A. K.

SHOEMAKER
Associate Research Consultant,
Notch-Ductility Transition of Structural
Applied Research Laboratory,
U. S. Steel Corporation,
Monroeville, Pa.
Mem. ASME
Steels of Various Yield Strengths
The notch-ductility transition of six structural steels, A 36, ABS-Class C, A 302-
Grade B, HY-80, A517'-Grade F, and HY-130, ranging in yield strength from 36 to 137
ksi, was studied with the use of 6 / 8 and 1 in. dynamic-tear (DT) test specimens.
The results were compared with previously published data for V-notch and fatigue-
cracked Charpy tests and dynamic fracture-toughness (Kin) tests. Energy, lateral-
contraction, and fracture-toughness values were compared. The results of this study
showed that the full-shear upper energy shelves in the Charpy V-notch and DT speci-
mens are the products of constant average plastic energy densities for each steel and the
plastic volume estimates for the fracture of the different specimens. The transition from
ductile to brittle fracture behavior is essentially the same in the fatigue-cracked Charpy
and D T specimens since, for each steel, the same lateral contraction was measured in
each specimen broken at a given temperature. This lateral contraction increased ex-
ponentially with temperature until a fidl-thickness shear fracture developed. However,
the maximum lateral contraction increased with increased test-specimen thickness, sug-
gesting thai the K0 values corresponding to fidl-shear fracture should also increase with
thickness. Using the proportionality found between the lateral contraction and the
values of KID^/^YDE for the brittle-fracture behavior of these steels, the Kc values are
estimated to be as much as 4.5 times greater than the Klc values at the same tempera-
tures. In general, the notch-ductility transition can best be quantitatively characterized
by the lateral contraction through KID and Kc values, whereas upper shelf energies are
related by constant plastic energy densities and plastic volumes which develop during
fracture.

Introduction best quantitative approach to interrelating stress and flaw-size

Nt UMEROus qualitative and quantitative methods
have been used to measure the notch toughness of a steel, t h a t is,
its resistance to the extension of a flaw. Comparing or inter-
'elating these different methods is at best difficult, since struc-
tural steels exhibit transitional fracture behavior from a brittle
low-energy fracture a t low temperatures to a ductile high-energy
fracture at higher temperatures. T h e difficulty in defining
toughness through this transition range is t h a t the transition be-
havior is sensitive to loading or strain rate, notch acuity, speci-
men thickness, and even specimen geometry.
Although some success has been achieved in qualitatively relat-
'ag different tests, few quantitative approaches exist that can be
Used for design purposes. Fracture mechanics has offered the
Contributed by the Pressure Vessels and Piping Division and
Presented at the First National Congress on Pressure Vessels and
''Ping, San Francisco, Calif., May 10-12, 1971, of THE AMERICAN
SOCIETY OF MECHANICAL ENGINEERS. Manuscript received at
4SME Headquarters, February 8, 1971. Paper No. 71-PVP-19.
behavior [ l ] . 1 To date, this analysis is restricted primarily to
the more brittle or frangible behavior where the fracture process
is confined within an analyzable elastic field. Thus, quantita-
tive determination can be made for some current commercial
high-strength steels t h a t do not exhibit transitional behaviors.
In addition, limited information is available on the application
of fracture mechanics to the brittle lower-shelf transition behavior
of intermediate and low-strength steels t h a t exhibit a fracture
transition [2].
Estimates of stress and flaw-size relationships for upper shelf
behavior have been proposed through the use of the ratio-
analysis-diagram (RAD). This analysis by Pellini [3] and Loss
[4] is related to fracture-mechanics concepts and correlations
between different types of fracture-toughness tests, the primary
test being t h e test with the dynamic-tear (DT) specimen. This
analysis is useful in defining relative levels of yield stress, operat-
ing stress, and toughness for preliminary stress—flaw size con-
siderations. I n most applications more detailed calculations are
required prior to the final design and materials selection.
1
Numbers in brackets designate References at end of paper.

