Designation: E 228 95
AMERICAN SOCIETY FOR TESTING AND MATERIALS
100 Barr Harbor Dr., West Conshohocken, PA 19428
Reprinted from the Annual Book of ASTM Standards. Copyright ASTM
Standard Test Method for
Linear Thermal Expansion of Solid Materials With a Vitreous
Silica Dilatometer1
This standard is issued under the fixed designation E 228; 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 (e) indicates an editorial change since the last revision or reapproval.
1. Scope
1.1 This test method covers the determination of the linear
thermal expansion of rigid solid materials over the temperature
range of 180 to 900C using vitreous silica push-rod or tube
dilatometers.
NOTE 1The temperature range for push-rod dilatometers can be
extended to 1600C by using high-purity alumina push-rod systems and
up to over 2500C using isotropic graphite systems. The precision and
bias of these systems is believed to be of the same order as that for silica
systems up to 900C. However, their precision and bias have not yet been
established over the relevant total range of temperature due to the lack of
well-characterized reference materials and the need for interlaboratory
comparisons.
1.2 For this purpose, a rigid solid is defined as a material
that, at test temperature and under the stresses imposed by
instrumentation, has a negligible creep or elastic strain rate, or
both, regarding significantly affecting the precision of thermal-
length change measurements. This includes metals, ceramics,
refractories, glasses, rocks and minerals, graphites, plastics,
cements, mortars, woods, and fiber, and other reinforced matrix
composites.
1.3 Many materials and certain material applications require
that detailed preconditioning and specific thermal test sched-
ules be followed for the correct evaluation of thermal expan-
sion. Since a general test method cannot cover all specific
requirements, details of this nature should be contained in the
relevant material specification.
1.4 The precision of this comparative test method is greater
than that of other push-rod dilatometery (for example, Test
Method D 696) and thermomechanical analysis (for example,
Test Method E 831) techniques but is significantly lower than
that of absolute methods such as interferometry (for example,
Test Method E 289). It is generally applicable to materials
having linear expansion coefficients above 5 m/mK and can
be used for lower expansion coefficient materials for which a
sufficient length of specimen is available.
1.5 Computer- or electronic-based instrumentation, tech-
niques, and data analysis systems equivalent to this test method
can be used. Users of the test method are expressly advised that
1
This test method is under the jurisdiction of ASTM Committee E-37 on
Thermal Measurements and is the direct responsibility of Subcommittee E37.05 on
Thermophysical Properties.
Current edition approved Sept. 10, 1995. Published December 1995. Originally
published as E 228 63 T. Last previous edition E 228 85 (1989).
1
all such instruments or techniques may not be equivalent. It is
the responsibility of the user to determine the necessary
equivalency prior to use. In the case of dispute only, the manual
procedures described herein are to be considered valid.
1.6 The values stated in SI units are to be regarded as the
standard.
1.7 This standard does not purport to address all of the
safety concerns, if any, associated with its use. It is the
responsibility of the user of this standard to establish appro-
priate safety and health practices and determine the applica-
bility of regulatory limitations prior to use.
2. Referenced Documents
2.1 ASTM Standards:
D 696 Test Method for Coefficient of Linear Thermal Ex-
pansion of Plastics2
E 220 Test Method for Calibration of Thermocouples by
Comparison Techniques3
E 289 Test Method for Linear Thermal Expansion of Rigid
Solids with Interferometry4
E 473 Terminology Relating to Thermal Analysis4
E 644 Test Methods for Industrial Resistance Thermom-
eters3
E 831 Test Method for Linear Thermal Expansion of Solid
Materials by Thermomechanical Analysis4
E 1142 Terminology Relating to Thermophysical Proper-
ties4
3. Terminology
3.1 DefinitionsThe following terms are applicable to this
test method and are listed in Terminologies E 473 and E 1142:
coeffcient of linear thermal expansion, thermodilatometry, and
thermomechanical analysis.
