top of page

It's Cracked! During working and Annealing

Updated: Oct 23, 2022

Cracking & Overworking

If metal is cold-worked without annealing, continued working will ultimately lead to failure by cracking.

A typical tensile curve (stress versus strain)

If a cylinder of metal is loaded in tension, the metal will respond as follows:

  1. The Elastic Regime – It will stretch and become longer for low loads. When the force is removed, the sample will return to its original dimensions. The strain is proportional to the force, i.e. Hooke's Law. The material behaves elastically.

  2. At a certain load or stress (force per unit cross-sectional area), the material begins to deform plastically. This deformation is permanent. Because volume is conserved during deformation, as it becomes longer, it becomes thinner. As we deform the material, the material becomes stronger and harder, known as work hardening.

  3. In the plastic region, the stress starts to fall at some point. This is due to necking. During necking, work hardening still occurs.

  4. If we continue, the specimen under test starts to crack up, and it eventually fractures.

Localized necking during tensile test. Source: Santa Fe Symposium

The cause of work hardening

As the material is deformed, the material becomes stronger and harder. This is called work hardening. While work hardening helps to harden jewelry, improving durability. It also ultimately leads to failure.

Deformation occurs through the motion of dislocations. More deformation requires more dislocations. As the number of dislocations increases, they impede the motion of one another, thereby increasing the necessary stress.

At some point, the necessary stress required for further dislocation motion and deformation exceeds the necessary stress for cracking, and so cracking occurs.

When we anneal the alloy, alongside an increase in grain size, recrystallization reduces the dislocation density. Removing dislocations restores ductility and allows further working.

Avoiding overworking

Overworking can occur in all forms of metal working – sheet and rod rolling, wire and tube

drawing, blanking, stamping, coining, spinning and raising, milling, turning and machining, or simply bending by hand.

The extent an alloy can be worked before failure depends on the microstructure. Materials in age-hardened condition or with hard second phases will have inherently lower ductility and tend to crack more easily on further working.

Cracking during stamping due to overworking. Source: Santa Fe Symposium

Materials should be annealed roughly after about a 70% reduction in thickness before annealing. However, there are considerable variations:

  • Nickel-white gold hardens rapidly. Annealing is necessary before a 40% reduction in thickness.

  • Fine gold and some of the high-karat golds can be worked well in excess of 90% reduction in the area before annealing becomes necessary.

Cracking & Incorrect Annealing

Residual internal stresses and Quench Cracking

Non-uniform deformation can often lead to residual stresses in wrought items. In many processes, there is a tendency for the deformation to be focused on the surface regions, and there is less work done in the center. Residual stresses can be circumferential (round the sample), longitudinal (along the length), or radial (pointing outwards from the center).

Fire cracking in 18K White Gold Tube. Source: Santa Fe Symposium

On annealing samples, they can sometimes spontaneously crack. This is attributed to residual internal stress. Residual stresses can be relieved by a low-temperature anneal without reducing the hardness. A common example of this effect is fire-cracking in Nickel-white gold.

Nickel-white golds work harden quite rapidly, so they require more frequent annealing. They must be quenched after annealing to prevent the formation of two immiscible phases and any resulting hardening effects due to the second phase.

Nickel-white and yellow karat-gold alloys, particularly those containing silicon additions, are prone to quench cracking. Repeated heating and cooling generate internal stresses as the outer layers cool more rapidly than the middle of the metal during quenching; the stresses generated can be large enough to cause cracking in such inherently brittle alloys.

To avoid the quench-cracking effect and also avoid the hardening effect induced by slow cooling in nickel-white golds, an intermediate cooling rate is performed using a variety of techniques:

  • Forced cooling in air

  • Cooling by placing on an iron plate

  • Quenching into hot water

  • Slow cooling until close to the critical temperature, at which the second phase begins to precipitate and then quench.

Individual manufacturers will have their own techniques for particular alloys and sizes of components but still will often face difficulties.

Hot Shortness

For investment-cast silicon-containing karat gold, the presence of second-phase particles with a low melting point at the grain boundaries leads to cracking. The residual stresses caused by cooling from too high a temperature lead to hot tearing and cracking.

Quench cracking in investment-cast silicon-containing 9K gold. Source: Santa Fe Symposium

Fire Cracking

Fire cracking commonly occurs in nickel-white golds during annealing or soldering. Heating, combined with residual stresses from working operations, is sufficient to fracture the component as the temperature increases (and the strength decreases).

The remedy is to slowly heat up to 300°C/575°F and hold at this temperature for a time to relieve the stresses before continuing onto the annealing or soldering temperature.

Overannealing and Orange Peel

Annealing at too high a temperature, for too long a time (or both!) can result in large, coarse grains.

Example of Orange Peel Effect. Source: Santa Fe Symposium

Subsequent deformation can lead to premature cracking and fracture (as well as an “orange peel” surface). This is particularly a problem with torch annealing, where the capability to control temperature is limited. Read more about orange peel effects here.


Spot a mistake?
Want something else?
Want to contribute?

Get in touch!

  • Facebook
  • Twitter
  • LinkedIn
  • Instagram

Thanks for submitting!

bottom of page