Properties of large 7079 aluminum alloy forgings in a cryogenic environment
Forged aluminum alloys for pump impeller and inducer applications in cryogenic liquid propellant rocket engines
SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Forged aluminum alloys for pump impeller and inducer applications in cryogenic liquid propellant rocket engines
Pump for hydrogen modified to remove moisture that condenses in impeller stage. Impeller-pump housing has circumferential groove leading to exit hole near high-pressure outlet. As impeller disk rotates, flings water droplets condensed in pump toward groove. Aerodynamic drag drives water around groove to exit hole.
The ratio of rocket engine thrust to weight is a limiting constraint in placing more payload into orbit at a low cost. A key component of an engine's overall weight is the turbopump weight, Reducing the turbopump weight can result in significant engine weight reduction and hence, increased delivered payload. There are two main types of pumps: centrifugal and axial pumps. These types of pumps can be further sub-divided into those with shrouds and those without shrouds (unshrouded pumps). Centrifugal pumps can achieve the same pump discharge pressure as an axial pump and it requires fewer pump stages and lower pump weight than an axial pump. Also, with unshrouded centrifugal pumps (impeller), the number of stages and weight can be further reduced. However. there are several issues with regard to using an unshrouded impeller: 1) there is a pump performance penalty due to the front open face recirculation flow; 2) there is a potential pump axial thrust problem from the unbalanced front open face and the back shroud face; and, 3) since test data is very linu'ted for this configuration, there is uncertainty in the magnitude and phase of the rotordynamics forces due to the front impeller passage. The purpose of the paper is to discuss the design of an unshrouded impeller and to examine the design's hydrodynamic performance, axial thrust, and rotordynamics performance. The design methodology will also be discussed. This work will help provide some guidelines for unshrouded impeller design. In particular, the paper will discuss the design of three unshrouded impellers - one with 5 full and 5 partial blades (5+5). one with 6+6 blades and one with 8+8 blades. One of these designs will be selected for actual fabrication and flow test. Computational fluid dynamics (CFD) is used to help design and optimize the unshrouded impeller. The relative pump performance penalty is assessed by comparing the CFD results of the unshrouded impeller with the equivalent shrouded impeller for a particular design. Limited unshrouded - versus - shrouded impeller data from the J-2 pump is used to anchor the CFD. Since no detailed impeller blade force data is available, axial thrust and rotordynamic force predictions are based on the CFD model. For the axial thrust, the impeller front flow passage axial force is integrated from the CFD results and compared to the equivalent shrouded impeller axial force. For the rotordynamics forces, the fluid reaction forces are computed from unsteady flow CFD results using a moving boundary method; the rotor- shaft is moved at several whirl-to-speed frequency ratios to extract the rotordynamics coefficients.
Visual observations to find flow distribution and coefficients through radial bladed centrifugal pump impeller
Phase 2 (development) which was concluded for the ultrasonic Doppler device and the light-pipe-reflectance device is reported. An ultrasonic Doppler breadboard system was assembled which accurately measured runout in the J-2 LOX pump impeller during operation. The transducer was mounted on the outside of the pump volute using a C-clamp. Vibration was measured by conducting the ultrasonic wave through the volute housing and through the fluid in the volute to the impeller surface. The impeller vibration was also measured accurately using the light-pipe probe mounted in an elastomeric-gland fitting in the pump case. A special epoxy resin developed for cryogenic applications was forced into the end of the fiber-optic probe to retain the fibers. Subsequently, the probe suffered no damage after simultaneous exposure to 2150 psi and 77 F. Preliminary flash X-radiographs were taken of the turbine wheel and the shaft-bearing-seal assembly, using a 2-megavolt X-ray unit. Reasonable resolution and contrast was obtained. A fast-neutron detector was fabricated and sensitivity was measured. The results demonstrated that the technique is feasible for integrated-time measurements requiring, perhaps, 240 revolutions to obtain sufficient exposure at 35,000 rpm. The experimental verification plans are included.
The perturbed flow in the leakage path between a shrouded-pump impeller and its housing is analyzed using experiences with the Space Shuttle Main Engine (SSME), high pressure fuel turbopump (HPFTP) wearing-ring seals. Analysis is based on a bulk-flow model which consists of the path-momentum, circumferential momentum, and continuity equations. The pressure oscillations in the leakage annulus are driven by a circumferential variation of the impeller discharge pressure. It is shown that the occurrence and nature of the pressure oscillations depend on the tangential-velocity ratio of the fluid entering the seal, the order of the Fourier coefficient, the closeness of the precessional frequency of the rotating pressure field to the first natural frequency of the fluid annulus, and the clearance of the wearing-ring seal. The results obtained may explain the internal melting observed on SSME HPFTP seal parts.
