Under Pressure – Understanding how stress and corrosion combine to create stress corrosion cracking

Stress corrosion cracking. It’s the bane of many a jeweler, even if they don’t know it’s happening. Ever wonder why a couple of perfectly good 14k white gold prongs failed for what appears to be absolutely no reason? Or why a 10k yellow gold chain link fails when all the others seem fine? Chances are, it’s because of stress corrosion cracking.

It’s not a new phenomenon. For as long as jewelry has been made, stress corrosion cracking has been happening and very often goes undiagnosed. It occurs across many product groups, though it seems more prevalent in jewelry made from lower-karat gold alloys, and the lower the karat, the more often it occurs. And these failures aren’t limited to just occurring in gold jewelry. Wire manufacturers have experienced low-karat gold wire stress corrosion cracking on spools, particularly with small gauges.

When these failures do occur, often the blame goes on the quality of the metal—there must be porosity or impurities present causing the failures. However, it may not be a metal quality issue, but what’s being done to the metal during the manufacturing process, and then the environment it is exposed to when worn. Higher incidences of stress corrosion cracking have been noted in hot and damp climates—the original name for stress corrosion cracking was “season” cracking. In fact, it was first investigated when brass rifle cartridges would spontaneously fail in monsoon season, when the environment was hot and humid.

But what exactly is it? Let’s take a look at what stress corrosion cracking is, what causes it, and what steps jewelers can take to prevent it from happening.

What Is Stress Corrosion Cracking?

Stress corrosion cracking can be described as the local rupture of an alloy under the effects of both corrosion and stress, so it’s actually all in the name. It’s cracking that occurs in metals when there’s both a stress component and corrosion present. Both of those factors must be present for the failure to technically be stress corrosion cracking, otherwise it’s a different type of failure. It occurs at stress and corrosion levels below those at which failure would normally occur if they were acting independently.

There’s also a time element involved in stress corrosion cracking. Failure can be immediate when metal is exposed to a corrosive element or can happen after a longer period of time. This will depend on a number of factors, including alloy composition, segregation/homogeneity, type and amount of stress present, corrosion levels, and even the geometry of the piece of jewelry.

A good example of stress corrosion failure is the 14k nickel white gold head setting on a ring. The stone is held in place by bending or rubbing the prongs over the girdle. This is cold work, which induces stress, even if only a small amount. When that ring is worn continuously, it often comes into contact with any number of corrosive agents, from ordinary household detergents to chlorinated swimming pool water. All the requirements for stress corrosion have been met and failure can occur.

Why Does It Happen?

To understand why stress corrosion cracking happens, let’s start at the beginning. Both 10k and 14k gold alloys aren’t really “gold” alloys. Sounds mad? Well, when looking at alloy compositions, jewelers usually look at weight percent for gold content. We know that 10k has 41.67 percent gold, and 14k has 58.33 percent gold by weight. This makes it easy for both jewelers and customers to gauge value. If a piece of 14k jewelry weighs 10 dwts, 58.33 percent by weight of it is gold, so it has 5.833 dwts of gold in it. This is the truth and very convenient.

However, when looking at the performance of alloys and how they interact with other metals and chemicals, metallurgists often look at the alloy from the viewpoint of the atomic percentage of the elements present. This is a measure of the actual number of atoms present, and it is a much better gauge of how a metal performs. Let’s look at a typical 10k yellow gold jewelry alloy and compare weight percent to atomic percent:

                              Weight %              Atomic %

Gold                       41.67                      20.41

Silver                     14.00                      12.52

Copper                 39.00                      59.21

Zinc                         5.33                         7.86

We can see that, atomically, only 20 percent of the atoms present are gold, with almost 70 percent of the remaining atoms being copper and zinc, so this is a copper-zinc alloy with silver and gold present, which is a brass. The same applies to 14k alloys. Although there is 58.33 percent by weight of gold, the number of gold atoms present is typically around 33 percent, with the majority of the remainder being copper and zinc, so a brass again. It’s only when you get to 18k gold that the number of atoms present gets to around 50 percent and we can metallurgically call it a gold alloy.

The reason for this is that gold is a dense and therefore relatively heavy atom for its volume. For a given amount of gold weight present, you need fewer actual gold atoms present. Since there is a corrosive element to this mode of failure, it tends to be more common in metals that are prone to corrosion, and alloys with high levels of copper and zinc present can fall into this category. Brasses are a huge family of alloys that have many different uses, and many are susceptible to stress corrosion cracking. This is where the earlier explanation of atomic percent vs. weight percent becomes important. As de facto brasses, 10k and 14k gold alloys—and any lower-karat alloys—are also susceptible to this mode of failure. (This is especially important to understand since the Federal Trade Commission issued new guidelines that make way for the sale of very-low-karat alloys as gold. Some of these alloys may be designed to be resistant to stress corrosion cracking, but many will not.)

How Does It Happen?

So now that we know the why, let’s look at how stress corrosion cracking happens.

It starts with corrosion, which is an electro-chemical reaction that takes place when the dissimilar metals that make up an alloy react with a reagent—a liquid—to form metal compounds in solution. The corrosion process involves the selective attack of the less noble metals in the alloy. When this happens, metal is removed from the parent body. Fine gold has a high resistance to chemical attack because of its nobility, so alloys containing a minimum of 50 atomic percent gold, which in the case of jewelry means 18k and 22k alloys, generally resist corrosion well.

Stress corrosion failures have two distinct phases: crack initiation and crack propagation. Crack initiation is the result of localized corrosion, generally at the site of a microscopic defect, such as porosity or pits. This is the beginning of a crack, and as corrosion continues and the defect gets bigger, the cracking potential grows greater until it begins to propagate due to the stress present. Crack propagation can be both trans-granular (through the grain) or inter-granular (via the grain boundary), the latter being most common in karat gold alloys.

