Before discussing the metallurgy of the nickel and cobalt alloys, it is important that certain metallurgical terms are understood. First and foremost, it should be understood that an alloy is a mixture of metals, possibly containing small quantities of non-metals, such as carbon. The predominant metal in the alloy is known as the base.
A solid solution is an alloy in the solid state having a single atomic structure or phase. Second phases are possible when the combined levels of alloying additions to the base exceed their solubility limits. So, as with a liquid solution, there are natural limits to how much can be dissolved in a metallic material of a given atomic structure, and as with a liquid solution, the higher the temperature, the more can be dissolved. Fortunately, it is possible to create supersaturated solid solutions by heat treating materials at high temperatures where the solubilities are higher, then rapidly cooling the materials to room temperature, or at least below 500°C, where the diffusion of atoms (the main catalyst for microstructural change) is no longer appreciable. Holding alloys at high temperatures, to dissolve unwanted second phases in their microstructures, is known as solution annealing. Rapid cooling, to lock in the high temperature microstructure, is known as quenching, and is best performed in cool water.
The problem with such supersaturated materials is that they are prone to second phase precipitation during excursions above 500°C, when diffusion becomes appreciable. Such excursions are common during welding, for example.
Unfortunately, precipitates tend to nucleate and grow at microstructural imperfections, such as grain boundaries. These then become prone to preferential corrosion attack.
Not all second phase precipitates are detrimental. Those that precipitate homogeneously (i.e. throughout the microstructure, rather than just at the grain boundaries) can be used to strengthen materials. This is known as precipitation-hardening or age-hardening. The heat treatments used to induce such precipitates often involve multiple steps in the temperature range 500°C to 800°C.
The microstructures of wrought and cast alloys comprise numerous grains, within which the crystal structure is aligned in a certain direction. However, these grains can sub-divide under the action of mechanical stress or temperature by a process known as twinning, whereby bands of material within a grain can realign.
Grain boundaries (of irregular geometry) and twin boundaries (which are straight and parallel) are very important microstructural features, since they are preferred nucleation sites for second phase precipitates.
The major alloying elements determine the general behavior of a material. However, minor alloying elements are also important. Some minor elements are there to ensure successful melting and processing; some are used to fine-tune performance in specific environments. Others are added to induce hardening precipitates.
Except in the case of the few precipitation-hardenable nickel alloys designed to resist aqueous corrosion, strengths are determined largely by the major alloying elements. These provide solid solution strengthening. Large atoms such as molybdenum are particularly effective strengtheners.
To maximize the corrosion resistance of the nickel alloys, many are deliberately overalloyed and reliant upon the previously mentioned process of solution annealing and quenching to optimize their microstructures. Even those that are not overalloyed are prone to second phases, due to the presence of insoluble residuals, such as carbon.
The purpose of this part of the manual is to provide information on the physical metallurgy of the four groups of wrought, corrosion-resistant, nickel alloy with which Haynes International is associated (i.e. the Ni-Cr, Ni-Mo, Ni-Cr-Mo, and Ni-Cr-Fe groups). For completeness, however, microstructures of materials from the other three groups will be shown and commented upon.
As already discussed, the atomic structure of pure nickel is face-centered cubic (FCC), otherwise known as gamma phase within the realm of nickel alloy metallurgy. Commercially-pure nickel (Ni), of which there are several grades, and the nickel-copper (Ni-Cu) alloys generally exhibit stable FCC microstructures, although (due to the insolubility of impurity carbon in wrought products from these two groups), carbides, or even free carbon, can be observed in grain boundaries or dispersed throughout these materials. The presence of carbides is evident in the following microstructures of annealed sheets of Nickel 200 and MONEL® alloy 400:
Incidentally, unless otherwise stated, the optical photomicrographs exhibited in this manual involved the following metallographic procedures:
This etchant is known as the “All Purpose Etch” at Haynes International.
It is appropriate to combine the discussions of alloys from the Ni-Cr and Ni-Cr-Mo groups, since some of the commercially-important, wrought, corrosion-resistant Ni-Cr alloys contain significant levels of molybdenum, so that the same phase diagrams are relevant.
The 850°C section of the Ni-Cr-Mo phase diagram, shown below, provides some indication of the solubilities of chromium and molybdenum in nickel, in the absence of other alloying elements. It indicates that, at this temperature, and with a chromium content of 30 wt.%, then the molybdenum content must be less than 10 wt.%, to maintain a single-phase (FCC) structure. Conversely, if the molybdenum content is 25 wt.%, then the chromium content should be less than 10 wt.%.
