle, a circular cross section
data on the effect of these treatments on moisture meter has its average MC at one sixth to one seventh of the diameter.
readings is available. The parabolic generalization generally does not pertain if
X2.2.4.2 Adhesives—Adhesives may cause abnormally high lumber has been dried in conditions that induce steep moisture
readings in reconstituted wood products. Before any particular gradients (such as in drying above 100°C) or if the lumber is
meter is used in moisture sensing of any particular product known or thought to contain wet pockets or streaks. This can be
containing adhesives, its calibration must be demonstrated on examined by driving insulated pins to mid-thickness. With
that product. Re-calibration must be carried out following any thick (deep) members, the difference between one fourth and
change in processing conditions. The calibrations must be one fifth of the depth can produce significant differences in MC
consistent with these test methods. when a gradient exists; testing is necessary to verify the correct
X2.2.5 Sampling: depth.
X2.2.5.1 Sampling Plan—The goals of sampling (for X2.3.2.2 Moisture Content Range—The range of moisture
example, estimating mean MC or MC variability, wet spot contents that can be detected is from a minimum of 6 or 7 %
identification, measurement of within piece gradients, etc.) and MC to a maximum of 25 to 27 % MC (nominal value of the
the intended use of the information (for example, kiln fiber saturation point). Meter scales extend above this limit
performance, specification adherence, effect of wood variables, only to permit temperature corrections of moisture contents up
etc.) directly influence a sampling plan. If the moisture and to the fiber saturation point and do not imply reliability of
wood characteristics, or both, vary significantly within a piece, readings above the fiber saturation point.
more than one location on a piece will need to be sampled. The (1) “Hot-Metering”—One use of conductance meters is for
size of the lot and the number of sampling locations both “hot metering” of kiln-dried lumber during which readings are
influence the confidence with which conclusions may be drawn taken to determine if the load has reached the desired endpoint
from the resulting data. MC. However, such readings are subject to considerable error
(1) Sampling Location—Selection of the location for sam- because of “edge-readings,” assumptions of wood temperature,
pling should consider wood characteristics and possible mois- unknown moisture gradients, and temperature effects on the
ture gradients. For example, in selecting the location on a piece meter circuitry. A further use is for moisture measurement of
of lumber, take the readings at least 500 mm from the end and dry lumber that is exposed to below-freezing temperatures. As
in the center of the face. Readings should be taken in areas that with hot lumber, considerable errors are possible due to
are reasonably straight grain and free of characteristics such as assumptions of wood temperature, unknown moisture
knots that affect the moisture level and influence the reading, or gradients, and temperature effects on meter circuitry. Any use
both. of these meters outside of the normal range is contingent upon
X2.2.5.2 Lot Size—The number of readings per sample or correlative tests and careful adjustments for the environment
per lot should be selected for consistency with the desired effects.
accuracy. Practice D2915 provides guidance on selection of lot X2.3.2.3 Moisture Gradients—Unless the moisture distribu-
size. tion and measuring techniques are well understood, readings
can be easily misinterpreted. Three special problems should be
X2.3 Technology-Specific Considerations for End-Use considered:
X2.3.1 The majority of hand-held meters employ either (1) Non-Insulated Electrodes—See X2.3.2.5.
conductance or capacitive-admittance operating technologies. (2) Non-Parabolic Gradients—See X2.3.2.1 (1).
The following sections contain observations on operating (3) High Surface MC on Sample—High surface MC of the
characteristics of, and recommendations for, these technologies material from condensation, wetting, and high relative humid-
in end-use. ity can cause excessively high readings that are not represen-
X2.3.2 Use of Conductance Meters: tative of the overall moisture content of the wood if non-
X2.3.2.1 Electrode Sensing Region—Conductance moisture insulated pins are used. If non-insulated pins are used, a higher
meters respond to the moisture content between the electrodes. surface than core MC can cause a misleading reading at the
They can be used to determine “point” moisture content depth of the pin tips. This can be tested by noting the indication
directly, if insulated pins are used, or average moisture content at initial contact and as the pins are driven in.
