Available online at www.sciencedirect.

com

Available online at www.sciencedirect.com


Procedia Engineering
Procedia 00 (2011)
Engineering 000–000
10 (2011) 2725–2730
www.elsevier.com/locate/procedia

ICM11

Mechanical Behaviour of SS 316 (N) Weld after Long Term

Exposure to Service Temperatures
B. Shashank Dutta, G. Sasikalaa*, G. Shanthia, S. Venugopala, M. Nani Babua,
Pradyumna Kumar Paridab, A.K. Bhaduri a a*
a
Materials Technology Division, bPhysical Metallurgy Division
Indira Gandhi Centre for Atomic Research, Kalpakkam - 603 102, India

Abstract

The composition of the welding consumable for the SS 316 L(N) components of prototype fast breeder reactor is
adjusted such that the weld deposit contains 3 to 7 % delta ferrite, to balance between resistance to hot cracking
during welding, and embrittlement of the weld metal during service. The operating temperature range for these
components is 623-823 K, which is the range for the so-called “748 K embrittlement”, noticed in cast duplex stainless
steel. Therefore, a study was taken up to assess the degradation, if any, in the mechanical properties of this weld due
to long term service exposure. Tensile, charpy-impact and fracture (J-R curves) toughness tests were done at ambient
temperature, in the as-welded condition and after thermal ageing at 643, 748 and 823 K for durations in the range
1000 to 20000 hours. A decrease in toughness, both charpy energy and J-R curves, was observed after ageing at 643,
748 and 823 K, after long ageing times. The variations in properties are discussed in the light of microstructural
changes in the delta ferrite phase observed using optical and TEM methods and the fractographic observations on the
fracture tested specimens.

© 2011 Published by Elsevier Ltd. Selection and peer-review under responsibility of ICM11

Key words: Ageing; fracture toughness; microstructure; fracture mechanisms

1. Introduction
Austenitic stainless steels of type AISI 316 and its variants have been chosen for high temperature
applications in nuclear power plants in view of their good high temperature mechanical properties, creep
resistance and ease of fabrication. A low carbon version of this steel (C≤ 0.03 wt.% ) alloyed with

* Corresponding author. Tel.: +91-44-27480500 – Extn- 22178; fax: +91-44-27480075.

E-mail address: [email protected].
.

1877–7058 © 2011 Published by Elsevier Ltd.

doi:10.1016/j.proeng.2011.04.454
2726 B. Shashank Dutt et al. / Procedia Engineering 10 (2011) 2725–2730
2 Shashank Dutt et al/ Procedia Engineering 00 (2011) 000–000

nitrogen (N - 0.06 to 0.08 wt.%), designated as SS 316L(N) is chosen for high temperature structural
components of prototype fast breeder reactor (PFBR) under construction at Kalpakkam, India. The
expected temperature range of normal operation of these components is 643-823 K.

The weld microstructure of these steels is specified to contain about 3-7 Vol.% delta-ferrite in austenite
matrix to prevent hot cracking during solidification [1]. Previous investigations [2-6] on the aging
behavior of this class of welds have shown that brittle phases such as sigma (), laves () and/or chi ()
form as decomposition products of the delta ferrite on aging at 823-1073 K. The effect of aging leading to
precipitation of different phases is also reported in cast duplex stainless steels [7-9] containing 10% or
more ferrite. Aging of duplex SS [10] and welds containing ~10% ferrite [11] at temperatures below 773
K leads to formation of Cr-rich α′ phase either by spinodal decomposition or by nucleation and growth of
phase depending on the composition of the steel and also, precipitation of Ni and Si rich G-phase, M23C6
and/or growth of existing phases at ferrite/austenite boundaries. W.J. Mills [12] has reviewed various
factors affecting the fracture toughness of SS 304 and SS 316 welds. There have not been any studies on
the effect low temperature ageing on the mechanical behaviour of austenitic welds with δ-ferrite contents
in the range 5-6 FN, presumably due to the belief that it is insignificant below 10% ferrite content [13].

A campaign was taken up to characterize the influence of ageing at 643, 748 and 823 K on the tensile
and fracture properties of the SS 316(N) welds at room temperature. In addition to standard charpy impact
testing, J-R curves for quasi-static loading conditions were established. The results are discussed in the
light of microstructural (optical, TEM) and fractographic (SEM) investigations.

2. Experimental

Weld pads of approximate dimensions 500×400×30 mm, were prepared by shielded metal arc welding
process using the consumable developed indigenously for PFBR applications. The chemical composition
of the weld metal is given in Table 1. The weld is designated as SS 316(N) since the carbon content is
about 0.05 mass %. The weld pads were subjected to radiography to ensure absence of any unacceptable
welding defects. Specimen blanks were subjected to thermal ageing at 643, 748 and 823 K (designated A,
B and C respectively) for 1000, 2000, 5000, 10000 and 20000 h (represented by suffixes 1 to 5).

