No single person discovered friction. The resistance you feel when dragging one surface across another was observed and exploited by humans long before anyone tried to write down rules for it, from ancient Egyptians lubricating sledges with water to Greek engineers greasing axles. The scientific study of friction, though, traces a surprisingly clear lineage through three figures: Leonardo da Vinci in the late fifteenth century, Guillaume Amontons at the turn of the eighteenth century, and Charles-Augustin de Coulomb a few decades later. Each built on the last, and the “laws” of friction that physics textbooks still teach today bear their combined fingerprints.
Leonardo da Vinci and the First Friction Experiments
Leonardo da Vinci is widely credited as the first person to study friction with anything resembling a scientific method. In the 1490s, he sketched experiments in his private notebooks showing blocks of different sizes being dragged across surfaces with weights hanging over pulleys. He observed two things that would later become formal laws: friction is proportional to the load pressing the surfaces together, and it does not depend on how large the contact area is. A wide block and a narrow block of the same weight, in other words, experience roughly the same frictional resistance.
The problem is that Leonardo never published any of this. His notebooks were written in mirror script, scattered after his death, and largely unknown to other scientists for centuries. The friction observations stayed buried until scholars rediscovered them much later, meaning they had no direct influence on the researchers who eventually formalized the same ideas independently.1Tribology Letters. Leonardo da Vinci’s Friction Experiments: An Old Story Acknowledged and Repeated This makes Leonardo a discoverer in hindsight rather than in practice. His work proves that the key insights about friction were within reach of a careful fifteenth-century experimenter, but the science of friction as a shared body of knowledge starts with the person who actually told other scientists about it.
Amontons and the Laws That Stuck
That person was Guillaume Amontons, a French instrument maker and physicist. In 1699, he presented a paper to the Académie Royale des Sciences describing experiments on friction in machines. But his widely cited December 1699 paper was not actually the first time he laid out his findings. Six months earlier, in a paper about a novel heat engine, Amontons described measurements of forces involved in glass polishing and stated clearly that friction force is independent of contact area and proportional to the load pressing the surfaces together.2Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology. A note on Guillaume Amontons and the laws of friction These two observations, essentially the same ones Leonardo had noted two centuries earlier, became known as “Amontons’ laws.” Because Amontons published them and discussed them before the French scientific establishment, they entered the public scientific record and could be tested, challenged, and refined by others.
Amontons also proposed that friction arises from surface roughness, from the interlocking of tiny bumps and ridges that resist sliding. This was intuitive and partly correct, though the full picture turned out to be more complicated.
Coulomb’s Deeper Investigation
Nearly a century after Amontons, Charles-Augustin de Coulomb, better known for his work on electrical charge, conducted the most thorough friction experiments of the era. His 1785 memoir expanded the understanding of friction in several ways. He drew a clear distinction between static friction, the force needed to start an object moving, and kinetic friction, the force needed to keep it moving once it has started. He also observed that static friction increases with the time two surfaces sit in stationary contact before sliding begins, and that kinetic friction can vary with sliding speed.3ScienceDirect. Parameter Identification and Monitoring of Mechanical Systems Under Nonlinear Vibration – Section: 5.4 Friction
Coulomb found, however, that for dry metal-on-metal surfaces, many of these subtleties vanished. The dependence on speed and on resting time was absent or negligible for bare metals touching each other. This led to the simplified version of friction that most people learn in school: a constant coefficient that depends only on the materials and the load. Coulomb added a third law to the two Amontons had stated, namely that kinetic friction is roughly independent of sliding speed. Together, the three statements are often called the “Amontons-Coulomb laws” of friction.
What the Classical Laws Actually Describe
Stripped of formulas, the three classical laws say the following. First, the frictional force resisting motion is proportional to how hard the surfaces are pressed together. Double the weight of a box on the floor, and you roughly double the push needed to slide it. Second, the friction force does not depend on how much surface area is in contact. A brick lying flat and the same brick standing on its narrow end resist sliding by the same amount, assuming the same material pairing. Third, once an object is sliding, the friction stays roughly constant regardless of how fast it moves.
These laws are remarkably useful approximations. Engineers still rely on them to design brakes, tires, and conveyor belts. But they are approximations, and the last century of research has shown where and why they break down.
Why Surfaces Really Resist Motion
The intuitive explanation for friction, that bumps on one surface catch on bumps on the other, is only part of the story. When viewed at the microscopic scale, even a polished metal surface is covered in tiny bumps called asperities. When two surfaces are pressed together, they do not make contact everywhere. Instead, they touch only at the tips of these asperities, so the real area of contact is a tiny fraction of the apparent area.4PubMed Central. Friction Behavior of Rough Surfaces on the Basis of Contact Mechanics: A Review and Prospects The overall friction you feel is the combined result of what happens at all of these tiny contact points.
