Page 31 - INTRODUCTION TO THE CALCULUS OF VARIATIONS
P. 31

18                                                        Preliminaries

                                                                                p
                          We now turn our attention to the notions of convergence in L spaces. The
                       natural notion, called strong convergence, is the one induced by the k·k L p norm.
                       We will often need a weaker notion of convergence known as weak convergence.
                       We now define these notions.
                                                n
                       Definition 1.15 Let Ω ⊂ R be an open set and 1 ≤ p ≤∞.
                                                                                   p
                          (i) A sequence u ν is said to (strongly) converge to u if u ν ,u ∈ L and if

                                                 lim ku ν − uk L p =0 .
                                                ν→∞
                                                                  p
                       We will denote this convergence by: u ν → u in L .
                          (ii) If 1 ≤ p< ∞,we say that thesequence u ν weakly converges to u if u ν ,
                            p
                       u ∈ L and if
                                        Z
                                                                           0
                                                                          p
                                    lim   [u ν (x) − u (x)] ϕ (x) dx =0, ∀ϕ ∈ L (Ω) .
                                   ν→∞
                                         Ω
                                                                  p
                       This convergence will be denoted by: u ν  u in L .
                          (iii) If p = ∞,the sequence u ν is said to weak ∗ converge to u if u ν ,u ∈ L ∞
                       and if
                                         Z
                                                                           1
                                     lim   [u ν (x) − u (x)] ϕ (x) dx =0, ∀ϕ ∈ L (Ω)
                                    ν→∞
                                          Ω
                                               ∗
                       and will be denoted by: u ν  u in L .
                                                       ∞
                       Remark 1.16 (i) We speak of weak ∗ convergence in L ∞  instead of weak con-
                                                                                      1
                       vergence, because as seen above the dual of L ∞  is strictly larger than L .For-
                                                         p
                       mally, however, weak convergence in L and weak ∗ convergence in L ∞  take the
                       same form.
                          (ii) The limit (weak or strong) is unique.
                          (iii) It is obvious that
                                                  ⎧             p
                                                  ⎨ u ν  u in L    if 1 ≤ p< ∞
                                              p
                                   u ν → u in L ⇒
                                                  ⎩     ∗
                                                     u ν  u in L ∞  if p = ∞ .
                       Example 1.17 Let Ω =(0, 1), α> 0 and
                                                   ⎧   α
                                                   ⎨ ν    if x ∈ (0, 1/ν)
                                           u ν (x)=
                                                   ⎩
                                                      0   if x ∈ (1/ν, 1) .
   26   27   28   29   30   31   32   33   34   35   36