Page 290 - Biaxial Multiaxial Fatigue and Fracture
P. 290

274                      A. BUCZWSH AND G. GLINKA

             The deviatoric and the actual stress components Sir and  doa at the end of each load increment
             are determined from eq.(14-15).










             where: n denotes the number of the load increments.
             The actual strain increments,   can finally be determined from the constitutive equation (3).


             CYCLIC PLASTICITY MODEL

             In order  to  predict  the  notch  tip  stress-strain response of  a  notched  component  subjected to
             multiaxial cyclic loading, the incremental equations discussed above have to be linked with the
             cyclic plasticity model. Several plasticity models are available in the literature. The most popular
             is the  model of  Ref.  [9] proposed by  Mroz. According to  Mroz  [9] the uniaxial stress-strain
             material  curve  can  be  represented  in  a  multiaxial  stress  space  by  a  set  of work-hardening
             surfaces.





             In the case of a two-dimensional stress state (01 la = 012'  = q3'=  0), such as that one at a notch
             tip, the work-hardening surfaces can be represented by ellipses in the coordinate plane for which
             the axes are defined  by the directions of principal stress components (Fig. 5). The equation of
             each work-hardening ellipse in the two-dimensional principal stress space is:




             The essential elements of the plasticity model can be presented in such a case graphically in a
             two-dimensional principal stress space.
              The  load  path  dependency effects are  modeled  by  prescribing a  rule  for  the  translation  of
             ellipses in the  d2a-03a  plane.  The translation of these ellipses is assumed to be caused by the
             sought stress increment, represented in the principal stress space as a vector. The ellipses can be
             translated with respect to each other over distances dependent on the magnitude of the stress/load
             increment. The ellipses move within the boundaries of each other, but they do not intersect. If an
             ellipse comes in contact with another, they move together as one rigid body.
              However,  it  has  been  found  that  the  ellipses  in  the  original  Mroz  model  may  sometimes
             intersect  each  other,  which  is  not  permitted.  Therefore,  Garud  proposed  [lo]  an  improved
             translation rule that prevents any intersections of plasticity surfaces. The principle idea of the
             Garud translation rule is illustrated in Fig. 6.

             a) The line of action of the stress increment, Ad',  is extended to intersect the next larger non-
               active surface, fi, at point B2.
   285   286   287   288   289   290   291   292   293   294   295