Simple Harmonic Motion

Pull a block on a spring 2 cm to the side and release it; time the round trip. Now pull it 8 cm — four times as far — and time it again. Same answer. The big swing and the small wiggle complete their cycles in exactly the same time, a property called isochronism, and it is strange enough that it took humanity until the 1650s to notice it, trust it, and build the first accurate clocks around it. Every mechanical watch since is a bet that this equality never fails. You arrive with both ingredients in hand: Newton's second law tells you how force makes acceleration, and Hooke's law hands you a force that grows with displacement and always points back toward equilibrium. Put them together and something remarkable happens — the equation picks the solution for you. Only one family of functions has an acceleration proportional to minus itself: sines and cosines. The mass must move sinusoidally, at a frequency $\omega = \sqrt{k/m}$ that the spring and the mass agree on before you ever touch the system. How hard you pull sets the amplitude and the top speed; it has no vote on the timing. And this is far more than a story about springs. Expand any smooth potential energy about a stable equilibrium and the leading behaviour is always quadratic — which means every stable system in nature, displaced gently, is this system. Molecular bonds, pendulums, quartz crystals, electrical circuits, the modes of a laser field: all of them run on the equation derived here. Master simple harmonic motion and you own the template for every oscillation in physics.

Key takeaways

A linear restoring force $F = -kx$ turns Newton's second law into $\ddot x + \omega^2 x = 0$ : acceleration proportional to minus the displacement.
The motion is sinusoidal, $x(t) = A\cos(\omega t + \phi)$ , with angular frequency $\omega = \sqrt{k/m}$ set entirely by the hardware.
The period $T = 2\pi\sqrt{m/k}$ is independent of amplitude (isochronism) — a small wiggle and a large swing take the same time.
The maximum speed $v_{\max} = A\omega$ occurs at the equilibrium crossing; the maximum acceleration $a_{\max} = A\omega^2$ occurs at the turning points.
Energy sloshes losslessly between potential and kinetic forms: $E = \tfrac12 kA^2 = \tfrac12 m v_{\max}^2$ , and over a cycle each form averages to exactly half the total.
Near a stable equilibrium every smooth potential looks parabolic, so every stable system oscillates harmonically for small displacements, with $\omega = \sqrt{U''(x_0)/m}$ .
Key idea: a linear restoring force $F=-kx$ gives $\ddot{x}+\omega^2x=0$ , whose sinusoidal solution oscillates at one amplitude-independent frequency $\omega=\sqrt{k/m}$ — and this is the universal small-displacement limit of any stable equilibrium.

The core insight

A linear restoring force $F=-kx$ gives the equation $\ddot{x}+\omega^2x=0$ , whose sinusoidal solution oscillates at one amplitude-independent frequency $\omega=\sqrt{k/m}$ — and this is the universal small-displacement limit of any stable equilibrium.

Where this lives today

The harmonic oscillator is the most manufactured physical system in history — you probably own dozens. Every quartz watch and every phone contains a tiny tuning-fork crystal oscillating at exactly 32,768 Hz ( $2^{15}$ cycles, halved fifteen times by digital logic, equals one tick per second); its stability comes precisely from isochronism, the period's refusal to care about amplitude as the drive level wanders. MEMS accelerometers and gyroscopes are literal micromachined spring–mass systems, and atomic force microscope cantilevers are springs whose resonant frequency $\sqrt{k/m}$ shifts measurably when a few atoms' worth of force acts on the tip. At larger scales, the 660-tonne tuned mass damper hanging in Taipei 101 is a giant pendulum-spring oscillator deliberately tuned to the tower's sway frequency so it drinks energy out of wind-driven oscillations. And aboard the ISS, where a bathroom scale reads zero, astronauts strap into a Body Mass Measurement Device — a chair on a calibrated spring — and their mass is computed from the measured period via $m = kT^2/4\pi^2$ : equation (4) of this article, run backwards, as a medical instrument.

The physical picture

Watch one full cycle of the archetype — a block on a frictionless spring. At the far right turning point, $x = +A$ , the block is momentarily at rest but the stretched spring is pulling its hardest — maximum force, maximum acceleration, zero speed. The block accelerates left, gaining speed as the pull weakens, and by the time it reaches the natural length the force has died to zero — but the block is now moving at top speed and inertia carries it straight through. On the far side the compressed spring pushes back, braking the block until it stalls at $x = -A$ , and the whole story runs in reverse. The motion is an endless argument between the spring's eagerness to restore and the mass's reluctance to stop, and neither ever wins.

