Program Speaker, Dr. Jill Glass, Brittle Material Fracture and Fractography
Text by Stephen and Nancy Attaway

Dr. Jill Glass, a ceramic scientist at Sandia National Laboratories, presented research on fracture and crack growth titled “Brittle Material Fracture and Fractography”. She covered information relating to her work on ceramic and glass failure analysis, and she correlated the data to correspond to the faceting of natural gemstones. Dr. Glass related her brittle fracture theory and showed how fractography, the study of fracture surface features, can help determine the different ways cracks can grow.

Dr. Glass began by explaining what constituted good, bad, and ugly fractures. Good fractures were mining and road construction, the cleaving of gem material, like diamonds, and removable valves. Good fractures also include the Echo Ampitheatre in northern New Mexico, as the rock there exhibits the same twist tackle features found in glass fracture. The bad fractures were the Fred Murrah Building in Oklahoma that was made of glass and was bombed, the fallen rock slabs in Yosemite National Park, and osteoporosis of human bones. Bad fractures also include the Liberty type of ships, which broke in half from the brittle metal used in their construction. Ugly fractures were car windshield cracks and the Fred Ward emerald.

Dr. Glass defined brittle fracture as happening with little or no warning. Brittle fracture shows little or no plastic deformation, as the atoms are unable to slide past one another. The pieces resemble a puzzle and match perfectly, if you can find them all. Brittle materials are everywhere, appearing in many types of glasses, intermetallics, foods, ice, metals, polymers, and semi-conductors. They can form as single crystals or as poly crystals.

Brittle materials fail under tensile stress from tension and compression. The presence of flaws produces local tensile stresses that appear from macroscopic compression. Most brittle materials exhibit linear elastic behavior.

Flaws are present on many scales. Vacancies, interstitials, and substitutions can appear in the atomic structure and lead to fractures. Rows of atoms may be dislocated and cause fractures. Crystallites or grains and anistrophy can produce fractures. Multi-grain levels, pores, chemical in homogeneities, and other such processing defects may also enable fractures to occur. Impacts and scratches when materials are carelessly handled may even cause fractures.

The worst flaws that stem from size and orientation in the worst location possible produce failure of the material. Stress, whether applied or residual, is necessary for fracture. The residual stresses are often hidden and are ready to help initiate material failure. The processing of materials, whether it was manufactured too fast or had too little time to cool, along with the damage from the subsequent handling, can produce residual stresses.

Dr. Glass pointed out that flaw size can be correlated with tensile strength to estimate the theoretical strength of a material. She showed equations that were developed to describe the concentrations of stress at a crack tip. The larger the flaw, the more stress is concentrated at the crack tip. The stress at the crack tip will be proportional to the square root of the crack size. Large flaws lead to low strength. Small flaws lead to high strength.

Material failure may be due to its fabrication and the presence of any flaws on the atomic level, such as voids, large grains, and inclusions in the material. Material fracture can be caused by machining and accidental damage from handling. Fractures may appear as radial and lateral cracks. Dr. Glass showed some very interesting mirographs that illustrated how chatter marks from machining can form along fracture surfaces.
As a model for how atoms are pulled apart, Dr. Glass gave an example of two boards nailed together with uniform spacing. The nails represent the bonds between atoms. For short cracks, there will not be much leverage applied to the most critical nail. For long cracks, the leverage will be greater, and the most critical nail will be under a great deal of tension.

Dr. Glass also pointed out how quantitative information can be obtained from indentation tests. She reviewed the Vicker’s Hardness test that uses a small diamond pyramid to indent and fracture a ceramic. By measuring the forces and indentation size, an estimate for hardness and fracture toughness can be made.
Dr. Glass introduced the concept of sub-critical crack growth. In normal conditions, a crack will need to have a critical stress applied before a crack will grow. Under some conditions, however, the crack will grow even when the stresses are well below the critical stress. Sub-critical crack growth is promiscuous, because a flaw grows slowly at stresses far below the expected failure load until the flaw is large enough to cause catastrophic failure.

