Minor Losses through Start Connectors in Microirrigation Laterals: Dimensional Analysis and Artificial Neural Networks Approaches

Abstract

In microirrigation systems, start connectors are used at lateral-manifold or microtube-lateral junctions, and minor losses occur when water flows through the start connectors. The connector’s geometry may cause contractions and/or expansions in the flow section, inducing local head losses that should be considered when designing irrigation subunits. The objective of this research was to analyze the experimental datasets available from previous studies in order to develop and validate mathematical models to estimate local head losses through start connectors using dimensional analysis and artificial neural networks. A dataset with 55,331 records of local head loss as a function of flow rate through several geometries of start connectors was evaluated. Four models based on dimensional analysis and artificial neural networks were developed and validated. The local head losses through start connectors depended on the fluid properties, operating hydraulic conditions, and geometric characteristics of the start connectors. For the full and simplified models based on dimensional analysis, 95% of discharge predictions had relative errors less than 13.0% and 14.3%, respectively. For the full and simplified models based on neural networks, 95% of discharge predictions had relative errors less than 12.2% and 12.7%, respectively. A web application was developed to facilitate the discharge predictions. Depending on the connector’s geometry, the minor losses through start connectors represent a significant percentage of the total head loss along the lateral.