- Email: [email protected]
A Study of Fracture Mechanisms in RAFM Steel in the Ductile to Brittle Transition Temperature Regime
Available online at www.sciencedirect.com
ScienceDirect Procedia Engineering 86 (2014) 258 – 263
1st International Conference on Structural Integrity, ICONS-2014
A Study of Fracture Mechanisms in RAFM Steel in the Ductile to Brittle Transition Temperature Regime A. Moitraa,*, Arup Dasguptaa, S. Sathyanarayanana, G. Sasikalaa, S. K. Alberta, S. Sarojaa, A. K. Bhaduria, E. Rajendra Kumarb and T. Jayakumara a
Metallurgy and Materials Group, Indira Gandhi Centre for Atomic Research, Kalpakkam-603102, India b Institute for Plasma Research, Bhat, Gandhinagar-382 428, Gujarat, India * E-mail ID: [email protected]
Abstract The fracture behaviour of a Reduced Activation Ferritic Martensitic Steel (RAFM) has been studied within the Ductile to Brittle Transition Temperature (DBTT) regime. The DBTT has been determined by ASTM E 1921 based reference temperature approach under dynamic loading condition. The dynamic reference temperature (T0dy) was found to be − 33.8 °C. The fracture mechanism has been studied by extracting TEM specimens precisely at the crack initiation sites using focused ion beam (FIB) technique in a high resolution dual beam scanning electron microscope. Detailed analytical TEM studies revealed that the morphology of carbides play a crucial role in the initiation of a crack. The larger ellipsoidal carbides, which were found to be Crrich, have been found to be responsible for dislocation piles ups. The shorter edge of these ellipsoidal carbides are areas of high stress concentration and were found to initiate cracks by decohesion of the particle-matrix interface. On the contrary, the iron rich carbides have been found to be smaller, more spherical, and thus less effective in blocking dislocation movement and therefore formation of pile ups. The results, which reveal an important mechanism towards crack initiation in ferritic-martensitic steels, will be presented in detail. © © 2014 2014 The The Authors. Authors. Published Published by by Elsevier Elsevier Ltd. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research. Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research Keywords: RAFM Steel, DBTT, TEM, Fracture Mechanics
1. Introduction Ductile to brittle transition temperature (DBTT) is an important design parameter to ensure the structural integrity of the first wall of the test blanket module of fusion reactor, which must retain adequate mechanical properties under intense neutron irradiation (~14.1 MeV energy) and high thermo-mechanical loads during reactor operation [1-3]. For this purpose, the Reduced Activation Ferritic Martensitic (RAFM) steel with alloying elements of W, Ta and V is the chosen material in order to overcome residual radioactivity, generally encountered in conventional ferritic-martensitic steels originating from the long-lived transmutation nuclides of Mo, Nb, Ni, N, B,
1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research doi:10.1016/j.proeng.2014.11.036
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
Cu, Co, Ti etc [4-6]. For evaluating DBTT under quasi-static loading situations, the reference temperature (T0) approach, as described in ASTM E 1921-10 [7], is a well-recognised approach. This concept has been further analysed by Odette et al. [8] in steels for fusion reactor application. Considering that the effect of dynamic loading would give rise to conservative DBTT, the reference temperature was evaluated under dynamic loading condition as T0dy by Moitra et al. [9-12] for 9Cr-1Mo family of steels. Further, fracture in the transition temperature regime is a crack initiation controlled event, as most of the energy to fracture would be associated with the crack initiation process rather than its propagation. Though various models of crack initiation are proposed, for the precipitation hardening material like RAFM steel the major emphasis lies in the dislocation-precipitate interaction, often leading to initiation of a critical crack either by cracking of the precipitate itself or by the decohesion of the precipitatematrix interface depending on the morphology, size and shape the precipitate. Though