Astm stp 466 1970 (1984)

In the first session, after the opening paper reviews the role that the various impact tests play in characterizing the toughness of materials, the remaining papers discuss various aspec

Trang 2

IMPACT TESTING

OF METALS

A symposium presented at the Seventy-second Annual Meeting AMERICAN SOCIETY FOR TESTING AND MATERIALS Atlantic City, N J., 22-27 June 1969

ASTM SPECIAL TECHNICAL PUBLICATION 466

ITI AMERICAN SOCIETY FOR TESTING AND MATERIALS 1916 Race Street, Philadelphia, Pa 19103

Downloaded/printed by

University of Washington (University of Washington) pursuant to License Agreement No further reproductions authorized.

Trang 3

(~) BY AMERICAN SOCIETY FOR TESTING AND ~IATERIALS 1970 Library of Congress Catalog Card Number: 74-97731

ISBN 0-8031-0038-8

NOTE The Society is not responsible, as a body, for the statements and opinions advanced in this publication

Primed in York, Pa

March 1970 Second Priming, Ba[timore, Md

October 1984

Trang 4

Foreword

The papers in the Symposium on Impact Testing of ~{etals were given

at the Seventy-second Annual 5Ieeting of the American Society for

Testing and 5Iater als held in Atlantic City, N J., 22-27 June 1969

The sponsors of this symposium were Committee E-1 on 5Iethods of

Testing, Subcommittee 7 on Impact Testing, and Committee E-24 on

Fracture Testing of 5Ietals D E Driscoll, Army 5Iaterials and Me-

chanics Research Center, presided as symposium chairman

Trang 5

Related ASTM Publications

Evaluation of Wear Testing, STP 446 (1969)

Trang 6

T h e Charpy I m p a c t T e s t - - I t s Accuracy and Factors Affecting Test Results

Measurement of Fracture Toughness by Instrumented I m p a c t T e s t - -

Influence of Inertial Load in Instrumental I m p a c t T e s t s - - s VENZI, A H

P R I E S T ~ A N D M J M A Y 165

Significance of the Drop-Weight Tear Test and Charpy V-Notch I m p a c t

Investigation of Transition Temperature Tests for Line Pipe M a t e r i a l s - -

Correlations Between K~o and Charpy V-Notch Test Results in the Transi-

D e v e l o p m e n t of a Pendulum-Type Dynamic Tear-Test Machine -

Trang 7

STP466-EB/Mar 1970

Introduction

Since the last ASTM Symposium on Impact Testing in 1955, there has

been considerable effort expended relative to the merits of impact testing,

particularly with regard to the Charpy V-notch test Yet, the Charpy

test continues to play an important part in many materials specifications

To gain a clearer understanding of what an impact test tells us, many

investigations have been conducted These range from changing the

configuration of the notch in the Charpy specimen to designing new

specimens and tests for measuring toughness

In impact testing, for example, the most recent advances are in the

areas of instrumentation of Charpy equipment and modification of

specimen geometry Both advances are aimed to provide a clearer under-

standing of the impact test itself or to attempt to find meaningful cor-

relations between the various fracture toughness criteria or both Not

to be overlooked are those efforts aimed at understanding the effects of

test and specimen variables on the resultant test values Coupled with

these efforts have been modifications of the various tests or the imple-

mentation of new tests, such as the dynamic tear (DT) test, which now

finds considerable application in the pressure vessel field Fracture tough-

ness investigations have been the cause of considerable discussion since

the 1955 symposium

Due to the interest and response to this year's symposium, four ses-

sions were required, and their classification best expresses the theme of

the symposium In the first session, after the opening paper reviews the

role that the various impact tests play in characterizing the toughness of

materials, the remaining papers discuss various aspects of the standard

Charpy test from the effects of material strength and thickness to the

accuracy of the test itself and the factors affecting test results The

second session is directed largely to the use of instrumentation to record

load versus time and aimed at measuring the various fracture toughness

parameters The last two sessions deal with the drop-weight and dynamic

tear tests The papers are quite diversified, ranging from applicable

equipment, to effects of test variables, to correlations between the various

tear tests or the Charpy test or both

Trang 8

2 IMPACT TESTING OF METALS

All of the above have contributed to the broadened scope of the cur-

rent symposium The end result is an excellent balance between theory

and experimental results for the various means of assessing toughness

D E Driscoll

Chief, Quality Assurance Division, Army Materials and :~'[eehanics Research Center, Watertown, Mass 02172;

Trang 9

W T Matthews ~

The Role of Impact Testing in Characterizing

the Toughness of Materials

REFERENCE: Matthews, W T., "The Role of I m p a c t Testing in

Characterizing the T o u g h n e s s of Materials," Impact Testing of Metals,

A S T M STP ~66, American Society for Testing and Materials, 1970, pp

3-20

ABSTRACT: The objectives of fracture toughness testing are to provide

information for design, screening, and acceptance of materials Several

tests are discussed in relation to an ideal design test possessing quantitative-

ness and generality: slow and impact-loaded Griffith-Irwin fracture me-

chanics, conventional Charpy, and drop-weight tear testing (DWTT) The

features and limitations of these methods are noted Fracture mechanics is

recommended for testing relatively brittle materials and DWTT and asso-

ciated procedures as the best available for tough materials Probable in-

creased development in quantitative fracture mechanics including impact

testing is discussed For screening and acceptance, conventional Charpy

testing is recommended provided that correlation with more basic tests has

been established

KEY WORDS: testing, toughness, design, impact tests, transition tempera-

ture, fractures (materials), fracture mechanics, loads (forces), evaluation,

tests

Since the last A S T M S y m p o s i u m on I m p a c t in 1955, there has con-

tinued to be considerable research effort in the field of fracture and

fatigue New approaches have been developed and existing methods

expanded I n order to assess the significance of this wide v a r i e t y of

methods and associated testing procedures for the prevention of fracture,

several excellent summaries have been published [1,2,3] ~ I t is appropriate

as an introduction to this s y m p o s i u m to follow a similar approach with

special a t t e n t i o n given to the role of impact tests The following dis-

cussion will draw upon the past summaries and, in addition, will con-

sider more recent developments, such as the use of a C h a r p y - t y p e

impact specimen for measuring fracture toughness.3

1 Mechanical engineer, Theoretical and Applied Mechanics Research Laboratory, Army Materials and Mechanics Research Center, Watertown, Mass 02172