During crack propagation, “new” metal that has not yet been subjected to the corrosive element is being continuously exposed at the crack tip. Passivation is a process that naturally occurs at the crack faces when all of the crack surfaces are occupied by atoms of the more noble elements—corrosion has removed all of the less noble atoms. This will occur after the removal of very little metal from the crack surface. If passivation occurs at a higher rate than exposure of new metal at the crack tip, the propagation process ceases and stress corrosion will not occur.

However, if the strain rate at the crack tip is high enough, new metal is exposed faster than passivation can occur and so failure progresses. This is especially likely to occur in lower-karat materials—there are fewer gold atoms present so it will naturally take longer to passivate.

After crack initiation, two conditions need to be filled in order for the crack to propagate. It must be energetically favorable (where more energy is released by the cracking process than it takes to make the crack in the first place), and the crack needs a mechanism by which it can propagate. Nature always wants to go to a state of least energy, so at every stage of crack propagation, energy stored in the material must be reduced for propagation to continue. When a crack begins propagation, as the faces part, a small amount of energy is released.

However, there is also an energy requirement for the crack to continue propagation and form new crack surfaces. The energy released by cracking is proportional to the square of the crack depth. For example, a crack 3 mm deep will release nine times the energy of a crack 1 mm deep, but it requires only three times the energy to form, so it is energetically favorable and cracking progresses. Once a critical crack length is reached where more energy is released by cracking than it takes to propagate the crack, spontaneous failure occurs. The grain boundary is the route or mechanism via which the crack can propagate. These boundaries are high-energy sites and as such have more energy to release.

Can It Be Prevented?

Published data and industry experience suggest that 10k and 14k nickel white golds are alloys that are particularly susceptible to stress corrosion cracking. (As a side note, there are no known instances of stress corrosion cracking with palladium white gold alloys.) Most failures occur in settings, and specifically in prongs due to their small, and often sharp, cross sections. Stone setting is also one of the last things done, so any residual stresses from bending the prong remain in the prong.

Published data also notes that square or rectangular cross sections are more likely to fail than round cross sections. This is due to the sharp corners of the section acting as stress raisers. When angular cross sections are bent, micro-cracking, also known as grinning, starts relatively quickly at the apex, inviting corrosion to initiate. Translated into real life, this means that settings with square or rectangular section prongs will be more susceptible to stress corrosion than basket-type settings with round prongs. There will always be exceptions to the rule, but as a generalization, this seems to be correct.

For nickel white gold alloys, it is important to know that gold and nickel do not mix well at lower temperatures, resulting in a segregated microstructure of nickel-rich and gold-rich phases. This stresses the bonding between them and creates internal stresses so the more that this segregation can be inhibited, the better. Quenching nickel white golds after annealing and soldering will help suppress the phase separation in this family of alloys. However, certain nickel white gold alloys are also susceptible to fire and quench cracking, so beware.

Nickel white gold alloys are relatively hard, so they tend to get worked less than they should be between anneals. The result of this is that the outer layers of the piece tend to be deformed and worked more than the inner section, creating a strain gradient and therefore internal stresses—ideal for stress corrosion cracking. To help prevent stress corrosion cracking, nickel white golds need to be cold worked as much as possible between anneals or before soldering. This minimizes the strain gradient effect.

10k and 14k yellow gold alloys can also suffer from stress corrosion but to a lesser degree than the nickel white golds, with 10k being more susceptible than 14k. Industry experience shows that for this family of alloys, the single-phase microstructure that results from quenching these alloys from higher temperatures makes them more susceptible to cracking, while the multi-phased microstructure that is generated from slow cooling makes them more resistant.

Some published sources claim that large grain sizes help prevent stress corrosion, but as jewelers, we typically don’t want this since it can lead to what is known as “orange peel” texture when the metal is bent. It is also true that hard-drawn or rolled microstructures do not easily stress corrosion crack because the grain boundaries are disrupted, removing the mechanism for inter-granular failure.

The main way to prevent stress corrosion cracking is to take away one of the elements required for it to happen—either the stress or the corrosion. The stress factors can be dynamic, static, or residual. Stress relieving by heat treatment is a great way to prevent stress corrosion cracking. For example, think back to the rifle cartridges mentioned earlier. Uneven residual stresses were present from the manufacturing process, but the cartridges would remain unaffected for months until influenced by corrosion due to the humid climate. The problem was very easily resolved once understood. A low temperature stress-relieving heat treatment during the cartridge manufacturing process solved the problem—it took away one of the elements needed. 

Plating can also help protect against stress corrosion. Provided there are no pits or micro-pores in the plated layer, rhodium plating of nickel white gold components will provide a degree of protection because it will not allow corrosion to begin, at least for as long as the plating lasts.

One of the best ways to prevent stress corrosion cracking is to clean jewelry regularly. Remember: This is a time-related phenomenon. Jewelry often has little nooks and crannies in the design that are perfect to entrap corrosive agents. Some also entrap dead skin cells, which can absorb corrosive agents, and thus provide the means for a long-term corrosive attack. If jewelry is cleaned regularly, it prevents this. While you can’t be responsible for what happens to your jewelry after it walks out the door, you can encourage your customers to regularly clean their pieces (or bring them to you for cleaning) and to avoid wearing them when using corrosive agents.

It’s important to note that not everything will stress corrosion crack, even if the requirements are fulfilled for it to take place. If, as a jeweler, you’ve never had an instance of failure described here, then you’re doing it right! Change nothing! But if you have had unexplained brittle failures in some of your jewelry, then maybe a few changes will help solve the problem.