Amontons's Law of Friction

Friction is one of the most fundamental and ubiquitous forces in our daily lives, yet its governing principles are often simplified into a few simple rules of thumb. At the heart of this classical understanding is Amontons's Law, a set of empirical observations made over three centuries ago that seem to elegantly describe how objects slide. However, this simplicity masks deep physical complexity. The central knowledge gap this article addresses is the fascinating paradox: why do these straightforward laws work so reliably on a macroscopic scale when they fundamentally fail at the microscopic level?

This article journeys from the classical world of sliding blocks to the atomic realm to uncover the truth behind friction. In the "Principles and Mechanisms" chapter, we will dissect the law itself, exploring the critical distinction between apparent and real contact area, the role of material plasticity and elasticity, and how statistical mechanics rescues the law from its nanoscale failures. Following that, the "Applications and Interdisciplinary Connections" chapter will demonstrate how modern tools like the Atomic Force Microscope are used to test these theories, reveal the stubborn emergence of Amontons's law in diverse physical systems, and even connect this centuries-old problem to the frontiers of machine learning. Prepare to discover that one of physics' simplest laws is a gateway to some of its most profound and complex ideas.

It’s one of the first things we learn about the physical world, probably through the frustrating experience of trying to push a heavy box across the floor. To get it moving, you have to push hard. If a friend sits on the box, you have to push even harder. The force needed to overcome friction seems to depend on how heavy the object is—its weight, or more precisely, the ​​normal load​​ pressing it against the floor. Sometime in the 17th century, the French physicist Guillaume Amontons studied this systematically and came up with two simple, almost childlike, laws. First, the friction force is directly proportional to the normal load. Double the load, and you double the friction. Second, and much more mysteriously, the friction force does not depend on the apparent area of contact.

Think about that second law for a moment. It’s deeply counter-intuitive. If you take a brick and lay it on its widest face, it has a large contact area. If you stand it on its end, the area is much smaller. Common sense might suggest that the wider brick, with more of its surface rubbing against the floor, should have more friction. But Amontons's law says no—the friction is exactly the same in both cases! For nearly three centuries, this remained a profound puzzle. How could it be? The answer, when it came, opened up a whole new way of looking at the surfaces we thought we knew.

The breakthrough came from a brilliantly simple idea, championed by the physicists Philip Bowden and David Tabor in the mid-20th century. They realized that surfaces, even those that look perfectly smooth and polished to the naked eye, are anything but. On a microscopic scale, they are rugged, mountainous landscapes. When you place one surface on another, they don't make contact everywhere. They touch only at the tips of the very highest "mountains," or ​​asperities​​, as scientists call them.

This means we have to distinguish between the ​​apparent area of contact​​ (the macroscopic footprint of the object, like the bottom of the brick) and the ​​real area of contact​​, $A_{\text{real}}$, which is the sum of the tiny, scattered areas where the asperity tips are actually touching. This real area is typically a minuscule fraction of the apparent area.

Friction, Bowden and Tabor argued, is an interfacial phenomenon. It happens where atoms are actually interacting and resisting being slid past one another. So, the friction force, $F_f$, shouldn't be proportional to the apparent area, but to the real area of contact. We can write this as a beautifully simple relationship:

Here, $\tau$ (the Greek letter tau) is the ​​interfacial shear strength​​, a property of the materials that represents the force required to shear a unit area of the interface. This one equation completely changes the game. If Amontons’s laws are true, it must mean that the real area of contact, $A_{\text{real}}$, possesses two remarkable properties: it must be directly proportional to the normal load, $L$, and it must be independent of the apparent area. The puzzle of Amontons's laws has been transformed into a puzzle about the mechanics of microscopic mountain ranges.

So, how does $A_{\text{real}}$ become proportional to the load? Imagine those microscopic mountain peaks. When you put a load on the object, the entire load is supported by these few tiny points. The pressure on them is immense—so immense, in fact, that they often can't support it elastically. They get crushed. They undergo ​​plastic deformation​​, like a thumbtack being pressed into a soft piece of wood.