Journal of Engineering for Industry FEBRUARY 1 9 7 2 / 299

Copyright © 1972 by ASME
Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 Table 1 Chemical composition of structural steels tested—percent

Type
Steel analysis C Mn s Si Ni Cr Mo V Ti B Cu Al* i'l** N*** 0
A 36 Ladle 0.23 1.08 0.010 0.022 0.025 ND ND ND ND ND N D ND ND ND ND ND
Check 0.24 1.06 0.007 0.019 0.024 0.032 0.057 0.013 ND ND N D 0.051 0.002 0.003 0.004 0.0014
ABS-C Ladle 0.19 0.60 0.013 0.021 0.22 ND ND ND ND ND N D ND ND ND ND ND
Check 0.20 0.60 0.009 0.016 0.23 0.02 0.07 0.01 ND ND N D 0.03 0.057 0.057- 0.006 0.0020
A302-B Ladle 0.19 1.37 0.011 0.010 0.18 ND ND 0.49 ND ND N D ND ND ND ND ND
Check 0.19 1.40 0.017 0.014 0.16 0.03 0.07 0.49 ND ND N D 0.03 0.007 0.010 0.010 0.0059
HY-80 Ladle 0.15 0.31 0.009 0.015 0.22 2.28 1.44 0.31 0.001 0.001 ND ND ND ND ND ND
Check 0.16 0.28 0.011 0.016 0.22 2.26 1.46 0.30 0.005 0.005 N D ND 0.015 0.016 0.003 0.0044
A517-F Ladle 0.16 0.79 0.010 0.016 0.23 0.87 0.53 0.43 0.04 N D 0.003 0.26 ND 0.031 ND ND
Check 0.17 0.78 0.012 0.017 0.23 0.88 0.56 0.42 0.036 ND 0.0033 0.26 0.033 0.035 0.005 0.0041
HY-130 Ladle 0.110 0.80 0.005 0.006 0.23 4.94 0.56 0.55 0.060 ND ND ND ND 0.025 ND ND
Check 0.110 0.85 0.009 0.007 0.23 4.91 0.58 0.58 0.050 ND ND ND 0.017 0.021 0.009 0.0033
* Acid-soluble.
** Total.
*** Kjeldahl determination.
N D — N o t determined.

Table 2 Longitudinal static room-temperature mechanical properties of structural steels tested

0.252-in. diameter
tension specimens
yield Elongation in 1 in., Reduction of area,
strength percent percent
Plate (0.2 percent tensile at at
thickness, offset), strength, maximum at maximum at
Steel in. ksi ksi load fracture load fracture
A 36 1V« 36 75 20.0 32.0 20.0 68.0
ABS-C 1 39 63 24.0 36.0 26.0 66.8
A302-B 1 56 88 11.0 26.0 17.4 67.0
HY-80 1 84 99 11.0 25.0 13.4 74.9
A517-F 1 118 129 8.0 19.0 11.8 65.4
HY-130 1 137 143 8.0 20.0 10.8 70.9

Because additional quantitative information is needed to inter-
relate results from different types of fracture-toughness tests,
results from D T tests of six structural steels, ranging in yield
strength from 36 to 137 ksi, were compared with previously pub-
lished static and dynamic crack-toughness, KIC} values [2], with
standard V-notch and fatigue-cracked Charpy-specimen values,
[5] and with nil-ductility transition (NDT) [2] values for these
same steels. The results used in the present study were confined
to those obtained for specimens with a maximum thickness of 1 in.
Materials and Results
For this study, 1 or P / s in. thick plates of the following six
structural steels were used: A36 (36 ksi lower yield strength),
ABS-C (39 ksi lower yield strength), A302-B (56 ksi lower
yield strength), HY-80 (84 ksi yield strength), A517-F (118 ksi yield
strength), and HY-130 (137 ksi yield strength). The chemical
composition and mechanical properties of these six structural
steels are shown in Tables 1 and 2, respectively. All steels were
manufactured by using production melting, rolling, and heat-
treating practices previously described [6].
The fracture energy results determined for the representative
steels with standard Charpy V-notch specimens, as reported by
Gross [5], are shown in Figs. 1 through 6, together with the re-
sults determined with machined-notch 6 /s in. thick D T specimens.
Test results determined with electron beam welded 1 in. thick
D T specimens are included for A302-B, A517-F, and HY-130
steels. The maximum lateral contraction was determined near
the end of the crack in fatigue-cracked Charpy specimens and
near the notch tip in 5 / 8 a n d 1 in. thick D T specimens, Figs. 7
through 11. This maximum contraction normally occurred at

