Crack growth in vitreous silica under dynamic loading

Abstract

To examine the fatigue phenomena observed in glass and ceramic materials, two methods are generally adopted. One method is a static fatigue test, in which a slow stable crack growth process is measured directly under a constant load or displacement [1-3] . The other is a dynamic fatigue test, in which changes in the fracture strength are measured by a tensile test at various loading rates [4]. Although the static fatigue test is useful for investigating the crack growth process in detail, it takes a long time to examine crack growth under a constant load. In contrast with the static test, the dynamic fatigue test is more efficient for studying the effect of environments and requires an increasing number of samples in order to make evaluations accurately [5]. On the other hand, recently, a slow strain rate loading technique has been used with respect to the study on stress-corrosion cracking observed in metallic materials [6]. This method is used to clarify the crack initiation and growth process in the imposing load at a very slow loading rate. It may be efficient to investigate fatigue phenomena by means of this slow strain rate loading method if a remarkable crack growth could be observed. This letter deals with the study of the stable crack-growth process in silica glass, which is widely used as a raw material in optical fibres for optical communications systems, under various dynamic loading conditions. The specimen is a wedge opening displacement (WOL) type compact tension (CT) specimen made of fused silica, which contains less than 2000 ppm impurities. The specimen configuration and precracking • procedure are described in [7]. The precracked specimen is mounted on a universal tensile testing machine with a temperature and humidity controlled chamber. The specimen is conditioned in the chamber at a constant environment for 2 h. Then, prior to the fatigue test, the pre-crack is further extended by about 2 mm with the testing machine in order to make the loading condition for stable crack growth suitable in the testing machine. In the present experiment, the following three loading methods are adopted: • 1. Slow strain rate loading: this loading was performed at a constant strain rate of 0.0005mm min -1. 2. Loading followed by holding a constant crosshead displacement: at first, a specimen is stressed at a constant strain rate of 0.005mm rain -a. When a crack begins to grow a little, the specimen is held with a constant crosshead displacement. 3. Unloading after loading: in this case, the load is applied to the specimen at a constant strain rate of 0.0005mmmin -1. When the crack begins to grow, the load is released by the inverse motion of the loading machine at the same rate. A very slow strain rate loading such as 0.0005 nlm min -1 applied to the specimen is achieved using a mechanism of one-tenth micrometer motion through a lever. The crack length is measured with an ITV camera and video recorder incorporated with a timer. A schematic diagram of the measurement system is shown in Fig. 1. The crack tip is observed using the ITV camera through a 135 mm telescope lens and recorded on the video tape recorder along with time and load data. A fibre light guide is used for photographing the specimen in the environment controlled chamber to contrast the crack tip. A video frame showing the photographed crack tip is shown in Fig. 2. As shown in Fig. 2, the time is recorded in the upper left side of the frame. The output from the load cell is also recorded in CH1 on the frame. The crack tip position in the frame is measured by the displacement analyser, i.e. the vertical cursor shown m Fig. 2, gives the transverse position in the frame, which is divided into 1024 vertical channels. The actual transverse length photographed in the frame is 10mm. Thus, the