A material has a limit to the pressure it can withstand before it deforms plastically. This limit is called its ​​hardness​​, denoted by $H$. So, as you increase the load, more asperities are crushed, or existing ones are crushed further, until the total real contact area is large enough to support the load. The average pressure across all the contacting asperities will be roughly equal to the hardness. We can write this as $L \approx H \times A_{\text{real}}$.

If we rearrange this, we find something wonderful:

The real area of contact is directly proportional to the load! The hardness, $H$, is the constant of proportionality. When we substitute this back into our friction equation, we get:

This is precisely Amontons's first law, $F_f = \mu L$. We have even found a physical meaning for the famous coefficient of friction, $\mu$: it’s the ratio of the interface's shear strength to the softer material's hardness, $\mu \approx \tau/H$. This model also elegantly explains the second law. The real area of contact depends only on the load you apply and the material's hardness, not on the apparent area over which the asperities are spread. Whether the brick is lying flat or on its end, the total area of the crushed asperity tips needed to support its weight is the same. It's a stunningly effective explanation.

For a long time, the plastic deformation model was the standard explanation. But what happens if the load is very light, or the surfaces are very hard, so that the asperities don't get crushed? What if they just deform elastically, like tiny rubber balls, and spring back when the load is removed?

This is the world of ​​nanotribology​​, the study of friction at the atomic and nanoscale, often explored with tools like the Atomic Force Microscope (AFM). In an AFM, a single, incredibly sharp tip—a single asperity—is dragged across a surface. If we treat this as a single spherical asperity pressing on a flat plane, the rules are governed by ​​Hertzian contact mechanics​​.

Hertz's theory tells us that for an elastic contact, the real contact area does not grow linearly with load. Instead, it follows a sublinear power law:

If friction is still proportional to this real area, then the friction force itself must scale in the same way: $F_f \propto L^{2/3}$. This is a profound result. It means that for a single elastic contact, Amontons's first law is fundamentally wrong! The friction force is not proportional to the load. If you were to calculate an "apparent" friction coefficient, you'd find it decreases with load ($\mu = F_f/L \propto L^{-1/3}$), a clear violation of the "constant" coefficient we're used to. This discovery showed that the simple laws we observe in our macroscopic world don't necessarily hold when you zoom in far enough.

We seem to have a paradox. If the fundamental building block of contact—a single elastic asperity—doesn't obey Amontons's law, how can a large surface, which is just a collection of many such asperities, obey it so well?

The answer lies in the magic of statistics, an insight first articulated by J.F. Archard and later formalized by J.A. Greenwood and J.B.P. Williamson. When you press two rough surfaces together, it's not just that the existing contact points grow larger (the $L^{2/3}$ effect). As you push harder, the surfaces move closer together, and a whole new population of previously untouched, shorter asperities are recruited into making contact.

The total real contact area is the product of the number of contacts, $N_c$, and the average area of each contact, $\langle A_i \rangle$. While the average area of individual elastic contacts grows sub-linearly with their individual load, it turns out that for many common types of random surfaces, the number of new contacts, $N_c$, increases almost directly in proportion to the total load, $L$.

The net effect of these two competing trends—sublinear growth of existing contacts and linear recruitment of new ones—is that the total real area of contact ends up being, to a very good approximation, proportional to the total load.

And just like that, Amontons's law is rescued! It’s an ​​emergent property​​, a statistical law that appears on a large scale from the collective behavior of countless non-conforming individuals. It's a beautiful example of how simple, robust laws can emerge from complex, messy microscopic physics.

Our story so far has only involved pushing things together. But at small scales, there are also forces that pull things together: ​​adhesion​​. The same van der Waals forces that allow a gecko to walk up a wall are present between any two surfaces brought close enough together.

These attractive forces mean that even when you apply zero external load ($L=0$), the surfaces are still pulled into contact, creating a finite real contact area, $A_{\text{real}}(0)$. According to our friction rule, $F_f = \tau A_{\text{real}}$, this leads to a startling conclusion: there must be a finite friction force even at zero load!

