Glass Fracture Mechanics: Griffith Theory and Stress Corrosion
The theoretical strength of glass materials is significantly higher than what is measured in real-world samples. This discrepancy is attributed to the presence of defects, primarily tiny cracks, within the material, which compromise its fracture strength. This phenomenon is explained by Griffith's theory.
Griffith's Theory and Stress Concentration
Griffith's theory posits that brittle materials like glass contain small cracks. A crack embedded within the bulk of a material is typically labeled with a length of 2a. However, for most glasses, defects originate from the surface, acting as the weak link. In this case, the effective crack length is a, meaning a surface crack of length a causes the same strength decrease as a bulk crack of length 2a.
When a macroscopic stress (labeled sigma infinity) is applied to a material with a crack, the stress field distribution shows a significant concentration of stress near the crack tip. This effective stress at the crack tip is much higher than the applied stress. Quantitatively, this is described by a stress concentration factor, which is proportional to a / a0, where a0 is the atomic spacing and a is the half-length of a bulk crack or the length of a surface crack. Since a (microns) is many orders of magnitude larger than a0 (angstroms), this factor leads to substantial stress concentration, making the crack tip the weakest link where failure preferentially initiates and propagates.
The fracture strength (sigma f) of a flawed material is given by 1/2 * sqrt(gamma * E / a), where gamma is the surface energy, E is Young's modulus, and a is the crack length. This formula shows an inverse relationship between fracture strength and crack length. For typical silica glass, a strength of 110 megapascals corresponds to a crack length of about 5 microns, demonstrating how even tiny defects can account for the large gap between theoretical and measured strength.
Experimental Validation and Water Condensation
Griffith's theory has been experimentally validated through techniques like atomic force microscopy (AFM), which visualizes surface morphology near cracks. AFM images show crack propagation and the resulting displacement fields, which align well with theoretical predictions, with variations on the order of 0.2 nanometers (about two angstroms), comparable to atomic spacing. This confirms the theory's accuracy in predicting stress/strain distribution and displacement fields around cracks.
Beyond stress amplification, cracks in glass materials also facilitate water condensation. Glass is typically hydrophilic, meaning water tends to spread on its surface. Within a crack, water vapor condenses, maximizing the interface area between water and glass while minimizing the free interface with air. This water can be problematic because it can break Si-O bonds in the glass network, forming Si-OH groups and accelerating the disintegration of the glass. In alkaline silicate glasses (like soda-lime glass), alkaline ions can leach into the aqueous solution, forming hydroxides that make the solution basic, further accelerating chemical corrosion. This process is known as stress corrosion.
Stress corrosion can lead to subcritical crack growth, where cracks slowly lengthen even if they are below the critical length predicted by Griffith's theory for catastrophic failure.
Stress Intensity Factor and Fracture Toughness
Because fracture strength depends on crack length (a sample-specific parameter), it is not a material constant. To characterize a material's inherent resistance to fracture, the stress intensity factor (Ki) is introduced. Ki = sigma_infinity * sqrt(pi * a). When Ki exceeds a critical value, K1c (the critical stress intensity factor), the material fails.
K1c is a material constant, depending only on surface energy and Young's modulus, not on crack length. This makes K1c a reliable measure of a material's ability to resist fracture and crack propagation. An equivalent energy-based quantity is the strain energy release rate (G1), where G1 = K1^2 / E. Failure occurs when G1 exceeds G1c.
Limitations of Griffith's Theory and Plastic Deformation
Griffith's theory assumes no plastic deformation during crack opening, which is a better assumption for brittle materials like glass but not entirely perfect. Even oxide glasses exhibit some plastic deformation, typically in the form of densification (increased density due to compression). This means Griffith's theory doesn't perfectly apply even to oxide glasses.
Amorphous metals, due to the non-directional nature of metallic bonds, can accommodate significantly larger plastic deformation compared to oxide glasses. This is evident in their fracture behavior.
Fracture Behavior of Amorphous Metals
Amorphous metals exhibit unique fracture surfaces. Unlike the smooth surfaces of fractured glass, as-cast amorphous metals show surface corrugations or "ridges." These ridges are a signature of local plastic deformation, often forming at a 45-degree angle where shear stress is maximized.
The plasticity of amorphous metals can be tuned. Annealing above the glass transition temperature can lead to thermal embrittlement by decreasing free volume, which reduces the capacity for plastic deformation. Annealed amorphous metals exhibit smoother, flatter fracture surfaces, resembling those of brittle materials. This ability to tune plasticity allows amorphous metals to exhibit different fracture behaviors compared to more brittle oxide glasses.
