Pioneer Anomaly: Thermal Recoil, Plan View and Force Budget (2D)

Why did two 1970s probes seem to be pulled toward the Sun by 8.7e-10 m/s²? Heat carries momentum. Here you see it from above and next to the other forces. Top view, Sun at the top; orange = heat photons (anisotropy drawn 10x), blue arrow = recoil toward the Sun; bars compare the forces on a log scale.
Thermal recoil acceleration-
Share of the reported anomaly-
Net aft fraction needed now-
= net directed heat power-
Solar gravity here-
Sunlight pressure here-
Velocity offset after N yr-
Position offset after N yr-
Heat power now-
Blue: recoil from the decaying RTG heat. Orange band: reported anomaly (8.74e-10 m/s^2, typical value).
Accumulated velocity offset (the quantity Doppler tracking sees): integral of the recoil acceleration, mm/s.
Independent numerical checks
Each line compares two separate calculations; a FAIL would show here.
The science

Photon recoil. Emitted radiation of power P carries momentum flux P/c. If a fraction ε of the heat leaves as a net beam along one axis (the rest is isotropic and cancels), the force is F = εP/c and a = εP/(mc). For P = 2500 W, m = 241 kg, ε = 2.4 % this gives ≈ 8.3×10−10 m/s², i.e. only about 60 W of net directed heat explains the reported value.

The anomaly. Doppler tracking of Pioneer 10 and 11 beyond about 20 AU showed a residual acceleration of roughly 8.7×10−10 m/s² toward the Sun (Anderson et al., 2002; quoted here as a typical value). Analyses of the spacecraft's thermal design and telemetry (for example Turyshev et al., 2012) found anisotropic heat emission can account for essentially all of it. No new physics is needed.

Known forces for scale. Solar gravity GM⊙/r² is about 3.7×10−6 m/s² at 40 AU, so the anomaly is ~0.02 % of it. Solar radiation pressure L⊙/(4πr²c) · A Cₓ/m (A = 5.9 m² dish, Cₓ = 1.2, illustrative) pushes outward by ~10 % of the anomaly here. The log-scale bars in this 2D view put these side by side.

Decay and Doppler. Heat from Pu-238 falls as 2^(−t/87.7 yr); the heat the dish sees follows the isotope decay (thermocouple degradation is not modelled). Doppler tracking measures the velocity offset, the time integral of the acceleration (second chart).

Simplifications. One effective ε, no view-factor ray tracing of the real bus, no thermal inertia, no solar-heat contribution, constant mass, 1-D radial coast. In the plan view the photon anisotropy is exaggerated 10× so the effect is visible; arrow length scales with the acceleration relative to the reported value. The probe is a generic Pioneer-style craft (public dimension ~2.7 m dish), not an engineering drawing; the sprite is a Blender render of the 3D page's model.

Checks. (1) A Monte-Carlo ensemble of 106 photons reproduces the closed form; (2) a leapfrog orbit with and without the extra force shows an offset ½ a t²; (3) RK4 integration of the decay law gives half the power after one half-life; (4) new in this 2D view: trapezoid integral of the decaying recoil equals the closed-form velocity offset.

Made with 🟠 Blender