This phenomenon, sometimes called ​​stiction​​, is another fundamental way in which nanoscale friction deviates from the classical Amontons's law, which demands that friction must vanish when the load is zero. If you plot friction force versus normal load for a nanoscale contact, the line doesn't pass through the origin. It hits the force axis at a positive value—a friction "intercept" determined by the strength of adhesion. For instance, modeling a typical silicon AFM tip sliding on an oxidized silicon surface, this adhesive friction intercept can be calculated to be on the order of a few nanonewtons, a tiny but measurable force that is a direct signature of these omnipresent sticky forces.

Putting these ideas together, we can now state the conditions under which the simple Amontons's laws are expected to work. They are an excellent approximation for the macroscopic world when two key conditions are met: the contact involves a very large number of asperities, and the forces from adhesion are negligible compared to the externally applied load. When either of these conditions fails—as in a high vacuum where surfaces are atomically clean and highly adhesive, or in a nanoscale experiment with a single tip—the beautiful simplicity of Amontons's laws gives way to a richer, more complex reality. At higher loads, the behavior changes again, as elastic deformation gives way to plastic flow and eventually wear, a transition governed by the material's hardness.

The final layer of our story reveals that even the "constant" in Amontons's law—the coefficient of friction $\mu$—is not truly constant. It depends, albeit weakly, on other factors like sliding speed and temperature.

Modern friction theories, known as ​​rate-and-state friction laws​​, introduce a "state variable," often denoted by $\theta$, which represents the average age or "maturity" of the asperity contacts. When surfaces are held in stationary contact, the bonds at the interface can slowly strengthen or rearrange—the contacts "age." When sliding begins, these mature contacts are sheared apart and replaced by new, "younger" ones. The friction force depends on both the instantaneous sliding velocity, $v$, and the current state, $\theta$, of the contact population. This framework leads to the prediction that friction depends logarithmically on sliding speed—a very weak dependence, but one that is critical for explaining phenomena from the screech of brakes to the stick-slip behavior of geological faults that causes earthquakes.

Furthermore, friction is not immune to heat. At the atomic scale, a sliding tip is essentially being dragged over a corrugated energy landscape. To move from one low-energy "valley" to the next, it must pass over an energy barrier. Thermal energy—the random jiggling of atoms, quantified by $k_{\text{B}}T$—can help the tip jiggle over these barriers, reducing the amount of external force needed. This process, called ​​thermal activation​​, leads to a reduction in friction that depends on temperature and, again, logarithmically on velocity.

What began as a simple observation about pushing boxes has led us on a journey deep into the microscopic world. We've seen that Amontons's "laws" are not fundamental edicts of nature but are emergent statistical approximations. The true story of friction is a rich tapestry woven from the threads of contact mechanics, plasticity, statistics, adhesion, and thermodynamics. It is a perfect example of how in physics, the simplest questions often lead to the most profound and beautiful answers.

Now, we come to a most delightful part of our story. In the last chapter, we laid out the basic principles of friction, culminating in Amontons' tidy little laws. You might be tempted to think, "Alright, friction is proportional to load. What's next?" One could be forgiven for thinking the subject is a bit, well, dry. But this is where the real adventure begins!

These simple empirical rules are not an end, but a beginning. They are like a treasure map. The map itself is simple, but it points to a vast and hidden landscape. By following this map—by testing where the law holds, where it fails, and, most importantly, why it fails—we are forced to burrow deep into the nature of matter. We will journey from the familiar world of sliding blocks into the fantastic, vibrating, sticky realm of the atom, and even connect to the frontiers of modern data science. Amontons' law is our guide, our foil, and our constant companion on a journey into the heart of how things touch.

For centuries, friction was studied by observing large objects. But what if we could "feel" the friction of a single, tiny point dragging across a surface? What would we find? Thanks to the invention of a marvelous device, the Atomic Force Microscope (AFM), we can do just that. An AFM is like a phonograph, but instead of a needle reading a record's groove, it has an exquisitely sharp tip at the end of a flexible cantilever that "feels" the landscape of a surface, atom by atom.

