This preview shows page 1. Sign up to view the full content.
Unformatted text preview: 2g
∂y Hence if the restriction of f to ∂D vanishes,
f ∆gdxdy = −
D ∇f · ∇gdxdy.
D Similarly, if the restriction of g to ∂D vanishes,
g ∆f dxdy = −
D ∇f · ∇gdxdy.
D Thus if f and g lie in V ,
f , ∆g = − ∇f · ∇gdxdy = g , ∆f .
D In particular, if (5.32) holds, then
µ f , g = f , ∆g = ∆f, g = λ f , g ,
(λ − µ) f , g = 0.
151 dxdy. It follows immediately that either λ − µ = 0 or f , g = 0, and the lemma is
Now let us focus on the special case in which D is a circular disk. Recall the
problem that we want to solve, in terms of polar coordinates: Find
u(r, θ, t), 0 < r ≤ 1,
so that ∂2u
r ∂r r ∂u
∂r u(r, θ + 2π, t) = u(r, θ, t), + 1 ∂2u
r2 ∂θ2 u well-behaved near r = 0, u(1, θ, t) = 0,
(r, θ, 0) = 0.
It follows from the argument in the preceding section that the general solution to the homogeneous linear part of the problem can be expressed in terms
of the eigenfunctions
u(r, θ, 0) = h(r, θ), fn,k (r, θ) = Jn (αn,k r/a) cos(nθ), gn,k = Jn (αn,k r/a) sin(nθ). Indeed, the general solution must be a superposition of the products of these
eigenfunctions with periodic functions of t of frequencies −λn,k /2π . Thus this
general solution is of the form
∞ u(r, θ, t) = a0,k g0,k (r, θ) cos(α0,k t)
k=1 ∞ ∞ + [an,k fn,k (r, θ) + bn,k gn,k (r, θ)] cos(αn,k t). (5.33) n=1 k=1 We need to determine the an,k ’s and the bn,k ’s so that the inhomoge...
View Full Document
- Winter '14