Paper reference: ECF 13 008 04371 X - 1 - EFFECTS OF GEOMETRY AND MATERIAL ON THE ENERGY DISSIPATION RATE W. Brocks1 and Th. Siegmund2 1) GKSS Research Centre, Geesthacht, Germany 2) Purdue University, West Lafayette, USA ABSTRACT The concept of energy dissipation rate toghether with a numerical model which allows for splitting into global plastic work and local work of separation is applied to different specimens and materials, which reveal an extremely different resistance behaviour against ductile crack growth, namely side-grooved C(T) and M(T) specimens of a ferritic steel and thin centre cracked panels with varying crack lengths of an aluminium alloy. In both cases, however, it could be shown, that the dissipated work is mainly due to global plastic deformation of the specimens, once again proving that JR-curves are not an appropriate measure of a material's fracture toughness. Though the dissipation rate is also geometry dependent, of course, scaling by limit load factors of the respective geometries is succesful in certain cases which opens perspectives for transfering experimental data from one geometry to another. The applied cohesive zone model is not only an effecitive tool for simulating crack growth phenomena but provides a method of characterizing the materials resistance against ductile crack extension by two parameters, namely the cohesive strength, σmax, and the separation energy, Γc, INTRODUCTION The concept of energy dissipation rate, R = dUdis/da, as proposed by TURNER [1] has brought some better understanding of ductile tearing resistance. The rate quantity R, which measures the increment of irreversible external work necessary to propagate the crack by some amount, Δa, is physically more meaningful for describing the resistance of a structure against ductile crack extension than the conventionally used cumulative J integral. Whereas JR-curves keep rising even for steady state crack extension, R was shown to decrease with Δa and approach a stationary value [2]. Nevertheless, the new concept has not yet found wide acceptance in application, since R, like J, depends on the geometry and the type of loading of the specimen or the structure. Hence, it does not solve the problem of missing transferability of resistance curves. It has been shown for a ferritic steel and two specimen types, namely a C(T) and a M(T) specimen, that this geometry dependence could be scaled by their respective limit load factors [3]. However, a detailed study on the geometry effects on R( Δa) curves for various materials and specimen geometries [4] exhibited limitations of this scaling. Classical elastic-plastic fracture mechanics suffers from the inherent inability of separating local work of fracture from global work of (remote) plastic deformation. Numerical simulations however make it possible to split up the total dissipated energy [5], within the limitations of any model, of course. This is illustrated on numerical simulations of experimental tests.
RkJQdWJsaXNoZXIy MjM0NDE=