conclusions may be derived based on models, it is of technological importance to experimentally validate the exact nature of the crack initiation mechanism operating in RAFM steels towards a better material design ensuring structural integrity. The conventional methodology of scanning electron microscopy (SEM) or the transmission electron microscopy (TEM) analysis is inadequate to address this issue as the former lacks in required resolution and the latter often misses the exact location of the cracking event leading to brittle fracture of the steel. A more appropriate method is to study the subsurface dislocation mechanisms at the exact location of crack initiation in a TEM specimen extracted using the state of the art Focused Ion Beam (FIB) technique. In the present study, while the DBTT of RAFM steel has been determined by T0dy approach, the precipitate-dislocation interaction in the determined DBTT regime has been investigated using the detailed TEM studies on specimens extracted at the crack initiation point by FIB. The results, which reveal an important mechanism towards crack initiation in ferritic-martensitic steels, have been discussed. 2. Materials The steel has been received from MIDHANI, India as a 12 mm rolled plate with chemical composition given as: Cr-9.15, C-0.08, Mn-0.53, V-0.24, W-1.37, Ta-0.08, N-0.02, O-0.002, Ni - 0.004, S - 0.002, Co-0.003, Al0.004, B < 0.001, Si <0.026, Ti-<0.002, Nb <0.001, Mo < 0.002, Cu < 0.002, P < 0.002, Fe-Balnace. 3. Determination of T0dy T0dy of this RAFM steel has been determined following ASTM E 1921 approach using dynamic fracture toughness (KJd) results obtained from pre-cracked Charpy (PCVN) specimens of nominal dimension of 10x10x55 mm. The loading rate in this study was optimized at 1.15 m/s to achieve a dynamic stress intensity factor rate in the range of 105 to 106 MPa.m0.5/s, while simultaneously reducing the oscillations in the load-displacement plots to ensure negligible error in load determination. The test temperature was selected to be −50 °C by a trial and error method to achieve an elastic-plastic fracture, i.e. a strain hardening regime before the brittle crack propagation, identified by a sharp drop of load. A typical load-displacement plot obtained from these tests, marked with strain hardening region, brittle crack initiation load and crack arrest load is shown in Fig.1. 7
RAFM Steel
Brittle Fracture Load
o Test Temp. -50 C, (Spe.No.B2)
6
Load, kN
5 4 3 Crack Arrest Load
2 1 0 0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
Displacement, mm
Fig.1 A typical load-displacement plot obtained from PCVN tests at -50 °C
259
260
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
The three parameter Weibull statistical model which is the basis of the Reference Temperature (T0) approach is given in Eqn. [1].
p f = 1 − exp{−[( K Jc − K min ) /( K 0 − K min )]b }
[1]
where pf is the cumulative probability of failure and KJc is the fracture toughness at quasi-static loading rate. The ‘three parameters’ in this equation are a) K0 - the scale parameter or characteristic value, b) Kmin = 20 MPa.m0.5 as the minimum toughness and c) ‘b’- the Weibull slope, also referred to as the shape parameter. The shape parameter ‘b’ is assigned a value of 4 for all ferritic steels. Following eqn.1, the dynamic fracture toughness results (KJd) obtained at 10 mm thickness has been converted to 1 inch thick equivalent as:
§B · K Jd (2) = K min + [ K Jd (1) − K min ] ¨ 1 ¸ © B2 ¹
1 4
[2]
where the subscript 2 refers to specimens of 25.4 mm (1 inch) thickness and the subscript 1 refers to the thickness of tested specimen (10 mm). The scale parameter (K0) is determined as: N
K 0 = [¦ ( K Jd (i ) − K min )4 /( N − 0.3068)]1/ 4 + K min
[3]
i =1
Kmin is 20 MPa.m0.5 and N is the number of tests included in calculation. Median KJd (KJd(med)) is the fracture toughness value having pf = 0.5 for the dataset determined at 1 inch thickness equivalence, and has been determined as: 1
K Jd ( med ) = ( K 0 − K min )[ln(2)]4 + K min
[4] 0.5
dy
The T0 , defined as the temperature at which KJd(med) is 100 MPa.m for 1 inch thick specimen, has been determined as:
T0 = T −
K − 30 1 ln[ Jd ( med ) ] 0.019 70
[5]
The T0dy thus determined is given in Table1.