(Raghavan et al, 1984)
Several Ni-Cr and Ni-Cr-Mo materials (for example HASTELLOY® C-4, C-22HS®, HYBRID-BC1®, Alloy 59, and G-35® alloys) are effectively ternary systems. All except C-4 alloy (which is well within the gamma phase field, and consequently relatively stable) are close to the gamma phase field boundary. This is due to a trend whereby designers of such materials have chosen to maximize the contents of chromium and molybdenum, for enhanced corrosion resistance, while using solution annealing and quenching to maintain an (albeit meta-stable) single-phase (FCC) microstructure.
Such an approach is limited by the need to avoid continuous precipitation of second phases (such as µ phase) in the grain boundaries during subsequent thermal cycles (for example, during welding).
To complicate matters, other second phases are possible in the Ni-Cr and Ni-Cr-Mo alloys. They can occur at temperature ranges below or (less likely) above 850°C. They can result from the presence of residual elements (notably carbon). Alternatively, they can be triggered by other alloying additions.
With regard to second phases associated with different temperatures, an ordered phase of the type A2B, or in this case Ni2(Cr, Mo), can occur in the range 300°C to 650°C by long-range ordering (Raghavan et al, 1982). The driving force for this ordering reaction depends upon the atomic ratios of the various alloying elements. The precipitation reaction is homogeneous, with no preferential precipitation at the grain boundaries or twin boundaries. At low temperatures within the range, the reaction is impeded by low diffusion rates; however, at temperatures closer to 650°C, the reaction can be strong enough to be used as a strengthening process (Pike et al, 2003). Indeed, this is the reaction used to strengthen C-22HS® alloy.
The most important second phase resulting from residual (unwanted) elements is M6C carbide, which is prevalent in the high-molybdenum alloys, even at very low residual carbon levels (0.005 wt.% or less). Hodge, 1973 indicates that M6C forms in the temperature range 650°C to 1038°C in C-276 alloy (as compared with a range of 760°C to 1093°C for µ phase). The same reference indicates that the kinetics of carbide formation are faster than those of m phase.
The microstructure of G-35® alloy, a representative of the Ni-Cr group (containing 33.2 wt.% chromium and 8.1 wt.% molybdenum) is shown in the following figure. It is generally free of second phase precipitates, although some are evident in selected grain boundaries. These are likely carbide precipitates, since the maximum carbon content in G-35® alloy is high (0.05 wt.%), relative to the wrought Ni-Cr-Mo alloys, for which the maximum carbon content is typically 0.01 wt.%.
To evaluate the propensity for second phase precipitation (i.e. thermal stability) in alloys of this kind, it is usual to subject them to long-term aging treatments, at different elevated temperatures (normally above 500°C, where diffusion of elements becomes appreciable). Such temperatures are well beyond the use temperatures of the corrosion-resistant nickel alloys, and the exposure times infinitely higher than those experienced during welding; nevertheless, they provide a glimpse of the “true” stability of the material. Below are microstructures (photographed using an optical microscope) of G-35® alloy, after aging for 8,000 hours at 538°C (1000°F), 649°C (1200°F), and 760°C (1400°F). Higher magnification, secondary electron images of the samples aged at 649°C and 760°C, taken on a scanning electron microscope, are also shown (Srivastava and Crook, 2016).
The optical photomicrograph of the microstructure after aging at 538°C indicates some second phase precipitation in the grain boundaries. The optical photomicrograph and secondary electron image of G-35® alloy aged at 649°C indicate more extensive grain boundary precipitation, together with some intra-granular, second phase precipitates. After long- term aging at 760°C, the alloy exhibits a large amount of inter- and intra-granular precipitation, and notably an array of acicular particles within the grains.
Energy-dispersive X-ray (EDX) analyses revealed that the second phases present in G-35® alloy, after long-term exposure at 649°C and 760°C, are:
The annealed microstructure of HASTELLOY® C-276 alloy is shown in the optical photomicrograph below. It features “clean” (i.e. precipitate-free, or undecorated) grain boundaries, and, like G-35® alloy, annealing twins. The small, black particles dispersed throughout the microstructure are most likely oxide impurities. Although photographs of long-term aged microstructures are unavailable for C-276 alloy, it is known that m phase and carbides occur in C-276, as already discussed.
