Abstract
The analog prediction system developed and investigated by R. L. Schafer and C. A. Reaves was used for predicting torques on rotary blades.
Three sizes of rotary blades with a similar configuration and cones as an analog tool were chosen to be in ratios of 1:1, 1:1. 25 and 1:1.61 relative to the smallest size. Tests were run in silt soil at three moisture levels. The tillage test was designed to run in three levels of depth and in one level of forward speed and peripheral speed of rotary blades. Torques on rotary blades, cone penetration forces, depths etc. were measured and average penetration forces from soil surface to each corresponding scaled depth were calculated.
Through the data analysis, the following results were obtained. The average prediction errors from two models of rotary blades were very low, as the result of offset account and accordingly their standard deviations were considerably large. As to the absolute values of the prediction errors, the larger model showed the average error of 13.5%, and the standard deviation of 7.61%. The result was not acceptable but exceeded expectation for such a powered one as a rotary implement, and the cause of these errors was investigated.
