Um estudo da conveniência da calibração das rugosidades absolutas e fatores de atrito de tubos de uma rede de distribuição de água com dados transientes

Abstract

This paper studies the adequacy of estimation of absolute roughness of pipes in an existing water distribution network on the basis of hydraulic heads observed during transient condition. The Inverse Transient Method of estimation using genetic algorithms employs the Method of Caracteristics in the solution of equations of motion for water flow in the pipes. An example network is used to study various calibration problems related to the determination of absolute roughness and friction factor. The network is composed of 11 pipes, 7 nodes and 1 reservoir. A twenty seconds record of transient hydraulic heads resulting from the operation a valvue in one of the nodes is used to effect the calibration. Four cases based on the degree of knowledge about the network are studied: three are related to the estimation of absolute roughness and the other calibrating flow-independent friction factors directly. It is found that the quality of estimation of the absolute roughness is improved as the knowledge about the kind of pipes is augmented. The calibration of absolute roughness shows itself to be more apropriate than the direct estimation of friction factor as the former considers the variation of friction factors during the transient event