Future space missions planned by NASA will be more complex in nature than current missions scheduled as part of the Space Shuttle Program. Examples of future missions include the construction and inhabitance of space stations, orbital transfers, and Lunar/Mars transfers and descents. The Space Transfer Vehicles (STV's) that will perform these missions must have the capability to provide deep-engine throttling thrust. To accomplish this, the turbopumps used to propel the STV's in space must operate at off-design flow rates to provide a wide range of flow outputs. Current state-of-the-art cryogenic fuel and oxidizer turbopump designs can not perform efficiently at off-design flow rates. This inefficiency is caused by flow separation and stall in the turbopump diffuser crossover. This thesis presents the results of the computational analysis for this problem of diffuser stall. An earlier study of this problem was done using a finite element based code, FIDAP, to develop a two-dimensional turbulent diffuser model. Conditions for this two-dimensional model were established allowing for flow separation and stall. Various rates and configurations of suction (removing decelerating fluid particles) and blowing (reenergizing decelerating fluid particles) were tested for their effectiveness in suppressing or eliminating the flow separation or stall at low off-design flow rates. Based upon the results of this previous work, a more complex three-dimensional turbulent diffuser model was created using FIDAP. The objective was to determine the effectiveness of suction as a boundary layer control device in eliminating flow separation. First, Rocketdyne's results with the single stage Water Tester were verified analytically. It was shown from the computational model that stall was introduced at the throat area of the diffuser when the flow rate was reduced to 76% of the design flow. This finding agreed with what Rocketdyne observed with their Water Tester. Then, the actual MK49-F diffuser was modeled using liquid hydrogen. Suction was applied only at the top of the diffuser at off-design flow rates, which is where the flow separation was observed. It was determined that suction applied in this manner at a rate between 3% and 5% of the inlet mass flow rate was very effective in reducing flow separation at 60% of design flow. It was observed that suction, used as a boundary layer control device, not only eliminates losses at off-design flow rates, but even improves diffuser performance over the design flow (10% improvement in Cp). This is explained by a significant amount of flow deceleration in the boundary layer at the design flow which does not cause appreciable flow separation, but does affect diffuser performance. It was also determined that there is a relationship between the off-design flow rate and the amount of suction required to effectively reduce stall and flow separation.

Library of Congress Subject Headings

Turbomachines--Diffusers; Turbomachines--Fluid dynamics; Turbine pumps; Space vehicles

Publication Date


Document Type


Department, Program, or Center

Mechanical Engineering (KGCOE)


Ogut, A.

Advisor/Committee Member

Mulay, S.

Advisor/Committee Member

Veres, J.


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: TJ267.5.D4 Y68 1992


RIT – Main Campus