ASM Metals HandBook Vol. 8 - Mechanical Testing and Evaluation

Подождите немного. Документ загружается.

where E

is the total available kinetic energy of the pendulum ( m · ) and:

= V

P · dt

(Eq 2)

where V

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

, maximum load; P

, general yield load; P

fast fracture load (generally cleavage); P

, arrest load after fast fracture propagation; t

, time to

maximum load; t

, time to general yield; W

, 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

, from the area under the load-time record follows the approach in Eq 1

and 2.

The calculated ΔE

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

, 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

curve by the Pressure Vessel Research

Committee and its inclusion in the ASME Code, Section III, created the need for dynamic initiation toughness,

, 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

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

, P

, and P

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

, characteristic value for onset of plastic deformation; P

maximum load; P

, load at the initiation of unstable crack propagation; P

, 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

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

is the specimen compliance, E is the Young's

modulus, S

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

(Eq 6)

where t

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

. 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:

≥ 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

. By controlling the impact velocity, a corresponding control of kinetic energy (E

) is inherent. The reduction

in striker velocity during the impact loading of the specimen should therefore be minimized. A conservative

requirement is:

≥ 3W

(Eq 8)

where W

is the system energy dissipated to maximum load P

. 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

, required only if 2.3τ ≤ t <3τ

Electronic curve fitting t

≥ 1.4τ

Energy criterion E

≥ 3W

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

for the elastic fracture, and t is the time to general yield, t

, 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

≥ 1.4τ is always possible, assuming that t ≥ 1.1t

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

32. “Plastic—Determination Of Charpy Impact Properties: Part 2—Instrumented Impact Test,” ISO-179-2,

International Organization for Standardization

33. Impact Testing of Metals, STP 466, ASTM, 1970

34. Instrumented Impact Testing, STP 563, ASTM, 1974

35. C.E. Turner, in Advanced Seminar on Fracture Mechanics, EURA-TOM-ISPRA Courses, 1975

36. Dynamic Fracture Toughness, The Welding Institute and The American Society for Metals, 1976

37. E.C.J. Buys and A. Cowan, Interpretation of the Instrumented Impact Test, Weld. World, Vol 8 (No. 1),

1970, p 70–76

38. Pressure Vessel Research Committee/Metal Properties Council Working Group on Instrumented

Precracked Charpy Testing, “Instrumented Precracked Charpy Testing: Report I—Recommended

Testing Procedure, Report II—Associated Test Program,” Westinghouse Research Laboratory,

Pittsburgh, PA, 1974

39. R.A. Wullaert, Ed., CSNI Specialist Meeting on Instrumented Precracked Charpy Testing, Nov 1980,

EPRI NP-2102-LD, Electric Power Research Institute, 1981

40. “ESIS Instrumented Charpy V-Notch Standard,” Proposed Standard Method for the Instrumented

Charpy-V Impact Tests on Metallic Materials, Draft 10, ESIS, Jan 14, 1994

41. E. Lucon, Instrumented Impact Testing of Sub-Size Charpy-V Specimens: The Activity of the ESIS

TC5 Working Party, ECF 11—Mechanisms and Mechanics of Damage and Failure, Vol 1, Elsevier,

1996, p 621

42. T. Kobayashi and I. Yamamoto, Progress and Development in the Instrumented Charpy Impact Testing

Method, Bull. Jpn. Inst. Met., Vol 32 (No. 3), 1993, p 151–159 (in Japanese)

43. W.L. Server, Impact Three-Point Bend Testing for Notches and Precracked Specimens, J. Test. Eval.,

Vol 6 (No. 1), 1978, p 29–34

44. D.M. Norris, D. Quiñones, and B. Moran, Computer Simulation of Plastic Deformation in the Charpy

V-Notch Impact Test, What Does the Charpy Test Really Tell Us?: Proceedings of the American

Institute of Mining, Metallurgical and Petroleum Engineers, American Society for Metals, 1978, p 22–

Impact Toughness Testing

Precracked Charpy Test

By inducing a fatigue precrack in the Charpy specimen, the notch acuity and depth restrictions are eliminated.

Early work concentrated on correlations with fracture toughness using only the total absorbed energy (i.e.,

uninstrumented testing). These energy values usually are normalized per unit area (A) below the fatigue crack;

the normalized energy values are designated as W/A.

Most of the correlations of W/A with fracture toughness have been conducted using slow-bend specimens. The

basic problem in reaching an impact correlation is the difference in loading rates between the Charpy impact

and the static K

tests, particularly for loading-rate sensitive materials (Ref 45). A general trend exists for a

correlation between /E and W/A, but the limited data and scatter make this difficult to utilize (Ref 16). A

better correlation with K

may be possible. The reason for using /E as the basis is the approximate

proportionality between /E and W/A, based on a presumed fracture mechanics relationship (Ref 45).

The precracked Charpy W/A values can also be used to estimate the nil-ductility transition temperature. The

typical technique defines an inflation point between lower shelf and transition region behavior as the estimated

nil-ductility transition temperature (Ref 46). Some exceptions have been noted to this approach (Ref 47).

