Page 337 - Schaum's Outline of Theory and Problems of Advanced Calculus
P. 337

328                            IMPROPER INTEGRALS                         [CHAP. 12


                                                           M        ð M

                                                           ð
                                                      2        x 2      y 2
                                                     I M ¼   e  dx     e  dy
                                                           0         0
                                                         ð  M  ð  M
                                                                 2  2
                                                              e  ðx þy Þ dx dy
                                                        ¼
                                                          0  0
                                                         ðð
                                                              2  2
                                                            e  ðx þy Þ dx dy
                                                        ¼
                                                         r M
                          where r M is the square OACE of side M (see Fig. 12-3).  Since integrand is positive, we have
                                                 ðð                  ðð
                                                      2  2       2        2  2
                                                    e  ðx þy Þ  dx dy @ I M @  e  ðx þy Þ  dx dy     ð1Þ
                                                 r 1                 r 2
                          where r 1 and r 2 are the regions in the first quadrant bounded  y
                                                      ffiffiffi
                                                     p
                          by the circles having radii M and M 2, respectively.
                              Using polar coordinates, we have from (1),
                                                          ffiffi
                                                         p
                             ð   =2  ð M           ð   =2  ð  M 2
                                       2       2              2                  D
                                   e    d  d  @ I M @     e     d  d   ð2Þ
                              ¼0   ¼0                ¼0  ¼0
                          or                                                     E             C
                                          M 2   2          2M 2
                                     ð1   e  Þ @ I M @  ð1   e            M√2
                                    4               4        Þ       ð3Þ
                          Then taking the limit as M !1 in (3), we find        M
                               2   2            p ffiffiffi
                           lim I M ¼ I ¼  =4 and I ¼   =2.
                          M!1
                                                                                 O             A    B
                                  ð                                                                   x
                                   1   2
                     12.32. Evaluate  e  x  cos  xdx.
                                   0                                                  Fig. 12-3
                                      ð
                                       1   2
                                        e  x  cos  xdx.  Then using integration by
                              Let Ið Þ¼
                                       0
                          parts and appropriate limiting procedures,
                                       dI  ð 1   x 2       1  x 2      1  ð 1   x 2
                                              xe  sin  xdx ¼ e  sin  xj 0      e  cos  xdx ¼  I
                                                                    1
                                       d   ¼  0            2           2  0              2
                          The differentiation under the integral sign is justified by Theorem 8, Page 314, and the fact that
                          ð
                           1    x 2                                                        x 2        x 2
                             xe  sin  xdx is uniformly convergent for all   (since by the Weierstrass test, jxe  sin  xj @ xe
                           0
                              ð
                               1   2
                          and   xe  x  dx converges).
                               0
                              From Problem 12.31 and the uniform convergence, and thus continuity, of the given integral (since
                                              ð
                             2         x 2     1   x 2
                             x
                          je  cos  xj @ e  and   e  dx converges, so that that Weierstrass test applies), we have
                                               0
                                       1  p ffiffiffi
                          Ið0Þ¼ lim Ið Þ¼  2   .
                                !0
                                                          ffiffiffi
                                    dI                   p             p ffiffiffi  2
                                                                             =4
                              Solving                                     e   .
                                          2
                                    d   ¼  I subject to Ið0Þ¼  2  ,we find Ið Þ¼  2
                                            ð              p ffiffiffi           ð
                                                     2                          2   2
                                             1                              1
                                               e  ðx  =xÞ    .(b) Evaluate   e  ðx þx Þ  dx.
                     12.33. (a)Prove that Ið Þ¼       dx ¼
                                             0              2              0
                                            ð
                                            1
                                                    2      2
                          (a)We have I ð Þ¼ 2  e  ðx  =xÞ  ð1    =x Þ dx.
                                      0
                                            0
                                 The differentiation is proved valid by observing that the integrand remains bounded as x ! 0 þ
                              and that for sufficiently large x,
   332   333   334   335   336   337   338   339   340   341   342