indirectly. All readings should be taken with the pins aligned so X2.3.2.4 Drift—Direct current conductance meters may
that the current flow is parallel to the grain. Average moisture show appreciable drift toward lower MC when readings are
content can be obtained through the thickness by integrating taken at the upper portion of the MC range. If such drift occurs,
moisture content versus thickness. Under the following condi- take the reading as soon as possible after the pins are driven in
tions it can also be inferred from a single point measure. and voltage applied. Digital and analog meters may respond
(1) Obtaining an Average MC Reading—Wood of rectan- differently; contact the meter manufacturer for recommenda-
gular cross section tends to develop a parabolic gradient during tions.
drying (assuming that the maximum moisture content is below X2.3.2.5 Electrodes—Preferred electrodes for the conduc-
fiber saturation point (FSP). From the geometry of a parabola, tance meter for solid wood measurements are of a two-pin
the point of average MC lies between one fourth and one fifth type, insulated except for the tips. If non-insulated pins are
of the total thickness. Therefore, if the pins are driven to this used, the wood must be tested for surface moisture content (see

5
 D7438 − 13
X2.3.2.3 (3)). If any other electrode is used, such as four-pin polystyrene or polyurethane. Electrode conformance with the
for wood or eight-pin for veneer, the readings must be adjusted surface may present a special problem with warped specimens,
as specified by the manufacturer or incorporated into the scale especially if the electrode is rigid. If measurements must be
corrections. Different pin configurations should not be used made on warped specimens, one option is to take readings on
interchangeably on the same meter without the appropriate opposite sides of the specimen and use only the higher value.
corrections. (2) Estimation of Measurement Sensing Region—If deter-
(1) Pin Corrections—Acceptable corrections for solid mination of the measurement region of the meter electrode
wood are:5 reading (two-pin) = 0.29 + 0.91 (reading four-pin), assembly is desired, a traditional method of testing for sensi-
or reading (four-pin) = 1.1 (two-pin) – 0.32. These pin tivity has been taking readings at two thicknesses: 20 and 40
corrections must be made after the temperature correction and mm. If a thickness sensitivity is detected, intermediate values
before the species correction. between 20 and 40 maybe determined with other thicknesses of
(2) Non-Insulated Pins—Non-insulated pins have been a wood. Alternatively, a sufficient number of pieces (to minimize
tradition in industrial use. If “normal” drying gradients exist read-through) of thin material may be stacked and compressed
(driest at the surface; least dry at the center of the piece), this for readings. For best results, rearrange the pieces for several
usage may be satisfactory for many applications. To assume readings that can then be averaged.
such a gradient without verification, however, may result in X2.3.3.2 Moisture Content Range—The normal range of
significant error. (See X2.3.2.1 (1) and X2.3.2.3.) Non- moisture sensing in capacitive-admittance meters is from 0 %
insulated pins will bias the reading toward the highest moisture to fiber saturation point. Semi-quantitative readings above fiber
content in contact with the pins. saturation point are possible with the capacitive-admittance
(3) Extension Electrodes—Extension electrodes maybe meter. The scale may be in arbitrary units or direct-indicating
necessary for some end-uses. Nails may be used; however, for a particular reference species if calibration for the reference
unless they are insulated except at the tips, the precautions for species has been carried out according to acceptable calibration
non-insulated pins apply. procedures. If the species scale is used, the range of density of
(4) Implanted Electrodes—Special precautions are neces- the calibration species must be given. When correction data are
sary to minimize errors caused by changing electrode contact not available, the meter may be calibrated in accordance with
pressure and electrode-wood contact resistance, particularly the procedures of this practice.
when the electrodes are implanted in green wood to monitor
drying. It is especially important that d-c voltage not be applied X2.3.3.3 Moisture Gradients—Unless the moisture distribu-
continuously in order to minimize the buildup of contact tion and measuring techniques are well understood, readings
interfacial resistance from ion migration. Meter systems with can be easily misinterpreted. Measurements with a capacitive-
low frequency a-c, intermittent d-c, or switched d-c voltages admittance meter must be qualified by the moisture gradient
can virtually eliminate irreversible ionic migration. (determined independently with a conductance meter or other
(5) Surface Moisture on Electrode—Surface films of mois- means), depth of electric field penetration, thickness of
ture on the electrode, particularly from condensation (insulated material, surface condition (such as rough or planed), and
pin holder), may cause large errors. Keep electrodes clean, and electrode contact pressure and conformance. In addition, the
store and use under non-condensing conditions. special problems of condensation on the electrodes (see
X2.3.3.4) and moisture on the wood surface should be consid-
X2.3.3 Use of Capacitive-admittance Meters: ered.
