Dilatancy: Principles, Mechanisms, and Applications

Have you ever wondered why running across wet sand feels firm, while standing still causes you to sink? This curious phenomenon is known as dilatancy, a principle with profound implications across science and engineering. While the experience is common, bridging the gap between this simple observation and the sophisticated theories needed to design safe structures or advanced materials presents a significant challenge. This article provides a comprehensive journey into the world of dilatancy. We will first delve into the fundamental "Principles and Mechanisms," uncovering the physics behind shear-thickening and the elegant language of plasticity theory. Following this, the "Applications and Interdisciplinary Connections" chapter will reveal how dilatancy is a critical concept in fields ranging from civil engineering and materials science to the development of cutting-edge computational models. Prepare to see how a stroll on the beach connects to the frontiers of modern engineering.

Now that we have been introduced to the curious phenomenon of dilatancy, let us take a journey to understand its inner workings. Like any good exploration of nature, we will start with things we can see and feel, and then use those observations as a guide to build a more profound and powerful picture of the underlying physics. We will find that what happens when you run across wet sand is connected by a beautiful thread of logic to the sophisticated theories engineers use to design tunnels and foundations.

Imagine you are at the beach, near the water's edge where the sand is wet and packed. If you stand still, your feet slowly sink. But if you run quickly across that same patch of sand, it feels surprisingly firm, almost solid, under your feet. You have experienced dilatancy. Another famous example is a mixture of cornstarch and water, often called "oobleck." You can slowly run your fingers through it as if it were a thick liquid, but if you punch it, your fist stops dead as if hitting a wall.

These materials are part of a class known as ​​shear-thickening​​ fluids. To understand what this means, let's first think about an "ordinary" or ​​Newtonian​​ fluid, like water. If you stir water, the resistance you feel—the ​​shear stress​​, denoted by $\tau$—is directly proportional to how fast you stir—the ​​shear rate​​, denoted by $\dot{\gamma}$. Double the speed, you double the resistance.

Shear-thickening fluids are different. Their behavior can often be approximated by a simple but powerful relationship called the ​​power-law model​​:

$\tau = K (\dot{\gamma})^{n}$

Here, $K$ is a constant related to the fluid's consistency, and the crucial number is the exponent $n$, the ​​flow behavior index​​. For water, $n=1$. But for our wet sand or cornstarch mixture, $n$ is greater than 1 ($n \gt 1$). What does this mean? It means the resistance grows faster than the rate of shear. If you double the shear rate, you might quadruple the stress, or more!

Let’s imagine a highly simplified model of a foot hitting the sand, where the top plate of a device moves across the wet sand mixture. If the impact speed ($v_r$) is 60 times greater than the slow deformation speed ($v_s$), the resisting force isn't just 60 times larger. With a typical flow index of, say, $n=1.7$, the force ratio would be $60^{1.7}$, which is over 1000 times larger! This dramatic increase in resistance is precisely what holds you up when you run.

We can define an ​​apparent viscosity​​, $\mu_{\text{app}}$, as the ratio of stress to shear rate: $\mu_{\text{app}} = \tau / \dot{\gamma}$. For a power-law fluid, this becomes:

$\mu_{\text{app}} = K \dot{\gamma}^{n-1}$

Now it's perfectly clear! If $n=1$ (water), the viscosity is constant. But if $n \gt 1$ (shear-thickening), then $n-1$ is positive, and the apparent viscosity increases as the shear rate $\dot{\gamma}$ increases. The faster you try to deform it, the "thicker" it gets. The opposite behavior, where viscosity decreases with shear rate ($n \lt 1$), is called ​​shear-thinning​​ (or pseudoplastic) and is common in things like paint and ketchup.

So we have a mathematical description, but why does this happen? The secret lies in the microscopic structure. Wet sand, like dry sand, is a ​​granular material​​—a collection of tiny, solid grains packed together. Imagine a box filled with marbles. If they are packed as tightly as possible, you can't slide one layer of marbles past another without them getting in each other's way. To move, the grains have to push each other apart, forcing the entire collection to expand in volume.

This volume expansion under shear is the very definition of ​​dilatancy​​.

When you step slowly onto wet sand, the sand grains have time to jostle and rearrange, and the water between them can flow out of the way. Your foot sinks. But when you strike the sand quickly, you are trying to shear the grains past each other instantly. They are forced to expand, creating tiny empty spaces (voids) between them. The water can't flow into these new voids fast enough. This creates a local suction effect, pulling the grains tightly together and dramatically increasing the friction between them. For a brief moment, the water-and-sand mixture acts like a rigid solid, supporting your weight.