To measure friction, we don't just move the tip up and down; we drag it sideways. As the tip sticks and slips across the atomic terrain, it experiences a lateral, or frictional, force. This force doesn't bend the cantilever up or down, but twists it. By bouncing a laser off the back of the cantilever and tracking the reflected spot, we can measure this minuscule torsion with incredible precision. The amount of twist tells us the frictional force, a technique we call Lateral Force Microscopy. For the first time, we have a pair of eyes—or rather, a single finger—to feel the frictional world at the humanly imperceptible scale of nanometers.

Of course, nature is a clever beast and doesn't give up her secrets easily. When you drag a tip across a surface, it might twist because of friction, but it also might twist simply because it’s going up or down a slope on the surface—a sort of "topographic cross-talk." Disentangling true friction from these geometric illusions requires ingenuity. Experimentalists have developed a beautiful trick: they scan a line forward (the "trace") and then backward (the "retrace"). The frictional force always opposes the motion, so it flips its sign on the return trip. The topographic signal, however, does not. By taking half the difference between the forward and backward signals, we can isolate the pure frictional force, creating what's known as a "friction loop".

With this tool, we can put Amontons' law to the test. And right away, we find something new. At the nanoscale, surfaces are wonderfully sticky. Van der Waals forces and other attractions create an "adhesion force" that pulls the tip onto the surface, even with no external load applied. This adhesion acts like an extra, invisible load. So, the old law gets a facelift. The friction force, $F_{f}$, is proportional not just to the normal load we apply, $F_N$, but to the total load, which includes the adhesion force, $F_{\mathrm{adh}}$:

$F_f = \mu (F_N + F_{\mathrm{adh}})$

This "modified Amontons' law" is our first great discovery on this journey. The simple law of the macro-world is still there, but it's been refined, made more precise by accounting for the stickiness that dominates the nano-world.

Now, let's get a bit mischievous. As physicists, our favorite game is to find a venerable law and see if we can break it. So, why should friction be proportional to load?

The modern understanding, pioneered by the great tribologists Bowden and Tabor, is that friction force, $F_f$, arises from shearing the "true" area of contact between two surfaces, $A_{\mathrm{real}}$. This can be written as $F_f = \tau A_{\mathrm{real}}$, where $\tau$ is the interfacial shear strength—a measure of how hard it is to shear the interface. For Amontons' law to hold, it must be that the real contact area, $A_{\mathrm{real}}$, is directly proportional to the normal load, $N$.

Is it? Let's zoom in on a single, perfectly smooth, elastic asperity (a tiny "hill" on our surface) pressing against a flat plane. The theory of contact mechanics, first worked out by Heinrich Hertz, gives us a very clear answer. For a spherical tip, the contact area does not grow proportionally with load. Instead, it follows a different rule: $A_{\mathrm{real}} \propto N^{2/3}$. And if we include the effects of adhesion using more advanced models like the JKR theory, the relationship becomes even more complicated and non-linear.

So, here is a beautiful paradox! At the most fundamental level of a single, clean contact, Amontons' law appears to be false! The very foundation seems to crumble. Even more dramatic is the difference in the shear strength, $\tau$. When we measure it with an AFM on a pristine single contact, we find it to be enormous, approaching the theoretical strength of the material. But when we measure the effective shear strength in a large, macroscopic experiment, it's hundreds or even thousands of times smaller. At the nanoscale, things are strong and sticky and follow complex rules. At the macroscale, they are weak and slippery and follow a simple, linear law. What's going on?

The resolution to our paradox is a profound lesson in physics: simple, macroscopic laws often emerge from underlying microscopic complexity. Amontons' law is a ghost that haunts any sufficiently complex sliding system, appearing through various mechanisms.