3.2 Definitions of Terms Specific to This Standard:
3.2.1 mean coeffcient of linear thermal expansion, amthe
average change in length relative to the length of the specimen
accompanying a change in temperature, between temperatures
T1 and T2 , expressed as follows:
am 5 @~L2 2 L1!/L0 ~T2 2 T1!# 5 ~DL/L0!/DT (1)
am is, therefore, obtained by dividing the linear thermal
2
Annual Book of ASTM Standards, Vol 08.01.
3
Annual Book of ASTM Standards, Vol 14.03.
4
Annual Book of ASTM Standards, Vol 14.02.

expansion (DL/L0) by the change of temperature (DT). It is
normally expressed as m/mK.
3.2.2 thermal expansivity, aTat temperature T, this is
calculated as follows:
1 limit ~L2 2 L1!
aT 5 L T2 T1 5 ~dL/dT!/Li ~T1 , T1 , T2!
i ~ T2 2 T1 !
and expressed as m/mK.
NOTE 2Thermal expansivity is sometimes referred to as the instan-
taneous coefficient of linear expansion.
3.2.3 vitreous silica dilatometera device that measures
the difference in linear thermal expansion between a test
specimen and the vitreous silica parts of the dilatometer.
3.3 Symbols:Symbols:
am 5 mean coefficient of linear thermal expansion,
(see 3.2.1), m/mK
aT 5 expansivity at temperature T (see 3.2.2), m/
mK
L0 5 original length of specimen at temperature T0,
mm
L1 5 length at temperature T1, mm
L2 5 length at temperature T2, mm
Li 5 length at a particular temperature Ti, mm
DL 5 change in length of specimen between tempera-
ture T1 and T2, m
(DL/L0)a 5 expansion indicated by the measurement trans-
ducer, m
T0 5 temperature at which initial length is L0, K
T1, T2 5 two temperatures at which measurements are
made, K
Ti 5 temperature at which length is Ti, K
DT 5 temperature difference between T2 and T1, K
m 5 measured expansion of the reference material,
m that might present difficulties to an individual not having the
t 5 true or certified expansion of the reference
material, m
s 5 assumed or known expansion of the vitreous
silica parts of the dilatometer, m
A,B 5 numerical constants (see 9.3.1).
4. Summary of Test Method
4.1 This test method uses a vitreous silica dilatometer of the
single push-rod or tube type to determine the change in length
of a solid material relative to that of the holder as a function of
temperature. A special variation of the basic configuration
known as a differential dilatometer is sometimes used. This
form is described briefly in Appendix X1.
4.2 The temperature is controlled either over a series of
steps or at a slow constant heating or cooling rate over the
entire range.
4.3 The linear thermal expansion and the coefficients of
linear thermal expansion are calculated from the recorded data.
5. Significance and Use
5.1 Coefficients of linear expansion are required for design
purposes and are used, for example, to determine thermal
stresses that can occur and cause failure of a solid artifact
composed of different materials when it is subjected to a
E 228
temperature excursion(s).
5.2 This test method is a reliable method of determining the
linear thermal expansion of solid materials.
5.3 For accurate determinations of thermal expansion, it is
absolutely necessary that the vitreous silica dilatometer be
(2) calibrated by using a reference material that has a known
reproducible thermal expansion. Table 1 and Table 2 contain
information on the current thermal expansion standard refer-
ence materials available for such purposes. Table 3 contains
information relating to other reference materials in current
general use.
5.4 The measurement of thermal expansion involves two
parameters: change of length and change of temperature. Since
measurements of the first parameter can be made by this test
method with good precision, it is essential that great attention
also be paid to the second. In order to ensure the necessary
uniformity in temperature of the specimen, it is essential that
the uniform temperature zone of the surrounding furnace or
environmental chamber be made significantly longer than the
length of the specimen. Temperature gradients depend on the
length/diameter ratio and insulation quality of the furnace.
