Pressure loss measurements and predictions for net structures perpendicular to the flow.
Master thesis
Permanent lenke
http://hdl.handle.net/11250/239482Utgivelsesdato
2010Metadata
Vis full innførselSamlinger
Sammendrag
A flow loop was designed, built and used in determining the pressure loss and pressure loss coefficient over five net structures suspended normal to the water flow. The basis was to validate pressure loss calculated by SINTEF CFD simulation, which used a modified Forchheimer equation for quadratic net structures in an oil containment boom and in addition to determine which models gave best results with the empirical data collected in this diploma. The purpose of the overlaying simulation and work is to produce a containment boom with higher efficiency then what is available at present and to get correct simulations in the design process. In the later it is imperative to have correct pressure loss calculations. Empirical test data are analyzed against eight different models,three of these being different variations of the Forchheimer equation. Rudi et al. gave best agreement with data from nets where solidity was approximately 30%, while Forchheimer sequation accounting for velocity change and acceleration showed best agreement with nets of low solidity, 8-9%, however the Forchheimer equation accounting only for velocity change gave adequate results for solidity 8-9%. The intermediate net with a solidity of 18% showed good agreement with Forchheimer equation accounting only for the velocity change. The Forchheimer equation accounting for velocity increase thorugh the net was chosen for solidities from 8% to approximatly 20%. The Forchheimer developed for the superficial upstream velocity used in the simulations seem to estimation a lower pressure loss then the data collected in this diploma, suggesting that an equivalent pressure drop could be achieved with lower solidity. From the data we can observed a decrease in the pressure drop coefficient for the highest solidity nets, i.e. 30%, this is discussed to be due to the increase on Reynolds number for the treads due to the restricted cross section area.There is presented a crude second order polynomial approximation for Cf for Forchheimerequation accounting for velocity and for the general resistance equation, both valid for 400 < Red <1700. The major error source in the flow loop is identified to be the flow meter which was used to calculate the superficial velocity in the test tube, resulting in aquadratic error due to the decision to use the general resistance law.