Figure

One cycle of the oscillator — force, speed, and energy trade places

The deepest way to see why the result is a sinusoid is to borrow a circle. Let a point travel at constant angular rate $\omega$ around a circle of radius $A$ , and shine light from above so the point casts a shadow on a horizontal axis. The shadow's position is $A\cos(\omega t + \phi)$ — exactly the oscillator's motion. Uniform circular motion viewed edge-on is simple harmonic motion: the radius is the amplitude, the rotation rate is the angular frequency, and one lap of the circle is one period of the oscillation. Every formula in this article can be read off that picture.

Figure

SHM is the shadow of uniform circular motion — radius A, rotation rate ω

Unroll one full period and plot all three kinematic quantities together, and the choreography of the cycle becomes explicit: displacement, velocity, and acceleration are the same sinusoid marching a quarter-cycle apart. Speed peaks exactly where the displacement — and with it the force — is zero, while acceleration peaks at the turning points, always pointing opposite to the displacement.

Figure

Displacement, velocity, and acceleration over one period — each a quarter-cycle out of step

The math: from force law to timetable

Take a frictionless horizontal track, displacement $x$ measured from the spring's natural length. The only horizontal force is the spring's, so Newton's second law reads

m\ddot x = -kx \qquad \Longrightarrow \qquad \ddot x + \omega^2 x = 0, \quad \omega \equiv \sqrt{\frac{k}{m}} \qquad (1)

where: $m$ — mass of the block (kg); $k$ — spring constant (N/m); $\omega$ — angular frequency (rad/s), fixed by the hardware before the motion even starts. (You will often see it written $\omega_0$ and called the natural frequency — the subscript earns its keep once damping and driving forces enter the story.) Read the structure before solving anything: the equation contains exactly one knob, $\omega^2 = k/m$ — stiffness over inertia, the spring's eagerness divided by the block's reluctance. No amplitude appears anywhere in it, which means the isochronism of the overview is already guaranteed at the level of the equation itself. Every measurable in this article — period, frequency, peak speed, peak acceleration — will be built from this single $\omega$ .

Derivation AFrom force law to sinusoid by the characteristic-root method

We are deriving the complete general solution of $\ddot x + \omega^2 x = 0$ — not guessing a cosine and checking it, but constructing the answer so we know nothing has been missed. The strategy is the standard weapon for every linear constant-coefficient equation in physics: try an exponential $e^{\lambda t}$ , find which exponents the equation permits, and superpose the allowed solutions. The step to watch is the very last one — counting constants. A second-order equation must carry exactly two arbitrary constants; if our final answer had one or three, we would know we had lost or invented solutions along the way.

Step 1. Try $x = e^{\lambda t}$ , so that $\ddot x = \lambda^2 e^{\lambda t}$ . Substituting into (1) gives $(\lambda^2 + \omega^2)\,e^{\lambda t} = 0$ , and since $e^{\lambda t}$ never vanishes, the bracket must: $\lambda^2 = -\omega^2$ . Why an exponential: differentiation maps $e^{\lambda t}$ to a multiple of itself, so a differential equation collapses into an algebraic one for $\lambda$ — this single trick solves every linear ODE with constant coefficients.

Step 2. The roots are purely imaginary: $\lambda = \pm i\omega$ . Why this matters physically: imaginary roots happen because $k > 0$ makes $+\omega^2$ positive — the force pushes back. Flip the sign of the force (a bead balanced on a dome, $F = +kx$ ) and the roots become real, $\lambda = \pm\omega$ , giving runaway exponentials instead of oscillation. The sign of the stiffness decides between vibrating and toppling.

Step 3. Superpose the two independent solutions and demand that $x$ be real. Euler's formula $e^{\pm i\omega t} = \cos\omega t \pm i\sin\omega t$ converts the complex pair into a real pair:

x(t) = C_1\cos\omega t + C_2\sin\omega t \qquad (2)

where: $C_1, C_2$ — real constants fixed by the initial position and velocity: $C_1 = x(0)$ and $C_2 = \dot x(0)/\omega$ . Why two constants: the equation is second order, so two numbers — where the mass starts and how fast it is moving — must pin down the entire future. Two constants for two initial conditions: the solution is complete, not a lucky guess.

