Many pump vaned diffuser designs are based on existing airfoil designs, with little attention given to the vane leading edge. There is a need for a vaned diffuser leading edge that helps resist flow separation and the resultant poor diffuser pressure recovery. Diffusers in pumps are often working with an incompressible fluid that makes potential flow methodologies — which have incompressibility as a boundary condition — attractive. The potential flow-based free-streamline analysis methods have been known to improve the aerodynamics of varied components at high incidence angles, such as diffusers, jet engine nacelles, and liquid rocket engine turbopump inducers.
The present innovation is based on the Stripling-Acosta free-streamline wake methodology. This method is applicable only to diffusers in a high-solidity (solidity >1) cascade, and expands the applicability of the free-streamline method to isolated airfoils such as those found in LSVDs.
A defining characteristic of the LSVD is the absence of a true aerodynamic throat. To simplify the analysis, it is assumed that this lack of a throat permits the diffuser vane to be modeled as a single, isolated airfoil. The Stripling and Acosta free-streamline wake theory assumes a cascade of infinitely long blades (and therefore infinite solidity). The theory may be applied to LSVDs, however, as the infinitely long blade assumption was made for ease of computation, eliminating the need for a nonlinear potential solution that describes cavity closure and the laborious algebra that results. The key mathematical term that describes the potential flow into the vanes is the cascade vortex and source strength. This term may be modified such that it describes a single airfoil as opposed to an infinite rectangular cascade, allowing the parameters of leading edge conditions and LSVD geometry to be input, thus producing the shape of the free-streamline at the leading edge. The shape of the free-streamline may then be used as the LSVD leading edge contour, providing a shape that is resistant to flow separation.
This analysis method is similar to the linear free-streamline analysis developed by Stripling and Acosta for the design of turbopump inducers. The premise is that there exists a free-streamline flow for cavitation number values below incipient where vapor exists on one side and liquid on the other. By shaping the blade leading edge to be at or below the thickness of the free-streamline contour, the suction performance is improved over blades with leading edge thicknesses that exceed the free-streamline height. Because the Stripling and Acosta free-streamline method predicts a free-streamline contour and has been implemented by the primary investigator in previous work, it is desired to explore if the theory may be adapted to the design of LSVDs as well.