Table 1. The KJd and the T0dy result for RAFM steel Spec. No. B1 B2 B3 B5 B6 B8 B10
Ttest, °C ņ 50 ņ 50 ņ 50 ņ 50 ņ 50 ņ 50 ņ 50
1” Corr. KJd, MPa.m0.5 51.6 125.1 47.2 70.3 58.4 71.5 125.1
KJd,(med), MPa.m0.5
T0dy °C
81.4
-33.8
4. Micro-Mechanism of Fracture 4.1 Extraction of Specimen using FIB Focused ion beam (FIB) technique towards extracting a TEM specimen was used in a FEI make Helios 600i high resolution dual beam scanning electron microscope. Specimen no. B5, exhibiting reasonable plasticity before brittle crack initiation was taken up for SEM-FIB operation. As a first step, the brittle crack initiation region just after the fatigue-precracked front was identified by tracing back the river pattern at lower magnification. Then
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
the observed region was further magnified to pin point the region of interest, as shown in Fig.2a. Then the region of interest was marked by platinum deposition, as shown in Fig.2b. The Pt layer also served to protect underlying fracture surface from Ga beam damage during milling operations. The initial foil for TEM sample preparation was extracted normal to the Pt-deposited line using a micromanipulator. The foil was then mounted onto a grid and insitu Ga+ milled up to ~50 nm thickness, suitable for observation in TEM.
Fig.2a Selected Region of Crack Initiation
Fig.2b Pt-deposited region for FIB operation
4.2 TEM Results From the extracted TEM specimen, the subsurface morphology has been examined in detail in a Philips make CM200 analytical TEM fitted with Oxford X-Max 80 SDD detector. Salient results of the TEM observation are given in the Fig. 3a to 3c.
Fig.3a Crack initiating points, as marked by arrows
Fig.3b Dislocation network around particles
Fig.3c Crack initiation at a particle, as marked by an arrow
Figure 3a shows the distribution of particles and dislocation network just below the fracture surface. It can be clearly observed that the cracking process has been initiated at the particle–matrix interface. No evidence has been obtained for any particle to have fractured itself in this region. Figure 3b shows that the dislocation network has been denser around a large ellipsoidal shaped particle as compared to the round shaped smaller particle. A further magnified image in Fig.3c clearly shows the debonding of the particle-matrix interface, as marked by an arrow. Generally two types of carbide morphologies have been observed in the TEM study, namely the globular particles with average size of ~ 100 nm and the ellipsoidal precipitates with average size ~200 nm. To generate an idea about the chemical composition of precipitates, energy dispersive X-ray spectroscopy (EDXS) was carried out with suitable tilt angles in the TEM from the particle as well as from the surrounding matrix region. Analysis of the EDXS patterns, show that the larger precipitates were rich in Cr as compared to the smaller ones, indicating that the particles were different types of carbide precipitates. The EDXS spectra obtained from a
261
262
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
large particle and that from the matrix is compared in Fig.4.