Step 4. Fold the two terms into a single cosine using $C_1 = A\cos\phi$ , $C_2 = -A\sin\phi$ , which gives the form used everywhere in practice, together with its derivatives:

x(t) = A\cos(\omega t + \phi), \qquad v(t) = -A\omega\sin(\omega t + \phi), \qquad a(t) = -A\omega^2\cos(\omega t + \phi) \qquad (3)

where: $A = \sqrt{C_1^2 + C_2^2}$ — the amplitude, the maximum displacement; $\phi$ — the phase constant, encoding where in the cycle the clock started. Note $a = -\omega^2 x$ holds at every instant — the acceleration is a live, continuously updated copy of the displacement, flipped and rescaled.

Step 5. Read off the period. The cosine repeats when its argument grows by $2\pi$ : $\omega T = 2\pi$ , so

T = \frac{2\pi}{\omega} = 2\pi\sqrt{\frac{m}{k}}, \qquad f = \frac{1}{T} = \frac{1}{2\pi}\sqrt{\frac{k}{m}} \qquad (4)

where: $T$ — period (s); $f$ — ordinary frequency (Hz), cycles per second, related to the angular frequency by $\omega = 2\pi f$ . Neither contains $A$ : the period is amplitude-independent. Verified: substituting (3) into (1) gives $-A\omega^2\cos(\omega t+\phi) + \omega^2 A\cos(\omega t+\phi) = 0$ identically; the two-constant count matches the order of the equation. Matches Morin Ch. 4 and Taylor Ch. 5.

Why does the amplitude cancel? Double the pull: the typical force doubles, so the typical acceleration doubles — but the block also has twice the distance to cover, and the two effects cancel exactly when the force law is linear. Bend the force law even slightly (a cubic spring, a large-angle pendulum) and the cancellation breaks, the period starts to drift with amplitude, and the system is no longer a good clock.

Which formula answers which question

you want	use	notes
position, speed, acceleration at time $t$	eq. (3)	need $A$ and $\phi$ from initial conditions
the timing — period or frequency	$T = 2\pi\sqrt{m/k}$ , eq. (4)	amplitude never enters
peak speed / peak acceleration	$v_{\max} = A\omega$ , $a_{\max} = A\omega^2$ , eq. (5)	one power of $\omega$ apart
speed at a given position	$v = \omega\sqrt{A^2 - x^2}$	from energy conservation, eq. (7)
stiffness from a stopwatch	$k = m\omega^2 = 4\pi^2 m/T^2$	eq. (4) run backwards
frequency of any stable system	$\omega = \sqrt{U''(x_0)/m}$ , eq. (11)	the universality result

Go deeper · Derivation B · 2 minMaximum speed two ways, and where the energy lives

Derivation BMaximum speed two ways, and where the energy lives

The target is the oscillator's complete energy story: the peak speed $v_{\max}$ , the conserved total $E$ , and how that total divides between kinetic and potential forms over a cycle. The strategy is deliberately redundant — we compute $v_{\max}$ first by calculus from the explicit solution, then again from energy conservation without ever touching the solution, and demand the two answers agree. Getting the same formula from two independent routes is the cheapest error-detector in physics; watch also for the factor $m\omega^2 = k$ , which quietly converts between the kinetic and potential bookkeeping throughout.

Step 1. Calculus route. From (3), $|v| = A\omega\,|\sin(\omega t+\phi)|$ is largest when the sine hits $\pm 1$ — which is exactly when the cosine, and hence $x$ , is zero:

v_{\max} = A\omega = A\sqrt{\frac{k}{m}}, \qquad a_{\max} = A\omega^2 \;\;\text{(at } x = \pm A\text{)} \qquad (5)

where: $v_{\max}$ — speed at the equilibrium crossing; $a_{\max}$ — acceleration magnitude at the turning points, where the force peaks. Speed and force peak a quarter-cycle apart — one is maximal exactly where the other vanishes.

Step 2. Energy route. At a turning point the block is at rest and the spring holds all the energy: $E = \tfrac12 kA^2$ . At the equilibrium crossing the spring is relaxed and the block holds all of it: $E = \tfrac12 m v_{\max}^2$ . Equate the two:

\tfrac12 kA^2 = \tfrac12 m v_{\max}^2 \;\;\Longrightarrow\;\; v_{\max} = A\sqrt{\frac{k}{m}} = A\omega \qquad (6)

Why this cross-check counts: differentiation and energy bookkeeping are logically independent arguments, and they land on the same formula — exactly the kind of two-route verification working physicists live by.