Most of the Ni-Cr-Mo materials are microstructurally similar to C-276 alloy. However, whereas C-276 alloy and others are not particularly prone to the A2B ordering reaction in the temperature range 300°C to 650°C, C-22HS® alloy is very prone, as already discussed. HYBRID-BC1® alloy, which is strictly a Ni-Mo-Cr alloy, since its molybdenum content is higher than its chromium content, is also very prone. Fortunately, optical photomicrographs and secondary electron images for long-term aged HYBRID-BC1® alloy are available, and illustrate the effects of A2B ordering extremely well, as shown below.
The presence of A2B in HYBRID-BC1® alloy after long-term aging at 649°C (1200°F) is shown to great effect in the above photomicrographs, particularly in the secondary electron image taken using a scanning electron microscope (SEM). The difference between the optical photomicrographs taken after aging at 538°C (1000°F) and 649°C (1200°F) is very striking, given that they were etched under the same conditions. Obviously, the driving force for the ordering reaction is very temperature-dependent, due to an exponential increase in diffusion rates over this temperature range.
Incidentally, the second-phase particle stringers evident in the mill annealed and 538°C aged microstructures of HYBRID-BC1® alloy are due to residual segregation (banding) effects. Banding is common in high molybdenum and/or tungsten alloys, since these elements diffuse slowly due to their large atomic sizes.
As regards the much larger second-phase precipitates evident in the grain boundaries, and within the grains, of HYBRID- BC1® alloy after long term exposure at 760°C (1400°F), these were identified by EDX as:
In summary, commercial wrought nickel alloys containing significant quantities of chromium and molybdenum (whether from the Ni-Cr or Ni-Cr-Mo groups) are typically metastable within their normal operating temperature range (i.e. R.T. to 427°C). By virtue of solution annealing and rapid cooling, they exhibit a microstructure that is predominantly gamma (FCC) phase, although minor amounts of precipitation are possible in grain boundaries, and small oxide particles can be seen, sparsely peppered throughout the material.
Most of these materials are designed such that short-term thermal excursions above 500°C (as encountered during welding) do not cause continuous precipitation of second phases (carbides or intermetallics) in the grain boundaries. However, long-term exposures reveal their equilibrium, multiple phase nature.
Turning now to the nickel-molybdenum (Ni-Mo) alloys, they are metallurgically complex, and prone to second phases that severely impair both their mechanical properties and resistance to stress corrosion cracking. With molybdenum contents between 28 and 31.5 wt.%, they can exhibit three phases, as indicated in the following binary phase diagram, if not fully solution annealed and rapidly quenched to lock in the FCC phase stable at temperatures above 900°C (which in this diagram is confusingly designated alpha phase, rather than the usual gamma phase). Beta is the ordered body- centered tetragonal intermetallic, Ni4Mo, and gamma is the ordered orthorhombic intermetallic, Ni3Mo. Beta and gamma phases are both very detrimental, if they form, causing loss of ductility and susceptibility to hydrogen embrittlement and chloride stress corrosion cracking. The formation of beta phase is much more rapid than the formation of gamma phase.
(Ref. Gutherie and Stansbury, 1961)
The microstructure of HASTELLOY® B-3® alloy is shown below. The straight parallel lines are twin boundaries (caused by realignment of the atomic structures within individual grains, during annealing) and the random lines are the alloy grain boundaries. The grey streaks that run across the photograph are etching effects due to remnant segregation from the cast ingot, and the fine black particles are oxide inclusions. Though not very evident, a few second phase particles (probably carbides that did not dissolve during the annealing process) are present. Preparation of the sample involved sectioning, mounting, and polishing, followed by etching in a mixture of chromic and hydrochloric acids.
With regard to the time required to induce deleterious Ni3Mo and Ni4Mo precipitates in the Ni-Mo alloys, the figure below shows the Time-Temperature-Transformation Chart for B-2 and B-3® alloys (Klarstrom, 1993). It demonstrates how small changes in the composition of Ni-Mo alloys can have significant effects upon the various possible transformations. The main compositional differences between B-2 and B-3® alloys are deliberate 1.5 wt.% additions of both chromium and iron to B-3® alloy, along with a 0.5 wt.% increase in molybdenum content. Interestingly, the original HASTELLOY® B alloy was less prone to the rapid precipitation of deleterious Ni3Mo and Ni4Mo precipitates, apparently as a result of its deliberate 5 wt.% iron addition, the intention of which was to allow the use of ferro-molybdenum in its manufacture. However, HASTELLOY® B alloy was more prone to carbide precipitation in the grain boundaries due to its higher carbon allowance, since it pre-dated the advent of the argon-oxygen decarburization process in the mid-1960’s, which enabled the attainment of very low carbon contents in the nickel-based corrosion-resistant alloys.