Instrumented Data

The types of data and test techniques used for instrumented precracked Charpy testing are the same as those

discussed earlier for instrumented Charpy impact testing. The 3τ criterion, which limits the impact velocity,

becomes more important for deeply cracked, brittle materials. The greatest advantage of precracking is the

transformation of the Charpy V-notch specimen into a dynamic fracture mechanics test piece. The direct

calculation of fracture toughness (within certain limitations) is now possible using the instrumented load-time

information. The following discussion presents the calculational aspects of these fracture toughness parameters.

If fracture is known to initiate at maximum load (as it usually does for cleavage initiation), the energy value of

(see Fig. 11) can be considered an initiation energy. However, W

includes contributions other than that

caused by the deflection of the specimen. Therefore, a compliance energy correction is needed to determine the

true specimen energy, E

(Ref 48). When the fracture is linear elastic (fracture before general yield; see the

first two load-time records in Fig. 11), the value of E

can be calculated directly:

(Eq 9)

where C

is the nondimensional specimen compliance (Ref 49). For a fracture occurring after general yield

(see Fig. 11), E

is obtained by correcting W

(Eq 10)

where C

is the total system compliance calculated at general yield and corrected for the decrease in velocity

through general yield:

(Eq 11)

This compliance correction is assumed to be linear with load, but the actual correction is not so simple.

However, the error in assuming a linear relationship results in a slightly smaller (conversvative) value of E

(Ref 43).

It is often desirable to partition the total fracture energy into initiation and postinitiation (propagation)

components. Assuming initiation occurs at maximum load, the propagation energy, E

, is:

= E

- E

(Eq 12)

where E

is the total fracture energy, as determined from a dial indicator, kinetic energy change (initial and

final velocity measurements), or ΔE

Linear Elastic Fracture Toughness. When the fracture is elastic (fracture occurs before general yield), the stress-

intensity factor, K

, can be calculated by applying linear elastic fracture mechanics:

(Eq 13)

where a is the crack length.

The ASTM size requirements for a valid K

are quite limiting, even if a dynamic yield strength is used.

However, the general specimen size requirements of ASTM E 399 may be too conservative for dynamic testing

of ferritic medium-strength steels (Ref 50). Therefore, if general yielding has not occurred, a linear-elastic value

of fracture toughness, K

, generally is calculated. The stress intensification rate is calculated as:

(Eq 14)

This loading rate reflects the dynamic aspect of the loading, because the lowest for impact loading of

precracked Charpy specimens is on the order of 11 × 10

MPa · s

-1

× (1 × 10

ksi · s

-1

Postgeneral Yield Fracture Toughness. When general yielding occurs, an energy-based value of the J-integral

can be used to obtain a measure of fracture toughness. The calculation of ductile fracture toughness, J

, is

contingent upon knowing the initiation point of fracture on the load-time record. For cleavage-initiated fracture,

this point generally corresponds to maximum load. However, for fibrous (ductile) initiation, maximum load is

generally a nonconservative assumption. When the initiation point is known or has been determined

experimentally (Ref 51) and when a/W ≥ 0.5 (Ref 52), then:

(Eq 15)

where b is the remaining ligament depth (W - a). A stress-intensity factor K

can be obtained from the J

value

as:

= (EJ

) 1/2

(Eq 16)

An average K can also be computed, as in Eq 16, by using a K

value. Validity criteria related to specimen

dimensions appear to be (Ref 43):

(Eq 17)

and

(Eq 18)

where σ

is the flow stress, defined as the average of the yield stress and the ultimate stress. For dynamic

loading, the yield stress, σ

, and flow stress, σ

, of standard Charpy V-notch specimens can be estimated for

postgeneral yield behavior as:

(Eq 19)

and

(Eq 20)

The general form of the equation results from slip-line field solutions for blunt-notch specimens, and the

constant 2.99 has been obtained from extrapolation of results from a slip-line field solution that included the tup

indentation at the center loading point (Ref 53). The constant of 2.99 reduces to 2.85 for sharp-notch specimens

with a fatigue precrack. The stress values obtained using this approach agree favorably with high rate tensile

test results (Ref 54).

Limitations. The test requirements and data analysis procedures described in this article were developed for

ferritic pressure vessel steels. A review of instrumented precracked Charpy testing can be found in Ref 55,

which discusses the theory and applicability of instrumented precracked Charpy testing. Not all of the

relationships and approaches presented in Ref 55 are universally accepted because standards or recommended

practices do not currently exist.

Related Test Techniques

Several attempts have been made to use the precracked Charpy specimen at loading rates beyond the limits

applicable to quasi-static analysis. The procedures described above assume a quasi-static situation for times

greater than the limiting values near 3τ. One such attempt for larger than Charpy-size specimens is described in

Ref 56, in which strain gages were mounted near the crack tip to avoid many of the spurious wave effects.

Other studies have been conducted using Hopkinson bar techniques (Ref 57) and the shadow optic method of

caustics (Ref 58). These studies indicate the need for dynamic analysis when using the instrumented Charpy

striker approach at high loading rates.