2 The italic numbers in brackets refer to the list of references appended to this

paper

3 In this paper "fracture toughness" refers to parameter defined by linear elastic

fracture mechanics

Trang 10

.~Iaterial toughness tests are conducted to provide information for

design and for acceptance and screening of materials In this paper the

emphasis will be almost entirely on methods of testing to provide in-

formation for design, since in design we are dealing with the end result

of the characterization of material properties, which therefore imposes

the most stringent requirements on the reliability of our design philosophy

and associated tests We will deal only briefly with the use of impact

tests for acceptance and screening of material

Material Toughness Testing for Design

We will trace the evolution of "design against fracture" philosophies

and associated tests in order to indicate the usefulness of various ap-

proaches to design Based on the capabilities of these approaches, we will

assess the role of impact testing in characterizing the toughness of mate-

rials for design Since this field is so broad, the discussion will be stream-

lined arbitrarily to limit attention to the most general procedures This

description will illustrate that, despite the development of new methods,

we are still unable to deal satisfactorily with many materials We will

begin with the standard textbook approach to design calculations

Ste~gth of Materials Approach

The classical strength of materials procedure calculates the load-

carrying capacity of a structural member on the basis of some percentage

of the gross, static yield stress of the selected material as it is measured

by a smooth uniaxial tension specimen The inherent toughness of the

material is counted upon to redistribute any large local stresses which

may occur

This method suffers from the inability of the smooth tension test to

reveal whether the material will display inadequate toughness when sub-

iected to combinations of tow temperature, high rates of loading, and

triaxial stress state as might be imposed by a sharp notch or crack

Hence, catastrophic failures can occur as a result of large local stresses

although the gross stress levels are small To overcome this difficulty,

new concepts and tests were devised to reveal the toughness of materials

when subjected to severe conditions

Tra~sition Temperature Approach

The transition temperature approach as applied to design deals only

with the behavior of the material The object of the method is to guarantee

that the material possesses sufficient toughness when subjected to severe

conditions to allow the load-carrying capability of a structural member

to be calculated by strength of materials methods without further regard

for the toughness of the material or the consequences of small flaws in

the structure It is necessary to devise a criterion which will ensure that

C o p y r i g h t b y A S T M I n t ' l ( a l l r i g h t s r e s e r v e d ) ; S a t D e c 5 0 9 : 4 7 : 0 3 E S T 2 0 1 5

D o w n l o a d e d / p r i n t e d b y

U n i v e r s i t y o f W a s h i n g t o n ( U n i v e r s i t y o f W a s h i n g t o n ) p u r s u a n t t o L i c e n s e A g r e e m e n t N o f u r t h e r r e p r o d u c t i o n s a u t h o r i z e d

Trang 11

MATTHEWS ON ROLE OF IMPACT TESTING 5

the material does possess this "sufficient" level of toughness For the

moment, in order to facilitate the description of the method, we will

consider only the simplest and most conservative criterion This require-

ment is that the service temperature should correspond to the full tough-

ness, 100 percent shear lip fracture, as measured in a notched bar impact

test

The familiar Charpy bar and test procedures are used for obtaining an

energy-absorbed or fracture appearance versus temperature curve which

usually shows a sharp transition between the relatively fiat upper and

lower shelf portions This use of the basic transition temperature con-

cept, with a Charpy V-notch specimen, and the upper shelf toughness

criterion has been successful in dealing with the toughness of materials

in design under severe conditions The materials used in these applica-

tions have been conventional low-alloy type with yield strengths less

than 100 ksi and newer high-toughness steels with yield strengths up to

150 ksi However, there are a number of difficulties with this procedure,

some of which are associated with the Charpy bar

The question of the effect of various dimensional changes on the per-

formance of the Charpy specimen has been studied extensively In this

paper, we will consider only the effect of the thickness of the Charpy

bar Since the standard bar has a maximum thickness of 0.394 in., it

does not represent the behavior of a material when that material is used

in a larger thickness in a structure The material is subjected to less

restraint by the thickness of the Charpy bar than by the thickness used

in the structure If the material possesses a low toughness in its structural

thickness and consequently fractures with a small percent shear lip,

then the Charpy bar result may significantly overestimate the toughness

of the material for that structural thickness As shown schematically in

Fig 1 for a particular service temperature, since the materi.~.l of the

Charpy bar is subjected to less restraint than in the structure, the

Charpy fracture displays a greater percent shear lip Several approaches

have been devised to overcome this difficulty

Trang 12

FIG 2 Drop-weight tear test specimen

One approach is to induce a greater portion of flat fracture by the use

of side notches or a brittle border of material [~] This approach can be

utilized if both the modified Charpy bar and the structural thickness are

subjected to plane strain restraint, but it cannot be applied generally

since it is impossible to match the amount of mixed mode restraint

occurring in the structure

Another approach makes use of the critical shear lip phenomenon [5,6]

Experimental observations have demonstrated that the thickness of the

shear lip in a plate of a given material is nearly constant at a particular

temperature regardless of the thickness of the section By employing

this constant critical thickness the percent of shear lip for specimens of

different sizes may be related by geometrical considerations

The most direct solution to the problem of the inadequate thickness of

the Charpy specimen is to simply increase the size of the test bar to the

full thickness of the material in the actual structure The drop-weight

tear test (DWTT) specimen, Fig 2, as described in ASTM Proposed

Method for Drop-Weight Tear Test of Ferritic Materials 4 is virtually

an enlarged Charpy bar in which the use of full thickness and increased

fracture path results in a highly reliable fracture appearance versus

temperature transition curve The original version of the DWTT 5 [7]

