Torsion Pendulum & Shear Modulus theory

In continuum solid mechanics, shear stress τ acts tangential to the surface of a material element, causing an angular distortion known as shear strain γ = tan(φ) ≈ φ (measured in radians). According to generalized Hooke’s law for isotropic materials, within the proportional elastic limit, shear stress is strictly proportional to shear strain: τ = η · γ, where η (also denoted by G) is the shear modulus or modulus of rigidity. While Young’s modulus Y quantifies a material’s resistance to longitudinal extension under tensile load, and the bulk modulus K quantifies resistance to isotropic volumetric compression, the shear modulus η quantifies resistance to shape distortion at constant volume. For isotropic homogeneous solids, these three fundamental elastic moduli are interrelated through Poisson’s ratio σ (the negative ratio of transverse contraction to axial extension): Y = 2 η (1 + σ) = 3 K (1 − 2σ). For typical structural steel, Poisson’s ratio is σ ≈ 0.28 to 0.30. Consequently, the shear modulus is related to Young’s modulus by η = Y / [2(1 + σ)] ≈ 0.39 Y. With Y_steel ≈ 200 GPa, the theoretical shear modulus is expected to lie near η ≈ 78 to 80 GPa, directly matching empirical torsion measurements.

To derive the restoring torque of a twisted wire of length L and circular radius r, consider the wire twisted through an angle θ about its central longitudinal axis. An imaginary coaxial cylindrical shell of radius x (where 0 ≤ x ≤ r) and thickness dx undergoes a shear strain given by the geometric arc displacement divided by length: γ(x) = (x θ) / L. By Hooke’s law in shear, the tangential shear stress acting on this shell is: τ(x) = η γ(x) = (η x θ) / L. The cross-sectional area of this infinitesimal shell is dA = 2π x dx. The infinitesimal tangential shear force acting on the shell is dF = τ(x) dA = [(η x θ) / L] · (2π x dx) = (2π η θ / L) x² dx. Because this force acts at a lever arm x from the central axis, the torque exerted by the shell is: dτ = x · dF = (2π η θ / L) x³ dx. Integrating from the center x = 0 to the outer boundary x = r yields the total restoring torque: τ = (2π η θ / L) ∫₀ʳ x³ dx = (2π η θ / L) · (r⁴ / 4) = [(π η r⁴) / (2 L)] · θ. Defining the torsional spring constant as κ = τ / θ gives Coulomb’s celebrated formula for torsional rigidity: κ = (π η r⁴) / (2 L). This rigorous derivation demonstrates that the outer layers of the wire contribute overwhelmingly to the restoring torque, creating the characteristic fourth-power dependence on radius.

The presence of the wire radius raised to the fourth power (r⁴) in the governing equation η = (8π I) / (S · r⁴) has profound consequences for experimental error propagation. Taking the natural logarithm of both sides gives: ln(η) = ln(8π I) − ln(S) − 4 ln(r). Differentiating yields the fractional error relation: dη / η = dI / I − dS / S − 4 (dr / r). By standard Gauss propagation of independent random uncertainties: (δη / η)² = (δI / I)² + (δS / S)² + 16 (δr / r)². The sensitivity coefficient multiplying the relative radius uncertainty is 4, which means that any fractional error in measuring the wire radius is magnified four-fold in the final value of the shear modulus. If the wire radius r = 0.600 mm is measured with an uncertainty of only δr = ±0.006 mm (a 1.0% error), this single measurement propagates into a 4.0% error in η! If measured with a cheap ruler (±0.05 mm, an 8.3% error), the resulting uncertainty in η would exceed 33%! This illustrates why high-precision metrology requires a digital micrometer screw gauge resolving to ±0.001 mm (0.17%), repeatedly measured across multiple orthogonal orientations to verify cylindrical circularity.