Table 1. Chemical (mass %) composition of SS 316 (N) weld

C Si Mn Cr Ni Cu Co N Mo P S Nb V

0.05 0.46 1.4 18.5 11.1 0.21 0.06 0.08 1.9 0.025 0.006 <0.07 0.075

Tensile tests were done on round specimens machined from the weld pads, having 28 mm gauge length
and 4.0 mm diameter as per ASTM E8 [14] standards, at a constant strain rate of the order of 3.1×10−4
s−1. Compact tension (CT) specimens of 20 mm thickness were fabricated with the notch placed such that
the crack plane lies in the weld and crack growth is along the welding direction, as per ASTM E 399-90
[15] specifications. These were subjected to fatigue pre-cracking at room temperature to obtain a sharp
fatigue crack ahead of the notch using a resonant fatigue machine. The crack length at the end of
precracking (initial crack length a0 for the fracture test) was targeted between 25 and 27 mm
 B. Shashank Dutt et al. / Procedia Engineering 10 (2011) 2725–2730 2727
Shashank Dutt et al/ Procedia Engineering 00 (2011) 000–000 3

corresponding to an ao/W in the range of ~ 0.50-0.54, where W = Width of the specimen. The pre-cracking
was carried out in K (stress intensity factor) decreasing mode; as the crack propagated, ∆K (stress
intensity factor range = Kmax −Kmin) was gradually reduced such that the crack was grown at loads (Pm) not
exceeding as prescribed by ASTM E 1820 [16]. Then the test specimens were side-grooved to 20% net
depth (2 mm deep groove on either surface along crack growth path with 45° flank angle and 0.25 mm
root radius) to restrict crack growth to the symmetry plane and to minimize the thumbnail crack growth. J-
tests were carried out using monotonic ramping at a constant stroke rate of 0.01 mm/s in a servo-
hydraulic machine, fully automated for test control and data acquisition. The load-line displacement
(LLD) was monitored using high resolution COD gages. A suitably calibrated direct current potential drop
(DCPD) device was used for online determination of crack length. High resolution load (P), LLD and
crack length (a) data were recorded using the data acquisition software integrated with the test application.
The tested specimens were held at about 673 K for about 30 minutes in order to heat-tint the fracture
surfaces to identify the crack extension regime in the fracture test. The samples were then pulled to
fracture at a higher pulling rate at room temperature. The initial and final crack lengths (ao and af) were
optically determined using the nine-point average method prescribed in ASTM E1820. Also linear-
interpolation method (from DCPD readings) was used to calculate the intermediate crack lengths and it
was observed that calculated crack lengths were comparable to that of measured values. The fracture
surface was examined in scanning electron microscope (SEM) to understand fracture mechanisms.
Standard Charpy V-notch specimens (10×10×55 mm3) were fabricated from weld pad in as-weld and aged
conditions. Charpy specimens were impact tested at room temperature using a 358 J capacity impact
machine at the maximum initial velocity of 5.12 ms-1.

3. Results and Discussion

The tensile properties are given in Fig.1a. A general decrease in yield strength (YS) values were observed
for ageing durations of 5000 h for all the ageing temperatures, compared to as-weld condition. Beyond
5000 h, YS values were nearly constant. However, the changes in UTS values with all the ageing
conditions were not significant. This is in agreement with the results on the effect of ageing on tensile
properties of this class of welds reported earlier [11,17, 18]. Variation of charpy V-energy with ageing is
presented in Fig. 1b. Ageing at 748 K resulted in continuous decrease in impact energy up to 20000 h,
compared to as-weld condition. For ageing at 643 K decrease in impact energy occurred up to 10000 h
followed by increase at 20000 h. Ageing at 823 K resulted in a moderate decrease of charpy energy up to
5000 h.

The J values were computed following ASTM E 1820 using the procedure described in Section
A2.4.2.2. For the initial crack blunting line slope, ASTM E1820 (Section A6) specifies a minimum of
M = J ∆a ⋅ σ f = 2 . A typical J-R curve is shown in Fig.2a for the as-weld condition. Values of J0.2
determined as the intersection of blunting line offset by 0.2 mm with the J-R curves are presented in
Fig.2b. The J0.2 value in as-weld condition was 207 kJ.mm-2. Ageing at 748 K resulted in increase of J0.2
values up to 1000 h, beyond which decrease of J0.2 values were observed. Ageing at 823 K also showed
increase in J0.2 up to 1000 h ageing, followed by decrease in J0.2 values. Ageing at 643 K resulted in
increase of J0.2 values up to 1000 h, followed by decrease of J0.2 values up to 5000 h. Also, compared to
as-weld condition, an overall decrease of J0.2 values were observed for all the ageing temperatures at
20000 h. The optical microstructure in the as-weld condition consisted of austenite matrix dispersed with
delta ferrite. No significant changes in microstructure could be observed at this magnification in all the
ageing conditions. The fractographic examination of the specimens tested in quasi-static loading
conditions revealed that the fracture surface features were dominated by ductile dimples indicating that in
all conditions the fracture is controlled by void nucleation and growth. The fractograph presented in Fig.3
2728 B. Shashank Dutt et al. / Procedia Engineering 10 (2011) 2725–2730
4 Shashank Dutt et al/ Procedia Engineering 00 (2011) 000–000