Step 3. Now prove the total really is conserved at every instant, not just at the two special points. Define $E = \tfrac12 m\dot x^2 + \tfrac12 kx^2$ and differentiate:

\frac{\mathrm{d}E}{\mathrm{d}t} = m\dot x\ddot x + kx\dot x = \dot x\,(m\ddot x + kx) = \dot x \cdot 0 = 0 \qquad (7)

Why the bracket vanishes: $m\ddot x + kx = 0$ is the equation of motion itself — energy conservation is not an extra assumption but a direct consequence of the force law being derivable from the potential $U = \tfrac12 kx^2$ . So $E = \tfrac12 kA^2$ at every instant, and solving $\tfrac12 mv^2 + \tfrac12 kx^2 = \tfrac12 kA^2$ for the speed gives $v = \omega\sqrt{A^2 - x^2}$ at any position.

Step 4. Finally, average over a cycle. Inserting (3) into each energy gives $U = \tfrac12 kA^2\cos^2(\omega t+\phi)$ and $K = \tfrac12 mA^2\omega^2\sin^2(\omega t+\phi) = \tfrac12 kA^2\sin^2(\omega t+\phi)$ , using $m\omega^2 = k$ . Over a full period $\langle\cos^2\rangle = \langle\sin^2\rangle = \tfrac12$ , so

\langle K\rangle = \langle U\rangle = \tfrac14 kA^2 = \tfrac12 E \qquad (8)

where: $\langle\cdot\rangle$ — the time average over one period. The oscillator splits its budget exactly in half between motion and configuration — the classical seed of the equipartition idea that later blooms in statistical mechanics. Verified: $\langle K\rangle + \langle U\rangle = \tfrac12 kA^2 = E$ as required; energy split matches Goldstein Ch. 6 and Landau & Lifshitz §21.

Notice one subtlety in the energy movie: each form of energy peaks twice per cycle — the block passes equilibrium once moving left and once moving right — so kinetic and potential energy oscillate at frequency $2\omega$ , twice the frequency of the motion itself.

The vertical spring — gravity moves home, not the clock

Hang the same block from the same spring. Gravity stretches it to a new equilibrium $x_0 = mg/k$ , and measuring displacement $y$ from that point, the net force is $-k(x_0 + y) + mg = -ky$ . The constant gravity term cancels exactly: the block oscillates about the shifted equilibrium with the same $\omega$ and the same period. A spring oscillator ticks at the same rate on Earth, on the Moon, and in free fall — which is precisely why astronauts measure their body mass with one.

Two more languages for the same motion

The sinusoid can be dressed in two other costumes, and both become indispensable one article downstream. First, the complex exponential. Since $\cos\theta = \mathrm{Re}\,e^{i\theta}$ , the whole two-constant solution compresses into a single complex number:

x(t) = \mathrm{Re}\!\left[C\,e^{i\omega t}\right], \qquad C = Ae^{i\phi} \qquad (9)

where: $C$ — the complex amplitude, packing both the amplitude $|C| = A$ and the phase $\arg C = \phi$ into one symbol; $i$ — the imaginary unit. The payoff is that differentiation becomes multiplication by $i\omega$ : the differential equation turns into algebra. For the undamped oscillator this is elegant; for the damped and driven oscillators it is the difference between a page of trigonometry and three lines. Second, the phase-space portrait. Divide the conservation law (7) by $E$ and the trajectory in the $(x, v)$ plane becomes

\left(\frac{x}{A}\right)^2 + \left(\frac{v}{A\omega}\right)^2 = 1 \qquad (10)

where: the motion traces a closed ellipse with semi-axes $A$ and $A\omega$ , traversed clockwise once per period. Closed curve = perfectly repeating motion = perfectly conserved energy; when damping enters, this ellipse will spiral into the origin, and the picture makes the energy loss visible at a glance.

Why the oscillator is everywhere

Here is the result that promotes this article from "springs" to "physics." Nature's potentials are not parabolas — molecular bonds, gravitational wells, and chemical landscapes are lopsided, saturating, complicated curves. But zoom in on the bottom of any smooth valley and it looks parabolic. That single geometric fact is why the harmonic oscillator is the most reused calculation in science.

Figure

Any smooth potential is a parabola near its minimum — the harmonic approximation

Go deeper · Derivation C · 1 minUniversality — every stable equilibrium oscillates harmonically

Derivation CUniversality — every stable equilibrium oscillates harmonically

We are proving the claim that licenses everything else in this article: a particle displaced slightly from any stable equilibrium of any smooth potential executes simple harmonic motion, with a frequency read off from the curvature of the potential. The strategy is a Taylor expansion of $U(x)$ about the equilibrium point, keeping terms honestly and watching each one die for a stated reason. The point to watch is the word stable: the argument needs $U''(x_0) > 0$ , and its failure modes — zero or negative curvature — are physically meaningful, not edge-case pedantry.