References cited in this section

16. “Rapid Inexpensive Tests for Determining Fracture Toughness,” National Materials Advisory Board,

National Academy of Sciences, Washington, D.C., 1976

43. W.L. Server, Impact Three-Point Bend Testing for Notches and Precracked Specimens, J. Test. Eval.,

Vol 6 (No. 1), 1978, p 29–34

45. T.M.F. Ronald, J.A. Hall, and C.M. Pierce, Usefulness of Precracked Charpy Specimens for Fracture

Toughness Screening Tests of Titanium Alloys, Metall. Trans., Vol 3, April 1972, p 813–818

46. G.M. Orner and C.E. Hartbower, Transition-Temperature Correlations in Construction Alloy Steels,

Weld. J., Vol 40 (No. 9), Oct 1961, p 459s

47. J.H. Gross, The Effect of Strength and Thickness on Notch Ductility, Weld. J., Vol 48 (No. 10), Oct

1969, p 441s

48. W.L. Server, D.R. Ireland, and R.A. Wullaert, “Strength and Toughness Evaluations from an

Instrumented Impact Test,” ETI TR-74-29R, Effects Technology, Inc., Santa Barbara, CA, Nov 1974

49. H.J. Saxton et al., Load Point Compliance of the Charpy Impact Specimen, Instrumented Impact

Testing, STP 563, ASTM, 1974, p 30–49

50. W.L. Server, R.A. Wullaert, and J.W. Sheckherd, “Verification of the EPRI Dynamic Fracture

Toughness Testing Procedures,” ETI TR 75–42, Effects Technology, Inc. Santa Barbara, CA, Oct 1975

51. W.L. Server, W. Oldfield, and R.A. Wullaert, “Experimental and Statistical Requirements for

Developing a Well-Defined KIR Curve,” EPRI NP-372, Electric Power Research Institute, Palo Alto,

CA, May 1977

52. J.D.G. Sumpter and C.E. Turner, Method for Laboratory Determination of J

, Cracks and Fracture, STP

601, ASTM, 1976, p 3–18

53. D.J.F. Ewing, Calculations on the Bending of Rigid/Plastic Notched Bars, J. Mech. Phys. Solids, Vol

16, 1968, p 205–213

54. W.L. Server, General Yielding of Charpy V-Notch and Precracked Charpy Specimens, J. Eng. Mater.

Technol. (Trans. ASME), Vol 100, April 1978, p 183–188

55. Committee on Safety of Nuclear Installations Specialist Meeting on Instrumented Precracked Charpy

Testing, EPRI NP-2102-LD, Electric Power Research Institute, Palo Alto, CA, Nov 1981

56. F.J. Loss, Ed., “Structural Integrity of Water Reactor Pressure Boundary Components,” Progress Report

Ending Feb 1976, NRL Report 8006, Naval Research Laboratory, Washington, D.C., Aug 1976

57. T. Nicholas, “Instrumented Impact Testing Using a Hopkinson Bar Apparatus,” AFML-TR-75-54, Air

Force Materials Laboratory, Wright-Patterson Air Force Base, Dayton, OH, July 1975

58. J.F. Kalthoff et al., “Measurements of Dynamic Stress Intensity Factors in Impacted Bend Specimens,”

Committee on Safety of Nuclear Installations Specialist Meeting on Instrumented Precracked Charpy

Testing, EPRI NP-2102-LD, Electric Power Research Institute, Palo Alto, CA, Nov 1981

Impact Toughness Testing

Drop Weight Testing

Because Charpy V-notch testing does not necessarily reveal the same transition temperature as that observed

for full-size parts, many other tests have been devised. Two such tests have achieved some degree of popularity.

These are the drop-weight test (DWT) and the drop-weight tear test (DWTT). Both of these tests were

developed at the United States Naval Research Laboratory. Both tests yield a transition temperature that more

nearly coincides with that of full-size parts. This has been described as the nil-ductility temperature (NDT).

Both tests have limited usage because of the required specimen sizes. There are three types of DWT specimens,

as shown in Table 4. The smallest of these measures 16 × 51 × 127 mm ( × 2 × 5 in.), and thus, when four to

eight specimens are required, a considerable amount of material is expended. Often parts are not of sufficient

size or are not shaped in such a manner to allow preparation of such specimens. A provision is made for

remelting and casting material to specimen size. Most DWT tests are made on plate that is 9.5 mm (⅜ in.) thick

or thicker. The DWTT is also a plate testing specification. This test requires a specimen 76 × 305 mm (3 × 12

in.) by full plate size.

Table 4 Standard drop-weight test (DWT) conditions

Drop-weight energy

for given yield strength

level

(a)

Type of

specimen

Specimen size, mm

(in.)

Span,

(in.)

Deflection

stop,

mm (in.)

Yield strength

level,

MPa (ksi)

ft · lbf

210–340 (30–50) 800

600

340–480 (50–70) 1100

800

480–620 (70–90) 1350

1000

P-1 25.4 × 89 × 356 (1 ×

3½ × 14)

305

(12.0)

7.6 (0.3)

620–760 (90–1650

1200