Potential of the component ρ_c (the proof notes (8′), per component):
2 L_c(x) = kC a c + wC c x (|x| ≤ a), 2 L_c(x) ≤ kC a c + wC c x (all x).
Proof: with t = a cos θ, x − a cos θ = −(a/(2e))(e − q₁)(e − q₂) on the unit circle, q₁q₂ = 1,
q₁ + q₂ = 2x/a; the Poisson formula for log|· − q| evaluates the balayage part and Jensen's
formula the
arcsine part. Outside [-a, a] the inequality reduces to ψ(q) ≥ 0 for q ≥ 1, proved by ψ′ ≥ 0.
theorem
Zeta32.Analytic.EnergyI.integral_Lx_fC
{a c : ℝ}
(ha : 0 < a)
(hc : 0 < c)
(x : ℝ)
(hL : IntervalIntegrable (Lx a x) MeasureTheory.volume 0 (2 * Real.pi))
:
∫ (θ : ℝ) in 0..2 * Real.pi, Lx a x θ * fC a c θ = ((1 / (8 * Real.pi) * ∫ (θ : ℝ) in 0..2 * Real.pi, Lx a x θ * PK (Complex.I * ↑(betaC a c)) θ) + 1 / (8 * Real.pi) * ∫ (θ : ℝ) in 0..2 * Real.pi, Lx a x θ * PK (Complex.I * ↑(-betaC a c)) θ) - c / uC a c / (4 * Real.pi) * ∫ (θ : ℝ) in 0..2 * Real.pi, Lx a x θ
Linear decomposition of ∫ L_x fC through the Poisson form of fC.