Fig.4 EDXS results from a large precipitate and matrix. 5. Discussion The concept of evaluation of ductile to brittle transition temperature via reference temperature approach can be summarized as a macroscopic engineering conclusion derived from local events in microstructural scale. The significant premises on which the concept is based are i) it appreciates the application of fracture mechanics concepts in terms of elastic-plastic assessment of the fracture toughness of the material under investigation, ii) it takes into account the inherent scatter in fracture toughness originating from the variation of the energy involved in the crack initiation process owing to the difference in size, shape and orientation of the randomly distributed carbides at the process zone. Further the material’s cleavage fracture stress, matrix strain and the level of triaxiality would ultimately determine the crack propagation process. On loading, at the region just ahead of the pre-existing fatigue crack of the specimen, the overall triaxiality situation limits the flow stress leading to a stress build up, often reaching the required cleavage fracture stress of the material. At this highly stressed region, commonly referred as the process zone, the initiation of the crack can occur at the carbides, either by cracking of the particle itself or by decohesion of the particle-matrix interface, depending on their relative strength vis a vis the stress build up. In the process of slip band and precipitate interaction behaviour the former process is expected when the carbides are of very large size or of easily brittle in nature. For the finer carbides, as observed in this steel, the idea of carbide cracking can be precluded and thus the latter process of crack initiation can be considered to be a preferred one. This concept has been vindicated by the TEM results, clearly showing that the cracks are initiating by particle matrix decohesion. The size and shape of the particles are expected to play a key role in the crack initiation process. As discussed earlier, in the TEM study, grossly two types of carbide morphologies have been observed. One is of round shaped with size ~ 100 nm with negligible amount of Cr and the other is of ellipsoidal type with size ~200 nm with significant amount of Cr. The dislocation networking has been found to be much less in the vicinity of round shaped particles in contrary to the ellipsoidal particles (Fig.3b). This implies that where the smaller and spherical carbides are providing less effective barriers for dislocations to prevent pile ups, the larger ellipsoidal carbides are subjected to high stress build up at the interface. It may be noted that the stress build up at the end of dislocation pile ups is proportional to the number of dislocations and the length of slip bands [13] . This stress can exceed the particle-
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
matrix bond strength and can lead to the debonding of their interface, initiating cleavage cracks. The Fig 3a-3c clearly establishes only the Cr-rich larger carbides are responsible for crack initiation in the present circumstances.
6. Conclusions 1. 2. 3.
The dynamic reference temperature (T0dy) has been determined to be − 33.8 °C. The crack initiation process in this steel is carbide-matrix decohesion. The Cr-rich ellipsoidal particles are of larger size and responsible for crack initiation process.
Acknowledgement The authors wish to express their gratitude to the UGC-DAE-CSR node at Kalpakkam for use of their experimental facilities. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Kohyama A., Kohno Y., Asakura K. and Kayano H., J. Nucl Mater., 212- 215 (1994) 684-689. Harrelson K. J., Rou S. H. and Wilcox R. C., J. Nucl. Mater., 141-143 (1986) 508-512. Baluc N., Schaublin R., Spatig P. and Victoria M., Nucl. Fis., 44 (2004) 56-61. Butterworth G.L. , J. Nucl. Mater., 179-181 (1991) 135. Cook I., Ward D. and Dudarev S., Plasma phys. Control. Fusion, 44 (2002) B121-B136. Laha K., Saroja S., Moitra A., Sandhya R., Mathew M.D., Jayakumar T., Rajendra Kumar E., J. Nu. Mat., 439 (2013) 41-50. ASTM E 1921-10 ‘”Test Method for Determination of Reference Temperature, To, for Ferritic Steels in the Transition Range”, ASTM, 2010. Odette G.R., Yamamoto T., Rathbun H.J., He M.Y., Hribernik M.L., Rensman J.W., J.Nu. Mat., 323 (2003) 313-340. Moitra A., Sreenivasan P. R., Mannan S. L. and Singh V., Mat. & Met. Trans A, 36 A, (2005), 2957 Moitra A., Sathyanarayanan S., Alber, S.K., Ramasubbu V., Samuel K.G., Ray S.K., Int. J. Pres. Piping, 85 (2008) 478-485, Sathyanarayanan S., Moitra A., Samuel K.G., Sasikala G., Ray S.K., Singh V., Mater. Sci. & Engg-A, 488 (2008) 519-528. Sathyanarayanan S., Basu J, Moitra A., Sasikala G. and Singh V., Mat. & Met. Trans A, 44A (2013) 21412155. Dieter, G.E., Mechanical Metallurgy, McGraw-Hill Book Company, Singapore, (1987) p.257.
