Mc. Burstow et al., THE EFFECTS OF MATERIAL STRENGTH MISMATCHING ON CONSTRAINT AT THE LIMIT LOAD OF WELDED 3-POINT BEND SPECIMENS, International journal of fracture, 89(2), 1998, pp. 117-142
This paper describes the results of a series of finite element analyse
s performed to investigate the suitability of the coefficient of the J
-CTOD relationship, d(n), as a parameter to quantify constraint. Analy
ses have been performed which employ the modified boundary layer solut
ion to demonstrate the relationship between the T-stress, Q and d(n) p
arameters. Analyses have also been performed to analyse the effects of
constraint in strength mismatched welded three-point bend specimens.
These results are compared with predictions of constraint made using v
alues of d(n) derived from slip-line field solutions. Material strengt
h overmatching is shown to cause a significant loss in constraint, whi
lst undermatching increases constraint. On the whole, predictions of t
he effects of constraint from slip-line field solutions are shown to a
gree with the measured constraint levels obtained using the finite ele
ment method, although the results from highly undermatched joints are
not as accurate as the others examined. This is shown to be due to the
effect of the base material outside the weld an the crack tip stress
fields. By employing a two-material idealisation of the modified bound
ary layer formulation, using elastic T-stresses to model the constrain
t due to the specimen geometry and the normalised load parameter, J/h
sigma(Yw), to control the size of the plastic zone relative to the thi
ckness of the weld material, it was possible to reproduce the complex
stress fields encountered in each of the specimens.