X2.3.3.1 Electrode Sensing Region—Moisture content read- (1) High Surface Moisture Content on Specimen—High
ings will be affected by the electrode configuration and the surface moisture content of the material from condensation,
configuration of the measurement region of the electrode. wetting, or high relative humidity can cause excessively high
Laboratory standardization and calibration may be based on a readings.
specific reference standard and on a specific electrode configu- (2) Low Surface Moisture Content on Specimen—All
ration. Use of an inappropriate standard reference, a different capacitive-admittance meters have a biased, greater response to
electrode configuration, or an erroneous assumption of effec- moisture nearer the electrode; therefore, for samples that
tive measurement region may yield erroneous results and contain moisture gradients, values determined even after spe-
require confirmation of the measurement results by further test, cies and temperature corrections are qualitative at best, and no
or both. limits of accuracy can be implied. However, an increase in
(1) Measurement Region Undefined—If the measurement accuracy may be obtained if corrected readings taken on both
region of the meter electrode has not been defined or is sides of the material are averaged.
uncertain for the wood variable under test, minimize the
X2.3.3.4 Electrodes—Electrode configurations will interact
possibility of reading adjacent material (read-through) by
with specimen size and test objectives. The depth of field
either supporting the specimen ends to create an air gap of at
penetration for capacitive-admittance meters should be known,
least 25 mm below the sample or placing the sample on a
at least qualitatively (see also X2.3.3.1). Contact pressure of
similar thickness of non-hygroscopic, low density foam such as
the electrode with the specimen may affect the meter response;
the manufacturer’s recommendations on contact pressure
5
should be followed. The electrode sensing area should be small
Cech, M. Y., and Pfaff, F. “Moisture Content Correction Tables for Resistance-
Type Moisture Meters,” Canadian Forestry Service, Eastern Forest Products enough that the field will not extend beyond the specimen edge
Laboratory, Forestry Technical Report 7, 1975. and large enough to provide reasonable integration. (For

6
 D7438 − 13
example, if a nominal 2 by 4 is the smallest product size (1) Surface Moisture on Electrode—Surface films of mois-
anticipated, the electrode could be sized to integrate over most ture on the meter, particularly from condensation on the
of the wide face.) A symmetrical configuration is preferred for electrode, may cause abnormally high readings. Meters should
electrodes in order to minimize the effect of grain direction. be stored and used under non-condensing conditions.
Note that the basis of the meter calibration should be known,
should be based on the measurement region of the electrode,
and readings confined to specimens of this size or larger.

X3. CONCEPTS OF CALIBRATION AND PREDICTION IN USE OF HAND-HELD MOISTURE METERS

X3.1 Concepts of Calibration spacial volume of measurement. The most technically correct
X3.1.1 Calibration of any measurement instrument assesses precision and bias statement for a hand-held meter should be
the bias of the instrument relative to an accepted standard. This based on measurements within the meter’s spacial volume of
bias then can be accommodated in any subsequent use of the measurement. On the other hand, for some applications, the
data developed by the instrument to adjust as closely as most practical interest may be in the predictive relationship
feasible to the accepted standard response. The concept of between a meter’s limited reading and some larger specimen
calibration assumes that the “repeatability” of the instrument (for example, a length of lumber).
has been assessed with a standardization process against an X3.1.4.1 Calibration tests have an element of prediction,
accepted reference. even under the most careful experimental conditions. This
element contains the errors of measurement. The distinction
X3.1.2 Referencing a Standard Method—Calibration of made in X3.1.4 is for the more gross predictive situations
hand-held meters are most usually conducted with Test Meth- wherein non meter-measured moisture content (for example,
ods D4442, “Standard Test Methods for Direct Moisture the moisture content of a 20-ft piece of lumber) is estimated by
Content Measurement of Wood and Wood-Base Materials.” a meter reading (for example, one reading taken 3 ft from the
This standard has several methods for direct measurement of end of the piece).