263
ScienceDirect Procedia Engineering 86 (2014) 258 – 263
1st International Conference on Structural Integrity, ICONS-2014
A Study of Fracture Mechanisms in RAFM Steel in the Ductile to Brittle Transition Temperature Regime A. Moitraa,*, Arup Dasguptaa, S. Sathyanarayanana, G. Sasikalaa, S. K. Alberta, S. Sarojaa, A. K. Bhaduria, E. Rajendra Kumarb and T. Jayakumara a
Metallurgy and Materials Group, Indira Gandhi Centre for Atomic Research, Kalpakkam-603102, India b Institute for Plasma Research, Bhat, Gandhinagar-382 428, Gujarat, India * E-mail ID: [email protected]
Abstract The fracture behaviour of a Reduced Activation Ferritic Martensitic Steel (RAFM) has been studied within the Ductile to Brittle Transition Temperature (DBTT) regime. The DBTT has been determined by ASTM E 1921 based reference temperature approach under dynamic loading condition. The dynamic reference temperature (T0dy) was found to be − 33.8 °C. The fracture mechanism has been studied by extracting TEM specimens precisely at the crack initiation sites using focused ion beam (FIB) technique in a high resolution dual beam scanning electron microscope. Detailed analytical TEM studies revealed that the morphology of carbides play a crucial role in the initiation of a crack. The larger ellipsoidal carbides, which were found to be Crrich, have been found to be responsible for dislocation piles ups. The shorter edge of these ellipsoidal carbides are areas of high stress concentration and were found to initiate cracks by decohesion of the particle-matrix interface. On the contrary, the iron rich carbides have been found to be smaller, more spherical, and thus less effective in blocking dislocation movement and therefore formation of pile ups. The results, which reveal an important mechanism towards crack initiation in ferritic-martensitic steels, will be presented in detail. © © 2014 2014 The The Authors. Authors. Published Published by by Elsevier Elsevier Ltd. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research. Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research Keywords: RAFM Steel, DBTT, TEM, Fracture Mechanics
1. Introduction Ductile to brittle transition temperature (DBTT) is an important design parameter to ensure the structural integrity of the first wall of the test blanket module of fusion reactor, which must retain adequate mechanical properties under intense neutron irradiation (~14.1 MeV energy) and high thermo-mechanical loads during reactor operation [1-3]. For this purpose, the Reduced Activation Ferritic Martensitic (RAFM) steel with alloying elements of W, Ta and V is the chosen material in order to overcome residual radioactivity, generally encountered in conventional ferritic-martensitic steels originating from the long-lived transmutation nuclides of Mo, Nb, Ni, N, B,
1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research doi:10.1016/j.proeng.2014.11.036
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
Cu, Co, Ti etc [4-6]. For evaluating DBTT under quasi-static loading situations, the reference temperature (T0) approach, as described in ASTM E 1921-10 [7], is a well-recognised approach. This concept has been further analysed by Odette et al. [8] in steels for fusion reactor application. Considering that the effect of dynamic loading would give rise to conservative DBTT, the reference temperature was evaluated under dynamic loading condition as T0dy by Moitra et al. [9-12] for 9Cr-1Mo family of steels. Further, fracture in the transition temperature regime is a crack initiation controlled event, as most of the energy to fracture would be associated with the crack initiation process rather than its propagation. Though various models of crack initiation are proposed, for the precipitation hardening material like RAFM steel the major emphasis lies in the dislocation-precipitate interaction, often leading to initiation of a critical crack either by cracking of the precipitate itself or by the decohesion of the precipitatematrix interface depending on the morphology, size and shape the precipitate. Though conclusions may be derived based on models, it is of technological importance to