Step 1. Expand the potential about the equilibrium $x_0$ :

U(x) = U(x_0) + U'(x_0)\,(x - x_0) + \tfrac12\,U''(x_0)\,(x - x_0)^2 + \cdots

Why we may: smoothness of $U$ is the only assumption — the expansion is Taylor's theorem, not physics.

Step 2. The constant $U(x_0)$ has no gradient and exerts no force — drop it. The linear term vanishes because $x_0$ is an equilibrium: the force there is $-U'(x_0) = 0$ by definition. The first term that survives with any influence is the quadratic one:

U(x) \approx \tfrac12\,k_{\mathrm{eff}}\,u^2, \qquad u \equiv x - x_0, \qquad k_{\mathrm{eff}} = U''(x_0)

where: $u$ — the small excursion from equilibrium; $k_{\mathrm{eff}}$ — the effective stiffness, which is nothing but the curvature of the potential at its minimum. Stability is the statement $U''(x_0) > 0$ — the valley curves upward — so $k_{\mathrm{eff}}$ is a genuine positive spring constant.

Step 3. Differentiate to get the force and apply Newton's law:

F = -\frac{\mathrm{d}U}{\mathrm{d}u} = -k_{\mathrm{eff}}\,u \qquad\Longrightarrow\qquad m\ddot u + U''(x_0)\,u = 0

Why this is the punchline: the equation is (1) all over again, with the potential's curvature standing in for the spring constant. Nothing about springs was used — only the existence of a smooth minimum.

Step 4. Read off the universal small-oscillation frequency:

\omega = \sqrt{\frac{U''(x_0)}{m}} \qquad (11)

where: $U''(x_0)$ — the second derivative of the potential evaluated at the minimum (N/m). Sharper valley, faster vibration; flatter valley, slower. When $U''(x_0) = 0$ (a quartic-bottomed valley) the harmonic term is absent, the leading force is cubic, and the period becomes amplitude-dependent — the approximation, not the oscillation, is what fails. Verified: for $U = \tfrac12 kx^2$ equation (11) returns $\omega = \sqrt{k/m}$ ; for the pendulum, $U = mgL(1 - \cos\theta)$ has $U''(0) = mgL$ and inertia $mL^2$ , giving $\omega = \sqrt{g/L}$ — both standard results. Matches Landau & Lifshitz §21.

One equation, many costumes

The same three-symbol equation runs systems that share not a single atom of hardware. In each row of the table, identify what plays "displacement," what resists change (the inertia), and what pushes back (the stiffness) — then every result of this article transfers wholesale.

system	"displacement"	"mass"	"stiffness"	frequency
block on spring	extension $x$	$m$	$k$	$\sqrt{k/m}$
simple pendulum (small angle)	angle $\theta$	$mL^2$	$mgL$	$\sqrt{g/L}$
diatomic molecule	bond stretch $r - r_0$	reduced mass $\mu = \frac{m_1 m_2}{m_1 + m_2}$	bond curvature $U''(r_0)$	$\sqrt{U''(r_0)/\mu}$
LC circuit	charge $q$	inductance $L$	inverse capacitance $1/C$	$1/\sqrt{LC}$

The molecular row is infrared spectroscopy in miniature: measure the vibration frequency of a bond and you have measured the curvature of the interatomic potential — a probe of quantum chemistry executed with this article's classical formula. The LC row is why radios tune: the circuit oscillates at $1/\sqrt{LC}$ , with charge sloshing between capacitor and inductor exactly as energy sloshes between spring and mass. Same equation, same solutions, same isochronism.

Limiting cases

Case	Result	Meaning
$A \to$ larger (linear regime)	$T$ unchanged, $v_{\max} \propto A$	isochronism: amplitude buys speed and energy, never time
stiffer spring, $k \to 4k$	$T \to T/2$	frequency doubles — stiffness accelerates the clock
heavier block, $m \to 4m$	$T \to 2T$	inertia slows the clock
vertical spring	same $T$ , equilibrium shifts by $mg/k$	gravity relocates the center of oscillation, not the period
any stable equilibrium, small $A$	SHM with $\omega = \sqrt{U''(x_0)/m}$	the parabola at the bottom of every smooth valley — universality
damping switched on, $\gamma \to 0$ limit	$\ddot x + 2\gamma\dot x + \omega^2 x = 0 \to$ eq. (1)	SHM is the idealized zero-friction limit; real amplitudes decay
nonlinear force $F = -kx - \beta x^3$	$T$ depends on $A$	isochronism is a property of linearity, not of springs