specimen is similar to that shown in Fig 2 with the exception that an

embrittled weld rather than a pressed notch is used as a crack starter,

in order to minimize initiation energy Thus, a transition curve is ob-

tained which reliably represents the crack propagation energy versus

temperature behavior that the material would display in structure This

test will be encountered in later sections in connection with other design

philosophies Assuming that an adequate representation of the toughness

versus temperature behavior has been obtained, we now ask what in-

formation is contained in this behavior which can be used to guarantee

that the material does have sufficient toughness for fracture considera-

tions to be neglected in calculating the load-carrying capacity of a struc-

tural member

It was mentioned previously that t h e criterion of upper shelf toughness

1969 Book of A S T M Standards, P a r t 31, p 1092

5 This test is now called the Dynamic Tear Test (DT) by its originators, W S

Pellini and P P Puzak

Trang 13

such as 100 percent shear lip fracture undoubtedly would be sufficient for

these purposes However, in the customary use of this philosophy a lower

level of toughness is often considered to be sufficient One method of

establishing the proper level for a particular application is by correlation

with actual service experience Apparently, this procedure has been

applied only in the case of World War II ship failures, obviously a most

undesirable method of accumulating data since it involves the actual

failure of the final product in service Such data are pertinent only to

the material, geometry, and history of imposed loading and environment

related to the particular structure Since data related even to the simu-

lation of the service loading and environment are usually not available,

other criteria are chosen These include the midpoint of the energy

absorbed, fracture appearance, or temperature difference between the

upper and lower shelf We will not consider the differences among these

criteria, but observe that, as a criterion the "midpoint" toughness is

used widely The basis for adopting this criterion is not clear It appears

that initially the midpoint criterion may have been chosen on the basis of

satisfactory service performance However, this correlation would pertain

only to a particular set of service conditions for a particular material

The limits of validity of the toughness criteria associated with the

transition temperature approach have not been established c!early With

reference to Fig 3, where a range of typical toughness versus temperature

behaviors for various materials are represented schematically, it is clear

from experience that the transition temperature approach works well

o YS 4O

Trang 14

for the top curve and is not applicable for high-strength, temperature-

insensitive materials represented by the lower curve In the region

between these extremes a reasonably smooth change in toughness be-

havior takes place Let us imagine that the transition temperature

approach is being applied successively to materials of decreasing tough-

ness After starting with a very tough material we would reach a material

where the midpoint criterion would no longer guarantee the toughness

required by the transition temperature concept Continuing this process

of application, we would reach a material whose maximum toughness is

not sufficient for the material to be used in design by the transition

temperature approach At these levels the conventional impact testing

procedures are no longer valid Thus, if these limits of validity were

known, the extent of the role of conventional impact tests in charac-

terizing the toughness of materials for design would be established We

will estimate these limits by relating results of the Charpy test to a struc-

turally based criterion of adequate toughness A qualitative guarantee

of fracture-safe performance requires that the material be capable of

arresting a running crack when subjected to gross stress levels equal to

the yield strength of the material A less conservative criterion can be

adopted only when the possibility of fracture initiation can be precluded

In lieu of direct measurements, estimates can be based on approximate

critical flaw sizes which have been obtained by Pellini and are shown in

Fig 30 of Ref 8 The pertinent information for this discussion has been

reproduced in Fig 4 of this paper Critical flaw depths of a surface crack

in a tensile sheet of 1 : 10 geometry which is loaded to stresses correspond-

ing to the yield stress of the material are determined approximately by

linear elastic fracture mechanics methods By correlation of the fracture

toughness with the upper shelf Charpy energy level the toughness of

various types of steels can be represented on the same figure The specifica-

tion of crack arrest toughness is roughly equivalent to requiring that the

material be capable of load carrying in the presence of a "large" flaw

when its toughness properties are related to dynamic (Charpy) condi-

tions At temperatures corresponding to the upper shelf toughness level,

optimum steels guarantee fracture-safe performance up to materials of

150 ksi yield point Conventional steels do not meet this criterion above

110 ksi yield point At temperatures corresponding to the midpoint

toughness it can be estimated that optimum steels do not guarantee

adequate toughness above 100 ksi and conventional steels above 80 ksi

yield point

There are various other situations in which it is not possible to guar-

antee that fracture considerations can be neglected when calculating the

load-carrying capacity of a structural member In addition to the pre-

viously discussed inapplicability to temperature-insensitive materials, a

similar situation arises when the service temperature corresponds to the

Trang 15

Yield Strength, ksi

FIG 4 -Critical flaw depths of surface cracks of 1:10 geometry for applied stress

equal to the yield strength of steel of I in thickness (from Pellini [8])

lower toughness shelf for any material Finally, the transition tempera-

ture approach cannot provide any useful information when a relatively

large flaw is formed by fatigue, stress corrosion cracking, or physical

damage, and the possibility of fracture at stresses in excess of yield level

is present In order to provide information in these areas where the

transition temperature approach is not applicable other methods have

been devised

Fracture Analysis Diagram

The fracture analysis diagram (FAD) is associated closely with the

transition temperature approach, since it deals with essentially the same

type of materials The FAD provides critical stress-flaw size information

as a function of temperature [9], whereas the aim of the transition tem-

perature approach is to remove consideration of fracture from design

calculations The actual toughness of the material is not dealt with

directly by the FAD method, but it is inferred from the critical stress-

Trang 16

flaw size relations The data for the diagram, Fig 5, is obtained in the

following manner The lowest curve (heavy line) represents the stress

levels at which a running crack will be arrested for a particular material

thickness as determined by a crack arrest test (CAT) [10,11] Since

complete fracture does not occur, the curve can be considered to repre-

sent the ability to resist a flaw of infinite size Information for a range of

critical flaw sizes has been obtained by averaging data from a variety of

structural tests at temperature levels corresponding to minimum tough-

ness Since fracture at this level is completely brittle, the results are

assumed to be independent of thickness These results for critical flaw

size are shown along the vertical axis at the corresponding levels of

critical stress expressed as a ratio of the yield level Information in the

transition region is obtained by extrapolation Constant flaw size curves

are faired along "parallel" to the CAT curve up to the yield stress level

Above the yield point the curves are continued by fairing in asymptot-

ically to the ultimate stress level The diagram can be obtained also by a

simpler but more approximate method [9], based on the ASTM Method

for Conducting Drop-Weight Test to Determine Nil-Ductility Transition

Temperature of Ferritic Steels (E 208-66 T) An analysis of service

failures of low-strength steels [9] has shown a good correlation with

FAD procedure, One problem which arises in applying this method in

design is the uncertainty concerning the type of stress which should be

associated with the ordinate of the diagram The structural tests upon

which the critical values are based involve different types of stress dis-

tributions In the diagrams used for correlating with service failures [9]

both gross and local stress, arising from stress concentrations or residual

stresses, have been associated with the ordinate of the diagram The

FAD does not recognize the difference in quality among materials at a

particular strength level, that is, all 80-ksi yield strength materials are

assumed to possess the same maximum toughness Despite these and

Trang 17

other approximations which have been questioned, the proponents of the method feel that the adequate service experience justifies the procedure The 50 percent toughness level of the D T has been related to the yield stress level on this diagram [8] Thus, the greater convenience of the

