The Physics of Muscles: How Molecular Motors Make You Move

The Motor Inside You

Lift your hand. Close your fist. Open it again. That sequence of movements — trivial, thoughtless, accomplished in under a second — required the coordinated firing of hundreds of motor neurons, the contraction and relaxation of dozens of muscles, and the collective action of roughly a trillion molecular motors, each one converting the chemical energy of a single ATP molecule into a mechanical pull of about five piconewtons over a distance of ten nanometres.

You are, from an engineering perspective, powered by molecular machines. Every movement you make — from blinking to sprinting to the beating of your heart — is driven by the same fundamental mechanism: a protein called myosin walking along a protein filament called actin, fuelled by ATP.

The physics of this system spans scales from the single-molecule (piconewtons, nanometres, milliseconds) to the whole-body (kilonewtons, metres, seconds). It involves thermodynamics, mechanics, fluid dynamics, and statistical physics. And it produces an engine that is, by several measures, competitive with the best machines humans have built.

Muscle Architecture: Macro to Nano

A skeletal muscle is a hierarchy of structures spanning six orders of magnitude in size:

The muscle (centimetres) is wrapped in connective tissue and attached to bones by tendons. The bicep, for example, is about 25 cm long and 5 cm across.

Each muscle contains thousands of muscle fibres (50–100 µm diameter, up to 30 cm long) — individual cells, each formed by the fusion of many precursor cells during development, containing hundreds of nuclei.

Each fibre contains hundreds of myofibrils (~1 µm diameter) — long, cylindrical bundles of contractile proteins, running the full length of the fibre.

Each myofibril is divided into repeating units called sarcomeres (~2.5 µm long at rest) — the fundamental contractile units. Sarcomeres are arranged in series along the myofibril, and it’s the shortening of each sarcomere that produces the overall muscle contraction.

Within each sarcomere, two types of protein filament interdigitate:

Thick filaments — bundles of ~300 myosin II molecules, each with two globular heads protruding from the filament backbone.

Thin filaments — helical polymers of actin monomers, decorated with the regulatory proteins tropomyosin and troponin.

The sarcomere shortens when the thick filaments pull the thin filaments inward — the filaments slide past each other without themselves shortening. This is the sliding filament model, proposed independently by Andrew Huxley (yes, the same Huxley who modelled the action potential) and Hugh Huxley (no relation) in 1954.

The Cross-Bridge Cycle: One Stroke at a Time

The molecular engine of muscle contraction is the cross-bridge cycle — the repeated attachment, pulling, and detachment of myosin heads on actin filaments. Each cycle consumes one ATP molecule and produces one mechanical stroke:

Step 1: Attachment. The myosin head, primed with the hydrolysis products ADP and inorganic phosphate (Pi) bound in its active site, binds to an exposed site on the actin filament. The binding is stereospecific — the myosin head fits into a complementary pocket on actin.

Step 2: Power stroke. Pi is released from the myosin head, triggering a conformational change — the lever arm of the myosin head rotates by about 60–70°, pulling the actin filament by 5–10 nm relative to the thick filament. ADP is released during or just after the stroke. This is the force-generating step.

Step 3: Detachment. A new ATP molecule binds to the myosin head, causing it to release from actin. Without ATP binding, the myosin head remains locked to actin — this is the molecular basis of rigor mortis: after death, ATP is depleted, and myosin heads remain permanently attached to actin, stiffening the muscles.

Step 4: Recovery stroke. The myosin head hydrolyses the ATP (ATP → ADP + Pi), and the energy is used to re-cock the lever arm back to its starting position. The head is now primed and ready for the next attachment.

The entire cycle takes about 50–100 milliseconds for slow (Type I) fibres and 25–50 ms for fast (Type II) fibres. A single myosin head is attached to actin for only about 2–5% of the cycle time — most of the time it’s detached, waiting to rebind. This low duty ratio means that many heads must work in parallel to produce continuous force.

In a single sarcomere, about 300–600 myosin heads per thick filament overlap with the thin filament zone. At any instant, perhaps 20–50 of these are attached and pulling. The others are cycling through detachment and recovery. The collective effect of hundreds of non-processive motors working in parallel is smooth, continuous force — like a rowing crew where each rower’s oar is in the water for only a fraction of the stroke, but there are always enough oars pulling to keep the boat moving.

Force Production: The Numbers

Individual myosin heads produce tiny forces: 2–5 piconewtons per power stroke, measured directly by optical trap experiments (for which Arthur Ashkin shared the 2018 Nobel Prize in Physics).