Common trap“Pull it farther and each cycle takes longer”

The single most tempting wrong intuition about oscillators. A bigger pull means more distance to cover — but it also means proportionally bigger forces and speeds everywhere along the path, and for a linear force the two effects cancel to the last decimal: $T = 2\pi\sqrt{m/k}$ contains no $A$ . Watch the companion traps too: $v_{\max} = A\omega$ but $a_{\max} = A\omega^2$ — mixing them up is a classic exam error — and maximum speed happens at $x=0$ , where the force is zero, not at the turning points, where the force is largest. Finally, keep the two frequencies straight: $\omega$ (rad/s) and $f$ (Hz) differ by $2\pi$ , and quoting one where the other belongs is the most common numerical slip in this chapter.

Worked examples

1 · The standard setup (confidence builder)

A 0.50 kg block on a frictionless table is attached to a spring with $k = 200\ \mathrm{N/m}$ , pulled $8.0\ \mathrm{cm}$ from equilibrium, and released. Find $\omega$ , $T$ , $f$ , and $v_{\max}$ .

Step 1. Angular frequency from the hardware: $\omega = \sqrt{k/m} = \sqrt{200/0.50} = \sqrt{400} = 20\ \mathrm{rad/s}$ .

Step 2. Period: $T = 2\pi/\omega = 2\pi/20 = \boxed{0.314\ \mathrm{s}}$ — about three cycles per second.

Step 3. Frequency: $f = 1/T = 3.18\ \mathrm{Hz}$ .

Step 4. Maximum speed: $v_{\max} = A\omega = (0.080)(20) = \boxed{1.6\ \mathrm{m/s}}$ .

Step 5. Energy cross-check: $E = \tfrac12 kA^2 = \tfrac12(200)(0.080)^2 = 0.64\ \mathrm{J}$ , so $v_{\max} = \sqrt{2E/m} = \sqrt{2(0.64)/0.50} = \sqrt{2.56} = 1.6\ \mathrm{m/s}$ ✓. Note what happens if the block is instead pulled 16 cm: the period stays $0.314$ s to the digit, while $v_{\max}$ doubles to 3.2 m/s.

2 · Reading the hardware off a stopwatch (exam level)

A 1.2 kg block hanging from an unknown spring completes 15 full oscillations in 9.0 s at amplitude $A = 5.0\ \mathrm{cm}$ . Find the spring constant, the maximum speed, the maximum acceleration, and the block's speed when it is 2.5 cm from equilibrium.

Step 1. Period from the timing: $T = 9.0/15 = 0.60\ \mathrm{s}$ , so $\omega = 2\pi/T = 10.47\ \mathrm{rad/s}$ .

Step 2. Invert equation (4) for the stiffness: $k = m\omega^2 = (1.2)(10.47)^2 = \boxed{132\ \mathrm{N/m}}$ . (That the spring hangs vertically is irrelevant — gravity shifts the equilibrium, not the period.)

Step 3. Maximum speed: $v_{\max} = A\omega = (0.050)(10.47) = \boxed{0.52\ \mathrm{m/s}}$ , at the equilibrium crossing.

Step 4. Maximum acceleration: $a_{\max} = A\omega^2 = (0.050)(10.47)^2 = 5.5\ \mathrm{m/s^2}$ — over half a $g$ , at the turning points.

Step 5. Speed at half amplitude, from energy conservation: $v = \omega\sqrt{A^2 - x^2} = 10.47\sqrt{(0.050)^2 - (0.025)^2} = 10.47 \times 0.0433 = \boxed{0.45\ \mathrm{m/s}}$ .

Step 6. Sanity check Step 5: at half the displacement the spring holds only a quarter of the energy ( $U \propto x^2$ ), so the kinetic energy is $\tfrac34 E$ and $v = v_{\max}\sqrt{3}/2 = 0.52 \times 0.866 = 0.45\ \mathrm{m/s}$ ✓ — the block is still moving at 87% of top speed halfway out.