They can be reduced by adjusting the furnace heater windings
such that they are closer at the ends than at the center. In
addition, for high-temperature operations, it is recommended
that a heavy metal liner and radiation shields be used. Tem-
perature control is best attained when the sensing element of
the controller is either the same as that for measuring the
specimen temperature (radiant heating furnace) or very close to
the heating element (resistive heated furnace).
5.5 This test method contains essential details of the design
principles, specimen configurations, and procedures for pro-
viding precise values of thermal expansion. It is not practical in
a test method of this type to try to establish specific details of
design, construction, and procedures to cover all contingencies
technical knowledge relating to the thermal measurements and
general testing practice. Standardization of the test method is
not intended to restrict further development of improved
methodology in any way.
5.6 The test method can be used for research, development,
specification acceptance, and quality control (QC) and quality
assurance (QA).
6. Interferences
6.1 Heating unannealed vitreous silica above 800C may
cause viscous flow and a time-dependent change in its thermal
expansion. The magnitude of these effects will depend on the
particular type of vitreous silica.
6.1.1 Recommended Annealing ProcedureAnnealing the
vitreous silica specimen and dilatometer parts in the following
steps may reduce any excessive lot-to-lot differences in linear
thermal expansion:
6.1.1.1 Heat at 100C/h to 1200C,
6.1.1.2 Hold at 1200C for 2 h,
6.1.1.3 Cool at 60C/h to 1000C,
6.1.1.4 Cool at 120C/h to 600C, and
6.1.1.5 Cool at 200C/h to room temperature.
6.2 Heating vitreous silica that is contaminated with alkali
compounds above 500C may cause it to crystallize. To prevent
2
 E 228
TABLE 1 Thermal Expansion of Various Standard Reference Materials
SRM 731 (Borosilicate Glass)A SRM 738 (Stainless Steel)A SRM 739 (Fused Silica)A
Temperature, K Expansion, 10 6
Expansion, 10 6
Expansion, 106
106 Expansivity 106 Expansivity 106 Expansivity
DL/L293 DL/L292 DL/L293
80 819 ... ... ... 1 0.07
100 771 2.64 ... ... 13 0.53
120 714 3.07 ... ... 22.5 0.38
140 649 3.43 ... ... 28.5 0.24
160 578 3.72 ... ... 32 0.10
180 501 3.97 ... ... 32.5 0.02
200 419 4.17 ... ... 31 0.13
220 334 4.34 ... ... 27.5 0.23
240 246 4.48 ... ... 22 0.32
260 155 4.60 ... ... 14 0.39
280 62 4.71 ... ... 6 0.45
293 0 4.78 0 9.76 0 0.48
300 34 4.82 69 9.81 ... ...
320 131 4.91 ... ... 13.5 0.53
340 230 4.99 466 10.04 24.5 0.56
380 432 5.11 872 10.28 47.5 0.60
420 638 5.19 1288 10.52 72 0.62
460 847 5.23 1714 10.76 97 0.63
500 1057 5.26 2149 11.00 122 0.63
540 1267 5.27 2593 11.23 ... ...
560 1372 5.27 ... ... 159 0.61
580 1478 5.27 3048 11.47 ... ...
600 1583 5.27 ... ... 183 0.59
620 1689 5.28 3511 11.71 ... ...
640 1794 5.29 ... ... 206 0.56
660 1900 ... 3984 11.95 ... ...
680 2007 ... ... ... 228 0.54
700 ... ... 4467 12.19 ... ...
720 ... ... ... ... 249 0.51
740 ... ... 4959 12.42 ... ...
760 ... ... ... ... 269 0.49
780 ... ... 5461 12.66 ... ...
800 ... ... ... ... 228 0.47
840 ... ... ... ... 307 0.44
880 ... ... ... ... 324 0.42
920 ... ... ... ... 340 0.40
960 ... ... ... ... 356 0.38
1000 ... ... ... ... 371 0.37
A
Available from NIST, Gaithersburg, MD. Values in table from relevant certificate for the reference material.