n r
or o
1000.- 100

j i

-100 -50 0 50 100 150 200 250

TEMPERATURE, F
Fig. 2 Energy absorption for CVN and 6 /a in. DT specimens of ABS-C
Fig. X Energy absorption for CVN and 6 /s '"• Dl specimens of A36 steel steel

300 / F E B R U A R Y 1972 Transactions of the AS ME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 1 i i i 1 1 1 1

CVN
160
- "
A
A A A
A
- 120 / • - &
-
Q

/~
CVN. ^

A /
o ,°
A. ^ 5 / 8 - I N C H DT
** 400 - 40
- -
AA^< A y°
]NDT
. A—!T o<"
0 &-r
-300 -250 -200 -150
I I
-100
1
-50
1 1

TEMPERATURE, F

s
Fig. 3 Energy absorption for C V N , */« in. DT, and 1 in. DT specimens Fig. 4 Energy absorpfion for C V N and / s in. DT specimens of H Y - 8 0
of A302-B steel stee|

r- 60 1 1 1 1 o 1 I l_
O

CVN
WT^-0^^
- 50
\ ~

ry// °^
* " * " - ^ C ]

o' /° /
P
- 40 A/ \
/ / / 1-INCH DT

- 30
A/A A Z
t/ ~
A Q/ / \ 5 / 8 - I M C H DT

- 20 _ A / / / ~"
A / /
A/ ° /

_ .^
A
/ u -
^ ^
o , NDT
0 1 • 1 ! 1 1 1 1 1
-150 -100 -50 0 50 100 150 200
TEMPERATURE,F

ENERGY ABSORPTION FOR CVN, 5 / 8 - I N C H DT, AND

I-INCH DT SPECIMENS OF A 5 I 7 - F STEEL Fig. 6 Energy absorption for C V N , 6 / s i n . DT, and 1 in. DT specimens of
H Y - 1 3 0 steel
Fig. 5 Energy absorption for C V N , 6 / s in. DT, and 1 in. DT specimens of
A 5 1 7 - F steel

1000 1 1 1 1 1 1
800 800 -
600 - - 600 -

400 - 400

200 - - 5 / 8 - I N C H DT
5 / 8 - I N C H DT

5 100 100
e 80 - 80
z
o
60
^£—1—*—* 60 AIB
A
tT
FATIGUE-CRACKED CVN
\-
u.
K
40 —
°#f £ FATIGUE-CRACKED CVN __
O
m 20 - ' A - i0 ~~ O
< /L\ A A
> 10
a - o / _ 10 - A /
8
2u
8
in
" _ 6 ?"
/ / ° 4
A

,° A

- 2
NDT NDT
1 1 i 1 1 1 1
1 1 _L X
-100 -50 50 100 150 200 250 50 100 150 200 250
TEMPERATURE,F TEMPERATURE,F

Fig. 7 Lateral contraction for sharply notched specimens of A 3 6 steel Fig. 8 Lateral contraction for sharply notched specimens of ABS-C steel

Journal of Engineering for Industry FEBRUARY 1 9 7 2 / 301

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 1000
800

50 100 150 200 250 150 -100 -50 0 50

TEMPERATURE,F TEMPERATURE.F

Fig. 9 Lateral contraction for sharply notched specimens of A 3 0 2 - B steel Fig. 10 Lateral contraction for sharply notched specimens of A517-F
steel