Amorphous metals can achieve high fracture toughness, a measure of energy absorption before fracture. While oxide glasses have high strength, they have low toughness because they absorb little energy. Bulk metallic glasses, however, can possess both high strength and high toughness due to their capacity for local plastic deformation, which absorbs significant energy and resists crack propagation. This makes them desirable for engineering applications, often residing in the top-right corner of yield strength vs. fracture toughness plots.
Crack Propagation Speed
The speed of crack propagation in glass is typically about one-third to one-half the speed of sound in glass. While a perfectly brittle material would propagate cracks at the speed of sound, the slight plastic deformation (densification) in glass slows down the crack, as energy is absorbed by this deformation.
Key takeaways regarding crack propagation: 1. Surface defects are critical: Glass typically breaks from surface defects, making surface quality crucial for strength. 2. Glass is not perfectly brittle: It can accommodate some plastic deformation, which influences crack speed.
Fractography: The Forensic Science of Broken Glass
Fractography is the study of fractured surfaces to determine the cause and mechanism of failure. This is used in forensic science (e.g., crime scene investigation) and was historically important for shaping stone tools.
When glass fractures, it often exhibits a "conchoidal fracture" pattern, resembling a seashell. This pattern is unique to brittle materials.
Features of a Fractured Glass Surface
A microscopic examination of a broken glass surface reveals distinct regions:
- Mirror Region: A flat, smooth, reflective region near the origin of the defect. This is where the crack initiates and propagates at a relatively low, stable velocity, leaving a well-defined surface.
- Mist Region: As the crack lengthens, stress at the crack tip increases, making propagation less stable. The crack begins to micro-branch, leading to surface roughness and a foggy appearance. This marks the onset of unstable propagation.
- Hackle Region: Beyond the mist region, the crack propagates with a very high energy release rate, creating significant surface roughness in the form of ridges that radiate from the origin of failure. These hackles are formed as the glass creates more fracture surfaces to accommodate the high energy release.
- Wallner Lines: These are tiny ridges that form the characteristic seashell pattern. They are caused by the interaction of the propagating crack tip with elastic waves (sound waves) that travel faster through the material. These elastic waves can be reflected or scattered by non-uniformities, creating secondary waves that interact with the crack tip, leading to periodic waviness and the formation of Wallner lines. The spacing between Wallner lines can indicate crack opening velocity; faster velocities lead to smaller spacing.
The 4R Rule for Impact Direction
For flat pane glass fractured by impact, the "4R rule" helps determine the side of impact: "Ridge lines on radial fractures are at right angles to the rear." Wallner lines on the rear surface (opposite the impact) will be perpendicular to the surface, while on the front surface (side of impact), they will be parallel.
Fractography of Tempered Glass
Tempered glass, which has surface compressive stress, exhibits a different fracture pattern. Defects usually initiate in the interior, where tensile stress is high. The mist and hackle regions are found in the center, with two symmetric Wallner line patterns propagating outwards. Near the surface, horizontal lines indicate the high compressive stress. Tempered glass breakage often shows primary radial cracks and secondary branching cracks, with a "butterfly-like" pattern at the center of the defect.
Static Fatigue and Stress Corrosion
Griffith's theory does not account for time-dependent failure. However, glass can fail over time even if the applied stress is below its immediate fracture strength. This phenomenon, called static fatigue, is due to stress corrosion caused by water vapor condensing at crack tips.
A universal static fatigue curve can be generated by normalizing time to failure (t) by the time to failure at half the fracture stress (t0.5). This curve shows that time to failure decreases logarithmically with increasing applied stress.
Microscopic Mechanism of Static Fatigue
The crack velocity (V) as a function of the stress intensity factor (K1) can be divided into three regions:
- Region 1 (Linear Increase): Crack velocity increases exponentially with
K1. Higher relative humidity significantly accelerates this process, confirming the role of water vapor. - Region 2 (Plateau): Crack growth rate becomes independent of
K1(and crack length). This is a mass transport-limited region, where the rate is controlled by how quickly water vapor can diffuse to the crack tip. - Region 3 (Expedited Growth): For sufficiently long cracks, crack velocity rapidly increases, leading to catastrophic failure as the applied stress reaches the fracture stress.
This demonstrates that glass can fail due to stress corrosion even in very dry conditions, as glass surfaces are highly hydrophilic.
Characterization of Fracture Behavior
Measuring fracture toughness often requires specialized sample geometries with predefined crack lengths, typically generated using static fatigue and monitored microscopically. This is a laborious process, often adhering to ASTM standards.
Simpler methods, like indentation, can infer fracture toughness. A stylus (e.g., Vickers indenter) is pressed into the material. Cracks typically form during the unloading cycle, not during loading. This is because compressive stress during loading causes elastic and plastic deformation. Upon unloading, the elastic regions try to restore their shape, while plastic regions retain deformation, creating tensile stresses that lead to cracking. The size and shape of these cracks can be used to estimate fracture toughness, though results may not always correlate perfectly with standard ASTM measurements.