this, clean the component by immersion in an aqueous solution
containing 10 % hydrofluoric acid for 1 min followed by a
thorough rinse with distilled water. To prevent contamination
with alkali compounds do not touch the vitreous silica with the
hands after cleaning.
6.3 Inelastic creep of a specimen at elevated temperatures
can often be prevented by making its cross section sufficiently
large.
6.4 Care is necessary to control the temperature gradient in
long specimens.
6.5 Avoid moisture in the dilatometer, especially when used
at cryogenic temperatures.
6.6 A closed specimen holder is required when the dilatom-
eter is immersed in a liquid bath.
6.7 Support or hold the specimen in a position so that it is
stable during the test.
6.8 The specimen holder and push-rod shall consist of the
same type of vitreous silica.
6.8.1 When applying the system calibration outlined in
Section 9, a test run with a vitreous silica specimen should not
result in an apparent mean coefficient of linear thermal expan-
sion greater than 6 0.3 m/mK.
6.9 Since the precision and accuracy of the length measure-
ments are fixed for a specific apparatus, to obtain the same
percent uncertainty in the coefficient, a larger temperature
range must be used for low-expansion materials than that for
high expansion materials. Conversely, the percent uncertainty
of the coefficient will be much larger for the low-expansion
material if the same temperature range is used.
6.10 Conditioning of specimens is often necessary before
reproducible expansion data can be obtained. For example,
heat treatments are frequently necessary to eliminate certain
effects (strain, moisture, etc.) that may introduce length
changes not associated with thermal expansion.
7. Apparatus
7.1 Push-Rod Dilatometer System, consisting of the follow-
ing:
7.1.1 Specimen Holder and Push-Rod or Tube, both made of
annealed vitreous silica. Illustrations of typical tube and
rod-type configurations are given in Fig. 1. Paragraph 6.1.1
contains a recommended annealing procedure for vitreous
silica.
7.1.2 Furnace, Cryostat, and Bath, used for heating or
cooling the specimen uniformly at a controlled rate over the
temperature range of interest but not below 180C or above
900C.
7.1.3 Transducer, for example, a digital encoder, differential

3
 E 228
TABLE 2 Thermal Expansion of Standard Reference Material
Kieselglas K001A
Temperature, Expansion, Expansivity, 106a,
C 106DL/L0 106a (t) K1
170 2.16 60.67 0.565 60.019 0.013 60.004
160 7.35 60.55 0.475 60.016 0.046 60.003
140 15.15 60.44 0.300 60.011 0.108 60.003
120 19.80 60.43 0.160 60.007 0.165 60.004
100 21.67 60.43 0.030 60.005 0.217 60.004
80 21.09 60.43 0.085 60.004 0.264 60.006
60 18.36 60.41 0.185 60.004 0.306 60.007
40 13.76 60.38 0.272 60.004 0.344 60.010
20 7.57 60.36 0.346 60.005 0.378 60.019
10 3.94 60.34 0.379 60.005 0.394 60.038
0 0.00 60.34 0.409 60.005 ... ...
10 4.23 60.33 0.436 60.005 0.423 60.038
20 8.71 60.33 0.46 60.004 0.436 60.019
40 18.38 60.35 0.504 60.004 0.459 60.010
60 28.81 60.37 0.538 60.004 0.480 60.007
80 39.86 60.39 0.565 60.003 0.498 60.006 FIG. 1 Common Forms of Tube Types
100 51.37 60.41 0.585 60.003 0.514 60.005
120 63.22 60.43 0.599 60.003 0.527 60.004 accordance with Test Method E 220 or fine-gage resistance
140 75.31 60.44 0.608 60.003 0.538 60.004
160 87.52 60.44 0.612 60.003 0.547 60.003
thermometers calibrated according to Test Methods E 644.
