Sunteți pe pagina 1din 6

ELSEVIER Journal of Nuclear Materials 212-215 (1994) 1682-1687

Current status of small specimen technology in Charpy

impact testing
H. Kurishita, H. Kayano, M. Narui, M. Yamazaki
The Oarai Branch, Institute for Materials Research, Tohoku University, Oarai, Ibaraki-ken 311-13, Japan


The current status of small-scale specimen technology in Charpy impact testing for ferritic steels is presented,
with emphasis on the effect of the notch dimensions (notch depth, notch root radius and notch angle) on the upper
shelf energy (USE) and ductile-to-brittle transition temperature (DBTT). The USE for miniaturized specimens,
normalized by Bb* or (Bbj3/* (B is the specimen thickness, b the ligament size), is essentially independent of notch
geometry and has a linear relationship with the USE of full size specimens, regardless of irradiation and alloy
conditions. The DB’IT of miniaturized specimens depends strongly on the notch dimensions; this dependence of the
DB’IT decreases as the DBTT of full size specimens increase due to neutron irradiation or thermal aging. These
results may be useful in determining the USE and DBTI for full size specimens from those for miniaturized

1. IntruduMion the notch dimensions (notch depth, notch root radius

and notch angle) on the USE and DBTT is presented.
It is well known that small specimen technology is
required in mechanical testing for the fusion reactor
materials development program, since it must rely on 2. Experimental
accelerator-based high energy neutron sources that are
quite limited in irradiation volume. In Charpy impact Four ferritic steels that have different Charpy im-
testing, with which the present paper is concerned, the pact properties were used, i.e., JFMS (Japanese fer-
major problems caused by the ‘use of miniaturized rite/martensite dual phase steel) in the unirradiated
specimens are that both the upper shelf energy (USE) and irradiated conditions, and two other high strength
and the ductile-to-brittle transition temperature ferritic steels that have lower values of USE and DBTI
(DB’IT) are considerably lower than those of full size than the unirradiated and irradiated JFMS. Full size
specimens. Many attempts to determine the USE and and miniaturized Charpy specimens were prepared in
DBTT of full size specimens have been made by using which the axis of the specimen lay in the rolling direc-
one third sized [l-9] or more miniaturized specimens tion. The miniaturized specimens were of four differ-
[lo-121. Although some successful results have been ent sizes, 3.3 by 3.3 by 23.6 mm, 2.0 by 2.0 by 20 mm,
reported, it is desirable to improve the methodology to 1.5 by 1.5 by 20 mm and 1.0 by 1.0 by 20 mm. They are
correlate the USE and DBTI of different sized speci- here called 3.3, 2, 1.5 and 1 mm specimen, respectively.
mens. Each of them had three or four variations in notch
The present authors had a particular interest in the dimensions; an example of the nominal and measured
effect of the notch geometry on Charpy impact test values is tabulated in Table 1 for unirradiated JFMS.
results of ferritic steels because the effect of the notch The measured values correspond to the average of all
may be controlled by varying notch dimensions. In this the measured values, with the range of measured val-
paper, the current status of the studies on the effect of ues given in parentheses. A notch root radius of ap-

0022-3115/94/$07.00 0 1994 Elsevier Science B.V. All rights reserved

LSD1 0022-3115(94)00280-2
H. Kurishita et al. /Journal of Nuclear Materials 212-215 (1994) 1682-1687 1683

