National Aeronautics and Space Administration

Glenn Research Center

Brendan Tracey Research Project

Optimization of a Flatback Airfoil Rotor Blade

Research Assistant: Brendan Tracey

Principal Investigator: Meng-Sing Liou

Background Information

In 2002, Sandia National Laboratories conducted a Blade System Design Study which examined wind turbine blade design through the combined perspective of aerodynamics and manufacturing. One of the concepts discovered by the study was the use of flatback airfoils for the inboard sections of the turbine blade. A flatback airfoil is an airfoil with a truncated trailing edge but thickened mid-sections to maintain an equivalent overall camber with the original airfoil. Due to the truncation of the trailing edge flatback airfoils have the possibility to create lighter, stronger, and cheaper turbine blades. The turbine blade is made lighter directly with the elimination of the material normally in the trailing edge of the airfoil and also indirectly by allowing thicker airfoil sections which increase the strength and thus decrease the weight of the blade structure. These airfoils are also much cheaper to manufacture as the blade section can be stopped at the structural box of the airfoil rather than needing to construct a trailing edge. Early studies have also shown that flatback airfoils may actually increase the generation of lift from the baseline airfoil, which should in turn increase the generation of power.

The drawback to flatback airfoils is an increase in drag due to the separation of flow at the trailing edge. It has been suggested that a plate of some sort may be added to the flat part of the airfoil to decrease the drag on the airfoil. This plate has the potential to greatly reduce the drag of the blade while maintaining the lower weight and material costs relative to a traditional airfoil.

My Role as a Research Associate

I will be working under Dr. Meng-Sing Liou in the Aeropropulsion Division (RT). Researchers at NASA developed the SWIFT Navier-Stokes computational fluid dynamics (CFD) solver to solve problems in turbomachinery design and analysis. I will be first validating the use of this code for wind turbine design by comparing the performance of the NREL Phase VI turbine with experimental results. SWIFT is a complicated program which has many ways of solving the Navier-Stokes equations. I will examine the effects of parameters such as grid generation method, turbulence model, and flow conditioning methods. The optimal choice for these parameters depends on the problem being solved and I will find which are optimal for analyzing wind turbine performance and design.

Once this has been accomplished, I will analyze the same turbine but with a modified blade using a flatback airfoil design to examine changes in both the loads experienced by the turbine blade and the torque generated by the new airfoil. There have been some preliminary studies using CFD to examine the 3-D effects of flatback airfoils on turbine blades, but results have been inconclusive. Through my analysis, I hope to have a conclusive answer to compare to previous findings and provide a new understanding of the 3-D effects on flatback airfoil performance.

I will also carry out an optimization of blade design through the use of evolutionary algorithms. I hope to improve upon the current flatback airfoil design by examining the use of a splitter plate or plates. I plan to vary the size, location and orientation of the plate and hope to decrease the drag on the airfoil while maintaining the high lift and low weight of the blade geometry.

Working alongside me will be Makoto Endo who is working at OAI this summer. He and I will both be using SWIFT to optimize blade geometry design, though the details of our optimization will differ.

I hope to be able to add to the body of knowledge on this new blade design concept. If the predicted improvements can be realized, it will be another step toward making wind turbines cheaper and better which will increase their deployment worldwide.

Brendan shows the mesh for an airfoil design.

Brendan shows the mesh for an airfoil design.