Comprehending head loss in pipe fittings is essential in fluid dynamics and hydraulic engineering to create effective piping systems. To further examine this topic, we performed a CFD simulation using ANSYS Fluent, taking inspiration from the research paper titled “Teaching turbulent flow through pipe fittings using computational fluid dynamics approach.” The objective of this study is to visually represent and measure the intricate flow patterns and changes in pressure that arise when fluid flows through different pipe fittings, including elbows, tees, and valves. This study aims to improve our comprehension of how various pipe-fitting geometries impact fluid flow and contribute to the total loss of system heads, offering valuable insights for optimizing the designs of piping systems.

Figure 1: Tee junction pipe fit

Simulation process

The reference paper gives all the T-junction geometrical dimensions. In this investigation, a grid independence study is also done, including four grids with 72532, 162918, 261270, and 494266 elements. As a result, grid #3 with 261270 cells is opted, regarding the changes in head loss coefficient.  The study is repeated for 3 different Reynlods. By the way, heat loss coefficient is calculated by pressure drop/0.5*density*V^2.

Figure 2: Mesh independence study

Post-processing

CFD simulation of head loss in a T-junction pipe fitting provides valuable insights into fluid behavior and energy dissipation. The velocity contour map displays a distinct pattern of flow speeds over the fitting, with the greatest velocities (shown by the color red) observed in the vertical part of the T-junction. This suggests a high level of flow energy in this specific location, most likely caused by the abrupt alteration in the flow direction. The horizontal portions exhibit a velocity gradient, characterized by slower velocities (blue and green) near the pipe walls and quicker flows (yellow and orange) in the center. This pattern is in line with the anticipated evolution of the boundary layer. At the point where two things meet, there is a distinct area with a low speed (shown by the color blue), indicating the presence of a region where the flow of fluid may recirculate or may be separate. This region is crucial for comprehending energy dissipation in the system. The given head loss coefficients for Reynolds numbers 10000, 15000, and 20000 (0.486, 0.412, and 0.389, respectively) exhibit a declining pattern as the Reynolds number increases. This suggests that as the flow becomes more turbulent, the efficiency of head loss slightly decreases. This could be attributed to the thinner boundary layer and reduced time available for heat exchange at higher velocities.

Figure 3: Head loss coefficient Vs. Reynolds number