(a) is for the as-welded condition showing dimples of nearly uniform size. No significant differences in
the morphology of dimples could be observed for ageing durations up to 10000 h. The fractograph of
specimen tested after ageing at 748 K for 20000 h in B5 condition (Fig.3b) showed dimple sizes relatively
smaller compared to those for as-welded condition, indicating a distribution of finer and/or more closely
spaced sites for void nucleation. Also, some microcracks were observed.
850 b SS 316(N) Weld
a YS, UTS 643 K 150
800 Tested at 295 K
748 K
750 Ageing Temperature
823 K
as weld As-welded
700

Charpy Energy, J
UTS 643 K
650
Strength, MPa

100 748 K
600 823 K
550
500 YS
450 50
400
350
300 0.0 3 4 4 4
0 2000 4000 6000 8000 10000 5.0x10 1.0x10 1.5x10 2.0x10
Ageing time, h Ageing Time, h

Fig.1. (a) Tensile properties of 316 (N) welds; (b) Comparison of charpy energy for SS 316 (N) weld subjected to
ageing at 643 748 and 823 K
550
a b as-weld
1000 500 Ageing at 643 K
as-weld
450 Ageing at 748 K
400 Ageing at 823 K
800
350
-2
-2

J0.2, kJ.m

300
J, kJ.m

600
250
200
400
150
100
200
50
0
0 0 5000 10000 15000 20000
0.0 0.5 1.0 1.5 2.0 2.5 3.0
∆a, mm Ageing time, h

Fig.2. (a) J-R curve of 316 (N) welds; (b) J0.2 comparison of 316 (N) welds at different ageing conditions

Fig. 3. (a) Fractographs of J- tested 316 (N) weld; (b) subjected to ageing at 748 K for 20000 h
 B. Shashank Dutt et al. / Procedia Engineering 10 (2011) 2725–2730 2729
Shashank Dutt et al/ Procedia Engineering 00 (2011) 000–000 5

Vitek et al. [11] reported increase in ductile-brittle transition temperature (DBTT), drop charpy-impact
energy in the upper shelf and fracture toughness(J1c) of SS 308 welds after ageing at 748 K for durations
up to 20000 h. The average ferrite number [11] in the as-weld microstructure reported was ~ 11. They
also observed that the changes in tensile properties with ageing were not significant. The decrease in
toughness after aging was attributed to spinodal decomposition of ferrite into iron rich () and chromium
rich ( ) phases. Also the formation of silicon-rich G phase within ferrite and carbide (M23C6) along
ferrite-austenite interfaces was observed. The decrease in toughness was explained [11] by the hardening
of ferrite due to G-phase and -  formation resulting in non-homogeneous strain during testing. In the
present investigation the formation of α′ by spinodal decomposition was observed after 1000 h ageing at
748 K as shown in Fig. 4 (a). TEM analysis revealed regions of mottled contrast due to spinodal
decomposition within ferrite. These mechanisms were also observed after 20000 h ageing at 748 K
condition as shown in Fig. 4(b). Also, features indicating strain localisation is evident in the specimens in
this study too (Fig.4c).

O’Donnel [19] reported the effects of long term ageing in both MMA and GTA welds of SS316 with
5-7 Vol.% ferrite. They observed precipitation of various phases after ageing at 673-823 K for 50000 h.
However, these potentially detrimental microstructural changes did not cause any significant reduction of
fracture toughness. This was explained by the lower ferrite contents, due to which the intermetallics
formed after ageing did not cause significant decrease in fracture toughness. The ferrite content in the
present work was similar to that reported by O’Donnel [19]. These differences may be explained based
on the combination of welding procedure and the resulting inclusion content, the delta-ferrite content and
the microstructural changes after aging. GTA welds generally are cleaner with lower inclusion contents,
whereas SMA welds which are used for higher productivity have higher inclusion content. The effect of
aging on the decrease in toughness was not predominant in SMA welds, since it was the inclusion content
that primarily controlled the toughness, and ageing is not expected to exert any influence. For clean welds,
as in the case of GTA, ageing-induced microstructural changes are expected to influence the toughness.
The fact that there is no reduction in the fracture toughness (both J1c and tearing modulus) of SS 308
GTA and SMA welds after ageing for 10000 h at 700 K [20] was attributed to sluggish kinetics of
spinodal decomposition and G phase precipitation at 673-723 K, leading to lowered embrittlement rates.