1000
800
600
formation t h a t can hopefully be applied in fracture-mechanics and
400 fracture-control considerations.
Upper-Shelf Energy Behavior. The plastic volume and energy de-
200 veloped during full-shear fracture have been described by Bluhm
[7] and later by Witt and Berggren [8]. Consider the slip and
plastic-zone development at the tip of the notch in the D T speci-
men for full shear, t h a t is upper shelf behavior, Fig. 12. Since
both the crack length and the ligament behind the crack (the
ligament is equal to the specimen width minus the crack length)
< are larger than the specimen thickness, slip for plastic deforma-
ce
f- tion develops through the thickness on approximately 45 deg
•21
O planes. Such a slip pattern was experimentally verified by studies
of Hahn and Rosenfield on an Fe-Si alloy [9]. The height of the
plastic zone is approximately equal to its width, t h a t is, the plate
thickness, t. The upper shelf energy level is then equal to the
average plastic-energy density times this volume.
As a first-order approximation for full shear behavior, the
plastic volume t h a t develops during the fracture of the D T
specimen is
NDT
1
j _
D T plastic volume » /i t2l (1)
_L I
-250 -200 -150 -100 -50 0 50 100 where t — thickness of D T specimen
TEMPERATURE,F I = specimen ligament behind crack
Fig. 11 Lateral contraction for sharply notched specimens of HY-130 For the 6 / 8 in. D T specimen (t = 0.625 in. and I = 1.125 in.), the
steel
plastic volume is 0.22 in. 3 , whereas the plastic volume of the 1 in.
D T specimen (t = 1.0 in. and I — 3.0 in.) is 1.5 in. 3 . Such an
approximation has definite limits of application. If the crack
distances from the notch tip of V2 to 1 times the specimen thick-
length or ligament becomes too short in comparison with the plate
ness. The lateral-contraction data for fatigue-cracked C harpy
thickness (such as that described below for the Charpy specimen),
specimens were used rather than the values for standard V-
slip lines will develop between the notch tip and the free surface,
notch specimens, because the fatigue-cracked specimens evidenced
and equation (1) will not govern. Likewise, if the ligament be-
a notch-acuity effect t h a t caused a shift of t h e curve to higher
comes too large in comparison with the specimen thickness, the
temperatures [5].
load developed during bending could cause excessive compressive
plastic deformation at the loading points that would invalidate
Discussion equation (1). Finally, if the specimen is so thick that shear lips
The data utilized in this study are discussed below in terms do not develop through the thickness, equation (1) would be in-
of upper shelf, transitional, and lower shelf behavior. The valid.
results obtained by different means of quantitatively measuring For the Charpy specimen, however, both the notch and liga-
toughness are examined, the primary emphasis being given to in- ment length are shorter than the specimen thickness, and slip

