Sensitization control in AISI 316L(N) austenitic stainless steel: Defining the role of the nature of grain boundary

Abstract

Two grades of AISI 316L(N) austenitic stainless steels differing only in copper content (0.083 and 0.521 wt.%), showed remarkable difference in resistance to sensitization and susceptibility to intergranular corrosion. Different thermal treatments were carried out with an overall objective of altering the nature of the grain boundary. An attempt was made to correlate the degree of sensitization (DOS) with various microstructural parameters such as grain size and grain boundary nature. No clear trend could be established between the individual parameters and DOS. Effective grain boundary energy (EGBE), which is a combined parameter showed clear trend with DOS. The presence of 0.521 wt.% of copper brings down EGBE remarkably leading to improved resistance to sensitization.

Introduction

Austenitic stainless steels have excellent resistance to uniform corrosion [1], [2], [3]. However, they are prone to localized corrosion such as intergranular corrosion, intergranular stress corrosion cracking, crevice corrosion and pitting corrosion. Intergranular corrosion and intergranular stress corrosion cracking are caused by ‘sensitization’ which refers to the formation of chromium – rich carbides along the grain boundaries and the concurrent depletion of chromium in the immediate vicinity. In this process, if the local chromium content drops below 12 wt.%, then the chromium depleted zones become prone to local corrosion [1], [2], [3], [4].

Sensitization control is attempted through control of alloy chemistry and heat treatment [1], [2], [3]. Another emerging approach is tailoring the nature of grain boundary, which in turn inhibits the formation of chromium – rich carbides and/or chromium depleted zones [4], [5], [6], [7], [8], [9], [10]. In this approach, there can be three issues involved: (i) processing steps to control nature of grain boundary, (ii) exact relationship between nature of grain boundary and degree of sensitization (DOS) and (iii) possible alloying additions (with preference for grain boundary segregation) to control DOS further. Issue (i) is typically attempted through combinations of plastic deformation and annealing [4], [5], [6], [7], [8], [9], [10], while issue (ii) requires representation of nature of grain boundary as a single or dominant parameter.

Based on the misorientation between adjacent grains, grain boundaries can be categorized as low and high angle boundaries, the latter being further classified as special and random [4], [11]. The coincident site lattice (CSL) is often used to define special boundaries [4], [11], [12]. Coincident site grain boundaries are boundaries where certain fraction of atoms in each grain coincide with positions corresponding to the extension of the crystal lattice of the other grain. A coincident site grain boundary would contain many atoms which are common to both grains, and show good atomic fit between the grains. In CSL boundaries, Σ refers to the reciprocal density of common lattice points for the two grains. e.g. a Σ5 boundary has 1 in 5 atoms at coincident sites. CSL boundaries have certain special properties, such as low boundary energy, less susceptibility to impurity or solute segregation, and greater resistance to intergranular degradation and are referred as ‘special boundaries’. In Grain Boundary Engineering, the material properties are sought to be improved by increasing the amount of CSL boundaries and decrease the non-CSL/‘Random’ boundaries. Usually, CSL boundaries from Σ1 (low angle boundary) to about Σ29 are analysed in Grain Boundary Engineering studies. Each CSL Σ boundary exhibits a specific misorientation angle–axis, and are well tabulated for cubic crystalline system. In EBSD technique, measured misorientation angle–axis at the grain boundary is compared with the ideal CSL set, to recognise the CSL (if the deviation is within the tolerance given by Brandon’s criteria Δθ = 15 Σ^−1/2). A CSL notation does not put forward deviation from ideal CSL nor does it distinguish between tilt and twist boundaries – though both can have significant effects on the nature and energy of grain boundary [4], [10], [11]. Attempts are usually made to link DOS with individual parameters of grain boundary nature [4], [5], [6], [7], [8], [9], [13], [14], [15], [16], [17]. Such parameters may range from grain size to CSL concentration and connectivity of special/random boundaries. Relationship between DOS and such individual parameters may not always be linear. The role of the individual parameters may also differ between materials and processes. Naturally, attempts are also made [4], [10] to link DOS with a combined parameter. Such a combined parameter, termed effective grain boundary energy (EGBE), can couple several aspects of the nature of grain boundary [9] and in turn, may serve as an effective microstructural index. It may be noted that DOS has been reported to increase with increase in EGBE, though beyond a critical EGBE significant drop in DOS has been observed [4], [10].

It may also be noted that certain alloying elements, e.g. cerium, has been reported to resist sensitization and can be effective for control of intergranular stress corrosion cracking [17]. This effect can be explained as selective segregation of over-sized atoms to grain boundaries. The problem is in the significant drop of workability, making a wide acceptability difficult. If other alloying elements, with preference for grain boundary segregation but with less detrimental effects on workability, can create similar effects remains an open-ended question.

Nitrogen alloyed AISI 316L(N) austenitic stainless steel is being used as the structural material in primary circuit of Prototype Fast Breeder Reactor which is under construction at Indira Gandhi Centre for Atomic Research, Kalpakkam, India. The present study was initiated through supply of two different grades of this material from two different sources. As shown in Table 1, except for differences in copper content, the grades had similar chemical composition. These two grades were subjected to a range of thermal processing [18] with an primary objective to define the role of nature of grain boundary in sensitization control of AISI 316L(N).