Research Horizons

Modeling Erosion in Hall-Current Thrusters

Hall-current thrusters (HCTs) are electric propulsion devices that augment chemical thrusters on satellites. In geosynchronous satellite applications, HCTs are used for a portion of the orbit insertion and for stationkeeping. This augmentation of the chemical propulsion system provides increased mass to orbit as well as tighter orbit control. With a high specific impulse (1500–3000 sec), these devices are becoming common in many commercial programs and will be used in the AEHF (Advanced Extremely High Frequency) program. Potential future applications include GPS III. The low-level thrust that these devices generate requires that they operate for long intervals (weeks) to adjust satellite orbits to achieve mission objectives. The advent of more efficient solar cells enables high-power HCTs to be used, decreasing orbit insertion time or enabling delivery of heavier payloads. The long-duration firing, however, makes an HCT susceptible to wall erosion, the slow, steady loss of material from the inner surface of its boron-nitride thrust chamber due to sputtering—the impact of a very hot and energetic xenon plasma. The wall erosion rate increases with thrust level; thus, it is critical to develop tools to study and characterize this phenomenon.

HCT wall erosion is an emerging field in electric propulsion research. Experimental studies of the process are difficult because of the long timescales involved, and theoretical investigations are limited by the problem's complexity. Thus, HCT wall erosion is best investigated with numerical models. Rostislav Spektor of the Propulsion Science and Experimental Mechanics Department is currently pursuing such a study—"Electric Propulsion Diagnostics and Modeling." As an added benefit, he expects that, when mature, the same numerical model can also be used to predict HCT long- and short-term performance, such as thrust and specific impulse, as a function of the supplied power; it may also help optimize thruster design and configuration.

During the first year of the project, Spektor's work focused on developing a one-dimensional numerical simulation of the plasma flow inside the acceleration channel of an HCT. He is currently developing a two-dimensional finite-volume solver for flow simulation. The finite-volume approach facilitates solution of the necessary differential equations as the thruster geometry changes with time (as a result of erosion). Spektor explains that "the plasma flow simulation takes a multifluid approach by modeling separately the flow of ions, neutrals, and electrons through the equations of mass, momentum, and energy conservation." In particular, he uses mass and momentum conservation equations to solve for the ion velocity and density distribution, while using energy conservation to obtain the electron temperature distribution. The mass conservation equation solves for the neutral flow.

Rostislav Spektor is developing numerical models to predict wall erosion—the steady loss of material from the inner chamber surface of a Hall-current thruster.

While a simplified one-dimensional flow model has been used to simulate a generic HCT with a rectangular cross-section acceleration channel geometry and purely radial magnetic field, the model ignores erosion processes. "Still," says Spektor, "the resulting electric potential, plasma density, and ion velocity distributions are in good quantitative agreement with typical measured HCT values. Additionally, thruster performance (specific impulse and thrust) matches the measured performance of a typical HCT. Furthermore, the model successfully simulates some of the instabilities observed during real HCT operation." In the coming months, he will extend this model to two dimensions.

Spektor will use the results of the flow model as inputs for the erosion model. In particular, he explains, "plasma density, ion velocity, and electron temperature distributions at the boundaries of the flow will partially determine the sputtering rate at the walls of the thruster. An erosion model, will, in turn, use the sputtering rate to influence the boundary conditions for the plasma-flow simulation."

During the second year of this project, Spektor will focus on improving the developed plasma-flow model and on working erosion into the simulation. Achieving this year's objectives will require completion of four tasks: improving determination of electron mobility within the acceleration channel to increase fidelity of the flow model; identifying the relevant physical equations governing the plasma interactions with solid surfaces, which will constitute the numerical erosion model itself; including electron secondary emission from the thruster walls into the flow model; and implementing the erosion model and integrating it with the plasma flow model.

Aerospace is well suited for future experimental investigation of high-power HCT erosion, but lacks numerical capabilities to study this effect. Specifically, an HCT erosion model like Spektor's can benefit future programs by estimating thruster lifetime. For example, future communications spacecraft may use a higher power HCT. At high power, thruster erosion may be the limiting factor in spacecraft operation. Consequently, his erosion model can go a long way in helping to qualify the thruster for future programs.

