Gravity by Coincidence Pendulum theory

In mechanical metrology, a Vernier caliper achieves sub-millimeter precision by comparing two spatial scales with slightly different graduation intervals (e.g. 9 mm divided into 10 divisions). The coincidence method operates on the exact identical principle in the time domain. When two periodic oscillators have nearly equal periods T and T₀, their phase difference Δφ evolves continuously at the beat frequency Δf = |1/T − 1/T₀|. At the moment of coincidence, their phase difference is an exact integer multiple of 2π. Between two consecutive coincidences in the same direction, the phase difference advances by exactly 2π radians (one complete cycle). Because this phase accumulation takes place over dozens or hundreds of oscillation cycles (a coincidence interval t_c of 20 to 100 seconds), any human timing error Δt in triggering the stopwatch at the coincidence instant is divided by the large number of elapsed oscillations N = t_c / T. Consequently, the relative uncertainty in the deduced period δT/T = (δt_c / t_c) · (T / t_c) is reduced by a factor of 100 to 500 compared to directly timing individual swings. This brilliant insight allowed 19th-century physicists such as Borda, Kater, and Bessel to determine absolute gravity g to five significant figures long before the invention of electronic frequency counters.

Consider a master clock pendulum of standard period T₀ and a test pendulum of period T, where T < T₀ (the test pendulum oscillates more rapidly than the master). Let both pendulums pass through the central equilibrium vertical line at time t = 0 traveling from left to right. After an elapsed time t, the number of complete cycles executed by the master pendulum is n₀ = t / T₀, and the number of cycles executed by the test pendulum is n = t / T. Since T < T₀, the test pendulum completes cycles faster and gains on the master clock: Δn(t) = n(t) − n₀(t) = t(1/T − 1/T₀). A coincidence occurs when the two pendulums again pass through the central equilibrium position traveling in the same direction. For the first subsequent coincidence in the same direction at time t = t_c, the test pendulum must have gained exactly one full cycle relative to the master: Δn(t_c) = 1. Therefore: t_c (1/T − 1/T₀) = 1. Dividing by t_c gives: 1/T − 1/T₀ = 1/t_c, which rearranges to: 1/T = (t_c + T₀) / (T₀ t_c). Taking the reciprocal yields the fundamental coincidence equation: T = (T₀ · t_c) / (t_c + T₀). In the alternative case where the test pendulum is longer than the master clock (T > T₀), the master pendulum gains one cycle over the test pendulum during t_c, giving 1/T₀ − 1/T = 1/t_c, whence T = (T₀ · t_c) / (t_c − T₀). Maintaining T strictly less than T₀ prevents mathematical singularity (t_c → ∞) and eliminates sign ambiguity.

For a simple pendulum undergoing small-amplitude planar oscillations in a uniform gravitational field, the period of oscillation is governed by the classical simple harmonic oscillator formula: T = 2π √(L / g), where L is the distance from the pivot axis to the center of mass of the bob and g is the local acceleration due to gravity. Squaring both sides yields the linear relation: T² = (4π² / g) · L. When multiple measurements of the coincidence interval t_c are recorded across a series of calibrated lengths L₁, L₂, ..., L_n, the deduced periods T_i allow construction of a linear plot of T² on the vertical axis against L on the horizontal axis. In accordance with the model y = S · x + b, the slope S obtained by ordinary least squares (OLS) regression corresponds to: S = 4π² / g. Solving for the acceleration due to gravity isolates g: g = 4π² / S. A major metrological advantage of the slope method over single-point calculations is that any small systematic offset in measuring the length (such as an undetected zero-point calibration error in the cathetometer scale ΔL_offset) shifts all x-values by a constant amount, altering the y-intercept b = −S · ΔL_offset but leaving the slope S = d(T²)/dL and the resulting value of g completely invariant.

The textbook equation T = 2π √(L / g) is derived by approximating the restoring gravitational torque τ = −m g L sin θ using the first-order Taylor expansion sin θ ≈ θ. For finite angular displacement amplitudes θ₀, the exact non-linear equation of motion d²θ/dt² + (g/L) sin θ = 0 yields an exact solution in terms of the complete elliptic integral of the first kind K(k): T(θ₀) = 4 √(L/g) K(sin(θ₀/2)) = 2π √(L/g) · [1 + (1/4) sin²(θ₀/2) + (9/64) sin⁴(θ₀/2) + ...]. Expanding for small angles in radians yields the classic Borda-Euler circular arc correction: T(θ₀) ≈ T_ideal · [1 + (1/16) θ₀² + (11/3072) θ₀⁴ + ...]. If a pendulum is released at an initial amplitude of θ₀ = 10° (0.1745 rad), the circular arc increases the period by (1/16)(0.1745)² ≈ 0.0019 (a 0.19% elongation), which would introduce an unacceptable 0.38% systematic error in g. In high-precision coincidence metrology, the initial swing amplitude is strictly restricted to θ₀ ≤ 2.5° (0.0436 rad), where the fractional period elongation is (1/16)(0.0436)² ≈ 0.000119 (0.012%). Furthermore, applying Borda’s correction factor [1 − (1/16) θ₀²] to the raw period eliminates this systematic error entirely, aligning the empirical measurement with ideal harmonic theory.

