Experimental investigations on stable crack growth in three-point bend specimens under mode I and mixed mode (I and II) loading

Abstract

In many ductile materials, there is evidence of stable crack growth prior to instability. The instability occurs at a load level higher than the crack initiation load. A design based on the initiation load may lead to an under utilisation of the material. In order to exploit the full potential of the material, it is necessary to characterise the stable crack growth. Experiments were conducted in an aircraft grade aluminium alloy (D16AT) to investigate crack initiation, stable crack growth and instability under quasi-static loadings. The three-point bend (TPB) specimen was chosen for the investigations. The ratio between initial crack length ( a 0) to width ( w) of the TPB specimen ( a 0/ w) was taken as 0.5. Different loading angles (β) of 45°, 60°, 75° and 90° were tested, where β = 90° corresponds to the mode I case. Crack opening displacement at the crack mouth (CMOD) was measured by a clip gauge for crack initiation and stable crack growth investigations. Data pertaining to load-displacement diagrams, crack initiation angle, crack initiation load ( P i ) and crack mouth opening displacement (CMOD) at different stages of crack growth, maximum fracture load ( P max ), crack front geometry etc. for different β were obtained. The CMOD at the initiation point remained almost constant within a range for the whole span of β. Initially the crack extends along a straight line for mixed mode cases, and is inclined at an angle with the initial crack. A substantial amount of tunnelling (i.e. crack front curving) is observed in all the experiments at all stages of stable crack growth. At the maximum load P max , the tunnelling is between 5.5 and 9.0mm for different β. The ratio between P max and P i ranges from 1.195 to 1.40. From these studies, it can be concluded that CMOD can be used as a characterising parameter to study stable crack growth in mode I and mixed mode (I and II) loading.