A study of design details of rocket engine swirl injection elements

The performance of a tangential-entry swirl injection element was studied based on the efficiency of the design in converting injector pressure drop into propellant momentum for mixing. For swirl injection, the most efficient use of the pressure drop favors configurations that maximize the transverse components of momentum as opposed to the axial component. Designs were tested that utilize up to 70% of the available pressure drop for producing dynamic pressure with up to 28% of the pressure drop converted into the transverse components of momentum. This compares quite favorably with the liquid flow of a shear-coaxial injector that uses nominally 15% of the pressure drop as axial dynamic pressure with 0% in the transverse components. For the geometries studied a turndown ratio of approximately 1.1 (ratio of upstream flow diameter to exit diameter) and a tangential inlet area 87% of the injector post upstream cross-sectional area produced the highest efficiency. These design parameters can be applied to any system by changing the inlet width to accommodate available system pressure drop without affecting overall efficiency with only small reductions in the transverse efficiency with reduced pressure drop.; Studies were conducted to determine the parameters that controlled element efficiency and the relative importance of each controlling parameter. These studies included both an analytical and experimental investigation. Experiments were conducted using water as a propellant simulant flowing into open air. The injector elements studied were constructed of acrylic to allow flow visualization and measurements of the internal flow structure. The analytical model extended inviscid theory to include wall friction through friction factors. As part of the model development, it was necessary to review and clarify discrepancies in the literature concerning the inviscid treatment of the problem. The inviscid theory of the “choked” behavior of the flow was supported in this research both by experiments and the viscous model results. The viscous model and test results showed that for representative cases that between 63 and 87% of the viscous pressure loss can be attributed to wall friction, with the balance attributed to “minor” losses.