A true mathematical simple pendulum assumes that the entire mass is concentrated at an infinitesimal point. In a physical laboratory apparatus, the bob is a solid sphere of finite radius r and mass M suspended by a wire of mass m_w. According to the parallel axis theorem, the moment of inertia of the spherical bob about the knife-edge pivot axis located at distance L from the sphere center is: I = I_cm + M L² = (2/5) M r² + M L². The equation of motion for small angular oscillations gives the period of a physical (compound) pendulum: T = 2π √(I / (M g L)) = 2π √[((2/5) M r² + M L²) / (M g L)] = 2π √[(L + (2/5) r² / L) / g]. Comparing this with the simple pendulum formula shows that the effective physical length is L_eff = L + (2/5)(r² / L), which exceeds the center-to-pivot distance L by the radius-of-gyration correction ΔL_bob = (2/5)(r² / L). For a brass sphere of radius r = 1.40 cm (0.0140 m) suspended at length L = 0.800 m, the correction is ΔL_bob = 0.4 × (0.0140)² / 0.800 = 0.000098 m ≈ 0.098 mm (~0.1 mm). While small, omitting this correction causes a systematic error in g of Δg / g = −ΔL / L ≈ −0.012%. In metrological work, L is adjusted to L_eff to guarantee exact consistency.

A pendulum oscillating in ambient air is subject to three distinct fluid dynamic effects: static Archimedean buoyancy, aerodynamic Stokes-form drag, and inviscid hydrodynamic added mass. First, Archimedes' principle reduces the effective gravitational weight of the bob: W_eff = (M − m_air) g = M g (1 − ρ_air / ρ_bob), where ρ_air ≈ 1.20 kg/m³ and ρ_brass ≈ 8400 kg/m³. This buoyancy reduction weakens the gravitational restoring torque by the factor (1 − ρ_air / ρ_bob). Second, as the sphere moves through air, it entrains a boundary layer and accelerates a volume of surrounding fluid. According to classical potential flow hydrodynamics, an accelerating sphere carries an added virtual fluid mass equal to half the displaced fluid mass: m_added = (1/2) m_air. The total inertial mass of the bob becomes M_inertial = M + (1/2) m_air. Combining buoyancy and added mass yields the effective frequency: ω² = [M g (1 − ρ_air/ρ_bob)] / [L (M + (1/2) ρ_air V_bob)] ≈ (g/L) [1 − (3/2)(ρ_air / ρ_bob)]. For brass in air, (3/2)(1.20 / 8400) ≈ 0.000214, meaning ambient air elongates the period by approximately 0.011% and depresses the uncorrected g by 0.021%. In absolute gravimetry laboratories, this correction is applied or pendulums are swung in high vacuum enclosures.

The legacy of pendulum gravimetry directly underpins modern geophysics, geodesy, navigation, and fundamental metrology. Until the mid-20th century, reversible pendulums (such as Kater’s and Cambridge pendulums) employing the coincidence method served as the world primary standards for absolute gravitational mapping, establishing the International Gravity Standardization Net (IGSN 71). Today, absolute gravimetry has evolved into laser-interferometric free-fall gravimeters (such as the Micro-g LaCoste FG5), which drop a retroreflector prism in an ultra-high vacuum chamber and track its trajectory with a helium-neon laser referenced to an atomic rubidium clock, achieving accuracies of 1 to 2 microgals (1 μGal = 10⁻⁸ m/s²). Furthermore, quantum gravimeters utilizing cold-atom atom interferometry drop laser-cooled rubidium or cesium atoms and pulse them with Raman lasers to measure quantum phase shifts proportional to g with extraordinary drift-free stability. Gravimetric measurements reveal tidal deformations of the Earth crust caused by the Moon and Sun, post-glacial rebound, tectonic magma migration beneath active volcanoes, and underground aquifer depletion. The core principles of timing synchronization, phase coherence, and gravitational field perturbation established through coincidence pendulum experiments remain foundational across modern physics and Earth observation satellites such as GRACE-FO.