The Two Postulates
All of special relativity follows from exactly two postulates: (1) the laws of physics are identical in all inertial frames, and (2) the speed of light in vacuum is the same for all observers regardless of the motion of source or detector. From these two statements alone — with no additional assumptions — every result in special relativity follows by pure algebra.
The key quantity is the Lorentz factor:
Lorentz Factor γ
γ = 1 / √(1 − v²/c²) = 1 / √(1 − β²)
β = v/c (velocity as fraction of light speed)
γ → 1 as v → 0 (classical limit)
γ → ∞ as v → c (no massive object reaches c)
At v = 0.87c: γ = 2.0 (clocks run at half rate)
At v = 0.99c: γ ≈ 7.1 (clocks run at 1/7 rate)
At v = 0.9999c: γ ≈ 70.7
Time Dilation & Moving Clocks
If a clock moves at velocity v relative to an observer, the observer sees it tick more slowly by factor γ. This is not an illusion or an artefact of signal travel time — it is a genuine geometric effect of moving through spacetime. GPS satellites travel at ~14 000 km/h and their onboard atomic clocks run slow by 7 microseconds per day due to special-relativistic time dilation (partially offset by the general relativistic gravitational blueshift of ~45 µs/day).
Time Dilation
Two animated clocks — Alice (stationary) and Boris (moving at β). Sidereal time vs proper time, Lorentz contraction visible, Minkowski worldline, 6 presets: ISS, GPS, muon, LEP, twin paradox, γ=2. Live γ readout.
Twin Paradox
Animated spacetime diagram: Alice ages at home while her twin rockets to a star at β and returns. β slider 0.1–0.99, distance 1–20 light-years. Proper-time ticks on the rocket worldline, live age panels — Earth twin vs rocket twin vs difference.
Time Dilation & Twin Paradox Formulae
Δt′ = Δt / γ (moving frame ticks slower)
Proper time on rocket: τ = 2d / (γv)
Earth time elapsed: T = 2d / v
Age difference: ΔAge = T − τ = T(1 − 1/γ)
Example — β=0.8, d=4 ly: T=10 yr, τ=6 yr → ΔAge = 4 yr
The twin paradox is not a true paradox. It appears paradoxical because symmetry seems to demand both twins age less than the other — but the situation is not symmetric. The travelling twin must accelerate to turn around; this breaks the symmetry and uniquely identifies which twin's worldline is shorter in spacetime. The spacetime interval — not coordinate time — is the invariant that resolves the apparent contradiction.
Spacetime Diagrams & Minkowski Geometry
Hermann Minkowski reformulated special relativity geometrically in 1908, one year before his death. In a Minkowski diagram, time is plotted vertically as ct (time multiplied by the speed of light, so both axes have units of length). The key insight: a Lorentz boost is a hyperbolic rotation of the coordinate axes — not a standard Euclidean rotation, which is why moving rulers shorten instead of appearing the same length.
Minkowski Spacetime Diagram
Frame S with ct×x axes. β slider tilts the moving frame S′ axes (ct′/x′). Lines of simultaneity, yellow 45° light cone, cyan worldlines (rest and moving). Live γ, time dilation, and Lorentz contraction readout.
Lorentz Contraction
Interactive length-contraction visualiser: β slider 0–0.999, purple contracted ruler + rest-length silhouette, green transverse ruler (unchanged), animation sweep, 5 presets. Live L′ = L₀/γ, γ, and % shortening.
Lorentz Transformation
t′ = γ(t − vx/c²)
x′ = γ(x − vt)
y′ = y, z′ = z
Spacetime interval (invariant): s² = (ct)² − x² − y² − z²
Length contraction: L′ = L₀ / γ (along direction of motion)
Transverse dimensions unchanged: y′ = y, z′ = z
One of the most striking results from the Minkowski diagram: two events that are simultaneous in frame S (same t) are generally not simultaneous in frame S′. In the diagram, the x′ axis — the line of constant t′ = 0 — is tilted relative to the x axis. Events on the same horizontal line in S lie on different ct′ values, meaning they occur at different times for the moving observer. Relativity of simultaneity is not a philosophical claim — it is a geometric fact about the Minkowski metric.
