Page 274 - Biaxial Multiaxial Fatigue and Fracture
P. 274

258                              M. END0


            Table 3.  Lower bounds of uniaxial fatigue limits, -1,   at R = -1 predicted for smooth nodular
            cast iron specimens.

                       Vickers                Graphite nodule   Microshrinkage cavity
                        micro-   Number of
              Material  hardness   specimens   Jarea,,    GI      6,
                                                                              Owl
                       of matrix               Pm        MPa        lun       MPa












            obtained by inserting the value of owl for 5 specimens into aw in Eq.(l I), since 3-6 specimens
            were used to determine each fatigue limit in the experiments. It is seen that the curve predicted
             for the cavities is the true lower bound, though the agreement with the curve for graphite nodule
             appears to  be better.  A  small change in G,,does  not  significantly affect the predicted
             results for  awl, since awl is proportional to the -1/6 power of Jarearnax, see Eq.(4). Since, in
             FCD700, the ferrite envelopes surrounding the  graphite nodules are thin  and their effect is
             negligible, the value of &,, only for graphite nodules was used for the prediction in this
             study. However, this approximation is not always valid, for example, for JIS FCD600 nodular
             cast iron  [24]  having a typical bull's  eye  structure in which the  ferrite regions are large in
             comparison to  the  graphite nodules.  A  detailed  discussion of  how  to  apply the  prediction
             method to such a microstructure is available in reference [24].


             Relation of  smooth and hole-containing specimens with respect to non-detrimental defects

             It is known that the pure torsional fatigue strength is insensitive to small defects or notches. For
             example, in the case of annealed 0.37  % carbon steel, a hole of  100 pm  in diameter which
             caused about 20 YO reduction in fatigue strength under uniaxial loading, under pure torsional
             loading  was  non-detrimental.  Gough  and  Pollard  [33]  proposed  the  following  empirical
             criterion,  called  ellipse  quadrant,  to  fit  the  data  for  smooth specimens of  ductile metallic
             materials
                                        (ET+(%)
                                                    2
                                                     =1



             where oWo is the smooth specimens uniaxid fatigue limit, rWo is the smooth specimens torsional
             fatigue limit, oa0 is the axial or bending stress amplitude, and  r,o  is the torsional shear stress
             amplitude at  the  fatigue  limit under  combined  loading.  The  experimental fatigue limits of
             smooth specimens of annealed 0.37 YO carbon steel were 145 MPa for   and 235 MPa for -0.
   269   270   271   272   273   274   275   276   277   278   279