experimentally validate the exact nature of the crack initiation mechanism operating in RAFM steels towards a better material design ensuring structural integrity. The conventional methodology of scanning electron microscopy (SEM) or the transmission electron microscopy (TEM) analysis is inadequate to address this issue as the former lacks in required resolution and the latter often misses the exact location of the cracking event leading to brittle fracture of the steel. A more appropriate method is to study the subsurface dislocation mechanisms at the exact location of crack initiation in a TEM specimen extracted using the state of the art Focused Ion Beam (FIB) technique. In the present study, while the DBTT of RAFM steel has been determined by T0dy approach, the precipitate-dislocation interaction in the determined DBTT regime has been investigated using the detailed TEM studies on specimens extracted at the crack initiation point by FIB. The results, which reveal an important mechanism towards crack initiation in ferritic-martensitic steels, have been discussed. 2. Materials The steel has been received from MIDHANI, India as a 12 mm rolled plate with chemical composition given as: Cr-9.15, C-0.08, Mn-0.53, V-0.24, W-1.37, Ta-0.08, N-0.02, O-0.002, Ni - 0.004, S - 0.002, Co-0.003, Al0.004, B < 0.001, Si <0.026, Ti-<0.002, Nb <0.001, Mo < 0.002, Cu < 0.002, P < 0.002, Fe-Balnace. 3. Determination of T0dy T0dy of this RAFM steel has been determined following ASTM E 1921 approach using dynamic fracture toughness (KJd) results obtained from pre-cracked Charpy (PCVN) specimens of nominal dimension of 10x10x55 mm. The loading rate in this study was optimized at 1.15 m/s to achieve a dynamic stress intensity factor rate in the range of 105 to 106 MPa.m0.5/s, while simultaneously reducing the oscillations in the load-displacement plots to ensure negligible error in load determination. The test temperature was selected to be −50 °C by a trial and error method to achieve an elastic-plastic fracture, i.e. a strain hardening regime before the brittle crack propagation, identified by a sharp drop of load. A typical load-displacement plot obtained from these tests, marked with strain hardening region, brittle crack initiation load and crack arrest load is shown in Fig.1. 7
RAFM Steel
Brittle Fracture Load
o Test Temp. -50 C, (Spe.No.B2)
6
Load, kN
5 4 3 Crack Arrest Load
2 1 0 0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
Displacement, mm
Fig.1 A typical load-displacement plot obtained from PCVN tests at -50 °C
259
260
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
The three parameter Weibull statistical model which is the basis of the Reference Temperature (T0) approach is given in Eqn. [1].
p f = 1 − exp{−[( K Jc − K min ) /( K 0 − K min )]b }
[1]
where pf is the cumulative probability of failure and KJc is the fracture toughness at quasi-static loading rate. The ‘three parameters’ in this equation are a) K0 - the scale parameter or characteristic value, b) Kmin = 20 MPa.m0.5 as the minimum toughness and c) ‘b’- the Weibull slope, also referred to as the shape parameter. The shape parameter ‘b’ is assigned a value of 4 for all ferritic steels. Following eqn.1, the dynamic fracture toughness results (KJd) obtained at 10 mm thickness has been converted to 1 inch thick equivalent as:
§B · K Jd (2) = K min + [ K Jd (1) − K min ] ¨ 1 ¸ © B2 ¹
1 4
[2]
where the subscript 2 refers to specimens of 25.4 mm (1 inch) thickness and the subscript 1 refers to the thickness of tested specimen (10 mm). The scale parameter (K0) is determined as: N
K 0 = [¦ ( K Jd (i ) − K min )4 /( N − 0.3068)]1/ 4 + K min
[3]
i =1
Kmin is 20 MPa.m0.5 and N is the number of tests included in calculation. Median KJd (KJd(med)) is the fracture toughness value having pf = 0.5 for the dataset determined at 1 inch thickness equivalence, and has been determined as: 1
K Jd ( med ) = ( K 0 − K min )[ln(2)]4 + K min
[4] 0.5
dy
The T0 , defined as the temperature at which KJd(med) is 100 MPa.m for 1 inch thick specimen, has been determined as:
T0 = T −
K − 30 1 ln[ Jd ( med ) ] 0.019 70
[5]
The T0dy thus determined is given in Table1.
