Volume 08 - Mechanical Testing and Evaluation Part 10 pdf

39. K. B. Yoon, A. Saxena, and D. L. McDowell, Influence of Crack-Tip Cyclic Plasticity on Creep-Fatigue
Crack Growth, Fracture Mechanics: Twenty Second Symposium, STP 1131, ASTM, 1992, p 367
40. A. Saxena and B. Gieseke, Transients in Elevated Temperature Crack Growth, International Seminar on
High Temperature Fracture Mechanics and Mechanics, EGF-6, Elsevier Publications, 1990, p iii–19
41. N. Adefris, A. Saxena, and D.L. McDowell, Creep-Fatigue Crack Growth Behavior in 1Cr-1Mo-0.25V
Steels I: Estimation of Crack Tip Parameters, J. Fatigue Mater. Struct., 1993
42. A. Saxena, Limits of Linear Elastic Fracture Mechanics in the Characterization of High-Temperature
Fatigue Crack Growth, Basic Questions in Fatigue, Vol 2, STP 924, R. Wei and R. Gangloff, Ed.,
ASTM, 1989, p 27–40
43. “Practices of Load Verification of Testing Machines,” E 4 94, Annual Book of Standards, Vol 3.01,
ASTM, 1994
44. A. Saxena, R.S. Williams, and T.T. Shih, Fracture Mechanics—13, STP 743, ASTM, 1981, p 86
45. “Test Method for Plane-Strain Fracture Toughness of Metallic Materials,” E 399, Annual Book of ASTM
Standards, Vol 3.01, ASTM, 1994, p 680–714
46. A. Saxena and J. Han, “Evaluation of Crack Tip Parameters for Characterizing Crack Growth Behavior
in Creeping Materials,” ASTM Task Group E24-04-08/E24.08.07, American Society for Testing and
Materials, 1986
47. H.H. Johnson, Mater. Res. Stand., Vol 5 (No. 9), 1965, p 442–445
48. K.H. Schwalbe and D.J. Hellman, Test Evaluation, Vol 9 (No. 3), 1981, p 218–221
49. P.F. Browning, “Time Dependent Crack Tip Phenomena in Gas Turbine Disk Alloys,” doctoral thesis,
Rensselaer Polytechnic Institute, Troy, NY, 1998
50. W.R. Caitlin, D.C. Lord, T.A. Prater, and L.F. Coffin, The Reversing D-C Electrical Potential Method,
Automated Test Methods for Fracture and Fatigue Crack Growth, STP 877, W.H. Cullen, R.W.
Landgraf, L.R. Kaisand, and J.H. Underwood, Ed., ASTM, 1985, p 67–85
51. P.K. Liaw, A. Saxena, and J. Schaefer, Eng. Fract. Mech., Vol 32, 1989, p 675, 709
52. P.K. Liaw and A. Saxena, “Remaining-Life Estimation of Boiler Pressure Parts—Crack Growth
Studies,” Electric Power Research Institute, EPRI CS-4688, Project 2253-7, final report, July 1986
53. P.K. Liaw, M.G. Burke, A. Saxena, and J.D. Landes, Met. Trans. A, Vol 22, 1991, p 455
54. P.K. Liaw, G.V. Rao, and M.G. Burke, Mater. Sci. Eng. A, Vol 131, 1991, p 187
55. P.K. Liaw, M.G. Burke, A. Saxena, and J.D. Landes, Fracture Toughness Behavior in Ex-Service Cr-
Mo Steels, 22nd ASTM National Symposium on Fracture Mechanics, STP 1131, ASTM, 1992, p 762–

789
56. P.K. Liaw and A. Saxena, “Crack Propagation Behavior under Creep Conditions,” Int. J. Fract., Vol 54,
1992, p 329–343
57. W.A. Logsdon, P.K. Liaw, A. Saxena, and V.E. Hulina, Eng. Fract. Mech., Vol 25, 1986, p 259
58. A. Saxena, P.K. Liaw, W.A. Logsdon, and V.E. Hulina, Eng. Fract. Mech., Vol 25, 1986, p 289
59. V.P. Swaminathan, N.S. Cheruvu, A. Saxerna, and P.K. Liaw, “An Initiation and Propagation Approach
for the Life Assessment of an HP-IP Rotor,” paper presented at the EPRI Conference on Life Extension
and Assessment of Fossil Plants, 2–4 June 1986 (Washington, D.C.)
60. N.S. Cheruvu, Met. Trans. A, Vol 20, 1989, p 87
61. R. Viswanathan, Damage Mechanisms and Life Assessment of High-Temperature Components, ASM
International, 1989
62. C.E. Jaske, Chem. Eng. Prog., April 1987, p 37
63. P.K. Liaw, A. Saxena, and J. Schaefer, Creep Crack Growth Behavior of Steam Pipe Steels: Effects of
Inclusion Content and Primary Creep, Eng. Fract. Mech., Vol 57, 1997, p 105–130

Impact Toughness Testing

Introduction
DYNAMIC FRACTURE occurs under a rapidly applied load, such as that produced by impact or by explosive
detonation. In contrast to quasi-static loading, dynamic conditions involve loading rates that are greater than
those encountered in conventional tensile tests or fracture mechanics tests. Dynamic fracture includes the case
of a stationary crack subjected to a rapidly applied load, as well as the case of a rapidly propagating crack under
a quasi-stationary load. In both cases the material at the crack tip is strained rapidly and, if rate sensitive, may
offer less resistance to fracture than at quasi-static strain rates. For example, values for dynamic fracture
toughness are lower than those for static toughness (K
Ic
) in the comparison shown in Fig. 1.

Fig. 1 Comparison of static (K
Ic

), dynamic (K
Id
), and dynamic-instrumented (K
Idi
) impact fracture
toughness of precracked specimens of ASTM A 533 grade B steel, as a function of test temperature. The
stress-intensity rate was about 1.098 × 10
4
MPa · s
-1
(10
4
ksi · s
-1
) for the dynamic tests and
about 1.098 × 10
6
MPa · s
-1
(10
6
ksi · s
-1
) for the dynamic-instrumented tests. Source: Ref 1
Because many structural components are subjected to high loading rates in service, or must survive high
loading rates during accident conditions, high strain rate fracture testing is of interest and components must be
designed against crack initiation under high loading rates or designed to arrest a rapidly running crack.
Furthermore, because dynamic fracture toughness is generally lower than static toughness, more conservative
analysis may require consideration of dynamic toughness.
Measurement and analysis of fracture behavior under high loading rates is more complex than under quasi-

static conditions. There are also many different test methods used in the evaluation of dynamic fracture
resistance. Test methods based on fracture mechanics, as discussed extensively in other articles of this Section,
produce quantitative values of fracture toughness parameters that are useful in design. However, many
qualitative methods have also been used in the evaluation of impact energy to break a notched bar, percent of
cleavage area on fracture surfaces, or the temperature for nil ductility or crack arrest. These qualitative tests
include methods such as the Charpy impact test, the Izod impact test, and the drop-weight test. Other less
common tests are the explosive bulge test, the Robertson test, the Esso test, and the Navy tear test (described in
the 8
th
Edition Metals Handbook, Volume 10, p 38–40).
This article focuses exclusively on notch-toughness tests with emphasis on the Charpy impact test. The Charpy
impact test has been used extensively to test a wide variety of materials. Because of the simplicity of the
Charpy test and the existence of a large database, attempts also have been made to modify the specimen,
loading arrangement, and instrumentation to extract quantitative fracture mechanics information from the
Charpy test. Other miscellaneous notch-toughness test methods are also discussed in this article.
Reference cited in this section
1. Use of Precracked Charpy Specimens, Fracture Control and Prevention, American Society for Metals,
1974, p 255–282