This intuitive idea of granular locking forms the basis for a more rigorous theory used in solid mechanics, known as ​​plasticity theory​​. This theory is essential for understanding the behavior of materials like soils, rocks, and even metals under high loads.

Imagine a "map" where the coordinates are not north and south, but different kinds of stress. For materials whose strength depends on confining pressure (like soil), it's useful to use two specific stress coordinates:

• ​​Mean Stress or Pressure ($p$)​​: This represents the average, all-around compressive stress, like the pressure a submarine feels deep in the ocean.
• ​​Deviatoric Stress ($q$)​​: This represents the shearing part of the stress that causes a material to change shape.

On this stress map, we can draw a line called the ​​yield surface​​, described by an equation $f(p, q) = 0$. As long as the stress state $(p,q)$ is inside this boundary, the material behaves elastically (it springs back). If the stress reaches the boundary, the material "yields" and begins to deform permanently, or plastically.

But in which direction does it deform? The theory of plasticity introduces another, profound concept: the ​​plastic potential​​, $g(p, q)$. The "direction" of the plastic strain increment ($d\boldsymbol{\varepsilon}^p$)—a tensor that describes the permanent deformation—is governed by the ​​flow rule​​:

$d\boldsymbol{\varepsilon}^p = d\lambda \, \frac{\partial g}{\partial \boldsymbol{\sigma}}$

Here, $d\lambda$ is a small positive number representing the amount of plastic flow, and $\boldsymbol{\sigma}$ is the stress tensor. The term $\frac{\partial g}{\partial \boldsymbol{\sigma}}$ is the gradient of the plastic potential. Geometrically, the gradient vector is always normal (perpendicular) to the surface of the potential function. So, the flow rule is a beautiful geometric statement: the direction of plastic deformation is normal to the surface of the plastic potential in stress space.

The part of this deformation that represents a volume change is given by the trace of the plastic strain tensor, $d\varepsilon_v^p = \mathrm{tr}(d\boldsymbol{\varepsilon}^p)$. A positive value means compaction; a negative value (in the standard geomechanics convention where compression is positive) means expansion—dilatancy!

The relationship between the yield surface ($f$) and the plastic potential ($g$) is at the heart of our story. There are two main possibilities.

Associated Flow: The Elegant Path

The simplest and most elegant assumption is that the plastic potential is the same as the yield function ($g=f$). This is called an ​​associated flow rule​​. It means the direction of plastic flow is normal to the yield surface itself.

Let's look at two examples. For metals, the yield strength is largely unaffected by hydrostatic pressure. Their yield surface, often described by the ​​von Mises criterion​​, is a vertical line on our $p-q$ map ($q = \text{constant}$). The normal vector to this surface is purely horizontal (it has no component in the pressure direction). This means the plastic volume change is zero! Metals are fundamentally non-dilatant; they deform at constant volume.

Now consider a granular material like soil, described by the ​​Mohr-Coulomb criterion​​. Its strength does depend on pressure—the higher the confinement, the stronger it is. Its yield surface is a sloped line on the $p-q$ map ($q = Mp + \text{constant}$). The normal vector to this sloped surface points both "out" (in the $q$ direction) and "up" (in the $-p$ direction). The upward component means that plastic shearing is inherently accompanied by a decrease in mean pressure, which corresponds to a volumetric expansion. An associated Mohr-Coulomb model naturally predicts dilatancy.

However, this elegant model has a problem: for real soils, it often predicts too much dilatancy. The predicted volume expansion is often significantly larger than what is measured in the laboratory.

Non-Associated Flow: The Pragmatic Path

To fix this, engineers take a more pragmatic approach. They decouple the rule for strength from the rule for flow by assuming the plastic potential is different from the yield function ($g \neq f$). This is a ​​non-associated flow rule​​.

We keep the Mohr-Coulomb yield surface ($f$) because it's good at predicting the material's strength. But for the flow, we introduce a new plastic potential, $g$, with a different slope:

$g(p, q) = q - M_p p$

The slope of this potential, $M_p$, is not determined by the friction angle $\varphi$ (which governs strength), but by a new, independent parameter called the ​​dilatancy angle​​, $\psi$. This angle provides a "dial" that allows us to control the amount of predicted volume change. If we set $\psi=0$, the plastic potential becomes $g=q$, which is independent of pressure. Just like in the case of metals, this leads to zero plastic volume change. By choosing a value of $\psi$ between 0 and $\varphi$, engineers can accurately model both the strength and the volume change behavior of the soil.