180 99.78 60.45 0.613 60.003 0.554 60.003 7.1.4.2 Type E and T are recommended for the temperature
200 112.01 60.45 0.610 60.004 0.560 60.003 range of 190 to 350C, and Types K, S, and N are recom-
220 124.16 60.45 0.605 60.004 0.564 60.002
240 136.18 60.46 0.567 60.004 0.567 60.002
mended for temperatures of 0 to 900C. If Type K is used
260 148.03 60.47 0.588 60.004 0.569 60.002 continuously, regular checking of the calibration should be
280 159.68 60.48 0.577 60.004 0.570 60.002 undertaken to ensure that contamination or phase change
300 171.10 0.50 0.565 0.004 0.570 60.002
A
phenomena due to alloy component migration from the junc-
Fused silica available from Physikalisch Technische Bundesanstalt, Braunsch-
weig, Germany. Values in table from relevant certificate for the reference material.
tion has not occurred during testing.
7.1.4.3 In all cases in which thermocouples are used, they
TABLE 3 Linear Thermal Expansion
shall be referenced to 0C by means of an ice water bath or
equivalent electronic reference system insulated from the
Platinum,A Aluminum,B
Temperature, C
106DL/L0 106DL/L0 effects of temperature variations in the immediate surrounding
ambient environment.
195 1756.66 ...
150 1420.6 3430 7.1.4.4 Additional information relating to the precise deter-
100 1024.09 2550 mination of temperature in dilatometry is contained in Appen-
50 607.96 1550
0 176.2 460
dix X2.
20 0 0 7.2 Measurement Instrument, such as an index micrometer
50 268.06 710 or calipers capable of reading to at least6 25 m in order to
100 722.38 1900
200 1654.6 4450
determine the initial and final lengths of the test specimen (and
300 2612.01 7130 other relevant components, where required).
400 3692.18 10 050
500 4596.55 13 230 8. Test Specimens
600 5628.65 16 760
700 6692.81 ... 8.1 The specimen length shall be such that the accuracy of
800 7793.27 ... determining DL/L0 is at least6 20 m/m. Where possible, the
A
See Refs (1-3). specimen should be at least 256 0.1-mm long and between 5
B
See Refs (4-8). and 10 mm in diameter.
NOTE 3For example, if the transducer is accurate to 62 m, the
or dial-gage transformer for measuring the difference in length length must be at least 0.1 m.
between the specimen and specimen holder or probe with a 8.2 The end surfaces of the specimen (as well as the
precision of at least 6 1.3 m. The transducer shall be contacting surface of the specimen holder or rods, or both)
protected or mounted so that the maximum temperature change should not be rougher than approximately 10 m rms and shall
observed in the transducer during a test will affect the trans- be uniform in shape, as shown in Fig. 2. Avoid pointed
ducer readings by not more than 1.0 m. specimen ends that may deform during a test.
7.1.4 Temperature Measurement System, consisting of a
calibrated sensor or sensors, together with manual, electronic, 9. Calibration
or equivalent readout such that the indicated temperature can 9.1 Calibrate the transducer by imposing a series of known
be determined to better than 60.5C. displacements with a precision screw micrometer, end-gage
7.1.4.1 Since this test method is used over a broad tempera- blocks, or equally accurate device. This step can be omitted for
ture range, different types of sensors may have to cover the absolute transducers such as a digital encoder.
complete range. The common sensors are fine gage (32 AWG 9.2 Calibrate the temperature sensor according to Test
or smaller wire) or thin foil thermocouples calibrated in Methods E 220 and E 644 or the procedure recommended by

4

FIG. 2 Suggested Shapes of Specimens and Push-Rod Ends
the National Institute of Standards and Technology (NIST)
(1).5
9.3 As a total system, calibrate the dilatometer by measuring
at least one reference material of known thermal expansion.
Select a reference material with expansion as close as possible
to that of the sample material. Refer to Tables 1-3 regarding
available materials.
9.3.1 Either one of the following two calibration constants
indicated in Section 1 may be used:
A 5 1DL/L0!m @~DL/L0!t 2 ~DL/L0!s #
or
B 5 ~DL/L0!t 2 ~DL/L0!m
9.3.2 Calibrate the dilatometer using the same test condi-
tions and procedures as those used for the test specimen, for
example, the same specimen length, temperature program, and
gaseous environment.