In the mid-twentieth century, researchers showed that friction between metals involves two processes happening simultaneously. At the contact points, the surfaces bond together briefly through adhesion, and sliding requires shearing those bonds apart. At the same time, harder asperities can dig into the softer surface and plough through it like a tiny blade. The total friction is the sum of these shearing and ploughing contributions, and it can be described in terms of the bulk physical properties of the metals involved.5Oxford Academic. Mechanism of Metallic Friction This adhesion-plus-ploughing model, developed primarily by researchers at Cambridge in the 1950s, replaced the purely geometric picture of interlocking bumps and gave engineers a way to predict friction from material properties rather than relying entirely on lookup tables.
The adhesion model also explains why Amontons’ area-independence law works even though it seems counterintuitive. When you increase the apparent contact area without changing the load, the pressure per asperity decreases, and the asperities deform less, so the real contact area stays about the same. The friction does not care about the footprint of the object. It cares about the real area where atoms are actually touching, which is governed by the load.
Static, Kinetic, and Rolling Friction
Coulomb’s distinction between static and kinetic friction remains fundamental. Static friction is almost always larger. You need a harder initial push to get a heavy dresser moving than to keep it sliding once it starts. The reason is that asperities in stationary contact have time to settle into each other and form stronger adhesive junctions. Coulomb noticed this time-dependence himself, and it is still an active area of research in earthquake science and industrial design.
Rolling friction is a different animal. When a ball or cylinder rolls across a surface, the dominant resistance comes not from adhesion or ploughing but from energy lost as the material deforms and recovers under the rolling load. This is sometimes called elastic hysteresis. The material compresses as the roller passes over it and springs back afterward, but not all the energy returns; some is dissipated as heat within the material. Early theoretical predictions of rolling friction based on simple hysteresis models turned out to underestimate the actual frictional force by a factor of two to three, suggesting that the deformation cycles during rolling are more complex than those in a straightforward compression test.6Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences. Hysteresis losses in rolling and sliding friction Rolling friction is much smaller than sliding friction for most materials, which is why wheels and ball bearings transformed transportation and machinery.
How Lubrication Changes Everything
Adding a fluid between two surfaces can reduce friction dramatically, but the way it does so depends on conditions. Engineers describe this with something called the Stribeck curve, which maps out three distinct regimes. In boundary lubrication, the surfaces are so close that they still touch at asperity tips, and the lubricant film is too thin to fully separate them. In mixed lubrication, an adsorbed fluid layer starts to dominate the contact. In full hydrodynamic lubrication, the surfaces are completely separated by a flowing film of fluid, and friction drops to levels determined by the fluid’s own internal resistance to flow.7Friction. Molecular dynamics simulation of the Stribeck curve: Boundary lubrication, mixed lubrication, and hydrodynamic lubrication on the atomistic level
These regimes matter practically because the same machine can pass through all three during normal operation. An engine bearing at startup, before oil pressure builds, operates in boundary lubrication and wears rapidly. At cruising speed, with a full oil film established, it operates in hydrodynamic lubrication and barely wears at all. Understanding where a given system sits on the Stribeck curve is central to designing lubricants, bearings, and maintenance schedules.
Friction in Biology
Living systems have solved friction problems that engineers still struggle with. Your knee joint, for example, supports enormous loads while allowing smooth motion for decades. The cartilage surfaces are lubricated by synovial fluid, a complex mixture of molecules including hyaluronan, lubricin, and phospholipids. Recent work has shown that these components work together synergistically, with lubrication at the boundary level maintained through a mechanism called hydration lubrication, where water molecules tightly bound to charged groups on the cartilage surface provide an extremely slippery interface.8PubMed. Lubrication of Articular Cartilage The friction coefficient in a healthy human joint is remarkably low, on the order of 0.001 to 0.01, which is better than most engineered bearings.
Geckos have solved the opposite problem: maximizing friction and adhesion on demand. Their toe pads are covered in millions of microscopic branching hair-like structures called setae, which create dry adhesion through van der Waals forces, the weak intermolecular attractions that operate at very short range. The clever part is not just sticking but unsticking. Setae detach when pulled at a critical angle, and geckos hyperextend their toes to peel them away from surfaces like peeling tape, allowing them to climb at speed without getting stuck.9PubMed. Frictional adhesion: A new angle on gecko attachment This “frictional adhesion” has inspired a generation of synthetic adhesive materials for robotics and medicine.