The overall success of Spektor's project can be measured by the ability of his model to predict HCT erosion and overall performance. In the future, the project will focus on flow simulation improvements and erosion model implementation for a generic thruster geometry. Thus, progress for this year may be assessed by successful predictions of overall erosion trends, like erosion rate and erosion pattern. Additionally, his model should predict generic thruster performance characteristics, such as thrust, specific impulse, efficiency, and the current-voltage curve. His long-term goal is to create a flexible HCT performance prediction tool.

Turbopump Cavitation Testing and Modeling

Cavitation in the fluid passages of a liquid-fueled rocket-engine turbopump can not only degrade pump performance, but also reduce system reliability through the generation of elevated engine and vehicle vibration. Turbopump cavitation has been identified as a source of several recent anomalous flight and ground-test vibration signatures in both booster and upper-stage engines and, more critically, as the cause of the 1999 flight failure of a Japanese H-II rocket. Cavitation-related vibration issues are currently addressed through expensive and time-consuming engine or turbopump test programs designed to assess engine and vehicle-component durability at elevated vibration levels. Often, these test methods provide insufficient data to completely assess risk because of the limited instrumentation access and the inability to adequately simulate flight conditions during turbopump operation. Both dedicated component test facilities and high-fidelity turbopump cavitation models are needed to improve the understanding of cavitation physics.

Existing design tools are incapable of accurately predicting cavitating flow fields in propulsion-system turbopumps. Consequently, cavitation-related problems are often not encountered until late in development testing or after propulsion-system deployment. While recent advances in the computational fluid dynamics of cavitating flows suggest it's now possible to model cavitating flows, these methods have not been validated for use in rocket turbomachinery and have not been implemented by the rocket-propulsion community. Dan Ehrlich of the Propulsion Department expects that "implementation and extension of these state-of-the-art methods in a practical modeling tool with subsequent validation through comparison with test data would provide a valuable means for anomaly resolution and new turbopump development."

The Aerospace test facility features a unique vertical feedline configuration.

In 2005, a research team led by Ehrlich began to address these problems with a testing and modeling program designed to investigate complex cavitation dynamics in turbopump inducers (the first stage of a two-stage propellant pump). A closed-loop water-flow test facility is being constructed to support this research.

The Aerospace facility overcomes a fundamental shortcoming of previous scaled-waterflow test facilities by employing a unique vertical feedline configuration. It can also operate over a wide temperature range not currently available to the national cavitation research community. This will enable the team to investigate and quantify temperature effects on turbopump cavitation.

A testing and modeling program was designed to investigate complex cavitation dynamics in a turpopump inducer and validate computer cavitation models for flight environments.

As Ehrlich explains, room-temperature water is a poor simulant for cryogenic rocket propellants, but it's commonly used to evaluate inducer cavitation performance. Heated water better simulates the behavior of cryogenic propellants. The Aerospace cavitation test facility can operate at boiling-water temperatures, and Ehrlich plans to use this capability to evaluate temperature effects on inducer cavitation. Working with coinvestigators John Murdock, Richard Welle, and Brian Hardy, he plans to determine the water temperature needed to simulate cryogenic turbopumps. In addition, the testing will map the impact of temperature effects on cavitation instabilities over inducer operating ranges encountered under flight conditions. This will be accomplished through a series of tests spanning the full range of available water temperatures. The team will evaluate temperature effects on both pump performance and cavitation instability.

The research team will also focus on acquiring high-fidelity data sets for the validation of computational fluid dynamic cavitation models for rocket turbomachinery applications. Current turbopump design and performance prediction practices often employ such models; however, a dearth of high-quality validation data for cavitating rocket turbopumps has left these models insufficiently validated, resulting in reduced confidence in predictive capability. John Schwille of the Fluid Mechanics Department has begun investigating the capability of commercially available models to predict cavitation in rocket turbopump inducers. The team will continue to evaluate the capability of commercial computational fluid dynamic cavitation models to predict inducer cavitation through comparison to Aerospace data. Knowledge gained through this effort may also support the development of advanced cavitation models at Aerospace to improve the capability to accurately predict cavitation in rocket turbopumps.

To Fall 2006 Table of Contents