10.07.2015 Views

EXERCISES PDE 23-25.01.13 1. Exercise Fourier analysis in ...

EXERCISES PDE 23-25.01.13 1. Exercise Fourier analysis in ...

EXERCISES PDE 23-25.01.13 1. Exercise Fourier analysis in ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

2 EX11for every N ∈ N. Comb<strong>in</strong><strong>in</strong>g this with (<strong>1.</strong>3) for every N ∈ N and every M ∈ N with M ≥ N , onehasM∑‖f M ′ − f N‖ ′ 2 2 = n 2 a 2 n ≤ 1 ∑MN 2 n 4 a 2 n ≤π2N 2 ‖f ′′ ‖ 2 2 .n=Nn=NWhen M goes to +∞ we therefore obta<strong>in</strong>√ π‖f ′ − f N ′ 1‖ 2 ≤2 N ‖f ′′ ‖ 2S<strong>in</strong>ce by Po<strong>in</strong>caré’s <strong>in</strong>equality‖f − f N ‖ 2 ≤ π‖f ′ − f N ′ ‖ 2we get that‖f − f N ‖ H 10 ((0,π)) ≤√ π2 (1 + π) 1 N ‖f ′′ ‖ 2as required.□2. <strong>Exercise</strong>Céa’s Lemma: Consider a symmetric, bounded, coercive bil<strong>in</strong>ear form B : H ×H → R on a Hilbertspace H , that isB(u, v) = B(v, u) , B(u, v) ≤ C 1 ‖u‖ H ‖v‖ H , B(u, u) ≥ c 0 ‖u‖ 2 H .(a) Fix a l<strong>in</strong>ear cont<strong>in</strong>uous functional F : H → R. Justify that there exists a unique u ∈ H suchthatB(u, v) = F (v)for every v ∈ H .(b) Fix additionally an N -dimensional subspace H N of H . Justify the follow<strong>in</strong>g fact: there exista unique approximate solution u N ∈ H N , that is a unique u N ∈ H N such thatfor every v N ∈ H N .(c) Prove the follow<strong>in</strong>g equalityB(u N , v N ) = F (v N )B(u − u N , u − u N ) + B(v N − u N , v N − u N ) = B(u − v N , u − v N ) (2.1)for every v N ∈ H N . Deduce Céa’s estimate:√C‖u − u N ‖ H ≤ 1c 0<strong>in</strong>f{‖u − v N ‖ : v N ∈ H N } . (2.2)(d) Consider now H = H0 1 ((0, π)) and H N := span{t ↦→ s<strong>in</strong>(nt) : n = 1, . . . , N}. For a(x) ∈C 1 [0, π] with 1 ≤ a(x) ≤ 2 for every x and f ∈ L 2 ((0, π)) setB(u, v) :=ˆ π0a(t) u ′ (t) v ′ (t) dt (2.3)and F (v) := ´ π0 f(t) v(t) dt. Check that u and u N as <strong>in</strong> (a) and (b), respectively, exist, and showthat u N → u <strong>in</strong> H 1 0 ((0, π)) as N goes to +∞. If additionally u ∈ H 2 ((0, π)) show that there existsa constant C such that‖u − u N ‖ H 1 ≤ C N ‖u‖ H 2 .Solution: (a) follows from the Lax-Milgram theorem. The same holds for (b) s<strong>in</strong>ce the restriction ofB to the Hilbert subspace H N × H N is clearly bil<strong>in</strong>ear, coercive and cont<strong>in</strong>uous.(c) By bil<strong>in</strong>earity and simmetry of the form B we haveB(u − u N , u − u N ) = B(u, u) − 2B(u, u N ) + B(u N , u N )B(v N − u N , v N − u N ) = B(v N , v N ) − 2B(u N , v N ) + B(u N , u N )B(u − v N , u − v N ) = B(u, u) − 2B(u, v N ) + B(v N , v N )

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!