increases the severity of the problem. In this work, tests that closely simulate the real life conditions of thermal shock encountered in thermal power station pressure components are completed. These tests are intended to develop data that will provide for realistic determination of lifetime to crack initiation and crack growth rates for service components as well as allow for an estimation of the conservatism of the current codes. Many influential external factors that are present in thermal power plant equipment are ignored in existing codes. Foremost is the combined effect of external primary loads and the environment in which the crack is growing. The external loads can be a direct result of the pressure or mechanical loading of the components, the effect of which is to open any cracks, exposing them to the environment. The environment is dependent on the process in which the component is being used, which in the case of thermal power station equipment is often aqueous in nature (including chemical treatment to control pH and oxygen levels) and will modify conditions at the crack tip. In this paper, selected results from crack growth tests using a unique test-rig arrangement are presented. Comparisons of the actual results with empirical prediction methods from current design codes are made. EXPERIMENTAL TECHNIQUE The testing completed in this investigation has been carried out on a thermal fatigue test rig that has been purpose built for the investigation of crack initiation and growth due to repeated thermal shock loading. Consisting of a convection furnace, static loading structure and quenching system it allows for the monitored growth of cracks for a wide variety of component geometries. The key advantages of this rig over those used in previous studies are: • The component is heated by convection, which means that there are no unwanted heat effects at the crack tip as may be the case for induction or resistance heating. • The component is quenched by room temperature pH and O2 controlled water. • The specimen size is representative of typical industrial components. • Large specimen size permits multiple simultaneous experiments. • An unloaded “control” specimen can be used. • Approximately one-dimensional conditions exist at any one crack because of the unique specimen design. A thorough analysis of the development of the test rig and specimen design, including a review of previous trends in the experimental investigations of thermal shock cracking can be found in [5]. Tests conducted with this rig have concentrated on identifying the effects of environment and primary loads on crack initiation and growth during repeated thermal shock. This was completed by simultaneously testing sets of two low carbon steel flat plate specimens (grade AS 1548-1995 [6]) placed side by side. One specimen is subjected to a 90MPa (13ksi) uniform tensile stress and the other left unloaded. The specimens, with a combination of 0.25mm and 0.1mm radius notches machined into the quenched faces, were fitted vertically in the furnace. The upper specimen temperature was limited to 370°C to remove any creep effects. Dissolved oxygen levels (D.O.) were varied between tests, while the pH was held steady at around 8.0. The first set of tests used fully oxygenated tap water with a D.O. level of around 8ppm. This water was vigorously pre-boiled in the second set of tests, driving off excess oxygen and reducing the D.O. level to around 2ppm. Each thermal cycle consisted of a slow heat to a central specimen temperature of 330°C followed by a 7s water quench. Cycle time was around 15 minutes. Due to the fact that the specimens were positioned vertically, the thermal gradient in the furnace prevented a uniform temperature from being achieved along the whole specimen length. Rather the temperature from top to bottom of the specimen varied linearly from 370°C to 290°C.
RkJQdWJsaXNoZXIy MjM0NDE=