moisture content. The method most often used for reference to
hand-held meters is Section 6—Method B—Oven-Drying X3.2 Calibration of Hand-Held Meters
(Secondary). In this method, the specimen size is “any conve- X3.2.1 The goal of calibrating a hand-held meter based on
niently sized piece of wood...” (see Test Methods D4442, measurement within the area of meter sensitivity is dependent
section 6.2). Consequently, Test Methods D4442 accepts any upon knowledge of this area and the ability to sample it for the
convenient sized specimen; thus, leaving the choice of the Test Methods D4442 measurement. An ideal calibration would
specimen to the user—presumably often the user of Test sample this area accurately; however, often some estimations
Methods D4444. are needed because the boundaries of the area may be
X3.1.3 Meter Response—Calibration of the hand-held me- imprecise, being affected by wood variables and the practical
ters meeting Test Methods D4444 criteria must reflect the issues of collecting the sample. The calibration resulting from
operating characteristics of the meter with special attention to this method relates directly the indication of the meter to the
the volume of a specimen within which the instrument actually MC determined from Test Methods D4442. Often this is an
responds to moisture and makes a “measurement.” This spacial oven dry reading that represents the average of the entire
volume of a specimen to which a meter will be sensitive can measurement region; however, the sample may be subdivided
vary widely between meters of different operating principles for gradient or other information.
(for example, capacitive-admittance versus conductance), me- X3.2.2 Area of Measurement Sensitivity—Defining the area
ters operating on the same principles but with differing of sensitivity of a meter is difficult, affected, as noted above, by
electrode sizes or spacings, meters with similar operating wood and sampling variables. The manufacturer may have
principles but differing parameters (for example, dielectric defined this area; if so, obtaining this information maybe an
meters operating at differing frequencies). optimum approach for most users. Note the restrictions placed
X3.1.4 Calibration versus Prediction—In the context of by the manufacturer upon the stated area: density, moisture
much moisture meter use, a distinction should be drawn content and other physical characteristics of the wood, equili-
between calibration of a hand-held meter, the determination of bration of specimens used in the determination, and so forth. If
the relationship between a meter reading and a Test Methods the manufacturer does not supply an area of sensitivity,
D4442 measurement—both within the meter’s spacial volume experiments can be conducted to ascertain this area. These
of measurement—and prediction, the inferential process of experimental procedures are beyond the scope of this practice;
estimating the moisture content of some volume of wood however, Appendix X2 of this practice briefly discusses the
different from and larger than, or both, that sensed by the meter. effective measurement regions of meter electrodes.
While both of these processes are legitimate goals, they are X3.2.3 Physical Sampling for D4442 Tests—To compare the
significantly different. Prediction, for example, introduces moisture content indicated by a moisture meter to that deter-
concepts of sampling which, in turn, can be complicated by mined by Test Methods D4442 in the measurement area of the
many wood variables that may not be within the meter’s meter, delicate sampling is required. The wood specimen will

7
 D7438 − 13
require some form of dissection, carefully preserving the sampling from the target population may also have a time
moisture within the dissected pieces. Including material from component, thus further complicating the sampling. One ex-
beyond the measurement region of the meter or, conversely, not ample is the sampling of wood recently removed from a dry
including all of the measurement region in the physical sample, kiln which has a continuously decreasing but still measurable
will bias the results of the comparison, perhaps in serious and moisture gradient.
an unknown quantitative degree, or both. Note the influence of X3.3.2.1 Examples of typical industrial target populations
insulated electrodes and other electrode geometries on sam- and sampling carried out to estimate moisture content are:
pling criteria. (1) Goal—Predict the moisture content of lumber leaving a
X3.2.4 Analysis—The calibration of a meter in the region of kiln while the lumber is still in the kiln.
measurement sensitivity is usually based on comparing the (a) Target Population—All the lumber in a kiln charge.
meter output to the Test Methods D4442 oven dry moisture Critical elements: All the lumber is one cross section (for
value as an average for the sensitivity region. If the region has example, 2×4); each kiln cart may have a different length; often
been subdivided for study of gradients or other variables, the a species group—making estimation of actual species ex-
reconstitution of the sample oven dry values to represent the tremely difficult; kiln charge dryness affected by imperfect
entire region must reflect the fact that the meter may not be baffling/air flow and heating; lumber is hot and each piece has
linear in response within the region. a moisture gradient.
