O. Ranestad et al., Quantification of geometry and material mismatch constraint in steel weldments with fusion line cracks, INT J FRACT, 99(4), 1999, pp. 211-237
Finite element analyses of an idealised steel weldment show that the constr
aint caused by geometry and material mismatch can be separated in a Modifie
d Boundary Layer (MBL) model. The MBL model was loaded by a K-I + T displac
ement field. Analyses of four fracture mechanics specimens revealed that th
e loss of constraint in the investigated weldments was different from the c
orresponding loss of constraint in a homogeneous reference case. Therefore,
it was not possible to use the homogeneous reference to predict the develo
pment of constraint in the investigated weldments. In the fracture mechanic
s specimens the analyses show that the mismatch constraint is slightly redu
ced together with the loss of geometry constraint as large scale yielding d
evelops in the specimens. By using the mismatch constraint determined from
the MBL model, good predictions of the constraint in the specimens were obt
ained. In order to predict the fracture toughness in steel weldments with v
arying material mismatch and geometry constraint, three failure criteria ha
ve been compared. The results show that the RKR failure criterion by Ritchi
e et al. (1973) is applicable to the inhomogeneous material in this study.
The study reveals that the mismatch effect on the failure predictions is in
fluenced by the critical threshold stress used in the failure criteria. For
all the investigated criteria, the mismatch effect on the predicted toughn
ess was amplified by an elevation of the threshold stress. The constraint d
escription has been used together with the RKR failure criterion to predict
the required toughness (J(ref)) as a function of J for all the investigate
d geometries and mismatch cases.