proximately 0.02 mm is the minimum value that can be scanning electron microscopy GEM) to obtain the lat-
achieved by machining. A very large notch root radius ter two measurements. The values of USE and DBTI
of 0.25 mm for specimen type no. 12 is specially intro- were determined from total absorbed energy versus
duced to examine the effect of less sharp notch on the temperature curves of all specimens tested. The DBTT
USE and DB’IT for the smallest-sized 1 mm speci- was determined as the temperature at which one half
mens. of the USE was absorbed during fracture.
Neutron irradiation was performed to a dose of
3 x 10z2 n/m2 (E > 0.1 MeV) at about 300°C in the
Japan Materials Testing Reactor (JMTR), in which it is 3. Results and discussion
possible to irradiate full size specimens as well as
miniaturized specimens at the same time. The effect of the notch dimensions on the tempera-
The specimens were subjected to instrumented ture dependence of the total absorbed energy in minia-
Charpy impact tests at temperatures between 93 and turized specimens is shown in Fig. 1 [ll], taking the 2
473 K using a specially designed electrically controlled mm specimen of the unirradiated JFMS as an example.
hydraulic machine. The details of the machine and the It is clear that the impact behavior is affected by each
procedures developed for testing were described ear- of the notch dimensions, i.e., notch angle, notch root
lier [13]. The span was set at 40 mm for full size radius and notch depth. The USE is dependent on
specimens, 15 mm for 3.3 mm specimens, 12.5 mm for notch depth, but not on notch angle or notch root
the other miniaturized specimens. The ratio of the radius, whereas the DB’IT changes significantly with
span to the specimen width was 4.0 for full size speci- all three notch dimensions, particularly notch depth
mens, 4.55 for 3.3 mm specimens, 6.25 for 2 mm and notch root radius. This trend was evident for other
specimens, 8.33 for 1.5 mm specimens and 12.5 for 1 specimen sizes.
mm specimens.
The DBTT was determined using three different 3.1. Upper shelf energy
measurements: energy absorption, ductility (i.e., the
amount of lateral expansion at the compression side of To correlate the USE of miniaturized and full size
the specimen), and fracture appearance (i.e., the per- specimens, the measured USES were divided by vari-
centage of fibrous fracture). The fracture surfaces of ous geometric factors and compared. The best correla-
both halves of each broken specimen were observed by tion, which was still not satisfactory, was obtained

Table 1
Notch dimensions for miniaturized and full size Charpy specimens of JFMS
Specimen Size Notch depth Notch root radius Notch angle
Type a (mm) P (mm) TV(deg)
Nominal Measured Nominal Measured Nominal Measured
1 3.3 mm 0.51 0.508 (0.502-0.515) 0.08 0.086 (0.080-0.093) 45 45 (44-46)
2 0.51 0.513 (0.510-0.520) 0.08 0.083 (0.079-0.086) 30 30 (28-32)
3 0.51 0.509 (0.500-0.518) 0.02 0.032 (0.029-0.035) 30 29 (28-30)

4 2.0 mm 0.40 0.404 (0.400-0.410) 0.08 0.079 (0.075-0.088) 45 45 (43-47)

5 0.40 0.406 (0.396-0.408) 0.08 0.073 (0.070-0.077) 30 30 (29-32)
6 0.40 0.407 (0.405-0.409) 0.02 0.028 (0.024-0.031) 30 29 (28-30)
7 0.60 0.604 (0.600-0.608) 0.02 0.031 (0.024-0.038) 30 29 (28-29)

8 0.305 (0.299-0.310) 0.02 0.030 (0.026-0.034) 30 31(30-32)

9 0.294 (0.291-0.298) 0.08 0.075 (0.069-0.079) 45 45 (44-46)
10 0.455 (0.445-0.460) 0.02 0.027 (0.025-0.030) 30 29 (29-30)

0.203 (0.197-0.207) 0.05 0.046 (0.041-0.049) 45 45 (44-47)

0.197 (0.193-0.200) 0.25 0.246 (0.234-0.259) _
0.300 (0.295-0.305) 0.02 0.024 (0.022-0.026) 30 31(30-32)

2.00 (1.992-2.008) 0.25 0.240 (0.233-0.243) 45 45 (45)

1684 H. Kurishita et al. /Journal of Nuclear Materials 212-215 (1994) 1682-1687

Temperature W)

Fig. 1. Temperature dependence of absorbed energy as a

function of notch dimensions for miniaturized Charpy speci-
mens (2 mm specimens) of unirradiated JFMS Ill].