Impact Toughness Testing

History of Impact Testing
Before fracture mechanics became a scientific discipline, notched-bar impact tests were performed on
laboratory specimens to simulate structural failures, eliminating the need to destructively test large engineering
components. The simulation of structural component failure by notched-bar impact tests is based on severe
conditions of high loading rate, stress concentration, and triaxial stress state. These tests have been extensively
used in the evaluation of ductile-to-brittle transition temperature of low- and medium-strength ferritic steels
used in structural applications such as ships, pressure vessels, tanks, pipelines, and bridges.
The initial development of impact testing began around 1904 when Considére discovered and noted in a
published document that increasing strain rate raises the temperature at which brittle fracture occurs. In 1905

another Frenchman, George Charpy, developed a pendulum-type impact testing machine based on an idea by
S.B. Russell. This machine continues to be the most widely used machine for impact testing. In 1908 an
Englishman by the name of Izod developed a similar machine that gained considerable popularity for a period
of time but then waned in popularity because of inherent difficulties in testing at temperatures other than room
temperature.
Impact testing was not widely used, and its significance not fully understood, until World War II when many
all-welded ships were first built (approximately 3000 of them). Of these 3000 ships, approximately 1200
suffered hull fractures, 250 of which were considered hazardous. In fact, 19 or 20 of them broke completely in
two. These failures did not necessarily occur under unusual conditions; several occurred while the ships were at
anchor in calm waters. In addition to ship failures, other large, rigid structures, such as pipelines and storage
tanks, failed in a similar manner. All failures had similar characteristics. They were sudden, had a brittle
appearance, and occurred at stresses well below the yield strength of the material. It was noted that they
originated at notches or other areas of stress concentration, such as sharp corners and weld defects. These
failures were often of considerable magnitude: in one case a pipeline rupture ran for 20 miles.
The Naval Research Laboratory, along with others, launched a study of the cause of these fractures. It was
noted that often, but not always, failures occurred at low temperatures. More detailed historical research
revealed that similar failures had been recorded since the 1800s but had been largely ignored. The results of this
study renewed interest, and further investigation revealed that materials undergo a transition from ductile
behavior to brittle behavior as the temperature is lowered. In the presence of a stress concentrator such as a
notch, it takes little loading to initiate a fracture below this transition temperature, and even less to cause such a
fracture to propagate. These transitions were not predictable by such tests as hardness testing, tensile testing, or,
for the most part, chemical analysis, which were common tests of the times. It was then discovered that a
ductile-to-brittle transition temperature could be determined by impact testing using test specimens of uniform
configuration and standardized notches. Such specimens were tested at a series of decreasing temperatures, and
the energy absorbed in producing the fracture was noted. The Charpy pendulum impact testing machine was
used. At first, test results were difficult to reproduce. The problem was partly resolved by producing more
uniformly accurate test equipment. The notch most often used was of a keyhole type created by drilling a small
hole and then cutting through the test bar to the hole by sawing or abrasive cutting. It was soon found that by
using specimens with sharper notches, better-defined transition temperatures that were more reproducible could
be determined. A well-defined notch with a V configuration became the standard. Steels in particular could then

be tested and the ductile-to-brittle transition temperature obtained.
Two problems remained. First, testing machines had to be standardized very carefully or the results were not
reproducible from one machine to another. The other problem was that the transition temperature found by
testing small bars was not necessarily the same as that for full-size parts.
Fortunately, the problem with standardization was resolved by the Army. They learned that impact testing was
a necessity for producing successful armor plate and gun tubes. Research at the Watertown Arsenal resulted in
the development of standard test specimens of various impact levels. The Army made these available to their
various vendors so that the vendors could standardize their own testing machines. This program was so
successful that such specimens were made available to the public, at a nominal charge, starting in the 1960s.
Next, the manufacturers of testing equipment were pressured into making equipment available that would meet
these exacting standards.
The problem of differing transition temperatures for full-size parts and test specimens was discovered when a
series of full-size parts was tested using a giant pendulum-type impact machine and these results were
compared with those determined using small standard test bars made from the same material. A partial solution
to this problem was the development of the drop-weight test (DWT) and the drop-weight tear test (DWTT).
These tests produced transition temperatures similar to those found when testing full-size parts. Unfortunately,
such tests are adaptable only for plate specimens of limited sizes and have not become widely used.
The Charpy V-notch test continues to be the most used and accepted impact test in use in the industry.
However, the restricted applicability of the Charpy V-notch impact test has been recognized for many years
(Ref 2). Charpy test results are not directly applicable for designs, and the observed ductile-to-brittle transition
depends on specimen size. Nonetheless, the Charpy V-notch test is useful in determining the temperature range
of ductile-to-brittle transition.
Reference cited in this section
2. C.E. Turner, Impact Testing of Metals, STP 466, ASTM, 1970, p 93

Impact Toughness Testing

Types of Notch-Toughness Tests
In general, notch toughness is measured in terms of the absorbed impact energy needed to cause fracturing of
the specimen. The change in potential energy of the impacting head (from before impact to after fracture) is

determined with a calibrated dial that measures the total energy absorbed in breaking the specimen. Other
quantitative parameters, such as fracture appearance (percent fibrous fracture) and degree of
ductility/deformation (lateral expansion or notch root contraction), are also often measured in addition to the
fracture energy. Impact tests may also be instrumented to obtain load data as a function of time during the
fracture event. In its simplest form, instrumented impact testing involves the placement of a strain gage on the
tup (the striker).
Many types of impact tests have been used to evaluate the notch toughness of metals, plastics, and ceramics. In
general, the categories of impact tests can be classified in terms of loading method (pendulum stroke or drop-
weight loading) and the type of notched specimen (e.g., Charpy V-notch, Charpy U-notch, or Izod). The
following descriptions briefly describe the key types of impact tests that are used commonly in the evaluation
of steels or structural alloys.
The Charpy and Izod impact tests are both pendulum-type, single-blow impact tests. The principal difference,
aside from specimen and notch dimensions, is in the configuration of the test setup (Fig. 2). The Charpy test
involves three-point loading, where the test piece is supported at both ends as a simple beam. In contrast, the
Izod specimen is set up as a cantilever beam with the falling pendulum striking the specimen above the notch
(Fig. 2b).

Fig. 2 Specimen types and test configurations for pendulum impact toughness tests. (a) Charpy method.
(b) Izod method
The Charpy V-notch test continues to be the most utilized and accepted impact test in use in the industry. It is
written into many specifications. While this test may not reveal exact ductile-to-brittle transition temperatures
for large full-size parts, it is easily adaptable as an acceptability standard on whether or not parts are apt to
behave in a brittle manner in the temperature range in which they are likely to be used.
The drop-weight test is conducted by subjecting a series (generally four to eight) of specimens to a single
impact load at a sequence of selected temperatures to determine the maximum temperature at which a specimen
breaks. The impact load is provided by a guided, free-falling weight with an energy of 340 to 1630 J (250 to
1200 ft · lbf) depending on the yield strength of the steel to be tested. The specimens are prevented by a stop
from deflecting more than a few tenths of an inch.
This is a “go, no-go” test in that the specimen will either break or fail to break. It is surprisingly reproducible.
For example, Pellini made 82 tests of specimens from one plate of semikilled low-carbon steel. At -1 °C (30 °F)

and 4 °C (40 °F), all specimens remained unbroken. At -7 °C (20 °F), only one of 14 specimens broke;
however, at -12 °C (10 °F), 13 of the 14 specimens broke. At temperatures below -12 °C (10 °F), all specimens
broke.
The drop-weight tear test (DWTT) uses a test specimen that resembles a large Charpy test specimen. The test
specimen is 76 mm (3 in.) wide by 305 mm (12 in.) long, supported on a 254 mm (10 in.) span. The thickness
of the specimen is the full thickness of the material being examined. The specimens are broken by either a
falling weight or a pendulum machine. The notch in the specimen is pressed to a depth of 5 mm (0.20 in.) with
a sharp tool-steel chisel having an angle of 45°. The resulting notch root radius is approximately 0.025 mm
(0.001 in.). One result of the test is the determination of the fracture appearance transition curve. The “average”
percent shear area of the broken specimens is determined for the fracture area neglecting a region “one
thickness” in length from the root of the notch and “one thickness” from the opposite side of the specimen.
These regions are ignored because it is believed that the pressing of the notch introduces a region of plastically
deformed material which is not representative of the base material. Similarly the opposite side of the specimen
is plastically deformed by the hammer tup during impact. The fracture appearance plotted versus temperature
defines an abrupt transition in fracture appearance. This transition has been shown to correlate with the
transition in fracture propagation behavior in cylindrical pressure vessels and piping.