The high-strength-to-weight ratio of titanium alloys, particularly at cryogenic temperatures, make them attractive for application in rocket engines - offering the potential of superior performance while minimizing component weight. This was particularly attractive for rotating components, such as pump impellers, where titanium alloys presented the potential to achieve a major advance in rotational tip speed, with a reduction in stages and resultant saving in pump weight and complexity. The investigation into titanium alloys for application in cryogenic turbopumps began in the early 1960's. However, it was found that the reactivity of titanium limited applications and produced unique processing challenges. Specialized chemical compositions and processing techniques had to be developed. A substantial amount of material properties testing and trials in experimental turbopumps occurred, ultimately leading to application in the Space Shuttle Main Engine. One particular alloy stood out for use at liquid hydrogen temperatures, Ti-5Al-2.5Sn ELI. This alloy was employed for several critical components. This presentation deals with the development effort, the challenges that were encountered and operational experiences with Ti-5Al-2.5Sn ELI in the SSME.
This paper reports the progress being made towards complete turbo-pump simulation capability for liquid rocket engines. Space Shuttle Main Engine (SSME) turbo-pump impeller is used as a test case for the performance evaluation of the MPI and hybrid MPI/Open-MP versions of the INS3D code. Then, a computational model of a turbo-pump has been developed for the shuttle upgrade program. Relative motion of the grid system for rotor-stator interaction was obtained by employing overset grid techniques. Time-accuracy of the scheme has been evaluated by using simple test cases. Unsteady computations for SSME turbo-pump, which contains 136 zones with 35 Million grid points, are currently underway on Origin 2000 systems at NASA Ames Research Center. Results from time-accurate simulations with moving boundary capability, and the performance of the parallel versions of the code will be presented in the final paper.
The unsteady fluid dynamic performance of a cavitating two-bladed oxidizer turbopump inducer was characterized through sub-scale water flow testing. While testing a novel inlet duct design that included a cavitation suppression groove, unusual high-frequency pressure oscillations were observed. With potential implications for inducer blade loads, these high-frequency components were analyzed extensively in order to understand their origins and impacts to blade loading. Water flow testing provides a technique to determine pump performance without the costs and hazards associated with handling cryogenic propellants. Water has a similar density and Reynolds number to liquid oxygen. In a 70%-scale water flow test, the inducer-only pump performance was evaluated. Over a range of flow rates, the pump inlet pressure was gradually reduced, causing the flow to cavitate near the pump inducer. A nominal, smooth inducer inlet was tested, followed by an inlet duct with a circumferential groove designed to suppress cavitation. A subsequent 52%-scale water flow test in another facility evaluated the combined inducer-impeller pump performance. With the nominal inlet design, the inducer showed traditional cavitation and surge characteristics. Significant bearing loads were created by large side loads on the inducer during synchronous cavitation. The grooved inlet successfully mitigated these loads by greatly reducing synchronous cavitation, however high-frequency pressure oscillations were observed over a range of frequencies. Analytical signal processing techniques showed these oscillations to be created by a rotating, multi-celled train of pressure pulses, and subsequent CFD analysis suggested that such pulses could be created by the interaction of rotating inducer blades with fluid trapped in a cavitation suppression groove. Despite their relatively low amplitude, these high-frequency pressure oscillations posed a design concern due to their sensitivity to flow conditions and test scale. The amplitude and frequency of oscillations varied considerably over the pump s operating space, making it difficult to predict blade loads.
This paper reports the progress being made towards complete unsteady turbopump simulation capability by using overset grid systems. A computational model of a turbo-pump impeller is used as a test case for the performance evaluation of the MPI, hybrid MPI/Open-MP, and MLP versions of the INS3D code. Relative motion of the grid system for rotor-stator interaction was obtained by employing overset grid techniques. Unsteady computations for a turbo-pump, which contains 114 zones with 34.3 Million grid points, are performed on Origin 2000 systems at NASA Ames Research Center. The approach taken for these simulations, and the performance of the parallel versions of the code are presented.
Liquid fluorine pump, impeller, and inducer corrosion resistance to cavitation flow
Novel bearing support for liquid-oxygen turbopumps protects against impact loads while avoiding a major disadvantage of earlier flexible supports. It allows controlled axial movement necessary for proper operation of pressure-operated pump impeller. While spring-loading rotor to midpoint of the axial movement to avoid impact-load damage to turbopump components. Support is made by machining azimuthal slots in cylindrical portion. Resulting structure permits controlled axial deformations.
