Limited, expensive and nonrenewable fossil energy resources have forced man to seek and exploit other and new forms of God granted energies. Wind energy, if employed through more efficient methods is one of such energies. Simplicity in design and suitability of windmill turbines for many parts of Iran where no intense wind is available along with, high cost of electricity transmission from power plants to scattered water wells, especially in remote areas, were among reasons to have this study initiated. Considering the lower efficiency of windmill turbines as compared to propelled ones, an introductory laboratory study was conducted to determine the optimized range of blade angle in relation to the number of blades in windmill turbines with close to 100% solidity lead to increase in efficiency and better exploitation of energy. A factorial experiment in a completely randomized design of three replications with three treatments of wind speed in three levels (7.5, 10, 12.5 m/s), blade angle in five levels (10, 20, 30, 40 and 50 degrees) and three levels of blade number (8, 16 and 24) was carried out. Power generated by the modeled turbine when made to operate in the wind tunnel of the Faculty of Aerospace Engineering lab in Amir Kabir University was recorded and analyzed. All blade sets were built flat from aluminum with an outer diameter of 256 mm. Analysis of variance showed that, at probability level of 1%, all factors including interaction combinations significantly affected on the produced power. For all three sets of blades, in the range of 20-30 degree blade angles the maximum power was produced. Also, for all sets of blades, up to near 40 degrees, the increase in produced power was accompanied by increase in number of blades, but at higher angle, settings, inverse results were obtained. Also, it was revealed that one can obtain a similar power with lower number of blades if the blade angle is set next after the optimum angle but before the angle in which the separation phenomenon (in this case around 40 degrees) occurs. One should expect different results in real situations where blades take their real dimensions and are not made flat, but it is expected that the trend and interpretation of results would follow the same, although a more detailed investigation is needed through incorporating simulation and dimensional analysis methods.
)