1. Introduction
High chromium ferritic heat resistant steels have been widely used as material for steam generator of thermal and nuclear power plants for its superior high temperature creep strength and good thermal conductivity1). Welding is the major fabrication process for steam generator. The creep rupture strength of weld joints of these steels is substantially lower than that of the base metal due to Type IV cracking at the fine grain (FG) or inter critical (IC) heat affected zone (HAZ)2,3). Improvement in the creep rupture life of high chromium steels with controlled addition of boron and reduced nitrogen level has been recently attempted4-8). Tabuchi et al observed the enhanced creep rupture life for 9Cr-3W-3Co-Nd-B steel by addition of boron6). Beneficial role of boron in suppressing Type IV failure in modified 9Cr-1Mo (P91) steel has also been reported by researchers7,8). Development of P91 steel with different levels of boron and nitrogen is in progress to mitigate the Type IV cracking problem2,7,8). To carry out shielded metal arc (SMA) welding of boron containing P91 steel, AWS E9016-B9 electrode has been modified with boron addition. Assessment of weld quality for the newly developed electrode is important, especially, with respect to impact toughness of the weld metal; as toughness of ferritic steel SMA joint is often very low due to the inclusions produced by the flux. As per specification of steam generator of fast breeder reactor (FBR), hydro testing is mandatory at various stages and this necessitates adequate impact toughness for the weld joints at ambient temperature (45 J at 20°C) to avoid unexpected failures during hydro test9,10).
Tungsten inert gas (TIG) welding produces cleaner weld with excellent mechanical properties. However, weld penetration in single pass TIG is limited to 3 mm11). Activated TIG (A-TIG) welding can be a better choice to improve mechanical properties and productivity. A-TIG process, which incorporates a thin layer of surface-active flux, is able to produce higher weld penetration based on arc constriction and reversal of surface tension in the weld pool11). Successful application of the A-TIG process for welding of ferritic steel components of nuclear reactors can cater the needs for high productivity as well as stringent mechanical property requirement.
In the present research, the influence of two welding processes namely SMA and A-TIG on boron added P91 steel has been studied by detailed microstructural investigations as well as mechanical property assessment of the weld joint by tensile, charpy impact toughness, impression creep and fracture toughness test. The study not only evaluates the suitability of employing these welding processes for fabrication of steam generator of nuclear reactor but also compares the advantages and disadvantages of A-TIG process over multi-pass SMA welding process.
2. Experimental Procedure
P91 steel (C: 0.1, Cr: 8.5, Mo: 1.04, Mn: 0.3, Si: 0.4, V: 0.23, Nb: 0.09, B: 0.01, N: 0.002 wt.%) plates (normalized and tempered at 1050°C/1 hour, 760°C/3 hours respectively) of dimensions 250 mm × 250 mm × 12 mm were joined by SMA and A-TIG welding processes. Multi-pass “V” groove welding was performed by SMA process (Fig. 1a) using AWS E9016-B9 electrode (boron modified) (welding current: 80-110 Amp, voltage: 23-26 V, speed: 40-80 mm/min). The electrode contains alloying elements as per American Welding Society (AWS) specification (C: 0.08-0.13, Mn: 1.2 max, Si: 0.3 max, P: 0.01 max, S: 0.01 max, Ni: 0.8 max, Cr: 8-10.5, Mo: 0.85-1.2, V: 0.15-0.3, Nb: 0.02-0.1, Cu: 0.25 max, Al: 0.04 max) with controlled addition of boron and nitrogen. Double side square butt A-TIG weld joint (Fig. 1b) was prepared using argon and helium as shielding gases in 70:30 ratio (current: 230 Amp, voltage:15 V, speed: 60 mm/min, gas flow rate:10 lit/min). For the A-TIG weld joint, welding was done with two passes; one from the top side and other from the bottom side of the plate with 6.5 mm depth of penetration for each pass and 1 mm overlapping height at the interface of two passes. Defect free weld quality was assured by radiographic examination. SMA and A-TIG weld joints were subjected to post weld heat treatment (PWHT) at 760°C for duration of 3 and 4 hours respectively. All microstructural studies and mechanical property assessment of the weld joints were performed after PWHT. The weld joints were characterized by optical microscopy, microhardness measurement at 500 g load, tensile test and charpy impact test. ASTM A370 test procedure was followed for tensile (E8) and charpy impact toughness (E23) testing. Impression creep testing12) was conducted using 1 mm diameter tungsten carbide cylindrical punch at 600°C, 10-6 mbar vacuum with 125 MPa applied load to assess the creep resistance of the weld metal and HAZ. Fracture toughness at room temperature for the base metal as well as SMA and A-TIG weld metals was determined using compact tension (CT) specimens (having 6.2 mm thickness, 31 mm length and 30 mm width). J-R curves were generated as per ASTM 1820 standard. Before fracture testing, the CT specimens were pre-cracked under fatigue loading with a target initial crack length of ~3 mm. After pre-cracking, the CT specimens were side-grooved to 10% of initial thickness. The initial and final crack lengths of the fracture tested specimens were measured by optical microscopy to determine fracture toughness (J1c).
