SUMMARY OF THE SESSIONS ON WELDMENTS – Chaired by Pål Efsing and Ian de Curières

Since the introduction/separation some six years ago, the weldment session(s) sees an increasing trend with respect to the number of offered presentations, and this year was no exception, being better represented than previous years. In fact a few weld-related presentations where included in the Ni-base alloys sessions too. The first session contained four papers, mostly connected to crack growth testing.

Juxing Bai (PSI, Switzerland) presented information from a growth and initiation project on Alloy 182 weld metal they are conducting in support to the Swiss Nuclear Safety Inspectorate (ENSI), respectively Swiss BWR plants. The electrochemical reality for these plants results in a working-ECP value that by large co-incides with the Ni/NiO phase transformation and this has been shown to be detrimental in other studies on PWR primary water chemistry. In these studies (see e.g. Morton or Was) it is clearly shown that there is a maximum with respect to CGR in the transition zone. However for BWR environment there is limited data. The study is an attempt to fill this gap somewhat, and shows that the CGR response for HWC (i.e., low ECP BWR chemistries) the results echo those for PWR which is yet another example of the transferability of understanding from one system to another and an indication of the phenomena to be part of a continuous unbroken connection.

Hajime Miyata (Hitatchi, Japan) discussed the current situation regarding the Japanese position on CGR for Alloy 82. The issue is highly interesting due to the increased number of defects found in, e.g., connecting weld between the support plate and support legs of the core shroud, and the RPV. It was suggested that only pure constant load samples are included in the analyses regarding a reference curve. Samples that were tested using periodical partial unloading in accordance to the Swedish test procedure were deemed being too conservative relative to the available Japanese data. However, it should be noted though that the actively loaded samples that where referenced and compared, had not been subjected to a PWHT that decreases the remaining residual stresses in the specimens unlike the samples tested in this study. Testing is proposed to be conducted to fill two areas of interest in the K vs. CGR regime that exhibits lack of data according to the presenter, near K = 50 and 80 MPa√m, respectively.

Tatsuya Kubo (Toshiba, Japan) continued the discussion on crack growth in the connecting welds for the core shroud. He presented a study where the fitness for service concept had been evaluated to support the continued operation of plants with defects in these areas. The result was a combination of CGR testing of material typical of the buttering layer (called cladding in the presentation) and testing of crack growth from the buttering layer into the LAS of the RPV using load shedding defined in a finite element analysis. The CGR in the Alloy 82 used for the buttering showed a significant lower response than the Alloy 82 that simulated the TIG-welded butt-weld. This is, according to Kubo, attributed to the lower yield strength of the material used for the buttering (citing other researchers who found the same response, in other materials, e.g., Alloy 600).

The first session was finalized by Ian de Courières, presenting a paper on behalf of Catherine Guerre (CEA, France). It was a summarizing the Ph.D. work of Elizabeth Chaumun on the influence on crack initiation in Alloy 82 in PWR primary water chemistry due to local grain boundary mechanical conditions and micro structural parameters. One of the tested weld heats has significantly higher resistance to initiation than the other tested weld heat. It has also been shown that strain is single handedly not a sufficient parameter to explain the different behavior as seen in this study. The study proposes a combination of causes for crack initiation to be dependent on both, mechanical parameters (such as local stresses and/or deformation discrepancy) and chemical parameters (such as sulfur content, intergranular oxides and grain boundary cohesion energy). What is clearly shown is the beneficial effect of heat treatment on the risk reduction regarding crack initiation of Alloy 82.

The weldments session continued on Friday morning. Welds are a critical location in components as they are a very specific stage of the manufacturing, sometimes more craftwork-like than industry. In such locations, EAC occurred and the assessment of the risk of EAC (SCC or CF) is difficult as the material undergoes strong local variations in chemistry, metallurgy and stresses. One cannot rely on studies performed on base metal to understand the behavior of such locations. In this session, three presentations explored this area.

The talks were dealing with homogeneous stainless steel welds (Donghai Du, SJTU, China) or dissimilar metal welds (Jianqiu Wang, IMR, China and Guangfu Li, Shanghai Research Institute of Materials, China). They presented very exhaustive characterizations of the manufactured welds used, from the chemical, metallurgical and mechanical points of view. These characterizations were followed by corrosion testing in conditions representative of the primary fluid, either nominal, or with deviations of oxygen or other impurities. These tests all acknowledged the fact that welds may be locally quite susceptible to EAC, essentially in the HAZ, even in low-carbon grades. However, this susceptibility, to become of real engineering significance, require some non-nominal chemistry conditions, a higher ECP seeming to be sufficient. In this case, crack initiation is favoured and CGRs in the range of 1 mm/year, or more, may be achieved. In case of dissimilar metal welds, the most susceptible area remains in the stainless steel HAZ. This highlights the fact that studies in purely nominal PWR primary water may not reproduce the actual behavior of stainless steel components.

On Friday morning, two further presentations were given after the coffee break in the final technical session of the meeting.

Qunjia Peng‘s (IMR, China) results indicate that stainless steel 308L is very resistant to SCC, a fact that he couples to the occurrence of a network of delta ferrite precipitated in the grain boundaries. The delta ferrite contains more chromium than the surrounding austenite and thus readily creates a passivating Cr₂O₃ film on the fresh crack surfaces. This also influences the corrosion fatigue response for 308L, resulting in consistently lower CGRs compared to 304 and 316 and in the same order as the CASS-type materials that have been tested. It was also shown that the CGR was considerably higher at oxygenated conditions compared to conditions containing lower DO/DH (<5 ppb).

The final technical paper at the meeting was presented by Tai-Cheng Chen (INER, Taiwan), who received a special prize for being the last presenter. The objective of this study was to investigate the influence of ageing on EAC of a weld between two CASS pipes. The cases studied were a weld between two un-aged pipes, and a weld between two accelerated aged CASS components. Accelerated ageing was conducted at 450°C, a temperature which the study claims would result in an acceptable microstructure compared to the actual case of service ageing. Despite introducing a notch in the welded portion of the samples, all of the failures during testing occurred in the CASS parts. Welding had little effect on the mechanical properties of the un-aged CASS material, whereas the aged material was strongly influenced such that the hardness of the HAZ was reduced to levels resembling the un-aged material. Furthermore, when testing the material in oxygenated water, there is a strong effect on the strain to failure during SSRT testing.

Please note that three further papers were presented during one of the summary/poster presentation sessions:

Wei Wang (USTB, China): “Characterization of microstructures of 316 stainless steel/Inconel 182 dissimilar welds and its corrosion properties in high temperature water”

Xinyuan Cao (USTB, China): “Investigation on microstructure and mechanical properties of thermal-aged stainless steel weld overlay cladding”

Yung-Lin Chen (National Tsing Hua University, Taiwan): “Stress corrosion cracking of welded type 316L stainless steels in simulated BWR environments”

Further comments from the session chairs:

A general observation at this meeting was that the variety of the presentations regarding materials and methods has increased, which is positive from a broadening of the area perspective. Testing of Alloy 52, which is indeed difficult and renders low crack growth rates without considerable acceleration in most cases, has limited value from the sake of learning something new. Thus the trend is very positive. It is also clear that the research area is of high interest, something that obviously is coupled to two facts: there are some 4000 welds in the primary coolant system alone in a normal LWR, and most failure cases are connected to welding or welded areas. This renders a higher interest from an operator’s and a regulator’s perspective. Therefore we look forward to many creative and interesting talks and discussions in the future weldments sessions.