when Bb2 or (Bbj312 was employed as a normalizing a lW
factor, where B is the specimen thickness and b the Fig. 3. Elastic stress concentration factor, K,, as a function of
ligament size (b = W- a). These volumetric parame- notch dimensions. Here, a is the notch depth, p is the notch
ters were originally employed by Corwin and coworkers root radius, W is the specimen width and 0 is the notch angle.
After Nishida [14].
[3] and are known to be related to the extent of plastic
deformation below the notch after general yield [5].
Fig. 2 shows the effect of the notch dimensions on the
In this case, for unirradiated JFMS, all but one data
USE normalized by (Bb) 3/2 for all the miniaturized
point (corresponding to a 1 mm specimen with a very
and full size specimens of unirradiated JFMS that were
large notch root radius of 0.25 mm) exhibits a normal-
considered in this work. Here, the elastic stress con-
ized USE that is constant to within &-15%, irrespective
centration factor, K,, was used to express the effect of
of K,. This means that the observed dependence of
the notch geometry, with the K, of the experimental
USE on notch depth resulted only from the depen-
data reported by Nishida [14] as shown in Fig. 3. The
dence of the USE on ligament size: the notch geometry
reason why the K, values by Nishida are used instead
is apparently unimportant for the USE. The higher
of the well known Neuber’s formula [15] was described
normalized USE of the 1 mm specimen indicates that
in a previous paper ill].
the estimated fracture volume used as a normalization
parameter, Bb* or (BbJ3/‘, is an underestimate and
that the actual fracture volume is larger.



3 5 Y.”
0 100 100 so0

Kl USE (Full Size) (J)

Fig. 2. Normalized USE against the elastic stress concentra- Fig. 4. Plot of the ratio, (Y, of USE/(Bb)3/2 of full size
tion factor, K,, for full size and miniature Charpy specimens specimens to that for third size specimens against the unnor-
of unirradiated JFMS [ll]. malized USE of full size specimens in ferritic steels.
H. Kurishita et al. /Journal of Nuclear Materials 212-215 (1994) 1682-1687 1685

On the other hand, for the irradiated JFMS and occurs by a larger reduction in specimen volume asso-
other unirradiated high strength ferritic steels, the ciated with fracture for full size specimens than pre-
normalized USE of full size specimens was always dieted using the volumetric parameters Bb* or (Bbj312.
lower than that of miniaturized specimens, and the To demonstrate this concept, the ratio, (Y, of the nor-
ratio of the full size to miniature normalized USE malized USE of full size specimens to that of miniatur-
decreases with decreasing USE (unnormalized) of full ized (third size) specimens was calculated for all the
size specimens [12]. This indicates that for full size ferritic steels reported to date in the literature, includ-
specimens of these ferritic steels the normalization ing the unirradiated and irradiated JFMS considered
parameter, Bb’ or (Bb)3/2, is an overestimate of the in the current work. The results are listed in Table 2,
fracture volume and that the actual fracture volume where the value of (Ygiven is for a normalization factor
should be smaller. These results suggest that the ob- of (Bb)3/2. Fig. 4 shows cx against the unnormalized
served difference in the effect of specimen size on the full size USE for the data given in Table 2. The
USE of the different alloy and irradiation conditions relationship between (Y and unnormalized USE of full