The power stroke distance is 5–10 nm, giving a work output per stroke of:

W = F × d ≈ 4 pN × 8 nm ≈ 3.2 × 10⁻²⁰ J ≈ 20 zJ

The free energy available from hydrolysing one ATP molecule under cellular conditions is about 54 kJ/mol, or roughly 9 × 10⁻²⁰ J per molecule. The single-molecule efficiency is therefore:

η = 3.2 × 10⁻²⁰ / 9 × 10⁻²⁰ ≈ 35%

(This varies with load; under optimal conditions it can reach 50–80%.)

Scaling up: a human bicep (cross-sectional area ~12 cm²) contains about 250,000 fibres, each with about 1,000 myofibrils, each with about 10,000 sarcomeres in series. The maximum isometric force is:

F_max ≈ 25 N/cm² × 12 cm² ≈ 300 N (about 30 kg)

The specific tension — force per unit cross-sectional area — is remarkably constant at 20–30 N/cm² across all skeletal muscles, across all mammals, and even across many invertebrates. This universality reflects the underlying molecular mechanism: the force comes from myosin-actin cross-bridges, and the density of cross-bridges per unit area is approximately constant because the packing of myofilaments in the sarcomere lattice is structurally constrained.

The Force-Velocity Relationship: A. V. Hill’s Equation

In 1938, the physiologist A. V. Hill — a Nobel laureate who was also an accomplished athlete — measured the relationship between the force a muscle produces and the speed at which it shortens. He found a hyperbolic relationship:

(F + a)(v + b) = (F₀ + a)b

where F is the force, v is the shortening velocity, F₀ is the maximum isometric force (at zero velocity), and a and b are constants.

The physics behind this curve is the cross-bridge cycle dynamics. At high loads (near F₀), the muscle shortens slowly because each myosin head must generate a large force per stroke, slowing the cycling rate. At low loads, the cycling rate is high and the muscle shortens quickly, but the force per head is low. At maximum velocity (v_max), the muscle produces essentially zero net force — the myosin heads are cycling as fast as possible but can’t overcome the viscous drag of the filament system.

The power output P = Fv peaks at roughly 30% of maximum velocity and 30% of maximum force. For a whole muscle, the peak power condition represents the optimal trade-off between force and speed — the condition that maximises the rate of mechanical energy output. This is why sprinters and cyclists are strongest (produce the most power) at intermediate speeds, not at maximum force or maximum velocity.

Hill’s equation, originally empirical, can be derived from models of cross-bridge kinetics — the rates of attachment and detachment as functions of filament displacement. This connection between a macroscopic muscle property and a molecular mechanism was one of the early triumphs of biophysical modelling.

Muscle Efficiency: 25% and Proud of It

The overall mechanical efficiency of skeletal muscle — useful work out divided by metabolic energy in — is about 20–25% for activities like cycling or running. The rest appears as heat.

Is this good or bad? Compare:

A petrol car engine: ~25–35% efficient An electric motor: ~85–95% efficient A steam engine (Watt era): ~5–10% efficient A human on a bicycle: ~25% efficient

Muscles are comparable to internal combustion engines — and they operate at body temperature (37°C), with water as the solvent, using proteins as mechanical components. The thermodynamic limit for a heat engine operating between 37°C (310 K) and the environment (~20°C, 293 K) is:

η_Carnot = 1 − T_cold/T_hot = 1 − 293/310 ≈ 5.5%

But muscles are not heat engines — they don’t operate on a temperature differential. They’re chemomechanical transducers: they convert the free energy of a chemical reaction (ATP hydrolysis) directly into mechanical work. The Carnot limit doesn’t apply. The relevant limit is the free energy of ATP hydrolysis (~54 kJ/mol), and the single-molecule efficiency of the cross-bridge can reach 50–80% of this.

The gap between the single-molecule efficiency (50–80%) and the whole-body efficiency (20–25%) is accounted for by overhead: calcium pumping (the sarcoplasmic reticulum Ca²⁺-ATPase consumes ~30% of the ATP budget during contraction), sodium-potassium pumps maintaining the membrane potential, metabolic pathway inefficiencies, and ventilation costs.

Muscle Types: Speed vs. Endurance

Not all muscle fibres are the same. The difference lies in which myosin isoform they express:

Type I (slow-twitch) fibres express slow myosin with a low ATPase rate. They contract slowly (~100 ms to peak tension), produce moderate force, but resist fatigue because they rely on aerobic metabolism — they’re packed with mitochondria, have a rich capillary supply, and contain myoglobin (an oxygen-storage protein that gives them a red colour). Postural muscles (back, soleus) are predominantly Type I.

