Page 310 - Fluid-Structure Interactions Slender Structure and Axial Flow (Volume 1)
P. 310

PIPES CONVEYING FLUID: NONLINEAR AND CHAOTIC DYNAMICS         29 1

                In  his  study,  Bourri2res considers  only  the  linear  case.  However,  his  approach,  if
              pursued  far  enough,  would  have  led  him  eventually  to  expressions  similar  to  those
              derived in Section 5.2.3.  In fact, the only difference between equations (5.47) and those
              of Section 5.2.3 lies in the absence of gravity and time-varying flow terms, not considered
              by Bourribres; this makes his work irreproachable.
                The remaining task would be to combine all the five equations of (5.47) into one, and
              to compare it with equation (5.39). This is not done here since it has effectively already
              been done by Rousselet & Herrmann (1981), to be discussed next.
              (b) Rousselet and Herrmann‘s work

              Rousselet & Herrmann (1981) derived the  equations of  motion  in  two  different ways:
              by  the  force-balance method  and  the  energy method. They  obtain  a  set  of  equations,
              fairly close to the one found in Section 5.2.3, but with some minor differences. Their first
              method follows closely Bourrikres’ work. Two differences are simply due to the addition
              of  gravity forces and the assumption that unsteady flow velocity effects may be present.
                Considering an element of  the  system (see Figure G.l), the application of  Newton’s
              law leads to
                     a                 a
                     -[(T  - P) cos 01  - -(Q  sin 0) + (m +M)g
                     as                as
                                 a2x              M u2             d6
                                                        sin0 - 2M U - sin@,
                       = (m + M) - +M U cos0 - -
                                 at2                R               dt
                     a                a                                           (5.48)
                     -[(T  - P)sin0] + -(Q  cos@
                     as               as
                                 a2z        d0        M u2
                                                           cos0 +M U  sine,
                       = (m + M) - + 2M U - cos0 + -
                                 at2        dt          R
              where R is the local radius of curvature. In these equations, (T - P) represents the tangen-
              tial forces and Q the shear force, and sin8 and cos6 are related to x and z  by
                                              az            ax
                                       sin 0 = -,    cos8 = -.                    (5.49)
                                              as            as
              By means of  the inextensibility condition and the definition of the curvature K, one can
              also prove that
                                     1  ax  aZ2       1 az   a2x
                                     R  as   -  a$’   R as   -  a$’               (5.50)
                Substituting (5.49) and (5.50) into (5.48), one obtains

                           ~((T-P):)  -:(a:)+(m+M)g
                           as

                                       aZx   +     a2 x       a2x       ax
                                        at2        asat       as2       as ’
                             = (m +M) - 2M U  ~        +M U2 - +M U -
                                                                                  (5.51)
                           as  ((T-P$)  +;  (e;)

                                       a2z         a2 z       a2z      az
                             = (m+M)-      +2MU-      +MU2 - +MU-.
                                       at2        asat        as2       as
   305   306   307   308   309   310   311   312   313   314   315