10. Procedure
10.1 Each step may not need to be followed in the sequence
given and, depending on the apparatus, all steps may not be
needed.
10.2 Measure the initial (room temperature) length of the
specimen.
10.3 Place the specimen in the dilatometer after making
certain that all contacting surfaces are free of foreign material.
Ensure good seating of the specimen in a stable position.
10.4 Place the temperature sensor at the mid-length of the
specimen. It should preferably be in intimate contact with the
specimen, but it shall be as close to the specimen as possible.
The temperature sensor shall not restrict movement of the
specimen in the dilatometer.
10.5 Ensure that the transducer is in stable contact with the
probe(s) and specimen.
10.6 Place the assembled dilatometer into the furnace,
cryostat, or bath, and allow the temperature of the specimen to
reach equilibrium.
10.7 Record the initial readings of the temperature sensor,
T0, and the transducer, L0.
5
The boldface numbers in parentheses refer to the list of references at the end of
this test method.
E 228
10.8 Measure the expansion (contraction) of the specimen
over the desired range of temperature to the highest tempera-
ture level required.
10.8.1 The most precise measurement is achieved by heat-
ing (or cooling) the specimen successively to a number of
regular incremental constant temperatures and allowing the
system to equilibrate until the transducer reading attains a
constant value (variation less than 62.6 m). At that point, the
indicated temperature of the specimen shall not vary by more
than 62C, and the temperature gradient in the specimen shall
not exceed 0.5C/cm. The hold time is a function of the total
thermal mass (thermal capacity) of the dilatometer and speci-
mens and will vary for different temperatures. Readings of
temperature Ti and changed specimen length Li need to be
recorded at each constant temperature Ti.
10.8.2 Alternatively, heat or cool at some constant rate less
than 3C/min. When using this procedure, the mean tempera-
ture of the specimen will probably differ from the measured
temperature (lower on heating and higher on cooling), but the
measured expansion of the test specimen will be accurate if the
system is calibrated correctly with a reference material. Read-
ings of temperature and change of length should be recorded
continuously.
10.9 A retest should be considered if the length of the
specimen at the end of the test differs from that at the beginning
by more than 20 m/m, or this permanent deformation should
(3)
be taken into account when reporting the expansion values.
11. Calculation
(4)
11.1 Depending on the calibration constant used, calculate
the linear thermal expansion of the test specimen as follows:
~DL/L0! 5 A~DL/L0!a 1 ~DL/L0!s (5)
or
~DL/L0! 5 ~DL/L0!a 1 B (6)
11.2 Using the calculated values of the linear thermal
expansion, calculate the mean coefficients by dividing by the
appropriate temperature range T1 T2.
am 5 ~DL/L0!DT (7)
11.3 The following instantaneous coefficients are deter-
mined from the slopes, dL/dT, of the length versus temperature
curve at each appropriate temperature. The slope may be
determined either graphically from a plot or numerically from
an equation fit to the data.
aT 5 ~1/Li !dL/dT (8)
11.4 In calculation of the relevant quantities, use all avail-
able decimal places for each input parameter, measured to its
actual level of precision, through to the final result, and report
the data to three significant figures.