a b c

Fig. 4. (a) TEM image of SS 316 (N) weld subjected to ageing at 748 K for 1000 h; (b-c) 20000 h
2730 B. Shashank Dutt et al. / Procedia Engineering 10 (2011) 2725–2730
6 Shashank Dutt et al/ Procedia Engineering 00 (2011) 000–000

From the above comparisons, it may be inferred that in the present investigation ageing to 20000 h for
all the ageing temperatures resulted a decrease in toughness (J0.2 ), despite lower specified ferrite contents
in the as-weld microstructure. Further electron microscopic investigations to ascertain the effect of ageing
on the microstructure (for all the ageing conditions) and their influence on fracture properties are being
conducted.

Conclusions

• A general decrease in resistance to crack initiation (J0.2 ) of SS 316 (N) welds was observed on
ageing at 643, 748 and 823 K for 20000 h. Decrease of charpy energy values subjected to ageing at
748 K for 20000 h. Changes in tensile properties was marginal.
• This decrease in mechanical properties, despite lower specified ferrite content in the weld
microstructure, indicates the effect of embrittlement at 643-823 K can be significant after longer
durations.
• The embrittlement is due to the formation of new phases in δ ferrite regions. α-α′ spinodal
decomposition is predominant at lower temperatures (643 and 748 K) causing strain inhomogeneities
leading to reduction in toughness, while σ phase formation is important at 823 K.

Acknowledgements

The authors acknowledge the contributions of Mr. Syed Meer Kaleem, Mr. R. Balakrishnan and Mr. S.
Sathyanarayanan of IGCAR Kalpakkam, India, for experimental support.

References
[1] Lundin CD, Chou CPD. Hot cracking susceptibility of austenitic stainless steel weld metals. WRC Bull 1983; 289: 1–80.
[2] Lewellyn DT. Steels: Metallurgy and Application. Butterworth-Heinemann, Oxford; 1992.
[3] Lin YP, Finlan GT, Steeds JW. Welding Metallurgy of Structural Steels, Metallurgical Society of AIME, Warrrandale; 1987,
p.531-541.
[4] Keown SR, Thomas RG. Metal Sci 1981; 386: 15.
[5] Lai JKL, Haigh JR. Weld.J.(Research supplement) 1979; p.1s-6s.
[6] Sasikala G , Mathew MD, Bhanu Sankara Rao K, Mannan SL. J. Nucl. Mater 1999; 273: 257-264.
[7] Tavassoli AA, Bisson A, Soulat P. Metals.Sci 1984; 18: 346.
[8] Auger P, Danoix F, Guttmann M. Mater.Sci. Technol 1990; 6: 301–313.
[9] Vrinat M, Cozar P, Meyzaud Y. Scripta. Metall 1986; 20: 1101–1106.
[10] Sahu JK, Krupp U, Ghosh RN, Christ HJ. Materials.Sc.Eng A 2009;508: 1-14.
[11] Vitek JM, David SA, Alexander D J, Keiser J R, Nanstad R K. Acta.Metal.Matter 1991; 39: 503-516.
[12] Mills WJ. International Materials Reviews 1997; 42: 45-81.
[13] Timofeev BT, Yu.K. Nikolaev. Int.J.Pres.Ves & Piping 1999; 76: 849–856.
[14] ASTM Standard Test Method for Tension Testing of Metallic Materials, ASTM E8-04.
[15] ASTM Standard Test Method for Linear-Elastic Plane-Strain Toughness KIC of Metallic Materials, ASTM E399-06.
[16] ASTM Standard Test Method for Measurement of Fracture Toughness, ASTME 1820-99a11.
[17] Chopra OK, Chung HM. Environtal Degradation of Materials in Nuclear Power Systems-Water reactors, In: Theus GJ,
Weeks JR,editors. TMS-AIME, Warrandale, Pa; 1988, p.359-370.
[18] Alexander DJ, Alexander KB, Nanstad RK. Recent Trends in Welding Science and Technology, In: David SA,
Vitek J.M, editors. ASM-Int, Mater.Park, Ohio; 1993, p.217-221.
[19] O’Donnell IJ, Huthmann H, Tavassoli AA. Int.J.Pres.Ves & Piping 1996; 65: 209-220.
[20] Mills WJ. J.Pressure Vessel.Technol (Trans ASME) 1987; 109: 440-448.