Impact Toughness Testing

Charpy Impact Testing
As previously noted, the specimen in the Charpy test is supported on both ends and is broken by a single blow
from a pendulum that strikes the middle of the specimen on the unnotched side. The specimen breaks at the
notch, the two halves fly away, and the pendulum passes between the two parts of the anvil. The height of fall
minus the height of rise gives the amount of energy absorption involved in deforming and breaking the
specimen. To this is added frictional and other losses amounting to 1.5 or 3J (1 or 2 ft · lbf). The instrument is
calibrated to record directly the energy absorbed by the test specimen.
Methods for Charpy testing of steels are specified in several standards including:
Designation
Title
ASTM E 23

Standard Test Methods for Notched Bar Impact Testing of Metallic Materials
BS 131-2
The Charpy V-Notch Impact Test on Metals
BS 131-3
The Charpy U-Notch Impact Test on Metals
BS 131-6
Method for Precision Determinations of Charpy V-Notch Impact Energies for Metals

ISO 148
Steel—Charpy Impact Test (V-Notch)
ISO 83
Steel—Charpy Impact Test (U-Notch)
DIN-EN 10045

Charpy Impact Test of Metallic Materials
These standards provide requirements of test specimens, anvil supports and striker dimensions and tolerances,
the pendulum action of the test machine, the actual testing procedure and machine verification, and the
determination of fracture appearance and lateral expansion.
The general configuration of the Charpy test, as shown in Fig. 3 for a V-notch specimen, is common to the
requirements of most standards for the Charpy test. Differences between ASTM E 23 and other standards
include differences in machining tolerances, dimensions of the striker tip (Fig. 4), and the ASTM E 23
requirements for testing of reference specimens. The most pronounced difference between standards is the
different geometry for the tip of the striker, or tup. The tup in the ASTM specification (Fig. 4a) is slightly flatter
than in many other specifications (Fig. 4b). From a comparison of results from Charpy tests with the two
different tup geometries, differences appeared more pronounced for several steels at impact energies above 100
J (74 ft · lbf) (Ref 3). From this evaluation, a recommendation was also made to use the sharper and smoother
tup (Fig. 4b) if the national standards are unified further.

Fig. 3 General configuration of anvils and specimen in Charpy test

Fig. 4 Comparison of striker profiles for Charpy testing. (a) ASTM E 23. (b) Other national and
international codes: AS1544, Part 2; BS 131, Part 2; DIN 51222; DS10 230; GOST 9454; ISO R148; JIS
B7722; NF A03-161; NS 1998; UNI 4713-79. Source: Ref 3
There are also three basic types of standard Charpy specimens (Fig. 5): the Charpy V-notch, the Charpy U-
Notch, and the Charpy keyhole specimen. These dimensions are based on specifications in ASTM E 23, ISO
148, and ISO 83. The primary specimen and test procedure involves the Charpy V-notch test. Other Charpy-
type specimens are not used as extensively because their degree of constraint and triaxiality is considerably less
than the V-notch specimen.

Fig. 5 Dimensional details of Charpy test specimens most commonly used for evaluation of notch
toughness. (a) V-notch specimen (ASTM E 23 and ISO 148). (b) Keyhole specimen (ASTM E 23). (c) U-
notch specimen (ASTM E 23 and ISO 83)
The Charpy V-notch impact test has limitations due to its blunt notch, small size, and total energy measurement
(i.e., no separation of initiation and propagation components of energy). However, this test is used widely
because it is inexpensive and simple to perform. Thus, the Charpy V-notch test commonly is used as a
screening test in procurement and quality assurance for assessing different heats of the same type of steel. Also,
correlation with actual fracture toughness data is often devised for a class of steels so that fracture mechanics
analyses can be applied directly. Historically, extensive correlation with service performance has indicated its
usefulness.
The keyhole and U-notches were early recognized (1945) as giving inadequate transition temperatures because
of notch bluntness. Even the V-notch does not necessarily produce a transition temperature that duplicates that
of a full-size part. Under current testing procedures, the Charpy V-notch test is reproducible and produces close
approximations of transition temperatures found in full-size parts. It is widely used in specifications to ensure
that materials are not likely to initiate or propagate fractures at specific temperature levels when subjected to
impact loads.
Equipment
Charpy testing requires good calibration methods. Machine belting should be examined regularly for looseness,
and broken specimens should be examined for unusual side markings. Anvils should also be examined for wear.
Testing Machines. Charpy impact testing machines are available in a variety of types. Some are single-purpose
machines for testing Charpy specimens only. Others are adaptable to testing Izod and tension impact specimens

also. They are offered in a range of loading capacities. The most common of these capacities are 325 and 160 J
(240 and 120 ft · lbf). Some machines have variable load capabilities, but most are of a single-fixed-load type.
When purchasing or using a machine, be sure that the available loading is such that specimens to be tested will
break with a single blow, within 80° of the machine capacity (as shown by the scale on the machine).
While loading capacity depends on the anticipated strength of specimens to be tested, the maximum value of
such specimens is the principal consideration. Very tough specimens may stop the hammer abruptly without
breaking. A number of such load applications have been known to cause breakage of the pendulum arm. On the
other hand, lower-capacity machines may be more accurate and more likely to meet standardization
requirements. For most ordinary steel testing applications, the machine with a capacity of 160 J (120 ft · lbf)
makes a good compromise choice. Testing of a large number of very tough specimens may require a machine
with a capacity of 325 to 400 J (240–300 ft · lbf).
Charpy impact machines are of a pendulum type. They must be very rigid in construction to withstand the
repeated hammering effect of breaking specimens without affecting the operation of the pendulum mechanism.
The machine must be rigidly mounted. Special concrete foundations are sometimes used, but at least the
machine must be bolted down to an existing concrete foundation, which should be a minimum of 150 mm (6
in.) thick. The pendulum should swing freely with a minimum of friction. Any restriction in movement of the
pendulum will increase the energy required to fracture the specimen. This produces a test value that is higher
than normal. There will always be small effects of this type, and they are usually compensated for, along with
windage friction effects, by scale-reading adjustments built into the equipment.
While the pendulum must be loose enough to swing freely with little friction, it must not be loose enough to
produce inaccuracies, such as nonuniform striking of the specimen. The components must be sturdy enough to
resist deformation at impact. This is particularly true of the anvil and pendulum. It is important that the
instrument be level. Some machines have a built-in bubble-type level. Others have machined surfaces where a
level can be used. In operation, the pendulum is raised to the proper height and held by a cocking mechanism
that can be instantly released.
ASTM E 23 specifies that tests should be made at velocities between 3 and 6 m/s (10 and 20 ft/s) and that this
is defined as “the maximum tangential velocity of the striking member at the center of strike.” When hanging
freely, the striking tup of the pendulum should be within 2.5 mm (0.10 in.) of touching the area of the specimen
where first contact will be made. The anvil that retains the test specimen must be made such that the specimen
can be squarely seated. The notch must be centered so that the pendulum tup hits directly behind it.

Most impact testing machines have scales that read directly in foot-pounds (scales also may read in degrees).
As noted, the scale can be adjusted to compensate for windage, pendulum friction, and other variations. The
scale should read zero when the pendulum is released without a specimen being present. Pendulum and anvil
design, configuration, and dimensions are important. It is also important that the broken specimens be able to
fly freely without being trapped in the anvil by the pendulum. Proper anvil design, such as that shown in Fig. 6,
can minimize jamming.

Fig. 6 Typical anvil arrangement with modification that reduces the possibility of jamming
Specimens. As previously noted, there are three commonly used standard Charpy impact test specimens, which
are similar except for the notch (Fig. 5). The V-notch bar is the most frequently used specimen, although some
specific industries still use the other types of test bars. The steel casting industry, for instance, uses the keyhole-
notch specimen more frequently. There are also many varieties of subsize specimens that should be used only
when insufficient material is available for a full-size specimen, or when the shape of the material will not allow
removal of a standard specimen.
It is important that specimens be machined carefully and that all dimensional tolerances be followed. Care must
be exercised to ensure that specimens are square. It is easy to grind opposite sides parallel, but this does not
ensure squareness. The machining of the notch is the most critical factor. The designated shape and size of the
notch must be strictly followed, and the notch must have a smooth (not polished) finish. Special notch-
broaching machines are available for V-notching. A milling machine with a fly cutter can also be used.
In preparing keyhole-notch specimens, the hole should be drilled at a low speed to avoid heat generation and
work hardening. Use of a jig with a drill bushing ensures accuracy. After the hole has been drilled, slotting can
be done by almost any method that meets specifications, but care should be exerted to prevent the slotting tool
from striking the back of the hole. In all cases it is desirable to examine the notch at some magnification. A
stereoscopic microscope or optical comparator is suitable for this examination. In fact, a V-notch template for
use with the optical comparator can be used to ensure proper dimensions.
Specimens must generally be provided with identification markings. This is best done on the ends of the
specimen. In preparing specimens where structural orientation is a factor (e.g., rolling direction of wrought
materials), such orientation should be taken into consideration and noted, because orientation can cause wide
variations in test results. If not otherwise noted, the specimen should be oriented in the rolling direction of the
plate (forming direction of any formed part) and the notch should be perpendicular to that surface (orientation

A in Fig. 7). This produces maximum impact values. All notching must be done after any heat treatment that
might be performed.