12. Report
12.1 Report the following information:
12.1.1 Description of the material, manufacturer, chemical
composition, and thermal and mechanical history;
12.1.2 Method of test specimen preparation, and axis orien-
tation, if material is anisotropic, together with details of any
subsequent thermal, mechanical, moisture, or other condition-
ing;
5
 E 228
12.1.3 Form and dimensions of the test specimen, including
initial length, L0, and reference temperature, T0;
12.1.4 Brief description of the apparatus, including dis-
placement and temperature measuring systems, estimation of
precision, heating and cooling rates, temperature controls, and
atmosphere;
12.1.5 Listing of the reference material(s) and procedure
used to calibrate the dilatometer system, including the trans-
ducer and temperature sensor;
12.1.6 Tabulation of the data, showing linear thermal ex-
pansion, test temperatures, and values for mean coefficient of
linear thermal expansion for selected temperature intervals;
12.1.7 Curves plotted as:
~DL/L0 versus T, am versus T, aT versus T
where:
am is computed from a common reference temperature to T;
12.1.8 Complete description of any unusual behavior of the
specimen, such as a permanent change in specimen length at
the reference temperature following testing, excessive oxida-
tion, scaling, discoloration, deformation, cracking, spalling,
etc., all of which may be of value in interpreting the test results;
and,
12.1.9 Any additional information required by a particular
material specification.
13. Precision and Bias
13.1 The precision and bias of determining both thermal
expansion and the mean coefficient of thermal expansion
depend on the simultaneous measurement of temperature and
relative length.
13.1.1 Random error is usually associated with the precision
and bias of repeated length and temperature measurements, but
other variables may also intrude on the measurements. For
instance, a specimen may change position repeatably or the
voltage applied to the measurement transducer may fluctuate.
13.1.2 Systematic error is usually larger and can result from
many sources. These include the following: accuracy of the
length and temperature measurements; deviation of the speci-
men mean temperature from that indicated by the sensor;
deviation from linearity of the transducer; temperature gradi-
ents between the specimen holder or probes, or both; difference
APPENDIXES
(Nonmandatory Information)
X1. DIFFERENTIAL DILATOMETER
X1.1 This variation of the basic configuration is undertaken
by the introduction of a second push-rod and a separate
reference specimen next to the test specimen. The transducer is
coupled to both push-rods such that it measures the differential
movement between them and not between one and the holder,
as in the regular configuration.
X1.2 The specimen and reference are made the same length
and are placed adjacent to each other, with one end of each
6
between the assumed or measured value of expansion of the
vitreous silica; and, effect of additional surface contacts be-
tween the specimen and the transducer. Little can be done to
improve the random errors once the transducer and temperature
sensor have been selected except to follow good experimental
practice. However, systematic errors can be reduced by careful
calibration of the individual components and the total system.
13.2 Repeat measurements on many different materials have
confirmed that the precision with which linear thermal expan-
sion (DL/L0) is measured with a vitreous silica dilatometer can
be estimated from the precision of the length and temperature
measurements. This estimate is determined from the following:
1
d~~DL/L0! 5 62/L0 @~dE!2 1 ~aT L0 d T!2#2 (9)
where:
dE and dT 5 precisions of the single measurements of
length and temperature. The error in measur-
ing L0 is usually so small that it does not
contribute to this estimate. For example,
L0 5 0.1 m, dE 5 61 m, dT 5 20 m/mC and dT 5 60.5C
The precision in determining (L/L0 ) is estimated to be as
follows:
2
d~DL/L0! 5 6 0.1 @~1!2 1 ~20 3 0.1 3 0.5!2#1/2 5 28 m/m (10)
It has been shown that the maximum difference to be
expected from repeated tests is estimated by the following:
2
d~DL/L0!max 5 6 L @dE 1 aT L0 dT# (11)
0
13.3 Tests of linear thermal expansion of borosilicate glass,
copper, and tungsten over the range of 25 to 400C have
indicated that a vitreous silica dilatometer can be accurate to
4 % at the 95 % confidence level when carefully calibrated.
This corresponds to a precision of approximately 0.8 % as
calculated from the example in 13.2.
14. Keywords
14.1 contraction; dilatometer; dilatometry; expansion; linear
thermal expansion; thermal expansivity; vitreous silica
contacting the lower surface in contact (or end plug, in a
horizontally oriented device) with the tube. Each has one
push-rod resting at the other end. The reference is usually
chosen to have expansion characteristics similar to those of the
test material so that the displacement difference between the
two push-rods is very small.