D T in relation to the CAT may be utilized

In common with the transition temperature approach, the FAD assumes that the ratio of the toughness to strength level of the material

is constant A particular flaw size is related always to the same ratio of applied stress to yield strength Therefore, this procedure has the same limits of application as the transition temperature approach Instead of attempting to guarantee high toughness, the FAD approach provides approximate quantitative, critical stress-flaw size information that can

be used on the lower shelf or through the transition region As a result, the load-carrying capability of a structural member of high toughness material may be approximated for the case of flaws arising from fatigue, stress corrosion cracking, or physical damage However, the inability to calculate the load-carrying capability still exists for structural members fabricated from low toughness, temperature-insensitive materials or for any material stressed into the plastic region Methods for dealing with these situations have been developed and will be discussed in the following section

Mechanics Methods for Design Against Fracture

The philosophy of the mechanics approach to design against fracture is

to deal with a quantitative measure of toughness and size of flaw in calculating the-toad-carrying capacity of a structural member We will consider first the method applicable to a high toughness material with gross plastic stresses

Plastic L i m i t Load A n a l y s i s - - T h e s e general theoretical methods have been developed in the theory of plasticity by assuming that the material

is ideally plastic [12] The general method is a limiting technique which provides an upper and lower bound to the maximum load that a structure can withstand For simple cases these bounds often converge to a single exact result Since we are interested primarily in a conservative result consistent with preventing fracture, we require only that the lower bound be obtained Therefore, we will deal only with a simplified aspect

of the general method The lower bound is found by assuming any stress distribution which does not exceed the yield stress and is in equilibrium with the applied loading A simple example is the case of a center-notched specimen, Fig 6 When a uniform yield level stress distribution is assumed

on both sides of the notch, the resulting lower bound for the limit load is:

P = ~ v s ( W - 2a)t (1)

Trang 18

P

t - Thick

~ - ~ "1 2a

P

Thus, an estimate of the maximum permissible flaw size in a region of

gross plastic stress may be obtained The second mechanics method to be

considered takes the toughness of the material into account directly

Linear Elastic Fracture Mechanics In order to fulfill the requirements

of the mechanics approach, a method must be useful for the complete

range of imposed loading and structural geometry for which strength

calculations are possible A method dealing with fracture must be capable

of defining and measuring material toughness in a manner which is

relevant to the mechanics approach for design The Griffith-Irwin linear

elastic fracture mechanics (LEFM) approach does have these features

Although originally expressed in terms of energy, an alternate stress

field form is usually more convenient The stress components at a point

of the linear elastic stress field surrounding a sharp, through crack in a

plate can be expressed independently of load and geometry in the

form [18]:

K

where:

~ = component normal to crack at the point,

K = constant, function of load and geometry, and

r,0 = length and direction of radius vector from tip of crack

The (a~)m,x occurs when 8 is zero Since f(O) then equals one, the equation

Trang 19

MATTHEWS ON ROtE OF IMPACT TESTING 13

For a particular loading and geometry of interest ~ , as found by linear elasticity, becomes infinite at the tip of the crack where r is zero The mathematical limiting process can be used to evaluate K for the particular imposed conditions,

ac = remotely applied uniform stress normal to the plane of the crack,

and

a - half crack length

For plane strain conditions, it is assumed that crack instability will occur for any configuration when the stress field reaches the same critical distribution Since Eq 2 shows that the stress field ahead of a crack always has the same 1/r ~/2 form, independent of loading and geometry, the critical value of the stress field will be constant Since the stress field

is always proportional to K, the value of K may be associated with the inception of critical cracking This parameter may be determined experimentally from crack instability tests and is designated as the fracture toughness for plane strain (K~o) Typical tests for measuring K~o use notched and fatigue cracked tension and slow bend specimens Extensive tests of various geometries and loadings have verified the assumption that K~r is a constant material parameter Thus, for center cracked plate

a designer could use Eq 5 with K~o a known material constant, and determine either the critical crack length for a given stress or the allowable applied stress for a particular crack length The necessary equations for

a wide variety of geometries and loadings are available [13]

The application of this approach is limited at present since, except for the antiplane shear deformation case, only linear elastic stress field solutions are available Theoretical complexities or excessively large computer requirements have thus far prevented the attainment of adequate elastic-plastic solutions Thus, the amount of plastic behavior in the real material must be quite small if the elastic field solution is to afford

a satisfactory approximation The range of validity, however, has been reasonably well defined: Ref 14 and ASTM Proposed Method for Test

of Plane-Strain Fracture Toughness of Metallic Materials.e

A second limitation is that to date techniques for measuring fracture toughness have been standardized only for the case of plane strain restraint Difficulties have arisen in the definition of the instability in the

s 1969 Book of A S T M Standards, P a r t 31, p 1099

Trang 20

| 4 IMPACT TESTING OF METALS

J::