Table 1. The KJd and the T0dy result for RAFM steel Spec. No. B1 B2 B3 B5 B6 B8 B10
Ttest, °C ņ 50 ņ 50 ņ 50 ņ 50 ņ 50 ņ 50 ņ 50
1” Corr. KJd, MPa.m0.5 51.6 125.1 47.2 70.3 58.4 71.5 125.1
KJd,(med), MPa.m0.5
T0dy °C
81.4
-33.8
4. Micro-Mechanism of Fracture 4.1 Extraction of Specimen using FIB Focused ion beam (FIB) technique towards extracting a TEM specimen was used in a FEI make Helios 600i high resolution dual beam scanning electron microscope. Specimen no. B5, exhibiting reasonable plasticity before brittle crack initiation was taken up for SEM-FIB operation. As a first step, the brittle crack initiation region just after the fatigue-precracked front was identified by tracing back the river pattern at lower magnification. Then
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
the observed region was further magnified to pin point the region of interest, as shown in Fig.2a. Then the region of interest was marked by platinum deposition, as shown in Fig.2b. The Pt layer also served to protect underlying fracture surface from Ga beam damage during milling operations. The initial foil for TEM sample preparation was extracted normal to the Pt-deposited line using a micromanipulator. The foil was then mounted onto a grid and insitu Ga+ milled up to ~50 nm thickness, suitable for observation in TEM.
Fig.2a Selected Region of Crack Initiation
Fig.2b Pt-deposited region for FIB operation
4.2 TEM Results From the extracted TEM specimen, the subsurface morphology has been examined in detail in a Philips make CM200 analytical TEM fitted with Oxford X-Max 80 SDD detector. Salient results of the TEM observation are given in the Fig. 3a to 3c.
Fig.3a Crack initiating points, as marked by arrows
Fig.3b Dislocation network around particles
Fig.3c Crack initiation at a particle, as marked by an arrow
Figure 3a shows the distribution of particles and dislocation network just below the fracture surface. It can be clearly observed that the cracking process has been initiated at the particle–matrix interface. No evidence has been obtained for any particle to have fractured itself in this region. Figure 3b shows that the dislocation network has been denser around a large ellipsoidal shaped particle as compared to the round shaped smaller particle. A further magnified image in Fig.3c clearly shows the debonding of the particle-matrix interface, as marked by an arrow. Generally two types of carbide morphologies have been observed in the TEM study, namely the globular particles with average size of ~ 100 nm and the ellipsoidal precipitates with average size ~200 nm. To generate an idea about the chemical composition of precipitates, energy dispersive X-ray spectroscopy (EDXS) was carried out with suitable tilt angles in the TEM from the particle as well as from the surrounding matrix region. Analysis of the EDXS patterns, show that the larger precipitates were rich in Cr as compared to the smaller ones, indicating that the particles were different types of carbide precipitates. The EDXS spectra obtained from a
261
262
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
large particle and that from the matrix is compared in Fig.4.