Type IIx (fast-twitch glycolytic) fibres express fast myosin with a high ATPase rate. They contract quickly (~25–50 ms), produce high force, but fatigue rapidly because they rely on anaerobic glycolysis for ATP — which produces ATP fast but in limited quantities and generates lactate and H⁺ as byproducts.

Type IIa (fast-twitch oxidative) fibres are intermediate — fast contraction, moderate fatigue resistance, using both aerobic and anaerobic pathways.

The physical basis of the speed difference is the myosin ATPase rate. Faster ATP hydrolysis means faster cross-bridge cycling, faster filament sliding, and faster shortening velocity. But it also means faster ATP consumption and earlier fatigue.

Elite sprinters tend to have a higher proportion of Type II fibres (~70–80% in the vastus lateralis). Elite marathon runners tend to have more Type I (~70–80%). This distribution is partly genetic and partly trainable — endurance training can convert Type IIx fibres to Type IIa (faster and more fatigue-resistant), though the conversion between Type I and Type II is limited.

Eccentric Contraction: Stronger While Lengthening

Here’s a counterintuitive fact: a muscle produces more force while being forcibly lengthened (an eccentric contraction) than while shortening or holding isometric. You can lower a heavier weight than you can lift.

The force enhancement during eccentric contraction — about 1.5–1.8 times the maximum isometric force — comes from the cross-bridge mechanics. When an active muscle is stretched, attached myosin heads are pulled beyond their equilibrium position. The elastic element in the myosin head (the lever arm acts as a spring) generates additional restoring force. Some heads are also forced into high-force states before they can detach.

Additionally, the giant protein titin — the largest protein known (3–4 million daltons, spanning half the sarcomere from Z-disc to M-line) — acts as a molecular spring that contributes passive force during stretch. Titin’s extensible regions unfold under tension, absorbing energy and contributing to the muscle’s elasticity.

Eccentric contractions are metabolically efficient (the muscle absorbs mechanical energy rather than producing it) but cause more muscle damage — the high forces literally rip some sarcomeres apart. This damage triggers the repair and hypertrophy response: eccentric exercise is the most effective stimulus for muscle growth. The delayed-onset muscle soreness (DOMS) you feel 1–2 days after unfamiliar exercise is primarily caused by eccentric damage.

Scaling: From Ants to Elephants

Muscle force scales with cross-sectional area (∝ L²), while body weight scales with volume (∝ L³). As animals get larger, the ratio of muscle force to body weight decreases:

F/W ∝ L²/L³ = 1/L

This is why an ant can carry 50 times its body weight (its muscles are large relative to its small mass) while an elephant can barely carry its own weight above what’s needed for locomotion. It’s why fleas can jump 100 times their body length (relative to body size) while humans can jump only about 1.5 times theirs.

The scaling also explains why large animals move differently than small ones. A mouse scurries; a horse trots; an elephant walks. The maximum sustainable running speed is limited by the capacity to dissipate heat (metabolic rate scales as ~M⁰·⁷⁵, but heat dissipation scales with surface area as ~M⁰·⁶⁷), and by the stresses on bones and tendons (which scale unfavourably with size). The physics of muscle is the same at all scales; the constraints on its use change with body size.

Your Molecular Engines

Stand up. Walk across the room. Sit down again.

In those few seconds, roughly 10¹² myosin heads in your leg muscles each went through the cross-bridge cycle several times — attaching to actin, pulling, detaching, resetting, and repeating. Each pull was about five piconewtons of force over ten nanometres. Each consumed one ATP molecule. Trillions of nanoscale events, coordinated by calcium signals and nerve impulses, producing a smooth, controlled movement at the human scale.

The machinery is astonishing in its simplicity. One motor protein. One track protein. One fuel molecule. The same design, with minor variations, drives the beating of your heart, the peristalsis of your gut, the dilation of your pupils, and the kick of a sprinter leaving the blocks. It works from −2°C (in Antarctic fish) to 47°C (in desert lizards). It has been refined by 600 million years of evolution and has never been surpassed by any artificial system in its combination of efficiency, reliability, scalability, and self-repair.

You are made of molecular motors. Trillions of them. And they’re running right now — even as you read this — keeping you upright against gravity, moving air through your lungs, and pumping blood to a brain that can contemplate the physics of its own machinery.