K~c

M

Thickness

mixed mode case In addition, the fracture toughness parameter for the

mixed mode case is a function of thickness, Fig 7, rather than a fixed

material parameter 7 The effect of the limitations of L E F M upon its

applicability will be dealt with later

Most of the measurements of fracture toughness thus far have been

conducted by slow loading at room temperature This is not a result of

limitations imposed by the L E F M approach but because it was felt that

the materials to which it was applied were nearly insensitive to tempera-

ture and rate effects Interest in accounting for these effects in measuring

fracture toughness and in the possibility of an extended range of L E F M

under dynamic conditions has lead to the use of impact techniques

in LEFM

Impact testing for fracture toughness to date has made use of a Charpy

impact machine and specimen and a standard KIr bend specimen

Fatigue precracking and side notches are employed in the Charpy

specimen in an effort to obtain flat, plane strain fracture Two methods

of conducting the tests and interpreting the results have been developed

The instrumented impact method measures the force versus time rela-

tionship of loading by applying strain gages to the striker head of the

force at the instant of crack instability is used in the appropriate L E F M

equation for a bar in bending An estimate of the kinetic energy of the

specimen has been made in order to account for any reduction in the

actual load imposed on the specimen below the value measured at the

brittle materials the correction for the energy absorbed reduces the

measured K~o considerably

area (A) directly to the fracture toughness expressed in energy form (G):

Gro = W / A

7 The discussion of t h e behavior of K at very small thicknesses will be omitted

Trang 21

The assumptions made in this procedure have been discussed previously

[14,17,18] This method assumes that all of the energy lost by the pendu-

lum of the impact machine is associated with the formation of new frac-

ture surfaces Therefore, the kinetic energy of the specimen is considered

to be negligible This assumption may not be realistic for brittle materials

in light of the instrumented impact results A second assumption is that

as the crack extends, the instantaneous value of W/A either remains

constant or varies in such a manner that the total W/A is equal to the

instantaneous value at the critical crack length It has been observed

that the more formal of these assumptions, that W/A remains constant,

is likely to be valid only for strain-rate insensitive materials [17]

The results of these impact methods have/been surprising Dynamic

KIr of a material obtained from precracked, side-notched, impact Charpy

specimens often has been larger than the value from static loading of the

material with the same type of specimen This is contrary to intuition

and past experience for many materials Impact loading of specimens is

used widely presumably to produce the most severe conditions that are

likely to be imposed on a material Whether the recent impact values

should be accepted is open to question since the validity of techniques

in these tests have not been investigated thoroughly As an example,

both impact procedures use side notches to promote gross plane strain

behavior in the specimen This is necessary since there would appear

to be no hope of measuring plane strain values by a pop-in or deviation

from linearity method under dynamic conditions Although widely used,

side notches have not been accepted generally because of the uncertainty

of their effect upon the stress field in front of the crack [14] Another

question of validity concerns the maximum level of Kit which can be

measured by the Charpy bar under dynamic conditions

A further point of interest with regard to these impact testing proce-

dures is identification of the source of the rate effects which are measured

As a consequence of their different techniques, the two methods for

obtaining dynamic KIr do not involve the same rate effect The instru-

mented impact test reflects the influence of loading rate at the inception

of crack instability The energy absorbed test is influenced primarily

by propagation effects, since fatigue precracking is used Therefore, the

movement of the crack front determines the rate effect which is measured

It is improbable that the rates introduced by these different sources

would coincide for a wide variety of materials Since the rates are meas-

ured in the impact test they are known to fall in the range shown in

Fig 8, taken from Eftis and Krafft [19] The energy absorbed tests are

likely to have a much wider range extending from very slow crack

velocity for tough materials to relatively fast rates for brittle materials

Apparently crack speeds have not been measured in small bend specimens

due to experimental difficulties It is likely, however, that crack speeds

Trang 22

would not reach the levels measured in wide plates, Fig 8 Thus, the

rate effects in dynamic KIe tests are somewhat uncertain, as well as their

influence upon Kic The foregoing discussion has illustrated that our

knowledge of dynamic and temperature effects upon fracture toughness

is limited and that impact tests can play an important role in this area

We have described the L E F M approach and its associated testing

procedure and indicated its usefulness for rather brittle materials It is of

interest to inquire, with reference to Fig 3, at temperature corresponding

to the upper shelf how far up into the toughness region is L E F M appli-

cable? It is certain, except when designing with very thick materials,

that static plane strain L E F M is applicable only to the lowest toughness

materials This limitation is imposed by lack of restraint rather than by

excessive plasticity The upper limit of plasticity can be expressed by

requiring the plastic zone size to be small relative to pertinent physical

dimensions such as the length of a crack or remaining net section There-

fore, if L E F M could realize its potential in the mixed mode case, its

usefulness would be extended to higher toughness materials except when

the structural dimensions are extremely small Since these requirements

involve actual physical dimensions, it is not possible to establish a gen-

eral toughness limit for the method As a final note, L E F M has shown the

additional capability of successfully correlating fatigue and stress cor-

rosion behavior for various types of loading and geometry [~0] In addi-

tion, the information is expressed in a very convenient form for calculating

critical crack sizes

In reviewing the applicability of all methods we have discussed thus

far, it is evident that a class of materials of intermediate maximum

toughness remains for which we do not have a method for designing

against fracture

Design Approaches for Material8 of Moderate Maximum Toughness

There are no standard, generally accepted methods for designing with

materials of moderate maximum toughness The region is not well defined

Trang 23

since it extends between areas where the application of transition temperature and fracture mechanics methods is uncertain However, many important structural materials may be included in this class, such as steels with yield strengths between 120 and 200 ksi, and many aluminum and titanium alloys

The most direct approach for dealing with this class of materials would appear to be the determination of the stress level associated with crack arrest at the service temperature This stress level would be used to establish the maximum load-carrying capacity of the structure Although the CAT is considered to be quite conservative when applied to low- strength materials, there is no assurance that this is also true for intermediate strength materials The stress distributions during the test are not known, and crack lengths of initiation or arrest are not dealt with quantitatively There is no evidence that this method has been used for design with materials of moderate maximum toughness

A method which has been applied uses the D T test The energy absorbed in the D T hat, been correlated experimentally with the strain levels associated with cr~,ck propagation in large-scale structural simulation type specimens [7] T h e purpose is to obtain the level of absorbed energy which correspond'~ to the requirement of yield stress levels to propagate

a crack The test used for structural simulation is either the explosion tear test (ETT) or a drop weight version of the test, Fig 9 In these tests strain gages are used to measure the stress levels remote from the crack If a material does possess sufficient toughness to require yield stress levels for propagation, then the calculations for load-carrying capability are carried out without regard for fracture considerations This method can be extended to cover large flaws in plastically loaded regions by the use of plastic limit analysis If the material does not possess sufficient toughness, some conservative approximation of the fracture toughness can be made 8 and the load-carrying capability calculated by LEFM However, this very conservative method is not appealing since it does not recognize the improved properties of the moderate toughness material