3. Results and Discussion
The objective of the present study is to assess the weld quality of boron containing P91 steel produced by SMA and A-TIG welding processes, as well as to explore the feasibility of employing these techniques for fabrication of steam generator of future nuclear reactors. Thus, it is important to ensure the required tensile strength and impact energy for the weld joints and the creep resistance of the HAZ. Retardation of the grain refinement and producing grain size similar to that of the base metal at FG and ICHAZ is the key to improve Type IV cracking resistance7). Microstructural studies of both the weld joints indicate retardation of grain refinement; more prominently in the A-TIG process as compared to the SMA weld joint (Fig. 2). Increase in prior austenite grain size in coarse grain (CG) HAZ (35 μm) of SMA weld joint is observed, whereas, grain sizes of FGHAZ and ICHAZ are comparable to that of the base metal (28 μm). The grain coarsening in CGHAZ is due to the repeated thermal cycles experienced by the multiple passes of SMA welding. The complex thermal gradient of SMA welding causes the variation in grain size in the different subzones of HAZ. In contrast, the increase in grain size is less predominant for CGHAZ of the A-TIG weld joint and grain sizes of three subzones of HAZ are almost similar (Fig. 2). It can be noted here that in A-TIG process, the oxides present in the flux dissociates during welding and captures the electron from the cooler peripheral regions of the arc towards the arc center. This results in increased current flow in the central region of the arc11). Hence, higher heat is concentrated in the central region causing higher weld penetration and base metal is less heated up in the HAZ region. Further, HAZ microstructure does not get modified by successive welding passes in case of A-TIG joint. This could be the reason for no significant variation in grain size in three subzones of A-TIG HAZ.
Fig. 2
Optical microstructure of (a) base metal and weld metal, CG, FG & ICHAZ of (b-e) SMA weld joint, (f-i) A-TIG weld joint
Complete tempering of the martensite matrix for both the weld metals is observed after PWHT (Fig. 2). The SMA weld metal also experiences tempering by the multiple welding passes; thereby requiring lesser duration for PWHT than the A-TIG weld joint. Presence of delta ferrite is observed in the fusion boundary of SMA weld joint (Fig. 1b), whereas for A-TIG weld joint delta ferrite is present close to the weld center at the interface of two overlapping passes (Fig. 1f). The volume fraction of delta ferrite is ~1% for both the weld metals. It is expected to observe more delta ferrite in SMA weld metal due to the presence of higher content of ferrite stabilizing (Cr, Mo) elements in the electrode. However, delta ferrite produced in a particular pass gets modified to austenite and subsequently to martensite by the successive passes of SMA welding. In A-TIG welding, the delta ferrite does not transform further. As delta ferrite has detrimental effects on ductile-brittle transition temperature (DBDT) and creep rupture strength of martensitic weld joints4,13), the effect of delta ferrite on mechanical properties of the SMA and A-TIG weld joints is carefully assessed and described in the subsequent sections.
