A study led by researchers from the Hebrew University of Jerusalem shows that a sudden fracture does not necessarily start with a crack that erupts immediately. First, a tiny fracture zone is formed that can progress very slowly; only after it propagates through the thickness of the material does the crack begin to accelerate dramatically.
When a glass screen cracks, a glass breaks, or a mechanical component fails, the event seems almost instantaneous. But new research shows that behind the rapid moment of breakage may be a much slower process, almost imperceptible to the eye.
Researchers Yuval Paz and Prof. J. Feinberg from the Institute of PharmacyPhysics At the Hebrew University, together with Meng Wang from the Beijing Institute of Technology and Mokhtar Ade-Badia from CNRS and ENS in Lyon, they traced the beginning of the formation of Cracks In three-dimensional materials.
The study, published inPhysical Review Letters, offers an extension to the classical theory of fracture mechanics and connects three stages: crack formation, its slow creep, and finally rapid fracture.
The missing step in classical theory
Linear elastic fracture mechanics, LEFM, describes well what happens to a crack that has already begun to move rapidly in a material.
One of the cornerstones of the field is an idea related to the work of Alan Griffith from the early 20th century: a crack has a critical size above which it may become unstable and continue to spread.
But the question of how the crack gets to this state in the first place remains more difficult.
Previous studies by Feinberg's group also found that the area where the fault begins is more complex than the models predicted. In 2021, we reported in Scientist on Research reveals unexpected behavior in the microscopic environment of the crack tip.
The new study focuses on an even earlier stage: the moment when the fractured region is born.
Not a line – but a tiny spot
The experiment revealed that the fracture does not necessarily begin as the linear crack we are used to imagining.
When the stress in a material reaches a critical value, it can form within it. Very small two-dimensional broken areaIn the researchers' terms, this is the nucleation stage – the initial formation of the fracture nucleus.
This region can appear smaller than the critical length that would be expected according to Griffith's classic description.
Instead of breaking out immediately, it begins to spread slowly.
Researchers call this stage CreepIts speed is orders of magnitude lower than the speed of Rayleigh waves in the material – the characteristic speed used as a scale forfractions Dynamic.
That is, for a significant part of the process, the fracture has already begun – but it still does not look like the rapid fracture we see at the end.
The thickness of the material is key.
The tipping point comes when the cracked area grows until its dimensions become similar to the thickness of the slab in which it develops.
Then a fundamental geometric change occurs.
The two-dimensional fractured region becomes a crack that actually goes through the thickness of the sample. From that moment on, the description becomes much more similar to the linear crack familiar from classical fracture mechanics.
And only then does the acceleration come.
The crack goes from slow creep to very rapid dynamic propagation through the material.
In other words, what appears from the outside to be a sudden moment of "breaking" is actually the final stage in a process that began earlier.
Griffith is not being abolished – it is being expanded.
The findings do not invalidate classical fracture mechanics.
On the contrary: after the crack crosses the thickness of the material and begins to move rapidly, the classical description continues to work well.
The innovation is the connection of the early stage to this picture.
The new model extends LEFM to account for the two-dimensional geometry of the cracked region and the final thickness of the material. The researchers report that the extended framework quantitatively describes both the crack initiation stage, creep, and the transition to dynamic fracture. ([APS Journals][1])
The fact that the thickness of the material itself participates in determining the transition point is particularly important, because real materials are not mathematical lines or surfaces but three-dimensional bodies.
Already in 2015, we described in a scientific paper how Experiments with gel sheets in Feinberg's lab make it possible to slow the propagation of cracks and study their physics.
Is it possible to identify a fracture before it collapses?
If a significant portion of the fracture process does indeed occur in a slow creep phase, this raises an interesting possibility: perhaps in some cases it will be possible to look for early signs of failure even before the crack reaches the dangerous acceleration phase.
This does not mean that the research already provides a practical warning system for bridges, aircraft or buildings. There is a big difference between a controlled experiment in a laboratory and a complex material in the real world.
But he changes the question that perhaps needs to be asked.
Instead of looking only for large cracks that are already approaching failure, it may also be important to understand how the tiny, slow-moving areas that precede them form and develop.
And what about earthquakes?
Feinberg's group has also been studying the physics of friction and the onset of slip events for years.
As early as 2010, we reported on A study from the Hebrew University that used laboratory "earthquakes" to understand how sliding between surfaces begins.
The current study deals with a tension fault and not an actual earthquake, so it should not be concluded that a mechanism for predicting Earthquakes.
However, the researchers note that there are physical similarities between crack formation and fracture and sliding processes along frictional interfaces. Therefore, the new framework may also help to understand how slip events initiate and develop.
Previous research by the group specifically addressedThe "nucleation point" of friction events and its possible connection to earthquakes.
The sudden breakup starts slowly.
The main significance of the study is a change in the intuitive picture of the fracture.
We see the moment when the material collapses – but this is not necessarily the moment when the failure process begins.
Before the rapid fracture, a tiny patch of material that has already fractured can form. It expands slowly, changing its geometry, and only when it crosses the thickness of the material does the violent and rapid phase begin.
The last few milliseconds are perhaps more dramatic.
But physically, the story began first.
Questions and Answers
Does a crack immediately start moving at high speed? Not necessarily. The study found tiny fractured areas that can go through a slow creep phase before transitioning to rapid dynamic fracture.
What is the length of a grip? This is a critical scale in classical fracture mechanics, related to the size at which a crack may become unstable and propagate. The new study shows that nucleation of a fracture zone can begin even below this length when the stress reaches an appropriate threshold.
What causes the crack to start accelerating? In the experiments, a key relationship was found to the material thickness. When the fractured area grows until it crosses the specimen thickness, a geometric transition occurs, after which the crack can transition to rapid propagation.
Does the research make it possible to predict earthquakes? No. It deals with tensile fracture under laboratory conditions. However, the researchers point to physical connections to sliding and friction processes, so the findings may help understand the onset of fracture events in other systems as well.
Scientific source: the article Nucleation, Creep, and Dynamic Tensile Fracture Are Described by Extended LEFM, by Yuval Paz, Meng Wang, Mokhtar Adda-Bedia, and Jay Fineberg, published inPhysical Review Letters, Volume 137, Article 106102, September 4, 2026.
More on the subject on the science website
- Surprising results in research on crack behavior
- Laboratory experiments allow us to study how cracks propagate in materials
- Friction as a key to understanding earthquakes
- The breaking point: how rapid cracks develop
For the original publication: Opening the original publication