Table 2
Values of USE and normalized USE for full size and third size specimens, together with the ratio, a, of USE/(Bbj3/* for full size
specimens to that for third size specimens in irradiated and unirradiated JFMS and other ferritic steels reported in the literature
Alloy Full size specimens Third size specimens
USE USE/(BbJ3” USE/Bb’ USE USE/(BbJ3” USE/Bb* a a
(J) (J/mm31 (J/mm3) (J) (J/mm3) (J/nun3)
9Cr-lMo-V-Nh 262 0.366 0.409 9.7 0.347 0.378 1.05
norm. and temp.
J-T [41
9Cr-lMo-V-Nh 200 0.277 0.313 8.8 0.315 0.343 0.879
norm. and temp.
TL t41
9Cr-lMo-V-Nh 111 0.153 0.173 5.7 0.204 0.222 0.750
LT [41
9Cr-lMo-V-Nh 72 0.101 0.113 5.1 0.183 0.199 0.551
LT 141
HT-9 [7] 129 0.180 0.202 6.0 0.209 0.227 0.861
A302 [51 62 0.086 0.097 3.4 0.134 0.150 0.642
A508 [51 118 0.165 0.184 6.5 0.253 0.284 0.652
A508 [S] 68 0.095 0.106 3.9 0.151 0.169 0.629
9Cr-1W [6] 259 0.362 0.405 10.8 0.387 0.420 0.935
9Cr-2W 161 245 0.342 0.383 9.8 0.351 0.381 0.974
9Cr-4W [6] 221 0.309 0.345 8.8 0.315 0.343 0.981
2.25Cr-W (V) [8] 245 0.342 0.383 9.4 0.328 0.328 1.043
224 0.313 0.350 9.7 0.340 0.338 0.921
278 0.389 0.434 9.6 0.335 0.335 1.161
272 0.380 0.425 9.7 0.339 0.338 1.121
5Cr-2W-V [81 245 0.342 0.383 10.0 0.349 0.349 0.980
9Cr-W (MO) [81 216 0.302 0.338 9.4 0.328 0.328 0.921
255 0.356 0.398 9.7 0.339 0.338 1.050
262 0.366 0.409 9.7 0.339 0.338 1.080
200 0.280 0.313 8.8 0.307 0.307 0.912
12Cr-W (MO) [8] 192 0.268 0.300 9.0 0.314 0.314 0.854
115 0.161 0.180 5.9 0.206 0.206 0.782
JFMS 190 0.265 0.313 8.2 0.295 0.333 0.898
Unirr. [ll]
JFMS 168 0.235 0.263 8.3 0.297 0.323 0.791
Irr. [12]
a (Y= [USE/(Bb)3/2]full size/[USE/(Bb)3/21third size.
1686 H. Kurkhita et al. /Journal of Nuclear Materials 212-215 (1994) 1682-1687

size specimens is linear and therefore so in the rela-

tionship between the unnormalized USE of full size
specimens and the normalized USE of miniaturized
specimens, as shown in Fig. 5. Accordingly, it is con-
cluded that the effect of specimen size on the USE can
be understood in terms of the specimen volume associ-
ated with fracture. In addition, the observed linear
relationship between the unnormalized USE of full
size specimens and the normalized USE of miniatur-
ized specimens can be used to determine the USE of
full size specimens from the USE of miniaturized spec- loo- _I

imens. The next step is to develop a theoretical analy- 0 2 4 6 8

sis to explain the observed linearity. Kt

Fig. 6. Plot of the DBTT based on energy absorption against
3.2. Ductile-to-brittle transition temperature the elastic stress concentration factor, K,, for full size and
miniature Charpy specimens of unirradiated JFMS [ll].
No significant difference was found between the
DBTT values determined on the basis of energy ab-
sorption, ductility, and fracture appearance [ll]. The
DBTT based on energy absorption is discussed in this with a notch giving a suitable value of K,. In order to
section. DBTT is plotted against the elastic stress con- examine this possibility, data from impact tests on the
centration factor, K,, in Fig. 6 [ll] for unirradiated neutron irradiated JFMS and other unirradiated fer-
JFMS. The DB’IT varies significantly depending on ritic steels of less ductility were compared. The K,
the notch dimensions, with some of the miniature dependence of the DBTT decreases as the DB’IT of
specimen DBTT exceeding that of full size specimens. full size specimens is increased by irradiation or ther-
Further, almost all the data points for miniaturized mal aging. There is also a relationship between the
specimens lie well on a single curve, though the two DBTT of full size specimens and the dependence on
data points for 3.3 mm specimens show a slight devia- notch dimensions of the DBTI’ for miniaturized speci-
tion from the curve: the DBTT for the miniaturized mens. This result excludes the above-mentioned possi-
specimens appears to be uniquely defined by K,. The bility for determining the DBTT of full size specimens,
DBTT for full size specimens, on the other hand, has but it provides an alternative way for determining the
the different dependence on K, and is approximately DBTT of full size specimens using miniaturized speci-
equal to that of the miniaturized specimens with the mens. The details of the technique to determine the
elastic stress concentration factor of _ 4.8 (indicated DBTT for full size specimens from the notch dimen-
by the arrow in the figure). This provides the possibility sion dependence of DBTT for miniaturized specimens
that the DBTT for full size specimens can be directly are described elsewhere [ 121.
obtained from the DBTT of miniaturized specimens