302 / F E B R U A R Y 1972 Transactions of the AS ME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 PLATE THICKNESS,
LENGTH OF PLASTIC ZONE
DYNAMIC-TEAR SPECIMEN
SPECIMEN
THICKNESS, b
TENSILE
PLASTIC ZONE COMPRESSIVE
PLASTIC ZONE
CHARPY SPECIMEN
Fig. 12 Schematic plastic-zone development for full-shear behavior in
DT and C V N specimens
should therefore develop in bending through the ligament behind
the notch, Fig. 12. This bending mode of deformation causes a
tensile plastic zone at the notch tip and a compressive plastic zone
at the back free surface. If similar average plastic flow and
plastic energy density are assumed for tensile and compressive
loading, then as a first-order approximation, the plastic volume
t h a t developed for full shear fracture in the Charpy specimen is
1
CVN plastic volume « /i bh*
where b = CVN specimen thickness
h = ligament behind notch
For a full-size Charpy specimen (6 = 0.394 in., h = 0.315 in.),
the calculated plastic volume is 0.0196 in. 3 , and the shelf energy
would be the product of this volume and the average plastic
energy densitj'.
If it is assumed t h a t all the material within the plastic volume
t h a t developed during full shear fracture has about the same
average plastic energy density (area under the stress-strain
curve), the same ratio of measured upper shelf energies should
occur for each steel and should be about the same as the ratio of
the calculated plastic volumes. The measured and calculated
ratios are shown in Table 3. A mean value of 9.2 for the mea-
sured plastic-volume ratio of the 6 /s in. D T shelf to the CVN shelf
is about 20 percent lower than the calculated ratio of 11.2. Con-
sidering the crude approximations of plastic-zone shape and
energy distribution between specimens of such significantly
different plastic deformation, a 20 percent difference would appear
reasonable.
If the through-thickness mode of slip is assumed for the Charpy
specimen instead of the bend mode, the calculated plastic-zone
ratio for 6 /s in. D T specimen to Charpy specimen is 9.0, which
would be in good agreement with the measured ratios. However,
such an agreement may be fortuitous. The plastic zone at the
notch tip is large compared with the ligament length and would
tend to cause general yielding in bending before fracture occurs.
This general yielding should suppress the through-thickness slip
mode. Likewise, subsize Charpy specimens (thickness less than
the standard value) generally scale in proportion to the thickness
rather than to the thickness squared, which supports the bending
slip system. In summation, because of the small size of the
Charpy specimen, the upper shelf behavior is probably a com-
bination of both types of deformation. The pertinent feature is
t h a t for the six steels examined the constant ratio of about 8.8 was
found.
The measured plastic-volume ratios for 1 in. and 6 /s in. D T
specimens for the six steels are in very good agreement with the
calculated ratio of 6.8. These ratios are somewhat lower than the
empirically determined value of 8.0 reported by Puzak and Lange
[10], As shown by the D T energy curves for the A36 steel, Fig.
1, and for the HY-130 steel, Fig. 6, the upper shelf energy may
occasionally decrease instead of remaining constant with increas-
ing temperature. T h e values reported in Table 3 reflect the
maximum upper shelf energy values and give ratios consistent
with the calculated ratios. As expected, the average plastic
energy densities of steels, Table 3, decrease with increasing yield
strength except for the highly alloyed HY-80 and HY-130 steels,
which have exceptionally high plastic energy density.
Similar correspondence in volumetric energy relationships
for specimens of this thickness range have been discussed b y
Witt and Berggren for an A533-B steel [8]. I n contrast, Pellini,
Loss, and co-worker have suggested the determination of the
ratios of the upper shelf energies in terms of area rather than
(2) volume, for specimens up to 12 in. in thickness [4]. The in-
consistency between methods is due to the use of thick specimens
(greater than about 3 in.), where the inability of materials
to slip and deform to the extremes necessary for shear-lip de-
velopment through the thickness of heavy sections causes the
through-thickness slip mode to be invalid. Witt and Berggren
have shown t h a t volumetric considerations still apply to heavy
sections if the shear-lip energy is separated from the flat-fracture
energy for thick sections. The good agreement in the data in
Table 3 for scaling the energies of thinner specimens on the basis
of plastic volumes certainly substantiates the use of such a cal-
culation.
Lateral Contraction for Fracture Transition and Upper Shelf Behavior.
For all the steels examined the lateral contraction appeared to
increase linearly with temperature on a semilog plot, Figs. 7
through 11. The lateral contraction was effectively independent

Table 3 Upper-shelf energy results for structural steels tested

verage plastic energy

density* for E /s in.
Upper-shelf energy, ft-lb Ratio of upper-shelf energies D T Specimens,
CVN Vs in. D T 1 in. D T Vs i n . / C V N 1 in./Vs in. 10" in-lb/in. 3
A 36 100 950 9.5 2.6
ABS-C 100 890 8.9 2.45
A302-B 85 740 5,000 8.7 6.75 2.03
HY-80 125 1030 8.2 2.83
A517-F 50 565 3,250 11.3 5.75 1.55
HY-130 100 880 6,000 8.8 6.83 2.42
Calculated 11.2 6.8
Ratio**
* Plastic energy density-energy/calculated plastic volume.
** Calculated ratio = ratio of calculated plastic volumes for each specimen.