Fig. 7 Effect of specimen orientation on impact test results
While correlation exists between full-size specimens and subsize specimens, such correlation is not direct.
Many specifications (ASTM and ASME, for example) specify differing acceptable values for various specimen
sizes (Table 1).
Table 1 Conversion table for subsize Charpy impact-test specimens
Minimum average impact
strength for three specimens
Minimum impact strength for one
specimens or for set of three specimens
Size of
specimens
(a)
, mm
J ft · lbf J
ft · lbf
10 × 10 (full size) 20.3 15.0 13.6
10.0
10 × 7.5 16.9 12.5 11.5
8.5
10 × 5 13.6 10.0 9.5
7.0
10 × 2.5 6.8 5.0 4.7 3.5
(a) Insofar as possible, full-size Charpy keyhole specimens should be used. However, where absolutely
necessary, it is permissible to use specimens with width (in direction of the length of the notch; see ASME
Section VIII, U-84, Unfired Pressure Vessels) reduced in accordance with the above tabulation.
Calibration. ASTM E 23 goes into considerable detail to ensure proper calibration of testing machines. Other
relevant standards for qualification or calibration of the test machines are:

ASTM E
1236
Standard Practice for Qualifying Charpy Impact Machines as Reference Machines
BS 131-7
Verification of the Test Machine Used for Precision Determination of Charpy V-Notch
Impact
BS-EN
10045-2
Charpy Impact Test on Metallic Materials Part 2: Method for the Verification of Impact
Testing Machines
ISO 148-2 Metallic Materials—Charpy Pendulum Impact Test Part 2: Verification of Test
Machines
These publications should be consulted for a basic understanding of machine calibration. Calibration and test
variables are also reviewed in Ref 4 and 5. These publications help identify causes of improper results.
Standard test bars for calibration can be purchased from: Director, Army Materials and Mechanics Research
Center, Attention AMXMR-MQ, Watertown, MA 02172 (formerly known as the Watertown Arsenal). Standard
specimens are tested as per instructions, and the results, along with a filled-out questionnaire and the broken
specimens, are returned. A report is then sent stating if the machine meets calibration standards and, if not, what
should be done to ensure qualification.
Test Method
Once the equipment has been properly set up and calibrated and the specimens have been correctly prepared,
testing can be done. Prior to each testing session, the pendulum should be allowed at least one free fall with no
test specimen present, to confirm that zero energy is indicated. Specimen identification and measurements are
then recorded along with test temperature. The pendulum is cocked, and the specimen is carefully positioned in
the anvil using special tongs (Fig. 8) that ensure centering of the notch. The quick-release mechanism is
actuated, and the pendulum falls and strikes the specimen, generally causing it to break. The amount of energy
absorbed is recorded (normally in foot-pounds), and this data is noted adjacent to the specimen identification on
the data sheet. The broken specimens are retained for additional evaluation of the fracture appearance and for
measurement of lateral expansion where required. The broken halves are often placed side by side, taped
together, and labeled for identification.

Fig. 8 Use of tongs to place a specimen in a Charpy impact testing machine for testing
The release mechanism must be consistent and smooth. Test specimens must leave the impact machine freely,
without jamming or rebounding into the pendulum; requirements on clearances and containment shrouds are
specific to individual machine types. The test specimen must be accurately positioned on the anvil support
within 5 s of removal from the heating (or cooling) medium; requirements for heating time depend on the
heating medium. Identification marks on test specimens must not interfere with the test; also, any heat treatment
of specimens should be performed prior to final machining.
A daily check procedure of the apparatus must be conducted to ensure proper performance. Verification of the
testing system is required using Army Materials and Mechanics Research Center (AMMRC) standardized
specimens; verification should be completed at least once a year or after any parts are replaced or any repairs or
adjustments are made to the machine. An operational testing sequence is recommended, as well as specifics on
dial energy reading, lateral expansion measurement (technique and measuring fixture), and fracture appearance
estimation.
Test Temperature. Specimen temperature can drastically affect the results of impact testing. If not otherwise
stated, testing should be done at temperatures from 21 to 32 °C (70–90 °F). Much Charpy impact testing is done
at temperatures lower than those commonly designated as room temperature. Of these low-temperature tests,
the majority are made between room temperature and -46 °C (-50 °F), because it is within this range that most
ductile-to-brittle transition temperatures occur. A certain amount of testing is also done down to -196 °C (-320
°F) for those materials that may be used in cryogenic service. Some additional testing (mainly research) is done
at the liquid helium and liquid hydrogen temperatures (-269 and -251 °C, or -452 and -420 °F). Such testing
requires special techniques and will not be discussed here. For testing at temperatures down to or slightly below
-59 °C (-75 °F), ethyl alcohol and dry ice are most commonly used. This combination solidifies at around -68
°C (-90 °F). A suitable insulated container should be used to cool the test specimens (a container insulated with
a layer of styrofoam works fine). A screen-type grid raised at least 25 mm (1 in.) above the bottom of the
container allows cooling liquid to circulate beneath the specimens. A calibrated temperature-measuring device,
such as a low-temperature glass or metal thermometer or a thermocouple device, should be placed so as to read
the temperature near the center of a group of specimens being cooled. The solution should be agitated
sufficiently to ensure uniformity of bath temperature. The cooling liquid should cover the specimen by at least
25 mm (1 in.). The specimen-handling tongs should be placed in the same cooling bath as the specimens. When

the specimens have been placed in the alcohol bath along with the tongs, chips of solid CO
2
(dry ice) can be
added and the solution agitated. Experience will dictate the amount of dry ice required to reach a certain
temperature. Once the temperature is reached, it seems to hold steady with only an occasional addition of a
small chip of dry ice. The specimens in a liquid bath should be held within +0 and -1.5 °C (+0 and -3 °F) of test
temperature for at least 5 min prior to testing. The specimens should then be removed one at a time with the
cooled tongs and tested within 5 s of removal from the bath. Watch the temperature between tests because the
tongs can raise the bath temperature if left out of the bath too long. The commercial cooling baths that are
available range from insulated stainless steel containers to containers with self-contained refrigeration units.
Also available are thermocouple devices that can be placed in the cooling bath and will give a digital
temperature readout. Dry ice cannot be stored for any length of time, but there is a device that produces
“instant” dry ice from a CO
2
compressed gas bottle. Testing between -59 and -196 °C (-75 and -320 °F)
requires a liquid medium that will not solidify at these temperatures. Various liquids are available. One that has
been successfully used is isohexane (adequate ventilation should be provided and care exercised to avoid
inhalation of the volatile organic fumes). Liquid nitrogen replaces the dry ice as a coolant material, and the
procedure is then similar to that for dry ice and alcohol. It is wise to keep handy a large, easy-to-handle piece of
metal to serve as a temperature moderator in case the temperature becomes lower than desired. It can be
plunged into the bath and, acting as a heat sink, can cause the temperature to rise quickly.
High-Temperature Testing. Occasionally, high-temperature impact testing is performed. This can be done using
an agitated, high-flashpoint oil (heat treating quenching oils may work) or other liquid medium that is stable at
the desired test temperature. The bath and specimens are then held at temperature in a furnace or oven for at
least 10 min prior to testing.
Test Results
Results of impact testing are determined in three ways. In the first method, already discussed, they can be read
directly from the testing machine (in joules or foot-pounds). This is the most commonly specified test result. It
is desirable to test three specimens at each test temperature; the average value of the three is the test result used.
If a minimum test value is specified for material acceptance, not more than one test result of the three should

fall below that value. If the value of one of the three specimens is about 6 J (5 ft · lbf) lower than the average, or
lower than the average value by greater than of the specified acceptance value, the material should be either
rejected or retested. In retesting, three additional specimens must be tested, and all must equal or exceed the
specified acceptance value. Since it is often required or important to determine the ductile-to-brittle transition
temperature, impact test results are plotted against test temperature. Somewhere in that transition zone between
the high-energy and low-energy values is an energy value that can be defined as the transition temperature.
When the transition is very pronounced, this value is easily determined. However, because the more common
case is a less sharply defined transition, an energy value may be specified below which the material is
considered to be brittle (below the ductile-to-brittle transition temperature). Such a value may vary with
material type and requirements, but the value of 20 J (15 ft · lbf) is often used as a specified value.
Fracture Appearance Method. Other methods of specifying ductile-to-brittle transition temperature are
sometimes presented along with the energy values obtained. The first of these auxiliary tests is the fracture-
appearance method. The fractured impact bars are examined and the fractures compared with a series of
standard fractures or overlays of such fractures. By this method the percentage of shear fracture is determined.
The amount of shear fracture can also be determined in another way. This is done by carefully measuring the
dimensions of the brittle cleavage exhibited on the specimen fracture surface (Fig. 9), and then referring to
Table 2. These methods are described in detail in ASTM A 370. The percentage of shear can be plotted against
test temperature and the transition temperature can be ascertained using the shear percentage value specified.
Table 2 Tables of percent shear for measurements made in both inches and millimeters for impact-test
specimens
Because these tables are set up for finite measurements or dimensions A and B (see Fig. 9), 100% shear is to be
reported when either A or B is zero.
Dimension A, in.
Dimension
B, in.
0.05 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38
0.40
0.05 98 96 95 94 94 93 92 91 90 90 89 88 87 86 85 85
84
0.10 96 92 90 89 87 85 84 82 81 79 77 76 74 73 71 69