Microhardness measurements across the weld joints (Fig. 3) indicate marginally higher hardness for SMA weld metal (~275 VHN) than the A-TIG (~250 VHN) weld metal in spite of repeated tempering of the former by multiple welding passes. This could be due to the oxidation loss of alloying elements (Cr, Mo) from A-TIG weld metal during welding (Table 1). The Cr and Mo contents of the SMA weld metal are observed to be higher than the base metal due to overmatching chemistry of the electrode (Table 1). However, hardness of FG and ICHAZ of SMA weld joint is lower than that of A-TIG weld joint (Fig. 3). Repeated thermal cycles experienced in the HAZ of SMA weld joint causes tempering of the HAZ microstructure, bringing down the hardness of the FG, ICHAZ even below that of the base metal. The lower hardness of FG and ICHAZ may have harmful effect on the Type IV cracking resistance of the SMA weld joint. For A-TIG weld joint, the decrement in hardness from weld metal to HAZ is gradual and hardness of ICHAZ is close to that of the base metal (Fig. 3).
Table 1
EDS analysis (wt%) for major alloying elements of SMA and A-TIG weld metal
| Cr | Mo | Mn | Si | Nb | V | |
|---|---|---|---|---|---|---|
| SMA weld metal | 10.2 | 1.1 | 0.4 | 1.2 | 0.5 | 0.4 |
| A-TIG weld metal | 7.9 | 0.9 | 0.4 | 1 | 0.3 | 0.5 |
Ultimate tensile strength (UTS) of transverse weld (cross-weld) specimens at room temperature and at 550°C, which is the intended service temperature of the steel, is 730 MPa and 520 MPa for SMA and 710 MPa and 465 MPa for A-TIG weld joint respectively (Table 2). Lower UTS for A-TIG weld joint could be attributed to the marginal oxidation loss of alloying elements during welding (Table 1). For all the transverse weld tensile tests, the location of failure is in the base metal, closer to the HAZ-base metal interface. The high dislocation density in the weld metal aided by the thermal and phase transformation stresses accounts for its high strength in both the weld joints14). The effect of delta ferrite is not manifested in the tensile test results due to lower volume fraction of it in the weld. Park et al. reported insignificant effect of delta ferrite on tensile strength below 8% volume fraction15).
Table 2
Mechanical properties of SMA and A-TIG weld joints of boron containing P91 steel
| Material | YS (MPa) | UTS (Mpa) | Impact Energy (J) | |||
|---|---|---|---|---|---|---|
| room temp | 550°C | room temp | 550°C | room temp | 0°C | |
| SMA weld joint | 615 | 550 | 730 | 710 | 53 | 12 |
| A-TIG weld joint | 475 | 430 | 520 | 465 | 66 | 18 |
The impact energies of the weld metals at room temperature and 0°C are listed in Table 2. The impact energies of both the weld metals (53 and 66 J respectively for SMA and A-TIG) are much lower than the base metal (180 J) due to the presence of delta ferrite in the weld microstructure. The deleterious effect of delta ferrite on impact toughness of martensitic steels is more pronounced at sub-zero temperature than at room temperature15). This is in agreement with lower toughness values observed for both the weld joints at 0°C. It is interesting to note that toughness of A-TIG weld metal is higher than that of SMA weld metal at all test temperatures. Further, the impact energies obtained for both the weld metals is adequate for the required application9,10). Fractographs of the weld joints also reveal ductile fracture (Fig. 4).
The creep resistance of the SMA and A-TIG weld joints is evaluated by impression creep testing and the results are plotted in Fig. 5. It is to be noted here that unlike uniaxial creep test, tertiary creep is not observed in impression creep testing and rupture life cannot be assessed by this test12). Hence, susceptibility of the joints to type IV failure cannot be determined by this test. However, it can be seen from Fig. 5 that the creep resistance of A-TIG weld metal in terms of impression depth and impression velocity, which are equivalent to the creep strain and creep strain rate of uniaxial creep testing12), is comparable to that of SMA weld metal. The creep resistance of A-TIG HAZ is better than that of SMA HAZ. This is attributed to minimal degradation in microstructure in the HAZ of A-TIG weld joint due to the lesser heat input of the process. Due to higher hardness, both the weld metals and the A-TIG HAZ possess enhanced resistance to recovery resulting in improved resistance to creep14). It may be further noted that as the width of HAZ is very narrow for both the weld joints; creep properties could not be evaluated for individual subzones of HAZ, especially for ICHAZ and the estimated creep strain and creep strain rate represents the overall properties of the HAZ. However, it is obvious from Fig. 5 that creep resistance of A-TIG HAZ is even better than that of the base metal. As this is a localized creep test technique, the deleterious effect of delta ferrite is not reflected in the test results.