Origins · 1 minChristiaan Huygens and how this was discovered

Christiaan Huygens

1629–1695 · The Hague & Paris

Huygens turned isochronism into technology. In 1656–57 he built the first pendulum clock, improving timekeeping accuracy roughly a hundredfold overnight, and in Horologium Oscillatorium (1673) — arguably the finest mechanics book before Newton's Principia — he analyzed oscillating systems with full mathematical rigor, deriving the pendulum period and inventing the cycloidal cheeks that make a pendulum truly isochronous at all amplitudes. In 1675 he devised the spiral balance spring, a literal Hooke's-law oscillator that made portable watches viable; Hooke claimed priority, and the ensuing quarrel echoed for decades. Every mechanical watch running today still beats to Huygens' spring–mass rhythm.

History

1581

— Galileo, timing a swinging chandelier in the Pisa cathedral against his pulse, notices that wide and narrow swings seem to take the same time.

1656

— Huygens builds the first pendulum clock; timekeeping error drops from minutes to seconds per day.

1660

— Hooke discovers the linear law of elasticity, the ingredient that makes spring oscillation exactly harmonic.

1673

— Huygens' Horologium Oscillatorium gives the first rigorous mathematics of oscillation, including the pendulum period.

1675

— Huygens invents the spiral balance-spring regulator for watches; Hooke bitterly disputes priority.

1687

— Newton's Principia supplies

F = ma

, turning Hooke's force law into the differential equation of harmonic motion.

1739

— Euler systematically solves linear differential equations with constant coefficients, delivering the general sinusoidal solution.

Going further

Texts. Morin, Introduction to Classical Mechanics, Ch. 4 (Oscillations) · Taylor, Classical Mechanics, Ch. 5 · Goldstein, Classical Mechanics, 3rd ed., Ch. 6 · Landau & Lifshitz, Mechanics (Vol. 1), §21–23.

Provenance. Equation of motion and solution from Morin Ch. 4 and Taylor Ch. 5; energy conservation and equipartition from Goldstein Ch. 6; universality of small oscillations from Landau & Lifshitz §21. A verifier should recompute: (1) $\omega_0=\sqrt{k/m}$ and $T=2\pi\sqrt{m/k}$ ; (2) that $x=A\cos(\omega_0t+\phi)$ satisfies $\ddot{x}=-\omega_0^2x$ ; (3) the energy split $\langle KE\rangle=\langle PE\rangle=\tfrac14kA^2$ using $\langle\cos^2\rangle=\langle\sin^2\rangle=\tfrac12$ ; (4) the pendulum and LC analogues $\omega_0=\sqrt{g/L}$ and $\omega_0=1/\sqrt{LC}$ . No empirical constants are quoted beyond standard textbook relations. Worked-example numbers on this page ( $\omega = 20$ rad/s, $T = 0.314$ s, $v_{\max} = 1.6$ m/s; $k = 132$ N/m, $v_{\max} = 0.52$ m/s, $a_{\max} = 5.5\ \mathrm{m/s^2}$ , $v = 0.45$ m/s at $x = A/2$ ) are illustrative arithmetic recomputed from the boxed formulas.

Builds on: Newton's Second Law · Taylor Series Expansion · Second-Order Linear ODEs. Leads to: Amplitude, Phase, and Angular Frequency · Energy in Harmonic Motion · Small Oscillations About Equilibrium · Complex Exponential Representation of Oscillations · The Simple Pendulum · Damped Harmonic Oscillator.

If the velocity is zero at the turning point, why doesn't the block just stay there?

Zero velocity is not zero force. At $x = +A$ the spring is at maximum stretch, so the restoring force — and with it the acceleration $a = -\omega^2 x$ — is at its largest magnitude of the whole cycle. The block is momentarily at rest but is being yanked back toward equilibrium as hard as it ever will be, so it immediately picks up speed again. Compare a ball thrown straight up: at the top its velocity is zero but gravity is still acting, and down it comes. "At rest" and "in equilibrium" mean different things — the block is only in equilibrium at $x = 0$ , which is precisely the one point where it is never at rest (it crosses at top speed).

Why doesn't the period depend on amplitude? Pulling farther really feels like it should take longer

Pull the block twice as far and it does have twice the distance to cover — but the linear force law hands it exactly twice the force, hence twice the acceleration and twice the typical speed, everywhere along the path. Twice the distance at twice the speed takes the same time; the cancellation is exact, not approximate. The proof is that $A$ simply never appears in $\ddot x + \omega^2 x = 0$ — the timing was decided before you chose how far to pull. But the cancellation is a property of linearity: for a large-angle pendulum or a stiffening spring the force law bends, the balance breaks, and the period does start to drift with amplitude.

What is the difference between ω and f, and how do I know which one a formula wants?

