Optimal blade placement for large turbofan balancing

Considers the problem of mounting n blades with different moments on a moving wheel, such that the resultant imbalance is minimized. In the absence of analytical solution, the authors present three stochastic optimization methods, iterative random search, genetic algorithm and simulated annealing, for the general turbofan balancing problem. Specific characteristics of the first two methods regarding this problem are discussed. Numerical results using these methods gave satisfactory results. It is observed that (1) random search is good enough for the off-line design purpose; (2) genetic algorithms have the ability to “memorize” some good substructures for the subsequent candidate solution selection, however proper parameter selection is necessary for great performance. (3) the neighbor set and parameters selection for the simulated annealing can be disadvantageous in this specific application