Fig. 5
(a) Impression depth-vs-time and (b) impression velocity-time plot for SMA and A-TIG weld joints
It is widely reported that crack initiation and propagation is easy in ferritic steel weld metal16) and hence, it is essential to determine fracture toughness of the weld metal. From the load-vs-crack opening displacement (COD) plots and crack lengths, J-R curves are determined as per ASTM 1820 standard17). Fracture toughness (J1c) values were determined from J-R curves (Fig. 6). The blunting line equation (Eq. (1)), used to determine fracture resistance, is as per ASTM 1820 (Section A9). Fracture toughness (J1c) was determined as intersection of the offset line (0.2 mm offset from blunting line equation) with the J-R curves (Fig. 6).
where σYS = Yield strength (YS), σf = flow stress = (YS+ UTS)/2, UTS= ultimate tensile strength, Δa= crack extension
The J1c values for the base metal, SMA weld metal are 297 and 181 kJ.m-2 respectively. In contrast, brittle fracture is observed in A-TIG weld metal (64 kJ.m-2) with negligible crack extension. Although, the tensile strength of the SMA weld metal (775 MPa) is much higher than that of the base metal (660 MPa); the higher ductility explains the better toughness of the base metal.
Unlike impact toughness, which is dominated by the energy for crack initiation, the fracture resistance or J-R curve for high strength ferritic steels mainly represents the energy consumed during both crack initiation (fracture toughness: J1c) and further crack extension. In a previous study, quantitative relationship between linear elastic fracture toughness (K1c) and impact toughness was established for various grades of high strength steels18). It is widely reported that fracture toughness (either J1c or K1c) is dependent on both strength and ductility. In this study, lower fracture toughness (J1c) of A-TIG weld, compared to the SMA weld metal can be explained based on the lower ductility of the A-TIG weld metal16). The peak temperature attained during A-TIG process being higher than SMA welding, carbon dissolution in martensite is more in case of A-TIG weld metal19). This leads to higher strength and lower ductility of A-TIG weld metal. The lower ductility of A-TIG weld metal is also attributed to the location of crack initiation between the two solidification fronts where more non-equilibrium segregation is expected. Prolong PWHT will reduce this segregation and may increase the toughness of A-TIG weld metal19).
Fractographic images of the CT tested specimens are shown in Fig. 7. The fractographs of the base metal (Fig. 7a) and SMA weld metal (Fig. 7b) reveal dimples of various sizes, indicating ductile fracture. Fractograph of A-TIG weld metal (Fig. 7c) reveals cleavage type of facets, secondary cracks at different locations and absence of dimples. From the fratographic observation of A-TIG specimen, reduced tendency of fracture resistance can be inferred.
Fig. 7
Fractographs of (a) base metal, (b) SMA weld joint, (c) A-TIG weld joint after fracture toughness test
The present study evaluates and compares the mechanical properties of weld joints produced by A-TIG and SMA techniques. In A-TIG weld joint, the damage caused due to the welding heat is minimum. As a result, the impact toughness and creep resistance is appreciable. However, SMA welding process is found to be beneficial to obtain weld metal free from delta ferrite and improved fracture toughness.
4. Conclusions
The research outcome of the present study can be summarized as follows:
1) The tensile strengths of SMA weld joints at room temperature and 550°C (730, 710 MPa) are higher than those of A-TIG weld joints (520, 465 MPa).
2) The charpy impact energies of A-TIG welds (66, 18J) at room temperature and 0°C are higher than SMA welds (53, 12J).
3) The tensile strength as well as impact toughness of both the weld joints is adequate for the intended application.
4) The lower total heat input in case of A-TIG welding cause minimum microstructural degradation in HAZ, leading to excellent creep properties of the weld and HAZ.
5) The fracture toughness of A-TIG weld metal is low (64 kJ.m-2) as compared to the SMA weld metal (181 kJ.m-2).
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