This completes our discussion of design approaches and their associated testing procedures which has indicated various areas of utility and some

8 We will not speculate how this approximation might be made in this paper

Trang 24

18 IMPACT TESTING OF METALS

basic limitations of impact testing methods We will now consider briefly the characterization of the toughness of materials in acceptance and screening tests

Acceptance and Screening Tests

When selecting a test for acceptance and screening it is of particular importance to consider the expense involved For routine acceptance testing of material related to a particular design, the Charpy test is attrac- tive because it is convenient and inexpensive The Charpy test result is related directly to design only for high-toughness steels with a thickness comparable to the Charpy bar or related by direct correlation with service failures of that design For general use the Charpy test must be correlated with one of the more basic tests A similar situation arises in screening or ranking tests The Charpy test is useful for investigating the suitability of various materials for a particular design or for studying the effect of processing or metallurgical variables upon toughness How- ever, it is necessary either to establish a correlation with a more basic test or to carefully restrict the application to a rather narrow range of material behavior To date, valid correlations of Charpy V-notch with basic tests have been limited to particular applications Recent results 9 indicate that more general correlations are possible The D W T T can

be used also for these applications Although its size is less convenient, the possible elimination of correlating tests may be advantageous

Summary of the Role of Impact Tests in Toughness

Testing of Materials

Impact tests of various types have been discussed in connection with several testing procedures We will now collect the observations related

to each particular test

The conventional Charpy V-notch test is not suitable in general for providing toughness information for design against fracture because of its limited thickness and inapplicability to low- and moderate-toughness materials However, when correlated with more basic information, the Charpy test is very useful for acceptance and screening purposes

The D T can be used for design of high-toughness materials in the conventional transition temperature approach or by association with the fracture analysis diagram In addition, this test can provide approximate information for design with moderate toughness materials by correlation with other tests It may be used also for acceptance and screening tests Impact testing for dynamic fracture toughness provides a convenient means for studying temperature and rate effects upon fracture toughness

of materials although techniques are still in the developmental stage

Trang 25

MATTHEWS ON ROLE OF IMPACT TESTING 19

A c k n o w l e d g m e n t

T h e a u t h o r is i n d e b t e d t o J I B l u h m f o r h e l p f u l d i s c u s s i o n s a n d s u g -

g e s t i o n s c o n c e r n i n g v a r i o u s p h a s e s of t h e p r e p a r a t i o n of t h i s p a p e r

R e f e r e n c e s

[1] Bluhm, J I., "Failure Analysis, Theory and Practice," presented at the William

Hunt Eiseman Conference on Failure Analysis, American Society for Metals,

New York, 1966

[2] "A Review of Engineering Approaches to Design Against Fracture," Subcom-

mittee on Prevention of Fracture in Metals, American Society of Mechanical

Engineers, 1965

[3] Adachi, J., "A Survey of Fracture Design Practices for Ordnance Structures,"

AMMRC MS 68-03, Army Materials and Mechanics Research Center, 1968

Also to be published in Fracture, Vol 5, Academic Press, New York

[~] Newhouse, D L and Wundt, B M "A New Fracture Test for Alloy Steels,"

Metals Progress, Feb 1961, p 81

[5] Bluhm, J I., "A Model for the Effect of Thickness on Fracture Toughness,"

Proceedings, American Society for Testing and Materials, Vol 61, 1961, p 1324

[6] Bluhm, J I., "Geometry Effect on Shear Dip and Fracture Toughness Transition

Temperature for Bimodel Fracture," Proceedings, American Society for Testing

and Materials, Vol 62, 1962

[7] Pellini, W S et al, "Review of Concepts and Status of Procedures for Fracture-

Safe Design of Complex Welded Structures Involving Metals of Low to Ultra-

High Strength Levels," NRL Report 6300, U.S Naval Research Laboratory,

1965

[8[ Pellini, W S., "Advances in Fracture Toughness Characterization Procedures

and in Quantitative Interpretations to Fracture-Safe Design for Structural

Steels," Bulletin, Welding Research Council, No 130, May 1968

[9[ Pellini, W S and Puzak, P P., "Fracture Analysis Diagram Procedures for the

Fracture-Safe Engineering Design of Steel Structures," NRL Report 5920,

U.S Naval Research Laboratory, 1963

[lOJ Robertson, T S., "Propagation of Brittle Fracture in Steel," Journal, Iron and

Steel Institute, Vol 175, 1953, p 361

[11] Feely, F J et al, "Studies of the Brittle Failure of Tankage Steel Plates,"

Welding Journal Research Supplement, Vol 34, No 12, 1955, p 596s

[12] Drucker, D C et al, "The Safety Factor of and Elastic Plastic Body in Plane

Strain," Transactions, American Society of Mechanical Engineers, Vol 73,

Journal of Applied Mechanics, p 371

[13] Paris, P C~ and Sih, G C M., "Stress Analysis of Cracks," Fracture Toughness

Testing, A S T M STP 381, American Society for Testing and Materials, 1964, p 3D

[141] Brown, W F., Jr., and Strawley, J F., Plane Strain Crack Toughness Testing of

High Strength Metallic Materials, A S T M STP ~10, American Society for Testing

and Materials, 1966

[15] Radon, J C and Turner, C E., "Fracture Toughness Measurements by Instru-

mented Impact Test," presented at the Second National Symposium on Fracture

Mechanics, Lehigh University, 1968, to be published in Engineering Fracture

Mechanics

[16] Orner, G M and Hartbower, C E., "Precracked Charpy Fracture Toughness

Correlations," paper presented at A S T M Symposium on Fracture Testing and

Its Applications, June 1964

[17] Irwin, G R., "Crack Toughness Testing of Strain-Rate Sensitive Materials,"