This seems like a perfect solution, but in physics, there is rarely a free lunch. The beautiful simplicity of the associated flow rule is not just a mathematical convenience. It can be derived from a deep physical principle known as ​​Drucker's stability postulate​​, which ensures that the material is stable in a certain sense.

When we use a non-associated flow rule, we are stepping outside this provably stable framework. It turns out that for such a model, it is possible for the material to perform negative work during certain infinitesimal stress changes, even while plastic deformation is occurring. This hints at the potential for material instabilities, like the formation of shear bands, which are not captured by the simpler associated theory.

And so, our journey from the beach ends with a profound lesson about science. We start with a simple observation, build an intuitive model, and then refine it into a rigorous mathematical theory. We find that the most elegant theories (associated flow) are beautiful and have deep foundations, but sometimes we must sacrifice some of that elegance for pragmatism (non-associated flow) to better match the complex reality of nature. The study of dilatancy shows us the ongoing, dynamic dance between physical intuition, mathematical beauty, and experimental reality.

In our exploration so far, we have unraveled the core principle of dilatancy: the curious tendency of certain materials to expand in volume when they are sheared or compressed. Like many profound ideas in physics, this one starts simply—think of the wet sand at the beach firming up under your feet—but its consequences ripple outwards, connecting seemingly disparate fields and forcing us to build more subtle and powerful theories. Now, we will venture beyond the principles and witness dilatancy at work, to see how this single concept shapes our understanding of the ground we build on, the materials we design, and even the virtual worlds inside our computers.

Nowhere is dilatancy more central than in geomechanics, the science of how soils, rocks, and other granular materials behave. For a civil engineer designing a foundation, a dam, or a tunnel, the question "Will the ground give way?" is not an academic one. To answer it, they rely on mathematical descriptions of the material’s strength, known as constitutive models.

At first glance, the task seems straightforward. We can create a "rule book" for the material, a yield criterion that tells us the combination of pressure and shear stress it can withstand before failing. Classic models like the Mohr-Coulomb or Drucker-Prager criteria do just this. They are what we might call phenomenological closures: they don’t describe every grain of sand, but they beautifully capture the macroscopic reality that squeezing a granular material (increasing its "confining pressure") makes it stronger and more resistant to shear. These models are not fundamental laws of nature, but wonderfully effective approximations for the quasi-static behavior of materials from sand to concrete under moderate pressure.

Here, however, we encounter a fascinating scientific puzzle. The most straightforward assumption we can make about what happens after the material starts to yield is that the direction of plastic flow, or deformation, follows the same rule as the yielding itself. This is called an ​​associated flow rule​​. It’s an elegant idea, suggesting that the material deforms in the most "natural" way in response to the forces causing it to yield. But when we test this beautiful idea against reality, it breaks down.

Consider a glassy polymer or a porous metal, whose yielding is also sensitive to pressure. If we calibrate an associated Drucker-Prager model to match their measured strength in tension versus compression, the model makes a startling prediction: when the material yields, it should expand dramatically. For a typical polymer, the model might predict a plastic volume increase that is 75% of the size of the plastic shear deformation. Yet, when we go into the laboratory and carefully measure the volume change, we find it to be tiny—perhaps only 5% of the shear deformation. The theory, in its simplest form, is wrong by more than an order of magnitude!

This is a wonderful moment. Our intuition has led us astray, and we are forced to be more clever. The resolution lies in a brilliant concept: the ​​non-associated flow rule​​. We abandon the idea that the rule for strength (the yield function, $f$) must be the same as the rule for motion (the plastic potential, $g$). We give the material two distinct "personalities": one that governs when it will yield, and another that dictates the direction in which it will flow once it has yielded.

This decoupling provides an independent "knob" to control the predicted dilatancy. The friction, which determines the strength, is governed by a parameter like $\alpha$, while the dilatancy, or plastic volume change, is governed by a separate parameter, often called $\beta$ or a dilatancy angle $\psi$. Now, we can go back to the lab, measure the small amount of actual dilatancy, and tune our "dilatancy knob" to match the experiment perfectly, without compromising the model's ability to predict the material’s strength. This "non-associated" approach is the cornerstone of modern modeling for almost all pressure-sensitive materials.

Armed with this more sophisticated non-associated model, we can now predict an even wider array of phenomena. One of the most visually striking is the formation of ​​shear bands​​. When a block of dense sand or stiff clay is compressed, it doesn’t usually fail by deforming uniformly. Instead, failure concentrates into narrow, well-defined bands. The angle of these bands is not random. It is a physical manifestation of the underlying constitutive law.