4. Conclusions

The USE depends on specimen thickness and liga-

ment size, but not on notch dimensions, and the de-
Fenitk Steels
300 / pendence of USE on specimen size varies with alloy
and irradiation conditions. This variation arises from
the difference in specimen volume associated with frac-
I A4 I
t ture because unnormalized USE for full size specimens
w zoo is uniquely related to the normalized USE for minia-
” t turized specimens, regardless of alloy and irradiation
conditions. The DBTT varies significantly depending
on notch dimensions, especially notch depth and notch
root radius. The dependence of DBTT on notch di-
mensions decreases as the DBTT of full size specimens
increases due to irradiation or thermal aging.
USE/(Bbp (Third-size) (J/mm’) These results may lead to techniques for determin-
Fig. 5. Plot of the USE of full size specimens against the ing both the USE and DBTT for full size specimens
USE/(Bb) 3/2 of third size specimens in ferritic steels. from impact test results on miniaturized specimens.
H. Kurishita et al. /Journal of Nuclear Materials 212-215 (1994) 1682-1687 1687

References [8] D.J. Alexander and R.L. Klueh, Charpy Impact Test:
Factors and Variables, ASTM-STP 1072, ed. J.M. Holt
[l] W.L. Hu and D.S. Gelles, Proc. Conf. On Ferritic Alloys (American Society for Testing and Materials, Philadel-
for Use in Nuclear Energy Technologies, Snowbird, UT, phia, 1990) p. 179.
1983, p. 631. 191A.S. Kumar, B.S. Louden, F.A. Garner and M.L. Hamil-
[2] W.L. Hu, HEDL-SA-2947-FP, 1983. ton, Small Specimen Test Techniques, ASTM-STP 1204,
131W.R. Corwin, R.L. Klueh and J.M. Vitek, J. Nucl. Mater. (American Society for Testing and Materials, Philadel-
122&123 (1984) 343. phia, 1993) p. 47.
[4] W.R. Corwin and A.M. Houghland, The Use of Small [lo] H. Kayano, H. Kurishita, A. Kimura, M. Narui, M. Ya-
Scale Specimens for Testing Irradiated Materials, mazaki and Y. Suzuki, J. Nucl. Mater. 179-181 (1991)
ASTM-ST 888, eds. W.R. Cotwin and G.E. Lucas 425.
(American Society for Testing and Materials, Philadel- 1111H. Kurishita, H. Kayano, M. Narui, M. Yamazaki, Y.
phia, 1986) p. 325. Kano and I. Shibahara, Mater. Tram JIM 34 (1993) 1402.
[S] G.E. Lucas, G.R. Odette, J.W. She&herd, P. McConnell 1121H. Kurishita, H. Kayano, M. Narui, M. Yamazaki, S.
and J. Perrin, The Use of Small Scale Specimens for Mizuta and I. Shibahara, to be submitted to Mater. Trans
Testing Irradiated Materials, ASTM-ST 888, eds. W.R. JIM.
Corwin and G.E. Lucas (American Society for Testing 1131H. Kayano, H. Kurishita, M. Narui and M. Yamazaki,
and Materials, Philadelphia, 1986) p. 305. Ann. Chim. Fr. 16 (1991) 309.
161 F. Abe, T. Noda, H. Araki, M. Okada, M. Narui and H. [141 M. Nishida, Stress Concentration, 5th ed. (Morikawa
Kayano, J. Nucl. Mater. 150 (1987) 292. Publishers, Tokyo, Japan, 1984) p. 572.
171 B.S. Louden, A.S. Kumar, F.A. Garner, M.L. Hamilton 1151 H. Neuber, Theory of Notch Stresses, 2nd ed. (Springer,
and W.L. Hu, J. Nucl. Mater. 155-157 (1988) 662. Berlin, 1958) p. 71.