Fig.4 EDXS results from a large precipitate and matrix. 5. Discussion The concept of evaluation of ductile to brittle transition temperature via reference temperature approach can be summarized as a macroscopic engineering conclusion derived from local events in microstructural scale. The significant premises on which the concept is based are i) it appreciates the application of fracture mechanics concepts in terms of elastic-plastic assessment of the fracture toughness of the material under investigation, ii) it takes into account the inherent scatter in fracture toughness originating from the variation of the energy involved in the crack initiation process owing to the difference in size, shape and orientation of the randomly distributed carbides at the process zone. Further the material’s cleavage fracture stress, matrix strain and the level of triaxiality would ultimately determine the crack propagation process. On loading, at the region just ahead of the pre-existing fatigue crack of the specimen, the overall triaxiality situation limits the flow stress leading to a stress build up, often reaching the required cleavage fracture stress of the material. At this highly stressed region, commonly referred as the process zone, the initiation of the crack can occur at the carbides, either by cracking of the particle itself or by decohesion of the particle-matrix interface, depending on their relative strength vis a vis the stress build up. In the process of slip band and precipitate interaction behaviour the former process is expected when the carbides are of very large size or of easily brittle in nature. For the finer carbides, as observed in this steel, the idea of carbide cracking can be precluded and thus the latter process of crack initiation can be considered to be a preferred one. This concept has been vindicated by the TEM results, clearly showing that the cracks are initiating by particle matrix decohesion. The size and shape of the particles are expected to play a key role in the crack initiation process. As discussed earlier, in the TEM study, grossly two types of carbide morphologies have been observed. One is of round shaped with size ~ 100 nm with negligible amount of Cr and the other is of ellipsoidal type with size ~200 nm with significant amount of Cr. The dislocation networking has been found to be much less in the vicinity of round shaped particles in contrary to the ellipsoidal particles (Fig.3b). This implies that where the smaller and spherical carbides are providing less effective barriers for dislocations to prevent pile ups, the larger ellipsoidal carbides are subjected to high stress build up at the interface. It may be noted that the stress build up at the end of dislocation pile ups is proportional to the number of dislocations and the length of slip bands [13] . This stress can exceed the particle-
A. Moitra et al. / Procedia Engineering 86 (2014) 258 – 263
matrix bond strength and can lead to the debonding of their interface, initiating cleavage cracks. The Fig 3a-3c clearly establishes only the Cr-rich larger carbides are responsible for crack initiation in the present circumstances.
6. Conclusions 1. 2. 3.
The dynamic reference temperature (T0dy) has been determined to be − 33.8 °C. The crack initiation process in this steel is carbide-matrix decohesion. The Cr-rich ellipsoidal particles are of larger size and responsible for crack initiation process.
Acknowledgement The authors wish to express their gratitude to the UGC-DAE-CSR node at Kalpakkam for use of their experimental facilities. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Kohyama A., Kohno Y., Asakura K. and Kayano H., J. Nucl Mater., 212- 215 (1994) 684-689. Harrelson K. J., Rou S. H. and Wilcox R. C., J. Nucl. Mater., 141-143 (1986) 508-512. Baluc N., Schaublin R., Spatig P. and Victoria M., Nucl. Fis., 44 (2004) 56-61. Butterworth G.L. , J. Nucl. Mater., 179-181 (1991) 135. Cook I., Ward D. and Dudarev S., Plasma phys. Control. Fusion, 44 (2002) B121-B136. Laha K., Saroja S., Moitra A., Sandhya R., Mathew M.D., Jayakumar T., Rajendra Kumar E., J. Nu. Mat., 439 (2013) 41-50. ASTM E 1921-10 ‘”Test Method for Determination of Reference Temperature, To, for Ferritic Steels in the Transition Range”, ASTM, 2010. Odette G.R., Yamamoto T., Rathbun H.J., He M.Y., Hribernik M.L., Rensman J.W., J.Nu. Mat., 323 (2003) 313-340. Moitra A., Sreenivasan P. R., Mannan S. L. and Singh V., Mat. & Met. Trans A, 36 A, (2005), 2957 Moitra A., Sathyanarayanan S., Alber, S.K., Ramasubbu V., Samuel K.G., Ray S.K., Int. J. Pres. Piping, 85 (2008) 478-485, Sathyanarayanan S., Moitra A., Samuel K.G., Sasikala G., Ray S.K., Singh V., Mater. Sci. & Engg-A, 488 (2008) 519-528. Sathyanarayanan S., Basu J, Moitra A., Sasikala G. and Singh V., Mat. & Met. Trans A, 44A (2013) 21412155. Dieter, G.E., Mechanical Metallurgy, McGraw-Hill Book Company, Singapore, (1987) p.257.
263