Transactions, American Society of Mechanical Engineers, Vol 86A, 1964, p 445

[18] Radon, J C and Turner, C E., "Note on the Relevance of Linear Fracture

Mechanics to Mild Steel," Journal, Iron and Steel Institute, Vol 204, Aug 1966,

pp 842-845

Trang 26

[19] Eftis, J and Krafft, J M., "A Comparison of the Initiation with the Rapid Propagation of a Crack in a Mild Steel Plate," Transactions, American Society

of Mechanical Engineers, Vol 87D, 1965, pp 257-263

[~0] Johnson, H H and Paris, P C., "Sub-Critical Flaw Growth," Engineering Fracture Mechanics, 1968, Vol 1, pp 3-45

[Zl] Shoemaker, A K and Rolfe, S T., "Static and Dynamic Low-Temperature Kx~ Behavior of Steels," presented at AWS-ASME Meeting, Chicago, Ill.,

2 April 1968, to be published in Transactions, American Society of Mechanical Engineers, Journal of Basic Engineering

Trang 27

J H Gross ~

EfFect of Strength and Thickness on

Notch Ductility

REFERENCE" Gross, J H., "Effect o f S t r e n g t h a n d T h i c k n e s s on

N o t c h D u c t i l i t y , " Impact Testing of Metals, A•TM STP ~66, American

Society for Testing and Materials, 1970, pp 21-52

ABSTRACT For a number of years, standards and code-writing bodies

have been attempting to specify fracture toughness in terms of the Charpy

V-notch rather than the Charpy keyhole test because various investigators

have shown that V-notch test results correlate much better with service

experience However, the change to V-notch specifications has been deterred

by uncertainty concerning the best criterion for establishing transition

temperatures and the effect of strength and thickness on transition tempera-

tures Therefore, five steels covering a wide range of yield and tensile strengths

(ABS-C:-39/63, A302-B: 56/88, HY-80: 81/99, A517-F: 121/134, and

HY-130: 140/148) were tested as quarter-, half-, single-, and double-width

(QW, HW, SW, and DW) Charpy V-notch specimens in the longitudinal

and transverse directions and with through-thickness and surface notches

Transition temperatures were determined for various energy-absorption,

lateral-expansion, and fracture-appearance criteria

The results showed that energy-absorption criteria for determining transi-

tion temperature should increase with strength to ensure a constant notch

ductility Thus the best method for determining transition temperature was

the direct measurement of lateral expansion Of the lateral-expansion cri-

teria evaluated, the 15 rail value agreed best with fracture-mechanics

considerations

The average increase in transition temperature was 60 F from QW to HW

specimens, 26 F from HW to SW specimens, and 2 F from SW to DW speci-

mens This indication of maximum constraint for the SW specimen was not

consistent with the effects produced when the standard V-notch was re-

placed with a fatigue crack Consequently, the size of the Charpy test speci-

men that should be used for evaluating thick plates has not been established

and requires additional study

The effects of strength and thickness on transition temperature were

much larger than the effects of testing direction, notch location, or notch

acuity

The results indicate that, of the various criteria for evaluating the Charpy

V-notch impact-test performance of structural steels, lateral expansion is the

best criterion for compensating for the important effects of steel strength

and plate thickness Moreover, its validity is supported by fracture-mechanics

concepts

1 Manager, Steel Products Development, Applied Research Laboratory, U.S

Steel Corp., Mouroeville, Pa 15146

Trang 28

K E Y W O R D S : steels, notch impact strength, tensile strength, notch tough-

ness, ductility tests, ductile brittle transition, fracture toughness, fracture

mechanics, evaluation, tests

For more than 50 years, standards and code organizations such as American Society for Testing and Materials (ASTM) and American Society of Mechanical Engineers (ASME) have specified fracture-toughness requirements for ferrous materials on the basis of energy absorption

in the Charpy keyhole-notch test Beginning in about 1950, however, Williams'[12] ~ classic study of ship-plate fractures, and subsequent studies by Pellini and co-workers I8,~,], showed that brittle failure of ship plates could be correlated with the Charpy V-notch characteristics of the plate but that the failures could not be correlated with the Charpy keyhole-notch characteristics Since that time, most experimental investigations have employed the Charpy V-notch rather than the keyhole notch when standard impact specimens were tested

Standards and code organizations also have been evaluating the feasi- bility of replacing the keyhole-notch specimen with the V-notch specimen

in specifications calling for impact tests The adoption of the Charpy V-notch test has been complicated by the numerous investigations that have shown that the 15-ft.lb transition temperature is applicable only to

a limited group of steels For example, the writer has recommended [5] that the Charpy V-notch energy-absorption transition-temperature criteria should be 11, 18, and 25-ft.lb, respectively, for steels with tensile strengths of 60, 100, and 140 ksi to ensure a constant notch ductility of

15 mils lateral expansion In accordance with this recommendation, the Boiler and Pressure Vessel Committee recently adopted Charpy V-notch impact testing "for weldments and all other materials for shells, heads, nozzles, and other vessel parts subject to stress due to pressure for which impact tests are required ," and increased the required energy absorption with increased tensile strength as shown in Table la (UG-84.1) These energy-absorption specifications lie between those recommended [5] for 15 and 20 mils lateral expansion, Fig 1 In Division 2 of Section VIII, toughness requirements for quenched and tempered steels are specified

on the basis of lateral expansion (AM-311.4) or the nil-ductility temperature (NDT) (A5~-312) The use of these criteria, as well as energy absorption and fracture appearance, indicate that the best method for specifying fracture toughness has not been established

The adoption of the Charpy V-notch test also has been complicated by the selection of criteria for subsize specimens Because of limited data, the UG-84 specification for subsize specimens scales the required energy absorption in direct proportion to the width (along the notch) of the specimen However, as shown in Table lb (UG-84.2), the test temperature

2 The italic numbers in brackets refer to the list of references appended to this paper

Trang 29

GROSS ON NOTCH DUCTILITY 23

section VIII, A S M E Boiler and Pressure Vessel Code)

(A) Table UG-84.1

M i n i m u m C h a r p y V - N o t c h I m p a c t E n e r g y R e q u i r e m e n t s

for C a r b o n a n d Low-Alloy Steels Listed in Table UCS-23 (Except SA-353 a n d SA-372)

