Numerical Methods for PDEs MATH 610 - 600 Alan Demlow Homework 1 Spring 2015 Homework 1 Exercise 1 30% Recall that for sufficiently smooth functions u : R → R, the following relation holds Z x u(x) = u(y) + u0 (s)ds. y 1 Let C ([0, 1]) be the set of function u : [0, 1] → R that are continuous and have continuous first derivative on [0, 1]. Consider the following subsets of C 1 ([0, 1]) : A = {u ∈ C 1 ([0, 1]) : u(0) = u(1) = 0}, B = {u ∈ C 1 ([0, 1]) : u(0) = 0}, C = C 1 ([0, 1]). Derive any five of the following six inequalities (6 points each) : Z Z 1 1 1 02 u dx, ∀u ∈ B, u2 dx 6 2 0 0 Z Z 1 1 1 02 u dx, ∀u ∈ A. u2 dx 6 4 0 0 Show that for u ∈ C the following inequalities are valid : Z 1 Z Z 1 2 1 1 02 2 u dx 6 u dx + udx ; 6 0 0 0 Z 1 max |u(x)|2 6 2u2 (1) + 2 u02 dx; x∈[0,1] Z 0 1 u2 (x)dx 6 2u2 (1) + 2 0 1 u02 (x)dx; 0 2 Z max |u(x)| 6 2 x∈[0,1] Exercise 2 Z 1 (u2 + u02 )dx. 0 30% We say that u ∈ L2 (a, b), a, b ∈ R, has a weak derivative in L2 (a, b) if there exists v ∈ L2 (a, b) satisfying Z b Z b 0 u(x)φ (x)dx = − v(x)φ(x)dx, ∀φ ∈ C0∞ ([a, b]). a a The function v is called the weak derivative of u and the space of functions in L2 (a, b) with weak derivative in L2 (a, b) is denoted by H 1 (a, b). 2 1. Consider the function u(x) = |x| 3 − 1 defined on the interval (−1, 1). Determine whether u ∈ H 1 (−1, 1) and if so, give an explicit formula for its weak derivative. 2. Let u1 ∈ H 1 (−1, 0) and u2 ∈ H 1 (0, 1). Show that if u1 (0) = u2 (0) then the function u1 −1 < x < 0; u(x) = u2 0 < x < 1; belongs to H 1 (−1, 1). Explain why if u1 (0) 6= u2 (0), then u 6∈ H 1 (−1, 1). Exercise 3 20% Let V be the closure of the set {v ∈ C 1 [0, 1] : v(0) = 0} in the usual H 1 norm kvkH 1 (0,1) = kukL2 (0,1) + ku0 kL2 (0,1) . Consider the variational formulation : Find u ∈ V such that Z 1 0 0 Z (k(x)u v + uv)dx − u(1)v(1) = 0 1 f vdx ∀v ∈ V. 0 Assuming that 2 ≤ k(x) ≤ 4, show that the corresponding bilinear form is coercive and continuous in V (equipped with the usual H 1 norm). Derive the strong form of this problem. Exercise 4 20% Consider the following second order boundary value problem : Seek u : (0, 1) → R such that (−k(x)u0 )0 + b(x)u0 + q(x)u = f, x ∈ (0, 1) supplemented with the boundary conditions u(0) = u(1) = 0. Here f ∈ L2 (0, 1) and k, q, b, b0 ∈ L∞ (0, 1) satisfy k(x) > 0 and q(x) − 12 b0 (x) > 0 for almost every x ∈ (0, 1). 1. Determine a weak formulation of the above strong problem. 2. Show that there exists a unique weak solution u ∈ H01 (0, 1).