X3.2.4.1 Precision and Bias—The measurements obtained (b) Sample—Exposed edges of stickered lumber within
in Appendix X3 provide a measure of the inherent measure- practical reach of kiln operator.
ment capability of a meter and can be used to estimate meter Critical elements: Assumption that pieces sampled along the
precision and bias. These characteristics provide the basis for edges of the loads represent the target in physical properties
the practical value of a meter; however, they often are not the affecting moisture measurement; sufficient sample size selected
ultimate values desired by many users. Section X3.3 discusses to include target characteristics.
the uses of meters to make moisture content predictions—for (c) Comments—A prediction method is required that re-
which most meters are used. flects the complexity of this problem—hot measurements must
be related to room temperature; edge measurement and gradi-
X3.3 Prediction of Moisture Content Using Hand-Held ents must be related to flat measurement and gradients. Study
Meters of the temperature and spacial effects can be carried out on the
X3.3.1 Practical interest in moisture content varies widely. same specimens; however, the overall goal will have a greater
Concerns may be focused small areas like the presence of “wet error due to the sampling attempting to represent the entire kiln
spots,” the influence of “kiln stickers” on moisture distribution charge. Note that the temperature correction is the only portion
in a piece of lumber, or the effect of rain penetration in a of this study that is directly related to meter performance;
structure. Often, however, the hand meter reading is expected virtually all of the other issues are related to sampling
to represent an area of wood much larger than that actually adequacy, geometry, and time factors. To address these, a
measured. One example is using one reading per specimen in carefully designed study is required.
a lot of lumber to assess conformance to a moisture specifica- (2) Goal—Predict the near maximum (specification) mois-
tion. Another example is taking a limited number of measure- ture content of a unit of lumber.
ments on a piece of lumber for correlation with the readings of (a) Target Population—A unit of 220 pieces of lumber
an in-line moisture meter that has a high frequency of reading selected by a third party to represent the mill production.
(see Test Methods D6782). If a piece of lumber were perfectly Critical elements: Assumption that the unit selected represents
uniform in moisture content and all physical variables in the mill production; this is not usually tested and is not a meter
wood were also perfectly uniform, the results of Section X3.2 issue. The specificity and statistical adequacy of the moisture
would apply directly to the larger sample—no major prediction specification are critical to adequate sampling and interpreta-
errors would be present. When the reality of the variability of tion of results. The uniformity of moisture measurement-
wood characteristics and moisture distributions are considered, affecting lumber characteristics throughout the population is
it is the wood sampling process, not meter characteristics, that important.
introduces the primary prediction errors. The desire to make (b) Sample—Pieces selected by the third party for test.
moisture predictions well beyond the measurement area re- Critical elements: The sample size from the unit and it’s
quires the use of predictive techniques to quantify the risks representativeness. The criteria used by the third party to select
associated with this essential process. the sample pieces and locations on the pieces to make moisture
X3.3.2 Defining the Sample—When the desire is to estimate measurements.
moisture content for wood beyond the measurement region of (c) Comments—A classic estimation study; meter perfor-
the meter, defining the sample to be measured for moisture— mance is a minor issue, assuming proper standardization and
the sample intended to represent the targeted moisture content calibration of the meter. The one exception is where the unit
of the larger, untested population—may be one of the more was selected in such a way that the meter sensitivity is biased
difficult steps in the calibration of a moisture estimating and uncorrected (for example, a density-related grading sys-
process. This target has technical characteristics that must be tem). If the specification is focused on a near maximum
appropriately replicated in the sample if the errors in estimation moisture content, rather than an average, estimation is more
are to be minimized. In many practical applications, the difficult. If the variability of moisture within pieces and

8
 D7438 − 13
between pieces is not known or well estimated and an measurement (Test Methods D4444) and a reference measure-
insufficient number of samples is selected, a simplistic analysis ment (Test Methods D4442). If the variability of one of the two
of sample results may have significant error. The assumption of significantly exceeds the other, the sample size will be deter-
the unit representing mill production is not tested. mined by the more variable of the two.