Journal of Engineering for Industry FEBRUARY 1 9 7 2 / 303

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 1/2 LATERAL CONTRACTION
FULL-THICKNESS PLANE-STRESS BEHAVIOR
SLANT FRACTURE

PLANE-STRAIN
r-
THICKNESS, t. B-
PARTIAL SHEAR FULL SHEAR
DEVELOPMENT DEVELOPMENT
Fig. 14 Schematic diagram of crack toughness for different plate thick-
Fig. 13 Schematic representation of lateral contraction and crack-open- nesses
ing displacement for slip development at a crack tip

increased ductile toughness with increased plate thickness will

of specimen thickness through the transition range of each speci-
men and reached different plateau values in the different
specimens examined. The maximum lateral contraction in-
creased with increasing specimen thickness for all steels examined.
I n the transition region, however, the concept that the
lateral contraction for a steel is a function of temperature only
and is independent of thickness is also consistent with the partial-
shear development for the through-thickness slip mode shown in
Fig. 13. Several investigators have previously shown that shear-
lip thickness for a steel is dependent only on temperature [11] and
is independent of specimen thickness until full-thickness shear
behavior is achieved. Since the shear-lip thickness is indepen-
dent of specimen thickness, the temperature dependence of the
shear-lip thickness should be controlled by the effect of tempera-
ture on some critical shearing strain of the steel. From geometry
considerations the size of partial-shear lips should be consistent
with lateral contraction and crack opening. Thus, in the transi-
tion range, the lateral contraction is dependent on temperature
and not on specimen thickness.
As shown schematically in Fig. 13, from geometrical considera-
tions the lateral contraction occurring from through-thickness
slip on 45-deg planes should be, as a first-order approximation,
equal to the crack-opening displacement. The crack-opening
displacement, 5, in turn can be related to the crack toughness,
Kc, [12]
be confined to section sizes in which full shear behavior can be
developed.
The concept of increasing toughness with increasing thickness
for ductile fracture has not been generally accepted. However,
application of these ideas to designing against ductile fracture in
thin sections could be of real engineering and economic signifi-
cance. For example, consider designing against long-running
shear fractures in gas transmission pipeline. This is a particularly
severe design application, since it is necessary to design for
sufficient toughness to stop a running crack, rather than simply
design against the unstable initiation of cracks as is done in normal
pressure vessel designs. If the toughness increases with the
square of the thickness (plastic volume is proportional to i 2 ), op-
erating at higher pressures with proportionally thicker pipe
walls should increase the resistance to long-running fractures (par-
ticularly in large diameter pipe). Another viewpoint would be
t h a t for higher pressures and proportionally thicker walls, the
toughness would increase as the thickness squared, whereas the
strain energy would only increase as the first power of the thick-
ness; consequently, thicker ductile sections would be beneficial.
Again, this behavior could be achieved only if the same plastic
energy density can be maintained for the thicker walls and if the
through-thickness mode of slip dominates the plasticity at the
tip of a running crack.

K> 50
(3) 1 1
o ABS-C
where <JY — yield strength at the appropriate temperature and • A302-B
40 —
- 0 A5I7-F —
strain rate V HY-130 0
E = Young's modulus 'o

Thus, the lateral contraction is related to the Kc behavior and z

o 30 — Ov —
equal to the right-hand side of equation (3). Carman and Irwin A 0 /

have verified this behavior with measurements of Kc and lateral < 0 /

contraction for through-thickness slip of fractures in center- /
z 20 — 0 /O D
—
cracked specimens [13]. As shown in the following section, equa- o
tion (3) requires a constant of proportionality for application to 0 „ /
0 V
the brittle lower-energy-shelf plane-strain behavior.
o / V
Since the lateral contraction, and therefore the crack-opening 10 — / —
displacement, increases with thickness for full shear behavior, / a
the crack toughness, Kel must also increase with thickness. This D
toughness increase with thickness occurs for thicknesses less than /
/ • 1 1
those corresponding to the maximum Kc value obtained in plane- 1.0 2.0 3.0
stress behavior (see left portion of Kc thickness curve shown
-3
schematically in Fig. 14). This behavior is predicted on the h*L JO inch
Etr
assumption t h a t full ductile behavior is achieved and t h a t the YD
increase in plate or section thickness is not accompanied by a
Fig. 15 Comparison of lateral contraction and caiculated crack opening
degradation of the actual metallurgical toughness. Therefore, for dynamic plane-strain fracture toughness

