Modeling flow, erosion and mass transfer in elbows

Erosion and corrosion in 90{dollar}spcirc{dollar} elbows were investigated using flow modeling, particle tracking and mass transfer modeling approach.; In the flow modeling of 90{dollar}spcirc{dollar} bends, a simultaneous variable solution procedure was extended to general curvilinear coordinates and turbulent flow calculations. Also, several turbulence models were evaluated for their capability to predict turbulent flow in 90 bends and elbows. The turbulence models evaluated include a mixing-length model, a low-Reynolds number two-equation turbulence model (k-{dollar}tau{dollar} model) and the standard k-{dollar}epsilon{dollar} model. The predictions using these turbulence models were compared with the experimental data for flow in a 90{dollar}spcirc{dollar} bend. Comparison of the results from flow modeling with the data indicated that all of the turbulence models, including the mixing-length model, resulted in similar velocity profiles. The results also indicate that the accurate prediction of the turbulent kinetic energy requires modification of the turbulence models to include the streamline curvature effects.; After a flow solution was obtained, individual sand particles were introduced and tracked by a Lagrangian approach. The sand particle-pipe wall impingement information obtained from particle tracking was then used to compute erosion rates in elbows. An efficient and accurate particle tracking method in general curvilinear coordinates was developed. This method is efficient and gives accurate prediction of particle trajectory in highly curved geometries. The erosion prediction method was verified by comparing the predicted elbow penetration rates with experimental data from four different sources in the literature. Results were obtained for short- and long-radius elbows. Based on the results, a simplified equation which defines the elbow radius effects on sand erosion in various flow conditions was developed.; Localized mass transfer coefficients along the walls of an elbow was modeled using a commercially available Computational Fluid Dynamics (CFD) software package called CFX. To verify the CFX capability to predict mass transfer coefficients in an elbow geometry, the predicted distribution of mass transfer coefficient along the walls of an elbow was compared with available experimental data. A fairly good agreement between CFX predictions and experimental data was observed at the outer wall of the elbow. Using the simulated results from CFX, an equation which describes the relation between the maximum mass transfer coefficient for an elbow and the flow parameters was developed. This relation can be used to predict corrosion in 90{dollar}spcirc{dollar} elbows.