The light cone is the boundary of causality. Events inside your past light cone can have caused you; events inside your future light cone you can influence. Events outside the light cone — spacelike-separated — are inaccessible: no causal link can exist between them, and different observers can disagree on which happened first. This is not a bug — it is how the universe protects causality from violation.
Mass–Energy Equivalence: E = mc²
Perhaps the most famous equation in science follows directly from special relativity. A body at rest has energy E₀ = mc². In motion, its total energy is E = γmc². The kinetic energy is the difference: K = (γ − 1)mc². For v ≪ c this reduces to the classical ½mv² — but the relativistic formula does not diverge; it simply requires an infinite force to reach c.
The practical consequence: nuclear binding energy. When protons and neutrons bind into a nucleus, the mass of the nucleus is less than the sum of its parts. This mass defect Δm corresponds to the binding energy E_b = Δm·c². For iron-56 — the most tightly bound nucleus — the binding energy per nucleon reaches ~8.8 MeV, around 0.9% of the nucleon rest mass.
Mass–Energy: E = mc²
Three modes: U-235 fission (→ Ba + Kr + 3n), D+T fusion (→ He + n), and the Bethe-Weizsäcker B/A binding energy curve. Animated approach, oscillation, split, and fly-apart phases; chain reaction with 21 nuclei; mass-defect histogram.
Nuclear Fusion Reactor
Tokamak PWR simulation: Lawson criterion nτT, D+T → He+n reaction. Animated plasma torus with 7×15 fuel rods, neutron particles, fission flashes, k_eff scale (subcritical / critical / supercritical).
Mass–Energy Relations
Rest energy: E₀ = mc²
Total energy: E = γmc²
Kinetic energy: K = (γ−1)mc²
Energy–momentum: E² = (pc)² + (mc²)²
Mass defect: Δm = (Σm_reactants) − (Σm_products)
Binding energy: E_b = Δm · c²
U-235 fission yield: ~200 MeV per event ≈ ~0.09% mass converted
D+T fusion yield: ~17.6 MeV per event ≈ ~0.37% mass converted
Gravitational Lensing
General relativity — Einstein's 1915 extension of special relativity to include gravity — predicts that massive objects curve spacetime and deflect the path of light. An observer aligned with a background source and an intervening mass sees the light bent into an Einstein ring. Off-axis alignment produces multiple arc-shaped images. The first observational confirmation came during the 1919 solar eclipse, when Eddington measured the deflection of starlight passing the Sun — matching Einstein's prediction to within 20%.
Gravitational Lensing
Einstein rings and multiple images from a point mass or galaxy cluster. Light-ray deflection via the GR formula α = 4GM/(c²b), interactive lens mass, live image positions and magnification.
Hubble Expansion
80 animated galaxies receding at v = H₀ · d. H₀ slider 50–100 km/s/Mpc, colour-coded redshift (blue→red), dashed Hubble sphere where v = c, velocity arrows, Planck/SH0ES/Hubble-1929 presets.
Gravitational lensing is now one of astronomy's most powerful tools. Weak lensing — the subtle statistical distortion of many background galaxies by foreground matter — maps the distribution of dark matter in clusters and on cosmic scales. Strong lensing occasionally produces multiple images of the same quasar separated by arcseconds, allowing astronomers to measure the Hubble constant independently from type Ia supernovae and the CMB.
Why Special Relativity Is Not "Counter-Intuitive"
Special relativity is often described as counter-intuitive. A better framing is that it is counter to low-velocity intuition — the intuition we built living at speeds far below c. The theory is perfectly self-consistent: no paradox exists in it. Every apparent contradiction dissolves when you draw the correct spacetime diagram and read off the invariants.
The Minkowski metric has a crucial sign difference from Euclidean geometry: ds² = (c·dt)² − dx² − dy² − dz². This minus sign — not a typo, not a convention — is why moving clocks tick slowly rather than quickly: maximising proper time means minimising spatial displacement, i.e., staying still.
Muons from cosmic rays confirm time dilation directly. Muons created in the upper atmosphere at ~15 km altitude travel at β ≈ 0.998 (γ ≈ 15.8). Their proper half-life is only 1.56 µs — enough to travel ~466 m before half decay at rest. Yet they reach sea level: from our frame, their clocks run slow by factor 15.8, so they "live" ~24.6 µs — long enough to travel 7.4 km. No other explanation fits the data.