304 / F E B R U A R Y 1972 Transactions of the AS ME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Lower-Energy-Shelf Plane-Strain Fracture Behavior. The lower-
energy-shelf brittle-fracture behavior of structural steels has re-
ceived the most attention from quantitative approaches to frac-
ture problems. The small plastic-zone sizes and large con-
straint enable this brittle behavior to be analyzed by applying
linear-elastic fracture mechanics concepts. Previous studies with
these same steels gave fracture toughness, Kla values and yield
strengths for both static and dynamic-loading conditions [2].
These data can now be compared with the impact data, since the
crack-tip strain rates associated with impact loading are within
an order of magnitude of the dynamic-strain rates obtained in the
KtD studies of these steels.
T h e relationship among lateral contraction, crack opening,
and KIc can now be examined for dynamic loading. Burdekin
has shown that even for brittle behavior the lateral contraction
at the crack tip is equal to the crack opening, <5 [12]. Therefore,
for the brittle-fracture behavior associated with the lower-energy
shelves of these steels, the crack opening, 5, in equation (3) is
replaced by the lateral contraction and must be proportional to
KID*
impact lateral contraction oc (4)
<JYDE
where KID = dynamic KIc value at the test temperature
<TVD = dynamic yield strength at the test temperature
The application of equation (4) to four of the structural steels
is shown in Fig. 15. Crack-toughness values were not previously
obtained for the A36 steel, and lateral-contraction measurements
were not made for the HY-80 steel. For each steel the data in-
clude the temperature range from —320 deg F to the N D T tem-
perature, since plane-strain K1D values were obtained in 1 in.
thick specimens for temperatures up to the N D T temperature.
The results for the four steels indicate a linear proportionality
between the lateral contraction and KjD2/trYDE; the HY-130
results have the greatest deviation from this proportionality.
From Fig. 15 it is apparent t h a t the lateral contraction is 20 times
greater than the value calculated from the quantity KID2/o-yDE.
For all the steels examined, a constant lateral contraction of about
0.030 in. occurred at the N D T temperatures. A constant lateral
contraction or crack opening at the N D T temperature is con-
sistent with previous results which showed t h a t the N D T tem-
perature is the upper boundary for dynamic plane-strain fracture
in 1 in. thick sections [2].
If the thickness is insufficient to measure plane-strain IC!e
values, the same lateral contraction and crack opening will give
Kc values greater than the KIc. If, in fact, equation (3) is correct
for ductile behavior, it should be possible to obtain plane-stress
Ke values as large as \ / 2 0 or about 4.5 times the Klc value.
Such an estimate of crack-toughness increase due to lack of con-
straint in thin sections, points out the high degree of conservatism
that is adopted when calculations of critical stress and flaw size
are made on the basis of KIc behavior for plate thicknesses that are
much less than those required for the constraint of plane strain.
Summary
The notch-ductility transition of six structural steels ranging
in yield strength from 36 to 137 ksi was studied with the use of V-
Uotch and fatigue-cracked Charpy specimens, 6 / s and 1 in. thick
| dynamic-tear (DT) specimens, and previously reported dynamic-
fracture-toughness, KID, values. The results of this study may
I be summarized as follows:
1 For temperatures in the range of the upper shelf in the D T
[ and Charpy specimens, the energies absorbed in fracture can be
' related through the plastic volume, giving a constant average
plastic energy density for fracture in each steel. The plastic-
Volume ratio between 6 /s in- thick D T and CVN specimens for
full-shear behavior was about 8.8, whereas the ratio between the
1 and the 6 /s in. thick D T specimens was about 6.8.
2 T h e lateral contraction increased linearly with tempera-
'ui'e, independent of thickness, as an exponential function up to a
