The influence of dynamic effects on the crack arrest process is investigated. For propagating and subsequently arresting cracks, actual dynamic stress intensity factors were measured applying a shadow optical technique in combination with a Cranz Schardin high-speed camera. The experiments were performed in wedge-loaded double-cantilever-beam (DCB) specimens machined from an epoxy resin (Araldite B). In the initial phase of crack propagation the measured dynamic stress intensity factors were found smaller; in the arresting phase, however, they were larger than the corresponding static values. After arrest the dynamic stress intensity factor oscillates with decreasing amplitude around the static stress intensity factor at arrest. Crack arrest toughness values determined according to a static analysis showed a dependence on the crack velocity prior to arrest, but the dynamic crack arrest toughness yielded a single value only, indicating that this quantity represents a true material property.

Crack arrest toughness Kla values are usually evaluated using a static stress intensity factor analysis. According to this procedure which is advocated by Crosley and Ripling [i], the process of crack arrest is treated as mathematically identical to the process of crack initiation, with only the time scale being reversed. Dynamic effects (stress waves, structural vibrations) and their possible influence on the process of crack arrest are not considered. This concept is contested by Hahn et al. [2]. These authors analyzed the energetics of fracture arrest and concluded that, as the crack accelerates in the initial phase, kinetic energy is built up in the specimen which is subsequently available to contribute to the crack driving force during the arresting phase of crack propagation.In this work the possible influences of kinetic energy are investigated by measuring the actual dynamic stress intensity factors for arresting cracks and comparing these to the corresponding static values.The crack arrest experiments were performed in wedge-loaded DCB specimens (321x127x10 mm) made from an epoxy resin (Araldit B). Notches with different root radii were used to achieve different initial crack velocities. Dynamic stress intensity factors for propagating and subsequently arresting cracks were determined using the shadow spot method of Manogg [3] in combination with a Cranz~Schardin high speed camera; the static stress intensity factors for the corresponding crack lengths were obtained using the conventional stress intensity factor formulae from deflection measurements at the loading point.Experimental results for cracks, initiated at different values of the stress intensity factor at initiation (KI_) and thus propagating at different initial crack velocities, are summarized indFig, i. Fig. la shows values of the dynamic stress intensity factor KIyn (experimental points in the upper part of the diagram) as a function of crack length a compared with the corresponding static stress intensity factor K~ tat curve. The measured velocities are given in the lower par t of the diagram additionally. The same results as in Fig. la, but plotted in a different form, are shown in Fig. lb. Simplified curves have been drawn through the measured data points. The very early phase,of crack propagation (dashed lines), being of less importance in this context, has not been investigated.The following characteristics of the crack arrest process can be drawn from Figs. la and Ib:i. At the beginning of the crack propagation phase the dynamic stress intensity factor K~ yn is smaller than the corresponding static value K~ tat. At the end of the crack propagation phase the dynamic stress intensity factor K~yn is larger than the corresponding static value K~ tat. Only after arrest does the dynamic stress intenInt Journ of Fracture 12 (1976) 318 2.factor K~yn approach the statically determined stress intensity sity factor at ~ arrest K stat Ia "The velocity of the propagating crack stays constant for a considerable period but decreases to smaller value...

For different crack arrest test specimens, the influence of dynamic effects on the crack arrest process is investigated. Utilizing the shadow optical method of caustics, actual dynamic stress-intensity factors directly before, at, and after arrest were measured. In the first part of the investigation, specimens made from the model material Araldite B were analyzed. It was found that the dynamic effects, that is, the difference between statically and dynamically determined crack arrest stress-intensity factors, are largest for longitudinal wedge-loaded rectangular double cantilever beam specimens, considerably smaller for machine-loaded tapered double cantilever beam specimens, and smallest for transverse wedge-loaded compact specimens. In the second part of the paper, the behavior of arresting cracks in high-strength steel specimens was investigated. The overall dynamic effects on the crack arrest process were found to be similar to those in Araldite B, but in steel the oscillation of the dynamic stress-intensity factor after arrest shows higher-frequency disturbances and is damped out only after much longer times.

No abstract

Les a p p l i c a t i o n s de l a procedure de mesure sont i c i demontrees en determinant l a thnaciti? sous impact de 2 a c i e r s d i f f e r e n t s 2i diverses temperatures d ' e s s a i .