|
|
|
| Abstract Title:
|
| Flow Control for Cooling of Turbine Blades
|
| Graduate Student Presenter:
|
Luis Alvergue
|
| Name of the Author(s) and Affiliation(s):
|
Luis Alvergue, Department of Electrical & Computer Engineering Louisiana State University (LSU); Dr. Guoxiang Gu, Department of Electrical & Computer Engineering LSU; Dr. Sumanta Acharya, Department of Mechanical Engineering LSU
|
In a gas turbine engine, the efficiency can be increased if the inlet temperature to the turbine is raised. Currently, the operating inlet temperature to the turbine is higher than the melting point of the material of the blades, and therefore if the blades are not cooled, they will melt. The current solution consists in injecting cool air through slanted holes on the blades so that a film of coolant layer forms on the blades and protects it from the heated crossflow. Hotspots may occur, and active flow control may be a strategy to enhance cooling and eliminate hot spots. To achieve real-time flow control, a reduced order model that captures the essential physics of the flow system is needed as well as a control strategy, in our case robust adaptive control, that is able to cope with changes of operating conditions and model uncertainties. Figure 1 provides a schematic of a simplified setup illustrating the problem.
To obtain a reduced order model, a Proper Orthogonal Decomposition (POD) technique is being used. POD provides a useful tool for efficiently approximating a large amount of data by capturing the dominant temporal and spatial structures of a flow, in turn providing a way to parameterize the model more efficiently and more reliably that is pertinent to applications in adaptive control. As an initial test of the approach, we have generated POD for a lid-driven cavity. The snapshots are obtained from DNS data. This approach will be next extended to the film cooling problem.
|
|
|
