Prediction of Cavitation Damage for Spillways

Abstract

Cavitation damage on a spillway is a very complex process. Although much research about cavitation and the resulting damage has been conducted for more than 30 years, existing methods for predicting cavitation damage on spillways are incomplete. This inadequacy results from cavitation damage limits (none to major) that are not exact points and from the magnitude and interaction of many factors that influence the damage. Using fuzzy mathematics, a model has been developed to predict cavitation damage on a spillway. Four factors that contribute to the damage are considered: cavitation index; flow velocity; material strength; and operating time. Fuzzy mathematics is used to set up five different damage levels from none to major and a comprehensive prediction model. The model had been tested with data from the left and right tunnel spillways at Glen Canyon Dam. The tunnels were subjected to serious cavitation damage in 1983. Using the operating data from both spillways, the model predicted damage levels, locations, and times close to the observed damages.