68
0.12 95 90 88 86 85 83 81 79 77 75 73 71 69 67 65 63
61
0.14 94 89 86 84 82 80 77 75 73 71 68 66 64 62 59 57
55
0.16 94 87 85 82 79 77 74 72 69 67 64 61 59 56 53 51
48
0.18 93 85 83 80 77 74 72 68 65 62 59 56 54 51 48 45
42
0.20 92 84 81 77 74 72 68 65 61 58 55 52 48 45 42 39
36
0.22 91 82 79 75 72 68 65 61 57 54 50 47 43 40 36 33
29
0.24 90 81 77 73 69 65 61 57 54 50 46 42 38 34 30 27
23
0.26 90 79 75 71 67 62 58 54 50 46 41 37 33 29 25 20
16
0.28 89 77 73 68 64 59 55 50 46 41 37 32 28 23 18 14
10
0.30 88 76 71 66 61 56 52 47 42 37 32 27 23 18 13 9
3
0.31 88 75 70 65 60 55 50 45 40 35 30 25 20 18 10 5 0
Dimension A, mm
Dimension
B, mm
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
10
1.0 99 98 98 97 96 96 95 94 94 93 92 92 91 91 90 89 89 88
88
1.5 98 97 96 95 94 93 92 92 91 90 89 88 87 86 85 84 83 82

81
2.0 98 96 95 94 92 91 90 89 88 86 85 84 82 81 80 79 77 76
75
2.5 97 95 94 92 91 89 88 86 84 83 81 80 78 77 75 73 72 70
69
3.0 96 94 92 91 89 87 85 83 81 79 77 76 74 72 70 68 66 64
62
3.5 96 93 91 89 87 85 82 80 78 76 74 72 69 67 65 63 61 58
56
4.0 95 92 90 88 85 82 80 77 75 72 70 67 65 62 60 57 55 52
50
4.5 94 92 89 86 83 80 77 75 72 69 66 63 61 58 55 52 49 46
44
5.0 94 91 88 85 81 78 75 72 69 66 62 59 56 53 50 47 44 41
37
5.5 93 90 86 83 79 76 72 69 66 62 59 55 52 48 45 42 38 35
31
6.0 92 89 85 81 77 74 70 66 62 59 55 51 47 44 40 36 33 29
25
6.5 92 88 84 80 76 72 67 63 59 55 51 47 43 39 35 31 27 23
19
7.0 91 87 82 78 74 69 65 61 56 52 47 43 39 34 30 26 21 17
12
7.5 91 86 81 77 72 67 62 58 53 48 44 39 34 30 25 20 16 11
6
8.0 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Fig. 9 Sketch of a fractured impact test bar. The method used in calculating percent shear involves
measuring average dimensions A and B to the nearest 0.5 mm (0.02 in.) and then consulting a chart
(Table 2) to determine the percent shear fracture. (Courtesy of ASTM)

Unlike Charpy energy, fracture appearance is indicative of how a specimen failed. It is therefore useful when
attempting to correlate results of Charpy testing with other toughness test methods that use different specimen
geometries and loading rates. However, the fracture-appearance method can also be subjective. In one round-
robin test survey of 20 specimens (Ref 6), results showed that agreement was best when operators are
experienced, samples are close to the fracture-appearance transition, and when simple, two-dimensional figures
are used for assessment.
Lateral-Expansion Method. The other auxiliary method of determining transition temperature is the lateral-
expansion method. This procedure is based on the fact that protruding shear lips are produced (perpendicular to
the notch) on both sides of each broken specimen. The greater the ductility, the larger the protrusions. This
lateral expansion can be expressed as a measure of acceptable ductility at a given test temperature. The broken
halves from each end of each specimen are measured. The higher values from each side are added together, and
this total is the lateral-expansion value. A minimum value of lateral expansion must be specified as a transition
value. These test results are then plotted against test temperature and a curve interpolated. The impact energy
(in joules or foot-pounds) is also reported. These methods are described in detail in ASTM A 370 and E 23.
Applications
Test criteria for Charpy V-notch impact testing usually involve:
• A minimum impact energy value
• Shear appearance of fractured test bars expressed in percent
• Lateral expansion
For steels, the minimum acceptable values most commonly specified for these three evaluation methods are,
respectively: 20 J (15 ft · lbf), 50% shear, and 1.3 mm (50 mil). As a general rule of thumb, Charpy V-notch
impact strengths of 14 J (10 ft · lbf) and lower are likely to initiate fractures. An impact strength of 27 J (20 ft ·
lbf) is likely to propagate brittle fracture once initiated, and values well above 27 J (20 ft · lbf) are necessary to
arrest fracturing once it has been initiated.
Charpy impact testing does not produce numbers that can be used for design purposes, but is widely used in
specifications such as ASTM A 593, “Specification for Charpy V-Notch Testing Requirements for Steel Plates
for Pressure Vessel.” Other applications are briefly described below.
Nuclear Pressure Vessel Design Code. For nuclear pressure vessels, the American Society of Mechanical
Engineers (ASME) Boiler and Pressure Vessel Code (Ref 7) and the Code of Federal Regulations (Ref 8)
currently use fracture mechanics principles that dictate toughness requirements for pressure vessel steels and

weldments. The specified toughness requirements are obtained using Charpy V-notch test specimens coupled
with the nil-ductility transition temperature (NDTT) per ASTM E 208. The actual approach involves a
reference temperature, designated RT
NDT
, and the reference fracture toughness curve, K
IR
. The reference
fracture toughness curve defined in Appendix G, Section III, of the ASME Code uses an experimentally
determined relationship between toughness and temperature that is adjusted along the temperature axis
according to an index reference temperature.
The reference toughness curve, K
IR
, is assumed to describe the minimum (lower bound) fracture toughness for
all ferritic materials approved for nuclear pressure boundary applications having a minimum specified yield
strength of 345 MPa (50 ksi) or less. The value of RT
NDT
is obtained by measuring the drop-weight nil-ductility
transition temperature and performing standard Charpy V-notch tests. The nil-ductility transition temperature is
determined initially, and then a set of three Charpy V-notch specimens is tested at a temperature that is 33 °C
(60 °F) higher than the nil-ductility transition temperature to measure the temperature, T
CV
, which ensures an
increase in toughness with temperature. Charpy energies of 68 J (50 ft · lbf) and lateral expansion of 0.89 mm
(35 mil) are used to ensure this condition.
The nil-ductility transition temperature becomes the RT
NDT
temperature if the Charpy results equal or exceed
the above limits. If the Charpy values at T
CV
or the nil-ductility transition temperature plus 33 °C (60 °F) are

lower than required, additional Charpy tests should be performed at higher test temperatures, usually in
increments of 5.6 °C (10 °F), until the requirements are satisfied and T
CV
is measured. The RT
NDT
temperature
then becomes the temperature (T
CV
) at which the criteria are met minus 33 °C (60 °F). Thus, the reference
temperature is always either greater than or equal to the nil-ductility transition temperature.
Steel Bridge Toughness Criteria. The American Association of State Highway and Transportation Officials
(AASHTO) has adopted Charpy impact toughness requirements for primary tension members in bridge steels
based on section thickness, yield strength, and expected service temperature. They are based on the fracture
toughness corresponding to the maximum loading rate expected in service (Ref 9).
Correlations with Fracture Toughness. Empirical attempts have been made to correlate the Charpy impact
energy with K
Ic
to allow a quantitative assessment of critical flaw size and permissible stress levels. Most of
these correlations are dimensionally incompatible, ignore differences between the two measures of toughness
(in particular, loading rate and notch acuity), and are valid only for limited types of materials and ranges of
data. Additionally, these correlations can be widely scattered. However, some correlations can provide a useful
guide to estimating fracture toughness; in fact, the preceding design criteria for nuclear pressure vessel and
bridge steels are partially based on such correlative procedures.
Some of the more common correlations are listed in Table 3 (Ref 9, 10, 11, 12, 13, 14, 15, and 16) with
appropriate units. Note that some of the correlations attempt to eliminate the effects of loading rate; the
dynamic fracture toughness, K
Id
, is correlated with Charpy energy. Other attempts have been made to improve
and explain some of the correlations (see, for example, Ref 17). A study has also been conducted using a
portion of the Charpy energy to separate initiation and propagation components in the Charpy test (Ref 18). The

results from this study for an upper-shelf J
Ic
correlation for pressure vessel steels were not significantly better
than the Rolfe-Novak correlation listed in Table 3. A statistically based correlation for lower-bound toughness
has also been developed for pressure vessel steels (Ref 19, 20). Thus, simple and empirical correlations can be
used as general guidelines for estimating K
Ic
or K
Id
within the limits of the specific correlation.
Table 3 Typical Charpy/K
Ic
correlation for steels
Correlation
Transition temperature regime
Barsom (Ref 9)
K
Id
2
/E = 5 (CVN)
Barsom-Rolfe (Ref 10)
K
Ic
2
/E = 2(CVN)
3/2

