Particle Transport Analysis of Sand Ingestion in Gas Turbine Jet Engines, 18-9481

Printer Friendly Version

Principal Investigator
Klaus Brun

Inclusive Dates:  07/01/04 – 01/01/06

Background - The purpose of this project is to develop and test a software tool that will allow SwRI to provide detailed analysis of the kinematic behavior of sand and other particulates inside a jet engine gas turbine. Using this tool's output information, SwRI can realize improved (intelligent) inlet air filtering techniques, engine maintenance, and component designs.

Approach - The project includes the analytical and numerical implementation of the particle transport model in a gas turbine jet engine (Task 1). Task 2 will be the experimental verification and calibration of this model using sand ingestion data from SwRI's high-speed turbomachinery test stand. Both tasks will be performed in parallel to assure proper feedback for the model calibration tasks.

Accomplishments - The sand kinematic behavior was studied to evaluate the relative influence of streamline curvature and particle size on the motion of particles in the airflow, as well as to test some basic particle kinematic theories. As anticipated, experimental results showed that smaller particles tend to follow the flow streamlines more closely than larger particles. In parallel to the experimental work, a simple two-dimensional computational fluid dynamics (CFD) model was employed to determine the streamlines (velocity and local curvature) of the flow around the airfoil. Combining CFD and experimental results showed that, within the uncertainty of the test results (~10 percent on velocity), the motion of a particle can be determined based on a simple model that balances inertial, centrifugal, and drag forces on the particle. When expanding the project-developed model to the rotating frame, particular care has to be given to the inertial terms — as the particles in the fluid will also be affected by global centrifugal and Coriolis forces.

Initial CFD work provided a baseline for experimental results. Only one stage of a two-stage compressor was modeled, as this single impeller will be the focus of experimental work. Using a commercial CFD code and expected experimental parameters, a series of CFD simulations were conducted to determine flow field characteristics at the leading edges of the impeller blades. A series of models were analyzed using various permutations of mesh density, boundary conditions, and turbulent flow models. The software package used for this analysis allowed for flow characteristics to be determined at specific locations. Of particular interest is the leading edge of the impeller blades. The nondimensional parameters used in this project were calculated based on the CFD results. For properties of sand, desert sand from Qatar was ascertained and analyzed. Average density and radius values were used for all calculations. Initial computations demonstrated that the simple inertial terms were dominated by centrifugal and Coriolis terms. While the centrifugal parameters did not vary significantly at small distances from the leading edge of the impeller blades, the Coriolis parameters showed significant dependence of position. The preliminary CFD results are ready for comparison with experimental data. After a mounting plate for an electric motor is machined, the compressor setup will be assembled and prepared for testing.

Initial CFD Results: Velocity Magnitude Plot and Vector Plot Near Leading Edge

2005 Program Home