residual stress. Prior to the tests, the specimens were kept under a 75% relative humidity condition at a temperature of 293 K for the purpose of adjusting the water content in the specimens, because the humidity condition is optimal for stable crazes growth according to our previous studies [4-5]. All the specimens had been stored in the humidity until the water content in the specimens saturated with the humidity, for more than a year [1]. Experimental procedure A fatigue testing machine, which can be installed on a scanning laser microscope (SLM) (made by Lasertec Co., Ltd.), has been developed to observe in-situ the crack tip shape on a side surface. The details of the machine are shown in Fig. 2. An ultrasonic motor drives a screw to apply load on a specimen. printer monitor video recorder scanning Laser microscope (SLM) load controller computer humidity generator tensile testing machine SLM objective lens crack propagation direction Figure 2: In-situ observation using scanning laser microscope(SLM) Prior to the crack propagation tests, pre-cracks of 1mm had been introduced to the specimens under a cyclic loading condition (at a mean stress intensity factor (Kmean) of 0.26 MPam1/2, at an R ratio of 0.1 and at a sinusoidal frequency of 1 Hz). Tensile load was applied to a specimen until the main crack started to grow and then the load was decreased until the crack stopped growing. This threshold stress intensity for a monotonic loading was defined as Kth, which was 0.78 MPam1/2, in this study. Seventy percent of Kth, 0.55 MPam1/2, was chosen for the maximum stress intensity factor (Kmax) for the fatigue crack growth tests. Fatigue tests were performed at Kmax of 0.55 MPam1/2- using triangle waves at a frequency of 0.01 Hz and at stress ratios of 0.1, 0.3, 0.5 and 0.7, to clarify the effect of closure on the formation of striations on fracture surfaces. The vicinity of fatigue cracks was observed on a side surface of a specimen with changing R-ratio and the images were recorded using the a video recorder. After the in-situ observations, the PMMA specimen was broke in two by overloading without unloading to observe the fatigue fracture surface. The fracture surfaces of the specimens were sputter-coated with
RkJQdWJsaXNoZXIy MjM0NDE=