X1.3 Historically, the latter feature popularized this type of
configuration when transducers and recording means were
 E 228
limited. However, the advantages became limited to special
applications with the advent of improved transducers. For
example, when it is difficult to replicate thermal cycles with
high accuracy and good reproducibility, it is a major advantage
of the calibration of the device not to be undertaken during one
cycle with the test itself during another. By having the
reference adjacent to the test specimen, both will experience
identical thermal excursions, and this will eliminate cycle-to-
cycle uncertainty. Modern temperature controls greatly dimin-
ish the value of this feature. However, it still offers desirable
characteristics in some instances, most notably in cases in
which a large number of specimens of closely similar materials
are compared (such as in production control, screening studies,
etc.).
X1.4 The disadvantages are the strict and limited specimen
requirements (which must be equal to that of the reference),
continued cycling of the reference, larger space needed, and a
greater difficulty in detecting faulty operation.

X2. TEMPERATURE DETERMINATION

X2.1 Special care must be taken when the sensor(s) is not
in direct contact with the specimen, for the measured tempera-
ture can still be in error even when it is in good thermal
contact. This is caused by heat being conducted away from or
toward the specimen or the thermocouple junction, or both, by
the thermocouple wires. This heat flow can be minimized by
reducing the temperature gradient along the thermocouple
wires near the junction. For example, some of the wire is
looped once or twice within an area that has a temperature very
close to that of the specimen. The temperature gradient is
thereby shifted to an area between an intermediate point on the
wire and the outside of the furnace.
X2.2 When the junction is not in direct contact with the
specimen, the temperature measurements may also be in error
when heating or cooling. The magnitude of this sensor lead or
lag is usually not great, but it will be dependent on its distance
from the specimen, the specimen size, the emittance and
thermal diffusivity of the material, and the heating or cooling
rate.
X2.3 However, the sensor lead and lag can be determined
in the following manner:
X2.3.1 Where the specimen to be measured is known or
shown to be reversible in its thermal expansion behavior, heat
the interferometer at a constant rate over the temperature range
of interest, and then cool through this region at the same rate.
A plot of the expansion versus apparent temperature will
exhibit a temperature displacement between the heating and
cooling data. One half of this displacement parallel with the
temperature axis is the lead and lag correction. Subtract this
correction from the apparent temperature on heating and add
on cooling. If the specimen is not reversible in thermal
expansion behavior, another material of similar emittance and
thermal diffusivity and known to be reversible may be substi-
tuted to determine the approximate lead and lag.

REFERENCES

(1) Hahn, T. A., and Kirby, R. K., Thermal Expansion of Platinum 293 to
1900 K, AIP Conference Proceedings, No. 3, 1972.
(2) Nix, F. C., and MacNair, D., Thermal Expansion of Pure Metals,
Physical Review, Vol 61, 1942, p. 74.
(3) White, G. K., Expansion Coefficients of Coppers at 283K and Oxygen
Atmospheres, Journal of Physics, Vol 26, 1972, p. 30.
(4) Nix, F. C., and MacNair, D., Thermal Expansion of Pure Metals,
Physical Review, Vol 60, 1941, p. 597.
(5) Simmons, R. D., and Bullutfi, R. W., Measurements of Equilibrium
Vacancy Concentrations in Aluminum, Physical Review, Vol 117,
1960, p. 52.
(6) Fraser, D. B., and Hollis-Hallet, A. C., Proceedings of the 7th
International Conference on Low Temperature Physics, 1961, p. 689.
(7) Altman, H. W., Rubin, T., and Johnson, H. L., Coefficients of Thermal
Expansion of Solids at Low Temperatures, Ohio State University
Cryogenic Laboratory Report OSU-TR-264-27, 1954.
(8) Hidnert, P., and Krider, H. S., Thermal Expansion Measurements,
Journal of Research National Bureau of Standards, Vol 48, 1952, p.
209.

The American Society for Testing and Materials 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 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, 100 Barr Harbor Drive, West Conshohocken, PA 19428.