Fracture Statistics: Weibull Distribution
Fracture toughness measurements are inherently statistical due to the random distribution of defect sizes. The ultimate strength of a glass sample is determined by its longest crack (weakest link). Therefore, a large number of samples must be tested to understand the statistical distribution of fracture toughness.
The Weibull distribution is commonly used to describe the cumulative failure probability (F), which is the fraction of samples that fracture at stresses below a given sigma. An ideal, defect-free material would show a step function, but real glass exhibits a sloped function. The Weibull modulus (m) characterizes the steepness of this distribution; a larger m indicates a tighter distribution and better quality control.
By plotting ln(ln(1/(1-F))) against ln(sigma), a linear relationship is often observed. The slope of this line gives m, and the x-intercept corresponds to the average failure stress (sigma0). Deviations from linearity, such as inflection points, can indicate bimodal defect distributions, suggesting inconsistencies in sample processing.
Strategies for Strengthening Glass Materials
The primary goal of glass strengthening is to increase fracture strength (sigma f). This can be achieved through three main strategies:
Increase
K1c(Improve inherent resistance to crack propagation):- Crystallization (Glass-ceramics): Forming crystalline phases within the glass can increase mechanical strength. Glass-ceramics can be denser and stronger than conventionally sintered ceramics. Surface crystallization can also induce compressive stress due to differential thermal contraction between the crystalline surface and the glass interior. However, crystallization can reduce transparency.
- Composites: Combining glass with other materials, like polymers, can leverage the high strength of glass and the ductility of polymers for improved durability.
- Engineered Glass Structures: Designing specific glass structures for enhanced strength.
Reduce Defect Density or Severity:
- Fire Polishing: Heating the glass surface above its glass transition temperature allows viscous flow to fill surface cracks and create a smooth, crack-free surface under surface tension (e.g., rounding edges of test tubes).
- Surface Etching: Removing the surface layer to eliminate surface defects.
- Surface Coating: Applying protective coatings. For example, tin oxide coatings increase abrasion resistance, and polymer coatings act as lubricants to prevent defect generation from wear and tear.
Introduce Surface Compressive Stress (Prevent crack propagation):
- Thermal Tempering: Rapid cooling of glass creates a surface layer under compression and an interior under tension. This compressive layer resists crack opening.
- Ion Exchange (Chemical Tempering): Exchanging smaller ions in the glass surface with larger ions from a molten salt bath at temperatures below the glass transition temperature. The larger ions pack more tightly, inducing compressive stress on the surface.
These strategies are crucial for applications requiring high-strength glass, such as solar panels, display screens, and even bulletproof windows in defense applications.
Takeaways
- Griffith's theory explains why real glass strength is far below theoretical values, attributing the reduction to surface cracks that concentrate stress at the tip.
- The fracture strength of a flawed glass varies inversely with the square root of crack length, so a 5‑micron surface crack can lower strength to about 110 MPa.
- Water vapor condenses in cracks, causing stress corrosion and subcritical crack growth, which allows glass to fail over time even under stresses below its nominal strength.
- The stress intensity factor K_I and the critical value K_Ic provide material‑specific measures of fracture resistance, while the strain energy release rate G relates to K through elastic modulus.
- Strengthening methods such as thermal tempering, ion‑exchange, fire polishing, and surface coatings increase K_Ic or reduce defect severity, enabling high‑strength applications like displays and bullet‑proof windows.
Frequently Asked Questions
Why does a surface crack of length a reduce glass strength as much as a bulk crack of length 2a?
Because Griffith's theory treats a surface crack as having an effective half‑length a, which produces the same stress intensity at the tip as a through‑thickness crack of length 2a; the stress concentration factor depends on the crack length relative to atomic spacing, making the two geometries equivalent in their weakening effect.
How does water condensation inside glass cracks cause stress corrosion and subcritical crack growth?
Water vapor condenses in the confined crack tip region, breaking Si‑O bonds to form Si‑OH groups; the resulting hydroxide solution attacks the glass network and, in alkaline glasses, leaches ions that raise pH, accelerating chemical attack. This stress‑corrosion mechanism allows cracks to grow slowly under stresses below the Griffith critical value, eventually leading to failure.
Who is MIT OpenCourseWare on YouTube?
MIT OpenCourseWare is a YouTube channel that publishes videos on a range of topics. Browse more summaries from this channel below.
Does this page include the full transcript of the video?
Yes, the full transcript for this video is available on this page. Click 'Show transcript' in the sidebar to read it.
Helpful resources related to this video
If you want to practice or explore the concepts discussed in the video, these commonly used tools may help.
Links may be affiliate links. We only include resources that are genuinely relevant to the topic.