The period of a torsion pendulum depends directly on the total moment of inertia I of the suspended system about the vertical axis of oscillation: I = ∫ r_perp² dm. For a uniform solid circular disc of mass M and outer radius R, integrating in cylindrical coordinates yields the textbook result: I_disc = (1/2) M R². In advanced laboratory investigations, the moment of inertia is systematically modified by adding auxiliary inertia bodies. For a concentric circular annular ring of mass M_ring, inner radius R₁, and outer radius R₂, the moment of inertia is: I_ring = (1/2) M_ring (R₁² + R₂²). Alternatively, two identical cylindrical masses, each of mass m and radius r_c, can be placed symmetrically on opposite sides of the disc at a radial distance d from the central axis. According to Steiner’s parallel axis theorem, the total added moment of inertia is: I_added = 2 · [(1/2) m r_c² + m d²]. By recording the period T as a function of the radial distance d, plotting T² against d² yields a linear relation whose slope directly verifies the parallel axis theorem and confirms the rotational kinetic energy formulation without requiring knowledge of the wire shear modulus.

A suspended torsion pendulum possesses multiple mechanical degrees of freedom beyond pure torsional twisting. These parasitic modes include: (1) Lateral pendulum swinging, where the disc swings as a simple pendulum about the top chuck with period T_swing ≈ 2π √(L / g); (2) Vertical longitudinal bouncing, where the wire acts as a vertical spring with period T_bounce = 2π √(M / k_ext); and (3) Angular rocking (wobble), where the disc tilts out of the horizontal plane. If the center of mass of the disc is not precisely aligned with the wire axis, or if the disc is released with an accidental lateral push, non-linear geometric coupling can transfer mechanical energy between the torsional mode and the lateral pendulum mode—a phenomenon known as parametric cross-talk or autoparametric resonance. To suppress parasitic modes, the disc must be dynamically balanced, released from rest using a symmetric two-finger twist or a burning thread trigger, and timed at low angular amplitudes (θ ≤ 15°) where coupling terms remain infinitesimally small.

While Hooke’s law assumes perfect elastic reversibility with zero energy dissipation, real crystalline metals exhibit microscopic anelasticity and internal friction. As the wire twists back and forth, cyclic shear strains cause microscopic dislocation motion, grain boundary sliding, and localized thermoelastic stress relaxation. Consequently, the mechanical energy of oscillation decays exponentially with time: θ(t) = θ₀ e^{−λ t} cos(ω_d t). The rate of energy dissipation is characterized by the logarithmic decrement δ = ln(θ_n / θ_{n+1}) = λ T_d = π / Q, where Q is the mechanical quality factor of the resonator. For high-tensile cold-drawn steel in an ambient draught shield, internal friction is low (Q ≈ 2000 to 5000, δ ≈ 0.001), meaning the damped oscillation frequency ω_d = √(ω₀² − λ²) differs from the undamped natural frequency ω₀ by less than one part in a million: ω_d / ω₀ = √(1 − 1/(4Q²)) ≈ 1 − 1/(8Q²) ≈ 1 − 10⁻⁸. Thus, viscous and internal damping exert a completely negligible effect on the period T, confirming that the ideal undamped harmonic model T = 2π √(I / κ) is valid to extreme metrological accuracy.

The torsion pendulum and torsion balance occupy a towering position in the history and contemporary practice of physics and engineering. In 1798, Henry Cavendish utilized an ultra-sensitive torsion balance designed by John Michell to measure the gravitational attraction between lead spheres, providing the first accurate determination of Newton’s gravitational constant G and the mean density of the Earth. In 1900, Roland von Eötvös improved the torsion balance to test the equivalence principle (the equality of inertial and gravitational mass) to one part in 10⁹, laying the empirical cornerstone for Einstein’s General Theory of Relativity. Today, the torsion oscillator principle is omnipresent in micro-electro-mechanical systems (MEMS). Vibrating micro-machined silicon torsion mirrors steer laser beams in automotive LiDAR, optical telecommunications switches, and augmented reality laser projectors. In aerospace navigation, MEMS tuning-fork gyroscopes use electrostatic torsion bars to detect Coriolis forces, measuring angular velocity in smartphones, drones, and spacecraft inertial guidance systems. In nanotechnology, torsional resonance atomic force microscopy (TR-AFM) vibrates cantilever tips in torsion to measure in-plane friction and shear stiffness of single graphene sheets and lipid bilayer membranes.