Friction and Earthquakes
Friction governs whether tectonic faults creep slowly and harmlessly or lock up and then rupture in a sudden earthquake. The behavior is governed by how friction on the fault changes with slip rate and contact history, a framework seismologists call “rate and state” friction. In computational models of faults under increasing fluid pressure, the same rock properties can produce wildly different outcomes depending on the frictional parameters: stable, smooth sliding in some cases, a sequence of many small slip events in others, or a single large seismic rupture after a long period of silent creep.10Geophysical Research Letters. Stick‐slip dynamics of flow‐induced seismicity on rate and state faults
This is particularly relevant to induced seismicity from underground fluid injection, such as wastewater disposal or geothermal energy extraction. The injected fluid raises pore pressure along a fault, reducing the effective normal stress and therefore the frictional resistance holding the fault in place. Whether the result is a harmless tremor or a damaging earthquake depends on friction parameters that are difficult to measure directly, which makes predicting induced earthquakes one of the harder problems in applied geophysics.
Superlubricity and the Quest for Near-Zero Friction
At the other extreme from earthquake-generating friction, researchers have achieved states where friction essentially vanishes. Structural superlubricity occurs when two crystalline surfaces are in contact but their atomic lattices are misaligned, so the tiny forces between atoms on opposite surfaces cancel out rather than adding up. The result is friction close to zero.11Lubricants. Structural Superlubricity of Two-Dimensional Materials: Mechanisms, Properties, Influencing Factors, and Applications
Two-dimensional materials like graphene are the stars of this field. In computer simulations, researchers have shown that structural superlubricity can be achieved in heterojunctions, pairings of two different layered materials, even when the layers are aligned, because the lattice mismatch between different materials inherently prevents the atoms from locking into a high-friction configuration.12PubMed Central. The Unusual Tribological Properties of Graphene/Antimonene Heterojunctions: A First-Principles Investigation Superlubricity is still mostly a laboratory and computational phenomenon, but it holds promise for micro-electromechanical systems and hard-drive components where even tiny frictional losses cause significant wear over billions of cycles.
Friction in the Vacuum of Space
On Earth, almost every metal surface is coated in a thin layer of oxide and adsorbed gases that acts as a natural lubricant. Remove that layer, and metals behave very differently. In the high vacuum of space, when two clean metal surfaces are brought into contact, they can bond together spontaneously in a process called cold welding. Experiments performed at vacuum levels below one hundred-millionth of a torr have shown that once surface oxides are disrupted or removed, metals adhere to each other readily at room temperature without any need for heat or diffusion.13Adhesion, or Cold Welding, of Materials in Space Environments. 2-3 Adhesion of Metals in High Vacuum
This is a practical headache for spacecraft designers. Moving parts on satellites, such as deployment hinges, antenna gimbals, and solar array mechanisms, must be designed so that metal surfaces never touch bare. Coatings, solid lubricants like molybdenum disulfide, and careful material pairing are standard countermeasures. The phenomenon also underscores something fundamental about friction: the oxide films and contaminant layers that we usually ignore on Earth are doing a lot of hidden work to keep surfaces from sticking to each other.
The Global Energy Cost of Friction
Friction is not just a physics concept; it is one of the largest energy drains on the global economy. Worldwide, roughly 100 million terajoules of energy are used every year simply to overcome friction, and that amounts to about one fifth of all energy produced. The biggest consumers are industry, at about 29 percent of friction-related energy use, and transportation, at about 27 percent.14FME Transactions. Global impact of friction on energy consumption, economy and environment Every engine piston sliding in its cylinder, every tire rolling on pavement, every conveyor belt moving goods through a warehouse is paying a friction tax in heat and wear.
This means that even small improvements in friction management have enormous cumulative effects. Better lubricants, lower-friction coatings, and improved bearing designs can save significant fractions of the energy consumed by vehicles and factories. It also means that the ancient question Leonardo da Vinci scratched into his notebooks, what governs the resistance between sliding surfaces, remains one of the most economically important questions in applied science.
Friction You Can Feel Through a Screen
One of the more surprising modern applications of friction science is in haptic touchscreens, the kind that give you a subtle texture or click sensation under your finger. These devices work by using ultrasonic vibrations to create a thin cushion of air between your fingertip and the glass, partially levitating the skin and reducing friction. By turning this effect on and off rapidly as your finger moves, the screen can simulate bumps, edges, and textures that are not physically there. Researchers have shown that the friction reduction is driven by acoustic radiation pressure and that the resulting friction coefficient can be predicted accurately from the mechanical impedance of the vibrating plate.15PubMed. Estimating Friction Modulation From the Ultrasonic Mechanical Impedance
This technology represents a full circle of sorts. For centuries, scientists studied friction as something to measure and overcome. Now friction is being precisely controlled to create sensory experiences, to make a flat piece of glass feel like sandpaper or silk depending on what the software demands. The same physical principles Amontons described in 1699, proportionality to load and independence from area, still underlie the engineering models. What has changed is our ability to manipulate friction at scales and speeds that the classical experimenters could never have imagined.