Page 542 - Handbook Of Integral Equations
P. 542

10.7-4. Integral Equation of Elasticity

               The integral equation (1) whose kernel is given via the Fourier cosine transform,
                                                   ∞  L(u)

                                           K(x)=          cos(ux) du,                      (18)
                                                  0    u
               frequently occurs in contact problems of elasticity. The function L(u) in (18) is continuous and
               positive for 0 < u < ∞) and satisfies the asymptotic relations

                                                   3
                                     L(u)= Au + O(u )         as  u → 0,
                                           N–1
                                                            
                              (19)
                                     L(u)=    B n u –n  + O u –N  as  u →∞,
                                           n=0
               where A > 0 and B 0 >0.
                   Formula (18) implies that the kernel is an even function: K(x)= K(–x).
                                                        –1
                   It is usually assumed that L(u)u –1  and u[L(u)] , treated as functions of the complex variable
               w = u + iv, are regular at the pole |v|≤ γ 1 and the pole |v|≤ γ 2 , respectively. It follows in particular
               that the kernel K(x) decays at least as exp(–γ 1 |t|)atinfinity.
                   Formulas (18) and (19) imply that K(x) has a logarithmic singularity at x = 0. Moreover, the
               representation (3) is valid with a n = 0 for n =1, 3, 5, ...
                   Thus, the kernel given by (18) has the same characteristic features as those inherent by assumption
               in the kernel of the integral equation (1). Therefore, the results of Subsections 10.7-2 and 10.7-3
               can be used for the asymptotic analysis of Eq. (1) with kernel (18) as λ →∞ and λ → 0.
                •
                 References for Section 10.7: I. I. Vorovich, V. M. Aleksandrov, and V. A. Babeshko (1974), V. M. Aleksandrov and
               E. V. Kovalenko (1986), V. M. Aleksandrov (1993).


               10.8. Regularization Methods

                 10.8-1. The Lavrentiev Regularization Method
               Consider the Fredholm equation of the first kind (see also Remark 3, Subsection 11.6-5)

                                         b
                                          K(x, t)y(t) dt = f(x),  a ≤ x ≤ b,                (1)
                                        a
               where f(x) ∈ L 2 (a, b) and y(x) ∈ L 2 (a, b). The kernel K(x, t) is square integrable, symmetric, and
               positive definite (see Subsection 11.6-2), that is, for all ϕ(x) ∈ L 2 (a, b), we have
                                           b  b

                                               K(x, t)ϕ(x)ϕ(t) dx dt ≥ 0,
                                          a  a
               where the equality is attained only for ϕ(x) ≡ 0.
                   In the above classes of functions and kernels, the problem of finding a solution of Eq. (1) is
               ill-posed, i.e., unstable with respect to small variations in the right-hand side of the integral equation.
                   Following the Lavrentiev regularization method, along with Eq. (1) we consider the regularized
               equation
                                            b

                                  εy ε (x)+  K(x, t)y ε (t) dt = f(x),  a ≤ x ≤ b,          (2)
                                           a


                 © 1998 by CRC Press LLC








               © 1998 by CRC Press LLC
                                                                                                             Page 524
   537   538   539   540   541   542   543   544   545   546   547