Dr. Glass and her co-workers at Sandia National Laboratories have identified one of the causes of sub-critical crack growth in glass. By reacting with silica, water allows fracture to occur more easily. Water and other reagents allow brittle fracture to occur at lower stresses than normal (under fast loading). This is known as sub-critical crack growth and static crack growth. Water decreases strength and changes the crack velocity. In the absence of stressed bonds, the rate of reaction between silica and water is very low. Straining the silica bonds increases the activity of the bond and allows water to break the bonds much more easily. At a crack tip, the atomic bonds will be highly stressed. Dr. Glass showed graphs where the crack velocity was greatly affected by this phenomena.

Dr. Glass is considered an expert at identifying the cause of fractures. For this forensic science, she uses what is known as fractography, and she explained that fractography is the study of fracture features. Fractography is the examination of a surface created by fracture, and it involves the interpretation of fracture markings seen on these surfaces. Fractography can be used both qualitatively (the direction of fracture propagation) and quantitatively (the stress at the time of failure). Fractography can provide factual information for comparison with eyewitness accounts. Well-established fractographic techniques are available for determining crack propagation direction, failure origin location, estimating the failure stress, identifying what types of flaws are present, and for identifying local events that initiated failure, e.g., impact, thermal shock, or sub-critical crack growth.

One of the primary tools used in fractography is the texture on the surface of the crack. Dr. Glass gave examples of mirror, mist, and hackle fracture features. It appears that fracture begins as a smooth mirrored surface. As it propagates away from the point of origin, however, a mist-like surface will develop. After further propagation, the surfaces become uneven and form what is know as a hackle. The helical mark will always point back to the origin of the crack. Eventually, the crack may branch out and form multiple cracks.
Since the dimensions of fracture features are related to the stress, Dr. Glass can estimate the magnitude of the stress that caused the fracture. A small mirror feature indicates a very high stress at failure. A large mirror feature is generated by a small failure stress. If the ceramic body is small and the failure stress is extremely low, then the mirror may extend over the entire fracture surface and the mist and hackle regions will be absent.

The Relevance to Faceters:

Dr. Glass recommends that we try to minimize stress and look for it (birefringence). She said that we can use birefringence to study gemstones stress. Some stones, like tourmaline, can have tremendous residual stresses. Stress can sometimes be seen using a polariscope.

When Nancy Attaway asked whether sub-critical crack growth could be the cause of a Mexican opal cracking, Dr. Glass stated that there was a good chance that it could explain the demise of more than one gemstone.
In single crystals, crack propagation is strongly influenced by cleavage tendencies. Even so, the mirror, mist and hackle regions will still be present. Dr. Glass advised that we minimize damage and look for it. Since the failure strength depends on the size of a crack, we should try to minimize the damage caused by the sawing and grinding processes.

Dr. Glass advised us to fingerprint our gemstones. The small flaws visible under the microscope are often the flaws that will grow to become cracks. If your gem rough is filled with these small inclusions, then you should expect it to be weaker than gem rough that is clean and contains no inclusions. She also said to keep our gem rough dry. Gem rough that has been preformed may have cracks that can grow by sub-critical crack growth. Keeping gemstones dry will help reduce the sub-critical crack growth rate.

Any stones that have been polished but show sub-surface damage could also be susceptible to sub-critical crack growth. Making sure that all the sub-surface damage has been removed by proper working of the grinding grits will help to insure that your stones pass the test of time.

Bill Swanter pointed out that while hardness and fracture toughness are fundamental to gem behavior, he knew of no one who had made such measurements. A good catalog of the fracture toughness of different gemstones would help us communicate to the public which gemstones would be best for ring stones and which would be better set in earrings and pendants.

More detail of Dr. Glass’s research can be found in her chapter, Ceramics (Mechanical Properties), in the Kirk-Othmer Encyclopedia of Chemical Technology, Forth Ed. Volume No. 5, 1993 John Wile & Sons.