K
Ic
K

Id
= psi
E = psi
CVN = ft · lbf
Sailors-Corten (Ref 11)
K
Ic
2
/E = 8 (CVN)
K
Id
2
= 15.873(CVN) 3/8
K
Id
= ksi
CVN = ft · lbf
Begley-Logsdon—three points (Ref 12)
(K
Ic
)
1
= 0.45 σ
y
at 0% shear fracture temperature
(K
Ic
)
2
From Rolfe-Novak Correlation at 100% shear fracture

temperature
(K
Ic
)
3
= [(K
Ic
)
1
+ (K
Ic
)
2
] at 50% shear fracture temperature
K
Ic
= ksi
σ
y
= ksi
Marandet-Sanz—three steps (Ref 13)
T
100
= 9 + 1.37 T
28
J
K
Ic
= 19 (CVN) 1/2
Shift K

Ic
curve through T
100
point
T
100
= °C, for which K
Ic
= 100 MPa
T
28
= °C, for which CVN = 28J
K
Ic
= MPa
CVN = J
Wullaert-Server (Ref 14)
K
Ic,d
= 2.1 (σ
y
CVN) 1/2
K
Ic,d
= ksi
CVN = ft · lbf
σ
y
= ksi corresponding to approximate
loading rate

Upper-shelf region
Rolfe-Novak—σ
y
> 100 ksi (Ref 15)
(K
Ic
/σ
y
)
2
= 5 (CVN/σ
y
- 0.05)
K
Ic
= ksi
CVN = ft · lbf
σ
y
= ksi
Ault-Wald-Bertolo—ultrahigh-strength steels (Ref 16)
(K
Ic
/σ
y
)
2
= 1.37 (CVN/σ
y
) - 0.045

K
Ic
= ksi
CVN = ft · lbf
σ
y
= ksi
1.0 ksi = 6.8948 MPa; 1.0 ksi = 1.099 MPa ; 1.0 ft · lbf = 1.356 J; CVN is the designation for
Charpy impact energy; σ
y
is the yield stress; and E is the Young's modulus.
As previously described, a lower-bound K
IR
toughness curve is shifted relative to a reference temperature,
RT
NDT
, and used to define the ductile-to-brittle transition. The RT
NDT
is a critical value and is defined very
conservatively in terms of Charpy and dynamic tear specimen results. Continued application of these
requirements is now a principal limitation to continued operation of several commercial nuclear power plants
(Ref 21). Recent work by ASTM Committee E-8 has proposed a method to obtain a new reference temperature
and a method to define, using a probabilistic approach, a median ductile-to-brittle transition curve from a set of
six properly tested small samples that would, in many cases, be precracked Charpy specimens. Statistical
confidence bounds would then be available for this median transition “master curve,” which would be specific
to the particular nuclear plant of interest and could be used to assure that the pressure vessel had adequate
toughness for continued operation.
A generalized prediction method to predict K
Ic
transition curves has also been developed with data from various

steels including 2.25Cr-1Mo, 1.25Cr-0.50Mo, 1Cr and 0.50Mo chemical pressure vessel steels, and ASTM A
508 C1.1, A 508 C1.2, A 508 C1.3 and A 533 Gr.B C1.1 nuclear pressure vessel steels (Ref 22). This method
consists of a master curve of K
Ic
and a temperature shift, ΔT, between fracture toughness and Charpy V-notch
impact transition curves versus yield strength relationship for T
0
, where T
0
is the temperature showing 50% of
the upper-shelf K
Ic
value. The K
Ic
transition curves predicted using both methods showed a good agreement
with the lower bound of measured K
Ic
values obtained from elastic-plastic, J
c
, tests.
References cited in this section
3. O.L. Towers, Effects of Striker Geometry on Charpy Results, Met. Constr., Nov 1983, p 682–685
4. J.M. Holt, Ed., Charpy Impact Test: Factors and Variables, STP 1072, ASTM, 1990
5. T.A. Siewert and A.K. Schmieder, Ed., Pendulum Impact Machines: Procedures and Specimens for
Verification, STP 1248, ASTM, 1995
6. B.F. Dixon, Reliability of Fracture Appearance Measurement in the Charpy Test, Weld. J., Vol 73 (No.
8), Aug 1994, p 39–46
7. “Rules for Construction of Nuclear Power Plant Components,” ASME Boiler and Pressure Vessel Code,
Section III, Division 1 , Appendices, Nonmandatory Appendix G, American Society for Mechanical
Engineers, 1983

8. Energy (Title 10), Domestic Licensing of Production and Utilization Facilities (Part 50), Code of
Federal Regulations, U.S. Government Printing Office, 1981
9. J.M. Barsom, The Development of AASHTO Fracture Toughness Requirements for Bridge Steels, Eng.
Fract. Mech., Vol 7 (No. 3), Sept 1975, p 605–618
10. J.M. Barsom and S.T. Rolfe, Correlations Between K
Ic
and Charpy V-Notch Test Results in the
Transition Temperature Range, Impact Testing of Materials, STP 466, ASTM, 1979, p 281–302
11. R.H. Sailors and H.T. Corten, Relationship Between Material Fracture Toughness Using Fracture
Mechanics and Transition Temperature Tests, Fracture Toughness, Proceedings of the 1971 National
Symposium on Fracture Mechanics—Part II, STP 514, ASTM, 1972, p 164–191
12. J.A. Begley and W.A. Logsdon, “Correlation of Fracture Toughness and Charpy Properties for Rotor
Steels,” WRL Scientific Paper 71-1E7-MSLRF-P1, Westinghouse Research Laboratory, Pittsburgh, PA,
July 1971
13. B. Marandet and G. Sanz, Evaluation of the Toughness of Thick Medium-Strength Steels by Using
Linear Elastic Fracture Mechanics and Correlations Between K
Ic
and Charpy V-Notch, Flaw Growth
and Fracture, STP 631, ASTM, 1977, p 72–95
14. R.A. Wullaert, Fracture Toughness Predictions from Charpy V-Notch Data, What Does the Charpy Test
Really Tell Us?: Proceedings of the American Institute of Mining, Metallurgical and Petroleum
Engineers, American Society for Metals, 1978
15. S.T. Rolfe and S.R. Novak, Slow-Bend K
Ic
Testing of Medium-Strength High-Toughness Steels, Review
of Developments in Plane-Strain Fracture Toughness Testing, STP 463, ASTM, 1970, p 124–159
16. “Rapid Inexpensive Tests for Determining Fracture Toughness,” National Materials Advisory Board,
National Academy of Sciences, Washington, D.C., 1976
17. What Does the Charpy Test Really Tell Us?: Proceedings of the American Institute of Mining,
Metallurgical and Petroleum Engineers, American Society for Metals, 1978

18. D.M. Norris, J.E. Reaugh, and W.L. Server, A Fracture-Toughness Correlation Based on Charpy
Initiation Energy, Fracture Mechanics: Thirteenth Conference, STP 743, ASTM, 1981, p 207–217
19. W.L. Server et al., “Analysis of Radiation Embrittlement Reference Toughness Curves,” EPRI NP-
1661, Electric Power Research Institute, Palo Alto, CA, Jan 1981
20. Metal Properties Council MPC-24, Reference Fracture Toughness Procedures Applied to Pressure
Vessel Materials, Proceedings of the Winter Annual Meeting of the American Society for Mechanical
Engineers, American Society of Mechanical Engineers, New York, 1984
21. J.A. Joyce, Predicting the Ductile-to-Brittle Transition in Nuclear Pressure Vessel Steels from Charpy
Surveillance Specimens, Recent Advances in Fracture, Minerals, Metals and Materials Society/AIME,
1997, p 65–75
22. T. Iwadate, Y. Tanaka, and H. Takemata, Prediction of Fracture Toughness K
Ic
Transition Curves of
Pressure Vessel Steels from Charpy V-Notch Impact Test Results, J. Pressure Vessel Technol. (Trans.
ASME), Vol 116 (No. 4), p 353–358

Impact Toughness Testing

Instrumented Charpy Impact Test
The use of additional instrumentation (typically an instrumented tup) allows a standard Charpy impact machine
to monitor the analog load-time response of Charpy V-notch specimen deformation and fracturing. The primary
advantage of instrumenting the Charpy test is the additional information obtained while maintaining low cost,
small specimens, and simple operation. The most commonly used approach is application of strain gages to the
striker to sense the load-time behavior of the test specimen. In some cases, gages are placed on the specimen as
well, such as for the example shown in Fig. 10 (Ref 23).