Journal of Engineering for Industry
Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 02/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
maximum plateau value that was dependent on the thickness.
3 The maximum lateral contraction increased with increas-
ing thickness, an indication that K„ values should increase with
thickness.
4 From geometry considerations, the crack-opening displace-
/ K* \
ment I S = —^- I should be equal to the lateral contraction.
However, for the brittle lower-energy-shelf behavior, the mea-
sured lateral contraction was 20 times greater than that estimated
from the plane-strain fracture-toughness values and KjD2/aYDE,
an indication that Kc values may be as much as 4.5 times larger
than the Klc values.
References
1 Fracture Toughness Testing and Its Applications, American
Society for Testing and Materials STP 381, 1065.
2 Shoemaker, A. K., and Rolfe, S. T., "The Static and Dynamic
Low-Temperature Crack-Toughness Performance of Seven Struc-
tural Steels," presented at the National Symposium on Fracture
Mechanics, Lehigh University, Bethlehem, Pa., June 1969.
3 Pellini, W. S., "Evolution of Engineering Principles for Frac-
ture-Safe Design of Steel Structures," Naval Research Laboratory
Report 6957, Washington D. C , Sept. 23, 1969.
4 Loss, F. J., "Dynamic Tear Test Investigations of the Frac-
ture Toughness of Thick-Section Steel," Naval Research Laboratory
Report 7056 HSSTP-TR-7, Washington D. C , May 14, 1970.
5 Gross, J. H., "Effect of Strength and Thickness on Notch
Ductility," Impact Testing oj Metals, American Society for Testing
and Materials STP 466, Mar. 1970.
6 Clausing, D. P., "Tensile Properties of Eight Constructional
Steels Between 70 and —320 F," Journal of Materials, Vol. 4, No. 2,
June 1969.
7 Bluhm, J. I., "A Model for the Effect of Thickness on Frac-
ture Toughness," Proceedings of the ASM, Vol. 6, 1961.
8 Witt, F. J., and Berggren, R. G., "Size Effects and Energy
Disposition in Impact Specimen Testing of ASTM A533 Grade B
Steel," Oak Ridge National Laboratory Report ORNL-TM-3030, Oak
Ridge, Tenn., Aug. 1970.
9 Hahn, G. T., and Rosenfield, A. R., "Local Strain Measure-
ment of Ship Steel," Battelle Memorial Institute Report to Bureau of
Ships, Department of the Navy, Project SR-164, Contract No.
NObs-88348, Jan. 15, 1965.
10 Puzak, P. P., and Lange, E. A., "Standard Method for the
1-Inch Dynamic Tear Test," Naval Research Laboratory Report
6851, Washington D. C , Feb. 1969.
11 Shoemaker, A. K., and Rolfe, S. T., "Static and Dynamic Low-
Temperature Kic Behavior of Steels," Journal of Basic Engineering,
TBANS. ASME, Vol. 91, Series D, No. 3, Sept. 1969.
12 Burdekin, F. M., "Initiation of Brittle Fracture in Structural
Steels," British Welding Journal, Dec. 1967.
13 Carman, C. M., and Irwin, G. R., "Characterization of the
Plane Stress Fracture Toughness by Crack Opening Displacement,"
presented at the Fourth National Symposium on Fracture Mechanics,
Carnegie-Mellon University, Pittsburgh, Pa., Aug. 1970.
DISCUSSION
2
E. A. Lange
The author has provided a wealth of information concerning
the fracture resistance characteristics of six commercial steels
of high interest. The source of the many problems encountered
when criteria for toughness are based upon Charpy test results
is clearly displayed in Figs. 1-6. A comparison of the tempera-
ture transition curves for C„ energy and D T energy for the differ-
ent materials shows t h a t there is no fixed displacement for the
transition region of the C„ energy curve. This requires a sepa-
rate calibration for C„ energy for each steel to adjust the transi-
tion region to t h a t for a sharp notch or crack.
In the "Discussion" section of this paper, an hypothesis con-
cerning the plastic volume for full shear behavior is proposed
t h a t is not substantiated experimentally. The proposed model
intimates t h a t plastic strain is uniform in the plastic zone, and
t h a t the plastic zone is limited to a zone with a square cross
section bounded by the plate surfaces. This assumption leads
2
Engineer, Department of the Navy, Metallurgy Division, Naval
Research Laboratory, Washington, D. C.
FEBRUARY 1 9 7 2 / 305