Experiments were conducted to evaluate the characteristics of the Ti-5Al-2.5Sn (ELI) alloy used in a fuel pump impeller at cryogenic temperatures. Tension, fracture toughness, and fatigue crack propagation data were collected determining the effect of frequency and load ratio on crack propagation. The results revealed that tensile strength increased significantly at 20 K compared to room temperature and fracture toughness was reduced at cryogenic temperatures. The fatigue crack growth rate was not sensitive to experimental conditions and there were only minimal crack orientation effects. Different frequencies produced no effect. At various temperatures and frequencies a load ratio increase resulted in higher crack growth rates. At low stress intensity levels the fatigue rate for both temperatures was the same; however, at high stress intensity levels the crack growth rate at 20 K increased because of the decrease in fracture toughness. The results correlated well with previous data.
An experiment was designed to measure the rotordynamic shroud forces on a centrifugal pump impeller. The measurements were done for various whirl/impeller speed ratios and for different flow rates. A destabilizing tangential force was measured for small positive whirl ratios and this force decreased with increasing flow rate.
The objective of the proposed paper is to develop a computational procedure that solves incompressible Navier-Stokes equations for pump flows. The solution method is based on the pseudo-compressibility approach and uses an implicit-upwind differencing scheme together with the Gauss-Seidel line relaxation method. The equations are solved in steadily rotating reference frames and the centrifugal force and the Coriolis force are added to the equation of motion. As a benchmark problem, the flow through the Rocketdyne inducer is numerically simulated. A coarse grid solution is obtained with a single zone by using an algebraic turbulence model. In multi-zone fine grid computation, a one-equation Baldwin-Barth turbulence model is utilized. Numerical results are compared with experimental measurements and a good agreement is found between the two. The resulting computer code is then applied to the flow analysis inside a two-stage fuel pump impeller operating at 80 percent, 100 percent, and 120 percent of design flow.
Swirling of a liquid in a container may prove to be a more desirable method of managing liquids in low gravity (space) environments than by rotating the entire container. By injecting a relatively high velocity liquid tangentially into the body of the fluid, swirl can best be started rapidly, however an estimate of the quantity and velocity of jetflow, or mechanical power of a pump impeller required to maintain a given radial acceleration (G force) is needed to assess the feasibility of such a method. While the key aspect of the problem is determining the rate of rotational energy dissipation by wall friction in the container, there are other considerations, and the present study investigates the possible additional effects of axial variation of tangential velocity and secondary (radial and axial) flow components within the rotating fluid.
The Space Shuttle has three main engines that are used for propulsion into orbit. These engines are fed propellants by four turbopumps on each engine. A main element in the turbopump is the bearings supporting the rotor that spins the turbine blades and the pump impeller. These bearings are required to spin at very high speeds, support radial and thrust loads, and have high wear resistance without the benefit of lubrication. The liquid hydrogen and oxygen propellants flow through the bearings to cool the surfaces. The volatile nature of the propellants excludes any conventional means of lubrication. Lubrication for these bearings is provided by the ball separator inside the bearing. The separator is a composite material that supplies a transfer film of lubrication to the rings and balls. New separator materials and lubrication schemes have been investigated at Marshall Space Flight Center in a bearing test rig with promising results. Hybrid bearings with silicon nitride balls have also been evaluated. The use of hybrid, silicon nitride ball bearings in conjunction -with better separator materials has shown excellent results. The work that Marshall has done is being utilized in turbopumps flying on the space shuttle fleet and will be utilized in future space travel. This result of this work is valuable for all aerospace and commercial applications where high-speed bearings are used.
Steady and unsteady flows for propulsion systems are efficiently simulated by solving the incompressible Navier-Stokes equations. The solution method is based on the pseudo compressibility approach and uses an implicit-upwind differencing scheme together with the Gauss-Seidel line relaxation method. Current computations use one equation Baldwin-Barth turbulence model which is derived from a simplified form of the standard kappa - epsilon model equations. The resulting computer code is applied to the flow analysis inside an advanced rocket pump impeller in steadily rotating reference frames. Numerical results are compared with experimental measurements. The effects of exit and shroud cavities with the leak-age flow are investigated. Time-accurate incompressible Navier-Stokes formulation with the overlapped grid scheme capability was evaluated by using MIT flapping foil experiment. The grid dependency, turbulence model effects, and the effect of order of differencing were investigated. Numerical results were compared against experimental data. The resulting procedure were applied to unsteady flapping foil calculations. Two upstream NACA 0025 foils perform high frequency synchronized motion and generate unsteady flow conditions to the downstream larger stationary foil. Comparison between unsteady experimental data and numerical results from two different moving boundary procedures will be presented.