抄録
The present paper concerns with the method of prediction of the maximum pit depth expected for a component based on the results of laboratory test program. The method is typically illustrated with an example of determination of pitting corrosion life of air cooled heat exchanger tubes.
Various scale component models are exposed to the condition simulating the service condition of the component. The maximum pit depth is measured for each model as a function of time. The results are analyzed by means of the extreme value statistics and the maximum pit depth expected in the actual component is estimated by the concept of return period and is plotted as a function of time. Curve fitting by the least square method yields the corrosion life curve of the component. Furthermore, safety factor is statistically determined and incorporated into the prediction, which can be useful in determination of the thickness of the tube to be employed. The corrosion life curve of an air cooled aluminum heat exchanger tube predicted by the laboratory test program is compared with its field performance. A safety margin is indicated in the prediction and is attributed to a higher corrosion aggressivity of the environment to which the model is exposed.
