"Donald R. Bates" wrote:

Hi Paul:

Finally waded through this paper you sent me. It's interesting but hardly much new. He uses a lattice vortex method of analysis which he calls the step method because it uses only one row of discrete vortices radially instead of two rows usually used and claims the difference is negligible. It probably cuts computation time considerably. In addition this method is based on the flow field at the propeller disk rather than the ultimate wake properties which the Goldstein-Theodorsen model in my program uses. The Goldstein-Theodorsen model is based on a radial loading optimized for minimum induced drag. The lattice vortex setup is intended to avoid the use of a specific loading function so that it can be used to analyze any shape blade which optimization to trial and error runs by the user to get there. Since the Goldstein-Theodorsen model is based on a single tip vortex with its optimized loading function, it is limited to straight blades. Thus the Goldstein-Theodorsen model st!

arts out with optimized aerodynamics and needs only to match the blade to the maximum airplane performance which my program does automatically.

Finally, Barnes paper gives a method of aerodynamic sources and sinks to simulate a symmetric body behind the propeller and its influence on propeller performance. My program simply simulates the slipstream drag and thus its effect on efficiency. Barnes compares his predictions with NACA test data on one propeller and claims good agreement and therefore success in all cases. Since my program is less general I can only measure success after the fact by the results obtained from each design. Some have been very good, others not so good especially when drag prediction is less certain. Barnes method requires approximate methods to determine inflow at the blade whereas the Goldstein-Theodorsen model is exact but Barnes points out that non zero CL is probably not possible at the tip as predicted by theory which would be a small difference. One big drawback of barnes equations is that they are normalized to non-dimensionalize them by the free stream velocity and therefore not able !

to calculate static thrust. Both have some flaws but I would expect a similar degree of accuracy from each. Both are complex in computational nature requiring sophisticated computer programs to use them.

That's it Paul. Thanks for sending the paper. It's one of the more interesting ones I have seen.

Don Bates

--
Bates Engineering
2742 Swansboro Road
Placerville, CA 95667
Ph/Fax 530-622-1886
E-mail:  bateseng@earthlink.net
Website:  http://home.earthlink.net/~bateseng

Your welcome Don.

Paul Lamar
 
The AirCraft Rotary Engine NewsLetter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org