Fracture resistance curves (R curves) for describing ductile material behav
iour play an important rule for integrity assessment analyses. The determin
ation of R curves is widely expressed by experimentally based standardized
procedures. But the considered standard test specimens are usually small si
zed and the problem of resistance transferability from small laboratory spe
cimens under large scale yielding to large sized structures under contained
yielding emerges. It will be presented in the following a new strategy bas
ed on a cohesive type model in combination with the conventionally finite e
lement method. It allows numerical simulations of stable (ductile) tearing
uniquely for small scale yielding (presented formerly in this journal) as w
ell up to fully plastic deformation states only by two constant material pa
rameters, With the respective choise of cohesive parameters it is possible
to generate synthetical resistance curves which allows a new systematic und
erstanding of the driving influences and their margins, especially the geom
etry (loading) induced variations. By changing the two cohesive parameters
an arbitrary material can be introduced where the correlation between mater
ial and resistance behaviour can be answered. Furthermore, the inhomogeneou
s behaviour of compounds, made of different material phases, can be investi
gated straightforward and opens new perspectives in the fields of engineeri
ng testing and characterization. Conventional methods as used for homogeneo
us cases are no more sufficient for many aspects of new composed materials
and structures.