Fig. 10 Charpy specimen with additional instrumentation at the supports
General Description
Instrumentation of the tup provides valuable data in terms of the load-time, P-t, history during impact.
Extensive efforts have been made to help determine the dynamic fracture toughness, K

Id
, over a range of
behavior in linear-elastic, elastic-plastic, and fully plastic regimes. An overview of these efforts is given in Ref
24.
Figure 11 schematically illustrates the change in Charpy behavior as a function of temperature for a medium-
strength steel. As shown, instrumentation clearly allows the various stages in the fracture process to be
identified. The energy value, W
M
, is associated with the area under the load-time (P-t) curve up to maximum
load, P
M
. This impulse value is converted to energy by using Newton's second law, which accounts for the
pendulum velocity decrease during the deformation-fracture process. This velocity decrease is proportional to
the instantaneous load on the specimen at any particular time, t
i
; the actual energy absorbed, ΔE
i
, simplifies to
(Ref 25):


(Eq 1)
where E
o
is the total available kinetic energy of the pendulum ( m · ) and:
E
a
= V
o
P · dt

(Eq 2)
where V
o
is the initial impact velocity, and m is the effective mass of the pendulum. The ability to separate the
total absorbed energy into components greatly augments the information gained by instrumentation. Load-
temperature diagrams can be constructed to illustrate the various fracture process stages indicative of the
fracture mode transition from brittle to ductile behavior (Ref 26).

Fig. 11 Load-time response for a medium-strength steel. P
M
, maximum load; P
GY
, general yield load; P
F
,
fast fracture load (generally cleavage); P
A
, arrest load after fast fracture propagation; t
M
, time to
maximum load; t
GY
, time to general yield; W
M
, energy absorbed up to maximum load
One of the primary reasons for the development of the instrumented Charpy test was to apply existing notch
bend theories (slow bend) to the dynamic three-point bend Charpy impact test. Obtaining load information
during the standard Charpy V-notch impact test establishes a relationship between metallurgical fracture

parameters and the transition temperature approach for assessing fracture behavior (Ref 27). Initial studies
concentrated on the full range of mechanical behavior from fully elastic in the lower Charpy shelf region to
elastic-plastic in the transition region to fully plastic in the upper shelf region (see Fig. 11).
Most studies have been performed on structural steels, with primary emphasis on the effect of composition,
strain rate, and radiation on the notch bend properties. Interest in instrumented impact testing has expanded to
include testing of different types of specimens (e.g., precracked, large bend), variations in test techniques (e.g.,
low blow, full-size components), and testing of many different materials (e.g., plastics, composites, aerospace
materials, ceramics). The many variations in test methods is a motivation for standardized test methods,
although standardization for instrumented Charpy testing has been slow (see the section “Standards and
Requirements” in this article).
Instrumentation
Instrumentation for a typical Charpy impact testing system includes an instrumented striker, a dynamic
transducer amplifier, a signal-recording and display system, and a velocity-measuring device. The instrumented
striker is the dynamic load cell, which is securely attached to the falling weight assembly. The striker has
cemented strain gages to sense the compression loading of the tup while it is in contact with the test specimen.
The dynamic transducer amplifier provides direct-current power to the strain gages and typically amplifies the
strain gage output after passing through a selectable upper-frequency cutoff.
The impact signal is recorded and stored either on a storage oscilloscope or through the use of a transient signal
recorder. Digital data from a transient recorder can be reconverted back to analog form and plotted on an x-y
recorder, or the digital data can be transferred to a computer for direct analysis.
Triggering is best accomplished through an internal trigger that has the ability to capture the signal preceding
the trigger; external triggering from the velocity-sensing device is often used instead of an appropriate internal
trigger. The velocity-measuring system should be a noncontacting, optical system that clocks a flag on the
impacting mass immediately before impact so that initial velocity measurements can be made. Velocities must
be determined for all impact drop heights used.
The impact machine and the instrumentation package must be calibrated to ensure reliable data. Calibration of
the Charpy pendulum impact machine is performed in accordance with ASTM E 23, as discussed previously in
this article in terms of periodic proof testing of AMMRC calibration specimens to ensure reliable dial energy
values.
Instrumentation calibration consists of a time base and load-cell calibration with a system frequency response

measurement. The time base calibration consists of passing a known time mark pulse through the system and
calibrating accordingly. The load-cell calibration is typically accomplished by testing notched specimens of
6061-T651 aluminum that are only slightly loading-rate sensitive over the range used (Ref 28). The load cell is
calibrated when the measured dynamic limit load is only slightly higher than the predetermined quasi-static
limit load (measured using the same loading arrangement and anvil dimensions) and when the dial energy (or
velocity-determined energy measurement) matches the integrated total energy. The relationship used for
obtaining total absorbed energy, ΔE
o
, from the area under the load-time record follows the approach in Eq 1
and 2.
The calculated ΔE
o
value will match the dial energy reading when the system is calibrated (in addition to the
limit load check). Because the aluminum limit load is fairly low (around 7.1 kN, or 1600 lbf), a check on load-
cell linearity at higher loads is also needed. To accomplish this, the integrated energy/dial energy requirement
for a quenched and tempered 4340 specimen (52 HRC) that has a higher fracture load (near 27 kN, or 6000 lbf)
is checked.
Low-energy AMMRC calibration specimens can be used for this procedure. If the energies match for the 4340
test at the same amplifier gain as for the aluminum calibration, the load-cell calibration is usually linear
throughout the usable load range. Static linearity checks can also be made if the static loading system exactly
duplicates the dynamic loading conditions. Daily test checks using the aluminum calibration specimens are
suggested to verify load-cell calibration.
The system frequency response is determined experimentally by superimposing a constant-amplitude sine wave
signal on the output of the strain gage bridge circuit (Ref 29). The peak-to-peak amplitude of the signal should
be equivalent to approximately half the full-scale capacity of the load transducer at a frequency low enough to
ensure no signal attenuation. The frequency of the sine wave is then increased until the amplitude is attenuated
10% (0.915 dB), and the response time, t
R
, is calculated as:

(Eq 3)
where f
0.915
is the frequency at 0.915 dB (10%) attenuation.
Standards and Requirements
Instrumented impact tests that generate P-t plots from instrumented tups require careful attention to test
procedures and analytical methods in order to determine dynamic fracture toughness values with the accuracy
and reliability required for engineering purposes. Extensive efforts have been made to standardize instrumented
impact tests, but many inherent difficulties in analysis and interpretation have impeded the formal development
of standard methods. Nonetheless, instrumented impact testing is an accepted method in the evaluation of
irradiation embrittlement of nuclear pressure vessel steels (Ref 30). Several instrumented impact tests have also
been developed for plastics (Ref 31) with the ISO standard 179-2 on instrumented Charpy testing of plastics
(Ref 32). The following discussions focus on requirements for steels, while more information on impact testing
of plastics and ceramics are addressed in the article“Mechanical Testing of Polymers and Ceramics” in this
Volume. For nonmetallic materials, such as plastics and ceramics, the application of available models involving
energy considerations may be necessary for arriving at the true toughness values (Ref 24).
Standard Methods. Extensive efforts in the development of instrumented Charpy tests began in the 1960s and
1970s with the advent of fracture mechanics and precracked Charpy V-notch specimens, when a series of
seminars and conferences in the 1970s (Ref 33, 34, 35, and 36) examined the role of instrumented impact
testing in the evaluation of dynamic fracture toughness (Ref 24). The International Institute of Welding first
attempted to standardize the instrumented Charpy test, but concluded that the test was not sufficiently
documented, and the effort was discontinued (Ref 37). A few years later, two significant events prompted
serious consideration of standardization. The development of the K
IR
curve by the Pressure Vessel Research
Committee and its inclusion in the ASME Code, Section III, created the need for dynamic initiation toughness,
K
Id
, data. Simultaneously, two other related groups began formulating procedures and conducting

interlaboratory round robins. The Pressure Vessel Research Committee/Metals Property Council Task Group on
Fracture Toughness Properties for Nuclear Components developed procedures for measuring K
Id
values from
precracked Charpy specimens (Ref 38).
The Electric Power Research Institute (EPRI) funded work to develop procedures known as the “EPRI
Procedures” (Ref 28, 39). This procedure is summarized in the following section, “General Test Requirements.”
Since that time, important theoretical and technical developments have occurred, as outlined in Ref 24. Efforts
have also been made in the development of standards. In 1992, the European Structural Integrity Society (ESIS)
formed a working party (formed within ESIS Technical Subcommittee 5) devoted to instrumented impact
testing on subsize Charpy-V specimens of metallic materials. In 1994, ESIS issued a draft of a standard method
for the instrumented Charpy V-notch test on metallic materials (Ref 40). This method allows one to estimate an
approximate value of the proportion of ductile fracture surface by one of the following formulas:


where P
GY
, P
M
, P
IU
, and P
A
are characteristic points on the load-time diagram shown in Fig. 12.