C h a r p y V - N o t c h I m p a c t Energy, ft.lb

Trang 30

TENSILE STRENGTH, kli

FIG 1 Energy absorpLion required for A~ME Boiler and Preesure Vessel Code

Specification UG-8~.I compared with previous experimental observations from Ref 5

is reduced by 5, 20, and 50 F for 3/4, 1/2, and 1/4-width specimens,

respectively, when the width of the test specimen is less than 80 percent

of the material thickness This reduction in test temperature recognizes

the lowering of the transition temperature which occurs when the speci-

men width is decreased and which would raise unfairly the energy

absorption if subsize specimens were used to evaluate thick materials

However, neither the energy level for establishing the transition temper-

ature for subsize specimens nor the appropriate reduction in test temper-

ature for subsize specimens has been confirmed conclusively

The present study was undertaken to investigate further the best

criteria for evaluating Charpy V-notch impact-test data and to establish

the effects of strength and thickness on transition temperatures selected

on the basis of the various criteria

M a t e r i a l s a n d E x p e r i m e n t a l W o r k

Materials

The five structural steels used in the present investigation were

obtained as 1-in.-thiek plates and had the chemical compositions shown

in Table 2

Experimental Work

For each steel, 0.252-in.-diameter tension specimens were machined

from the quarterthickness location in the longitudinal and transverse

Trang 33

GROSS ON NOTCH DUCTILITY 27

directions, and the specimens were tested at room temperature Similarly,

longitudinal and transverse, double-, single- (standard), half-, and

quarter-width, surface- and through-thickness-notched Charpy V-notch

impact specimens were machined from the quarter-thickness location,

and the specimens were tested over a range of temperatures to establish

the ductile-to-brittle transition on the basis of energy absorption, lateral

expansion, and fracture appearance In addition, a series of longitudinal

through-thickness-notched fatigue-cracked Charpy V-notch impact speci-

mens was tested for each steel Full-size (P-1 type, ASTM E208) drop-

weight specimens were machined in the longitudinal and transverse

direction and tested to establish the NDT Subsize (P-3 type) drop-

weight specimens were machined from the broken halves of the P-1

specimens (preserving the original plate surface) and tested as for the P-1

specimens

R e s u l t s a n d D i s c u s s i o n

Tensile Properties

The longitudinal and transverse tensile properties of the steels inves-

tigated are listed in Table 3 The yield strengths ranged from about 40 to

140 ksi and the tensile strengths from about 60 to 150 ksi The ratio of the

yield strength to the tensile strength increased as the strength increased

and as the microstructure changed from ferrite and pearlite for the

ABS-C and A302-B steels to tempered bainite or tempered martensite

or both for the HY-80, A517-F, and HY-130 steels, which were quenched

and tempered The only significant difference between the longitudinal

and transverse properties was the slightly lower ductility in the trans-

verse direction

Fracture-Toughness Results

Drop-weight and Charpy V-notch test data were plotted to permit

determination of the nil-ductility transition temperature and various

energy-absorption, lateral-expansion, and fracture-appearance transition

temperatures To maintain brevity in the present paper, only the transi-

tion temperatures have been reported herein However, because of the

general usefulness of the 65 transition-temperature plots for the five

steels, they have been published in Welding Research Council Bulletin

No 147, Jan 1970, "Transition-Temperature Data for Five Structural

Steels." Figure 2 illustrates typical Charpy V-notch curves for A302-B

steel from which the various transition temperatures were determined for

the conditions indicated

steels investigated are summarized in Table 4 The agreement between

the N D T values for the standard-size (P-1 type) and the subsize (P-3

Trang 34

N o T E - - W h e r e two N D T t e m p e r a t u r e s are listed, t h e y represent the highest t e m -

p e r a t u r e a t which a " G o " or failure was observed a n d the lowest t e m p e r a t u r e a t which

a " N o G o " or no failure was observed T h e former is t h e usual measure of t h e N D T

t e m p e r a t u r e

Trang 35

GROSS ON NOTCH DUCTILITY 29

type) specimens in both the longitudinal and transverse directions was

excellent for the ABS-C, A302-B, and A517-F steels In contrast,

scatter in the N D T values was observed for the HY-80 and the

HY-130 steels The N D T values for these two steels were very low

compared with those for the other steels In addition, the change from

all failures to no failures (Go to No Go) occurred over a much broader

range of temperatures Consequently, the N D T values for the standard

and subsize specimens and for the longitudinal and transverse directions

cannot be compared reliably The relation between the N D T values and

Charpy V-notch transition temperatures will be discussed in subsequent

sections

peratures for various Charpy V-notch criteria are summarized in Tables 5

through 9 and Figs 3 through 7 A comparison of the average 15-ft.lb

transition temperature for the standard-size (single-width) specimens

with the N D T temperature shows that the disagreement between the

15-ft-lb value and the N D T increased as the strength of the steel increased

The average 15-ft.lb transition temperature, the average NDT, and t h e

difference were as follows:

Transition Temperature, deg F

These results indicate that the energy at which the transition temper-

ature is determined should increase with the strength of the steel This

effect is illustrated also by the energy absorption at the NDT, which rose

from 22 to 36 ft.lb as the strength of the steel increased The temperature

difference and N D T fix for HY-80 steel follow the pattern, but the changes

were much larger and were inconsistent with the average increases for

the other four steels These results strongly suggest that the N D T may

be as ultraconservative for certain other steels as for HY-80 steel The

results also illustrate the problem that has confronted standards and

code bodies in the selection of energy-absorption values for various steels

on the basis of the N D T fix For example, on the basis of the effect of

strength, an argument can be presented for greatly increasing the energy-

absorption requirement for ABS-C, A302-B, A517-F, and HY-130 to be

Trang 39

15 ff'-Ib ~ TN

15 f l - l b - SN

Tiêu đề	Impact Testing Of Metals
Người hướng dẫn	D. E. Driscoll, Symposium Chairman
Trường học	University of Washington
Thể loại	Báo cáo hội thảo
Năm xuất bản	1970
Thành phố	Atlantic City

Định dạng
Số trang	322
Dung lượng	7,14 MB