They count the same motion in different units: $f$ counts cycles per second (Hz), $\omega$ counts radians of phase per second, and one cycle is $2\pi$ radians, so $\omega = 2\pi f$ . The circle picture makes it concrete: $f$ is laps per second, $\omega$ is the angular speed of the point running the circle. Formulas born from the differential equation — $\omega = \sqrt{k/m}$ , $v_{\max} = A\omega$ , $a_{\max} = A\omega^2$ — always want $\omega$ , because derivatives of $\cos(\omega t)$ pull down factors of $\omega$ , not $f$ . Lab instruments and musical pitch quote $f$ . The classic exam slip is computing $\omega = 20$ rad/s and reporting "20 Hz"; the honest value is $f = 20/2\pi \approx 3.2$ Hz.

How do we know x = A cos(ωt + φ) is the whole answer and not just one solution among many?

Because of the constant-counting argument, which is the quiet backbone of the characteristic-root derivation. A second-order equation needs exactly two pieces of initial data — position and velocity — to determine the future, so its general solution carries exactly two free constants. The exponential method produced two independent solutions ( $e^{i\omega t}$ and $e^{-i\omega t}$ , repackaged as cosine and sine), and the combination $A\cos(\omega t + \phi)$ has precisely the two knobs $A$ and $\phi$ . Any initial condition can be matched, and the uniqueness theorem for linear ODEs guarantees that a solution matching given initial data is the solution. Nothing is missing.

Why does hanging the spring vertically not change the period? Gravity is obviously pulling on the block

Gravity is a constant force, and a constant force cannot change the stiffness — it can only relocate the point where forces balance. The new equilibrium sits at $x_0 = mg/k$ , where spring pull equals weight; measure displacement $y$ from there and the mathematics returns exactly $m\ddot y = -ky$ , the gravity term having cancelled against the pre-stretch. The block oscillates about the new home at the old frequency. The general lesson: adding any constant force to a harmonic oscillator shifts the center of oscillation and changes nothing else. This is also why the ISS body-mass chair works identically in orbit — the period never cared about the local gravity in the first place.

What's the point of the complex exponential form if the position is a real number?

It is a bookkeeping device that converts calculus into algebra. Differentiating $e^{i\omega t}$ just multiplies it by $i\omega$ , so a differential equation becomes a polynomial equation, and the two real constants ( $A$ , $\phi$ ) travel together inside one complex number $C = Ae^{i\phi}$ . You do the whole computation in complex form and take the real part at the end — legitimate because the equation is linear with real coefficients, so real and imaginary parts separately satisfy it. For the undamped oscillator the saving is modest; for damped and driven oscillators it is decisive, which is why the complex-exponential article is the next stop on this track. Electrical engineers call the same trick phasors and run all of AC circuit theory on it.

At x = A/2 the block is halfway out — why is the kinetic energy 3/4 of the total and not 1/2?

Because potential energy grows with the square of displacement, not linearly. At half the amplitude the spring stores $U = \tfrac12 k(A/2)^2 = \tfrac14 \cdot \tfrac12 kA^2$ — a quarter of the total, not half. Conservation hands the remaining three quarters to the kinetic term, so $v = v_{\max}\sqrt{3}/2 \approx 0.87\,v_{\max}$ : the block still moves at 87% of top speed halfway to the turning point. The lesson generalizes: the oscillator spends most of its time near the turning points, where it crawls, yet carries most of its speed through a narrow zone around equilibrium. The quadratic energy landscape, not linear intuition, sets the split.

Real springs aren't perfectly linear — so is SHM ever exactly true, and when does the approximation break?

For a real spring, exactness is an idealization; the honest statement is that SHM is the leading behaviour for small enough displacements, with corrections controlled by the next terms of the Taylor expansion of the potential. How small is "small enough" depends on the system: a pendulum at $20°$ has its period wrong by only about $0.8\%$ (the correction is $\theta_0^2/16$ ), while a molecular bond stretched a few percent of its length is already visibly anharmonic. The breakdown signature is always the same — the period starts depending on amplitude and the motion picks up overtones at $2\omega$ , $3\omega$ . The anharmonic pendulum article on this track quantifies exactly this failure.

Simple Harmonic Motion

Where this lives today

The physical picture

The math: from force law to timetable

Which formula answers which question

The vertical spring — gravity moves home, not the clock

Two more languages for the same motion

Why the oscillator is everywhere

One equation, many costumes

Limiting cases

Worked examples

1 · The standard setup (confidence builder)

2 · Reading the hardware off a stopwatch (exam level)

Going further

Before this

Where this leads