Predicting turbulent flowfield and pressure drop in complex internal flows, such as valves, pumps, and turbines, is of great interest in many areas of technology. Accurate prediction of three-dimensional (3D) flowfield in complex devices is critically dependent on the accuracy of the turbulence model used to describe the turbulence energy production and dissipation processes within the flowfield region of interest.
For practical applications, the popular two-equation k – ε and k – ω models are frequently considered for both external and internal flows in an effort to compromise between the accuracy and computational time relative to simpler one-equation models and more advanced models. Although the k – ω model is known to suffer from sensitivity to freestream turbulence intensity levels, the relative performance of the k – ε and k – ω models for complex internal flow applications has not been fully established.
Three-dimensional flowfield in a pneumatic valve, through which gaseous nitrogen flows at a constant flow rate and inlet temperature, was numerically investigated by solving the governing Navier-Stokes equations. The inlet flow branches into ten secondary tubes of reduced diameters that ultimately carry the fluid to the outlet as a single stream. Finite element meshes and flowfield solution were obtained with the aid of ANSYS/FLUENT software.
It was found that the k – ω model yields a more satisfactory solution of the adverse pressure gradient region upstream of the stagnation region of the valve, while at the same time predicting the pressure drop across the valve to within 2% of that computed by the alternate k – ε model. The results suggest that the k – ω model is more satisfactory for prediction of flowfield in complex valve configurations.
This work was done by Max Kandula and Michael Harris of Kennedy Space Center. KSC-13755