First, there is the "bed of nails" argument of Bowden and Tabor. Real surfaces are not single smooth spheres; they are rough, like two mountain ranges grinding against each other. When you press them together, only the very highest peaks (asperities) touch. As you increase the load, these peaks flatten a bit, and more, smaller peaks are recruited into contact. Even though each tiny contact might not follow Amontons' law, the statistical effect of adding more and more contacts makes the total real contact area become roughly proportional to the load. The simple law is an illusion created by the statistics of a very large number of tiny events.

But that's not the only way the ghost can appear! Consider a completely different mechanism of friction: plowing. If you drag a hard diamond tip across a soft polymer, a significant part of the friction comes from the energy needed to plastically deform the material and carve a groove. This isn't shearing an interface; it's digging a trench. The physics is entirely different, governed by the material's yield stress. And yet, if you work through the model, you find a stunning result: the plowing force, under common assumptions, scales linearly with the normal load. Amontons' law rises from the ashes of a completely different physical process!

As if that weren't enough, consider what happens in most real-world sliding systems. They are messy. Between the two original surfaces, a complex layer forms—a mish-mash of wear debris, oxides, and surface films. Tribologists call this the "third body." This layer is not a simple solid; it flows and churns like a dense, granular sludge. Modern physics provides powerful frameworks to describe such behavior, such as the rheology of amorphous solids or the "$\mu(I)$" laws for granular flow. These are sophisticated, cutting-edge theories. And what do they predict for the relationship between shear stress and confining pressure in this complex layer? They predict a relationship that gives rise, yet again, to a friction force proportional to the normal load. The simple rule of thumb is a deep consequence of the chaotic dance of debris.

Having seen how stubbornly Amontons' law reappears, let's find a situation where we can truly banish it. To do so, we need to create a perfectly ordered, atomically clean interface. Imagine sliding two perfect, rigid crystals over each other. If their crystal lattices are mismatched—say, a hexagonal grid sliding over a square one—there is no unique way for them to lock together. As you slide, the upward forces experienced by some atoms are cancelled by the downward forces on others. The net potential energy landscape becomes incredibly flat. The result is a state of "structural superlubricity," where static friction almost vanishes.

This state, however, is delicate. If we begin to press down on the sliding layers, we squeeze them closer together. As the separation decreases, the interatomic forces get much stronger, and the energy landscape becomes more corrugated. At the same time, if one of the layers is compliant, it costs some elastic energy to strain it to match the other lattice. A battle ensues: the elastic energy cost of deforming versus the interfacial energy gain of snapping into a low-energy, commensurate (locked-in) state. As we increase the pressure, the interfacial gain grows until, at a critical pressure, it overcomes the elastic cost. "Click!" The interface locks into registry, superlubricity is lost, and a large static friction force suddenly appears. This is a beautiful example of a phase transition, where friction can be switched on and off not by a knob, but by pressure.

This vast array of phenomena—adhesion, elasticity, plasticity, granularity, and atomic registry—all play a role. How can we make sense of it all in a real experiment? This is where our story takes a turn and connects to one of the most powerful tools of the 21st century: machine learning. We can perform complex experiments where we vary load, humidity, temperature, and sliding speed, and measure the resulting friction. The data can be fed into a machine learning model, which can learn the intricate relationships without any a priori assumptions. But a "black box" model is unsatisfying. We want to know why. Using techniques like SHAP (Shapley Additive exPlanations), we can interrogate the model and ask it: "How much did you attribute the friction to the load? How did the lattice mismatch affect your prediction?" In doing so, we can verify whether the model has discovered known physical laws (like the modified Amontons' law) or uncovered new, unexpected relationships in the data. It is a new kind of dialogue with nature, mediated by an algorithm.

What began as a simple observation by Guillaume Amontons three centuries ago has become a gateway. The quest to understand this one "law" has led to the design of incredible instruments, forced us to develop new theories of contact and flow, revealed the surprising emergence of simplicity from complexity, and has now brought us to the doorstep of artificial intelligence. It shows us, in the most brilliant way, that there are no "boring" subjects in physics. There are only treasure maps waiting to be followed.