(Ref: Klarstrom, 1993)
With regard to the Ni-Cr-Fe and Ni-Fe-Cr materials, the most relevant phase diagram is the 800°C section of the Ni-Cr-Fe ternary system, constructed by Raynor and Rivlin (shown below).
(Raynor and Rivlin, 1981)
However, it should be noted that the Ni-Cr-Fe corrosion-resistant alloys typically contain significant quantities of molybdenum as well, rendering this diagram less relevant. What can be stated is that iron limits the solubility of molybdenum in wrought nickel-based alloys, and promotes the presence of sigma (s) phase in overalloyed materials.
In general, the wrought, Ni-Cr-Fe and Ni-Fe-Cr materials can be thought of in the same way as the Ni-Cr and Ni-Cr-Mo alloys. That is, they can be slightly overalloyed (super-saturated) to maximize their corrosion-resistance, but care must be taken during solution annealing and rapid cooling to ensure an optimum (predominantly gamma phase, but possibly metastable) microstructure within the service temperature range. Also, welding should not result in a continuous grain boundary precipitation of one or more second phases; otherwise, preferential grain boundary attack can occur in certain solutions. Ni-Cr-Fe and Ni-Fe-Cr alloys intended for corrosion service also benefit from ultra-low carbon and silicon contents, since these insoluble, residual elements can cause the precipitation of deleterious carbides and intermetallics, respectively.
To enable a microstructural comparison between G-30® alloy (from the Ni-Cr-Fe group) and G-35® alloy (from the Ni-Cr group), annealed and aged microstructures of the former are shown in the following photomicrographs.
EDX analyses indicate the presence of three second phases in G-30® alloy, after long-term aging at 649°C and 760°C. One contains about 80 wt.% chromium, and is believed to be alpha chromium. The second contains approximately 40 wt.% chromium, 25 wt.% nickel, 14 wt.% molybdenum, 7 wt.% tungsten, and 12 wt.% iron. The third, most likely a carbide, contains about 4.5 wt.% carbon, 65 wt.% chromium, 10 wt.% molybdenum, and 5 wt.% tungsten.
To complete the section on metallurgy of the wrought, corrosion-resistant, nickel-based alloys, it is appropriate to illustrate the mill annealed microstructure of a Ni-Fe-Cr alloy. A photomicrograph of 825 alloy is therefore shown below. Its fine grain size is either a result of special sheet processing, and/or the presence of an appreciable carbon content, resulting in a titanium carbide dispersion, which would encourage grain boundary pinning.
The unique metallurgical characteristics of the wrought, corrosion- (and wear-) resistant cobalt-based alloys were outlined earlier in this manual, so only a few additional details will be provided here. Considering first the tungsten- bearing, high carbon alloys, the microstructure of HAYNES® 6B (STELLITE® 6B) alloy is illustrated below. Note the large, discrete carbide particles; their size and morphology not only provide the wrought product with significantly enhanced ductility (as compared with cast equivalents), but also result in higher resistance to low stress abrasion (if the abrading particles are larger than the carbides, and can “skate” over them). According to Antony and Silence, 1979, the weight percentage of carbides in 6B alloy is approximately 12.5, of which about 90% is M7C3 and about 10% is M23C6.
With regard to the cast equivalents, these are used, more often than not, in weld overlay form. The microstructure of such a weld overlay is shown below. Note that the carbides are much finer, and dispersed in a hypoeutectic fashion.
The microstructure of wrought ULTIMET® alloy (a representative of the molybdenum-bearing, low-carbon cobalt alloys) is very similar to that of the Ni-Cr and Ni-Cr-Mo alloys, in the solution annealed and quenched condition, as shown below. This is by design, to provide optimum resistance to corrosion. Its outstanding wear characteristics derive not from the presence of hard particulates in the material, but to microstructural changes (under stress) at the atomic level (twinning and the formation of HCP platelets); these induce high work hardening rates, and provide high resistance to sub-surface fracture.