Fig. 12 Load vs. time record showing the definitions of the various load points used in various models to
estimate the percent shear fracture; P
GY
, characteristic value for onset of plastic deformation; P
M

,
maximum load; P
IU
, load at the initiation of unstable crack propagation; P
A
, load at the end of unstable
crack propagation.
The working group also performed round-robin testing to help develop the state of knowledge on the dynamic
behavior of miniaturized impact specimens (Ref 41). In 1992, a formal committee also was formed for
development of a possible JIS standard for evaluation of dynamic fracture toughness by the instrumented
Charpy impact testing method. Problems to be resolved before the standardization of the instrumented Charpy
impact test method are pointed out in Ref 42.
General Test Requirements. Only subtle differences exist between the “EPRI Procedures” (Ref 28) and the
Pressure Vessel Research Committee procedures for measuring K
Id
values from precracked Charpy specimens
(Ref 38, 43). The following test requirements are taken from the EPRI procedures.
The load signal obtained from an instrumented striker during an impact test oscillates about the actual load
required to deform the specimen. Therefore, the signal analysis procedure employed should minimize the
deviation of the apparent load from the actual specimen deformation load. A simplistic view of the impact event
allows three major areas for test specification to be identified: initial loading, limited frequency response, and
electronic curve fitting.
The impact loading of a specimen will create inertial oscillations in the contact load between striker and
specimen, and a time interval between 2τ and 3τ is required for the load to be dissipated, where τ is related to
the period of the apparent specimen oscillations and can be predicted empirically for a span-to-width ratio of 4
by:


(Eq 4)
where W is the specimen width, B is the specimen thickness, C

s
is the specimen compliance, E is the Young's
modulus, S
o
is the speed of sound in the specimen, and τ is typically 30 μs for standard Charpy steel specimens.
When any time, t, is less than 2τ, it is not possible to use the striker signal to measure the portion of the
specimen load caused by inertial effects. An empirical specification for reliable load and time evaluation is:
t ≥ 3τ


(Eq 5)
Control of t is obtained by control of the initial impact velocity. The constant 3 in Eq 5 may be as low as 2.3
without adversely affecting the test results, if the curve-fitting technique described below is followed. A value
of 3 was chosen for the case of “unlimited” frequency response. The original EPRI procedures corresponded to
the 2.3 factor and included the selective filtering for curve fitting (Ref 28). Computer simulations of the Charpy
test have approximately verified the value of τ and the 3τ criterion (Ref 44).
The potential problem of limited frequency response of the transducer amplifier is avoided by specifying:
t ≥ 1.1t
R


(Eq 6)
where t
R
is defined as the 0.915 dB response time of the instrumentation, as indicated in Eq 3. Inadequate
response results in a distorted signal response. It is important to note that the electronic attenuation must be
representative of a resistance-capacitance circuit for Eq 6 to apply.
The curve fitting of the oscillations is achieved by specifying a minimum t
R
. The amplitude of the observed

oscillations is therefore reduced such that the disparity between tup contact load and effective deformation load
is minimal. For the best test, it has been empirically found for resistance-capacitance circuit systems that:
t
R
≥ 1.4t


(Eq 7)
is adequate for the electronic curve fitting without altering the overall curve, when t ≥ 2.3τ. When t ≥ 3τ, it is
not necessary to electronically curve fit because the disparity between the contact load and the specimen
deformation load is less than approximately 5%.
The requirements for obtaining acceptable load-time records (in particular, Eq 5) result in the need to control
V
o
. By controlling the impact velocity, a corresponding control of kinetic energy (E
o
) is inherent. The reduction
in striker velocity during the impact loading of the specimen should therefore be minimized. A conservative
requirement is:
E
o
≥ 3W
M


(Eq 8)
where W
M
is the system energy dissipated to maximum load P
M

. This requirement ensures that the tup velocity
is not reduced by more than 20% up to maximum load. This requirement is seldom a problem for full-impact
Charpy V-notch tests; Eq 8 may not be met, however, when precracked Charpy tests are conducted for very
tough materials. The test requirements for reliable load measurement are summarized as follows:
Inertial effects t ≥ 3τ
Limited frequency response

t ≥ 1.1t
R
, required only if 2.3τ ≤ t <3τ

Electronic curve fitting t
R
≥ 1.4τ
Energy criterion E
o
≥ 3W
M

The time t corresponds to the shortest time required for measurement after the specimen has been impacted;
that is, t is the time to maximum load t
M
for the elastic fracture, and t is the time to general yield, t
GY
, in the
postgeneral yield fracture (See Fig. 11). The specification for electronic curve fitting is only required if 2.3τ ≤ t
< 3τ. Because it is often difficult to ensure that t ≥ 3τ and because the filtering has no adverse effect when t ≥
2.3τ filtering at t
R
≥ 1.4τ is always possible, assuming that t ≥ 1.1t

R
.
Limitations on Testing. Violation of any of the general test requirements presented above will invalidate the
data obtained from instrumented Charpy V-notch tests. Limitations of this testing technique are the same as
those for standard Charpy testing. The effects of small size relative to typical component size, the rounded
machine notch, and shallow notch depth restrict general applicability and usefulness of the Charpy test. Note
that the notch depth for the Charpy V-notch specimen is too shallow to prevent yielding across the gross section
of the specimen.
Instrumentation has allowed separation of energy components and measurement of applied loads throughout the
fracture event, but direct determination of the initiation component is not directly possible for ductile
(microvoid coalescence) initiation from the instrumented test record. Some of these limitations have been
addressed by fatigue precracking the Charpy specimen, which eliminates the notch effects and makes it a small
fracture-mechanics-type specimen.
References cited in this section
23. W. Schmitt, W. Böhme, and D Z. Sun, New Developments in Fracture Toughness Evaluation,
Structural Integrity: Experiments, Models, Applications—European Conference of Fracture (ECF) 10,
Vol 1, Engineering Materials Advisory Service, 1990, p 159–170
24. P.R. Sreenivasan, Instrumented Impact Testing—Accuracy, Reliability, and Predictability of Data,
Trans. Indian Inst. Met., Vol 49 (No. 5), Oct 1996, p 677–696
25. B. Augland, Fracture Toughness and the Charpy V-Notch Test, Br. Weld. J., Vol 9 (No. 7), 1962, p 434
26. G.D. Fearnehough and C.J. Hoy, Mechanism of Deformation and Fracture in the Charpy Test as
Revealed by Dynamic Recording of Impact Loads, J. Iron Steel Inst. Jpn., Vol 202, 1964, p 912
27. R.A. Wullaert, Application of the Instrumented Charpy Impact Test, Impact Testing of Metals, STP 466,
ASTM, 1970, p 148–164
28. D.R. Ireland, W.L. Server, and R.A. Wullaert, “Procedures for Testing and Data Analysis,” ETI Report
TR-75-43, Effects Technology, Inc., Santa Barbara, CA, Oct 1975
29. D.R. Ireland, Procedures and Problems Associated with Reliable Control of the Instrumented Impact
Test, Instrumented Impact Testing, STP 563, ASTM, 1974, p 3–29
30. L.E. Steele, Ed., Radiation Embrittlement of Nuclear Pressure Vessel Steels: An International Review,
STP 1011, ASTM, 1989

31. S. Kessler, G.C. Adams, S.B. Driscoll, and D.R. Ireland, Instrumented Impact Testing of Plastics and
Composites Materials, STP 936, ASTM, 1986