(3) Goal—Predict the average moisture content of a piece
of lumber. X3.3.4 Developing the Predictive Relationship—Sampling
(a) Target Population—A piece of lumber. and measuring moisture content following the procedures of
Critical elements: The moisture content may vary significantly X3.3.2 and X3.3.3 can yield either mean or near-maximum
within the length and cross section of the piece. The piece may values that characterize the population; however, these are
have varying physical characteristics that affect moisture meter single numbers and, aside from providing a ratio of meter to
measurement. oven-dry, for example, may not provide the insights desired on
(b) Sample—One or more locations along the piece. these two methods relate. To explore this relationship, the
Critical elements: Location of measurement relative to local individual data collected can be examined using regression and
wood characteristics such as compression wood, knots, slope analysis of variance techniques. This requires dispersion in the
of grain, wet spots, etc. Moisture gradients through the cross data; that is, there must be sufficient range in the variables of
section. The specification may be critical (that is, when an interest (moisture content, density, etc.) to make an effective
average is the goal, are wet spots included in the calculation?). analysis. Again, the issue is one of both meter performance and
Selecting an adequate number of measurement locations de- estimation technique. For example, if the goal is to examine the
pends on good estimates of moisture variability. effect of density on the relative performance of a meter versus
(c) Comments—Procedures that specify the number of oven-dry, a regression of the ratio of the two measurements
measurement locations may be arbitrary; that is, not based on taken at the same locations versus density could be instructive;
measurement or good estimation of within-piece variability. however, there must be sufficient variability in density over the
Experience with the moisture history of the piece may be range of measurement locations to make the comparison
critical (for example, the slowly declining gradient in recently meaningful. If the relationships of wood variables are only
kiln-dried material versus the possibility of post-dry wetting in moderately robust (for example, R2 ~ 0.6) care in selecting the
field exposure). X3.3.2.1 (3) is important for field calibration sample is needed to ensure the effect of the variables can be
of in-line moisture meters as well as in meeting acceptance adequately examined. Stratified sampling techniques may be
specifications. needed. Correct design of the experiment anticipates the
X3.3.3 Determining the Sample Size—Once the physical analysis to be used.
sample characteristics are defined, the sample size—the num- X3.3.4.1 Precision and Bias—The ASTM terms of precision
ber of specimens or measurement locations—is determined. If and bias apply directly to the performance of a meter only in
a mean value is to be estimated, the precision and bias to be the area of its actual measurement. A majority of moisture
associated with the estimate must be chosen. An example is to meter uses go beyond this measurement and apply some
accept a bias of 60.5 % MC with 90 % confidence. To make concept of predictive function, although not often so stated. In
the appropriate estimate, the expected variation of the mea- Appendix X3, we have introduced the often-applied concepts
surement is needed; if no prior estimates are available, pre- of prediction to which a meter is most often linked—the
liminary tests may be necessary. Practice D2915, section 3.4, “measurement” of the moisture content of an entire piece of
provides guidance for this calculation. If a near maximum lumber from a few actual measurements is, in actuality, not a
value is to estimated (for example, the upper 5th percentile of “measurement,” but a prediction ; the “measurement” (and
the population) so that an estimate may be made about the acceptance) of the moisture content of a unit of lumber from
percent of the population above a specification limit, a non- the actual measurement in a few locations on a small sub-
parametric estimate of the percentile is most often used. To sample of pieces. All of these practical uses depend on both the
make this estimate, Practice D2915, section 3.4.3, offers inherent precision and bias of the meter, but, often more
methods using a tolerance limit concept. importantly, on the nature of the predictive relationship.
NOTE X3.1—The Practice D2915 text usually refers to the lower Standard statistical methods use parameters, such as R, stan-
tolerance limit, whereas with moisture measurement it is more common to dard error, COV, etc., to demonstrate the effect of variables on
need the estimate of the upper tolerance limit. The same concepts apply. predictive relationships; however, it is not correct to confuse
X3.3.3.1 Both X3.3.2 and X3.3.3 emphasize identification the ASTM use of precision and bias of the instrument by
and selection of samples in order to obtain a credible estimate incorporating those more broad concepts of sampling and
of moisture content. These concepts apply to both a meter analysis into the measurement capability of the instrument.

9
 D7438 − 13

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org). Permission rights to photocopy the standard may also be secured from the ASTM website (www.astm.org/
COPYRIGHT/).

10