In a beautiful synthesis of forces and kinematics, the orientation of a shear band emerges as a compromise between two competing influences: the material's internal friction (governed by the yield function) and its dilatancy (governed by the plastic potential). Increasing the dilatancy angle $\psi$—that is, making the material push outwards more strongly as it shears—literally rotates the failure plane, making the shear band steeper relative to the direction of compression. This is a profound connection: a microscopic tendency to expand dictates the macroscopic geometry of collapse.

But there is a deeper, more dramatic consequence of non-associated flow. The very fact that the material flows in a direction that is not "normal" to the yield surface introduces a kind of internal conflict. This lack of "association" makes the underlying mathematical equations that govern the material's response non-symmetric, and this asymmetry can lead to ​​instability​​. For a given amount of friction, there exists a critical value of dilatancy. If the material's dilatancy parameter exceeds this critical threshold, the governing equations can "lose ellipticity"—a mathematical term for a catastrophic loss of stability. At this point, the material becomes exquisitely sensitive to tiny imperfections and can fail abruptly in a highly localized manner. This theoretical insight is crucial for understanding and predicting phenomena like landslides and the sudden collapse of over-stressed ground.

The principles we've developed for soils and rocks have a surprisingly universal reach, appearing in many other classes of materials.

​​Ductile Fracture in Metals​​: We tend to think of metals as being incompressible when they deform plastically. For a fully dense metal, this is an excellent approximation, described by classical theories like von Mises plasticity. But many advanced materials, such as those made by powder metallurgy, contain microscopic voids. When such a porous metal is stretched, it deforms, and these tiny voids grow and link up. This growth of internal voids is nothing but a plastic increase in volume—it is dilatancy at the micro-scale! This process of void growth is the fundamental mechanism behind ductile fracture. Models like the Gurson-Tvergaard-Needleman (GTN) criterion explicitly incorporate pressure and void volume fraction into the plastic potential, thereby capturing the dilatancy that is absent in classical metal plasticity.

​​Shear-Thickening Fluids​​: The term dilatancy is also used in fluid mechanics to describe a related, but distinct, phenomenon: shear thickening. Consider a dense suspension of cornstarch in water. At rest, it flows like a liquid. But if you try to stir it quickly or punch it, it becomes almost solid. This is dilatancy. The rapid shearing forces the particles to jam together, increasing the resistance to flow (viscosity). The application of this idea to a power-law fluid flowing in a pipe reveals that the velocity profile becomes much "blunter" than the classic parabolic profile of a simple Newtonian fluid like water. This is because the fluid near the walls, where the shear rate is highest, becomes more viscous and slows down, forcing more of the flow to occur in the center. This strange behavior is being harnessed to create "smart fluids" for advanced shock absorbers and even liquid body armor that stiffens on impact.

In the 21st century, much of engineering design is performed not in a physical lab but in a virtual one, using powerful computer simulations like the Finite Element Method (FEM). But how do we teach a computer about the subtleties of non-associated dilatancy? It turns out to be a tremendous challenge.

The very nature of dilatancy—a coupling between volumetric change and shear deformation—can cause problems for standard low-order numerical methods. In situations of near-incompressibility, these simple methods can suffer from "volumetric locking," where the simulated material becomes artificially stiff and produces nonsensical results. While dilatancy introduces some volumetric compliance, which can alleviate locking, the non-associated nature of the flow creates a new headache: the underlying system of equations becomes non-symmetric, complicating the solution process and potentially affecting stability.

This challenge has spurred innovation. To accurately capture non-associated plastic flow, computational scientists have developed sophisticated "mixed finite element" methods. These methods introduce additional variables, like pressure or even the plastic multiplier itself, as independent unknowns in the simulation. This allows them to more robustly handle the volumetric constraints and control the plastic flow. Specialized stabilization techniques are also required, with parameters that must be carefully scaled not with the elastic properties of the material, but with the algorithmic tangent stiffness, which properly reflects the plastic dilatancy. In this way, a deep physical principle—dilatancy—directly drives progress in the seemingly distant field of computational mathematics.

From the shifting sands to the subtle dance of atoms in a yielding polymer, and from the geometry of a landslide to the algorithms in a supercomputer, the principle of dilatancy reveals the beautiful and unexpected unity of the physical world. It is a perfect example of how grappling with a simple observation forces us to refine our theories, discover new connections, and ultimately build a deeper and more powerful understanding of nature.