Abstract
A new black box approach is suggested based on running a minimum number of balancing trials and using regression analysis to relate the vibration amplitude at one or two rotor bearings to the balancing variables (weight, mass and location angle). A powerful optimization technique is used to minimize an objective function combining the vibration levels at both bearings with pre-assigned weighting constants subjected to constraints on the balancing variables. The technique can be applied to both static and dynamic balancing of rigid rotors and a computer program is outlined to facilitate the application of the suggested approach. The preliminary results are encouraging; a reduction in vibration level up to 83.7% has been achieved. Although the approach is applied to rigid rotor balancing problems, it may be of interest in other dynamic problems such as rotors having non-linear support characteristics, fine balancing, variable speed rotors and flexible rotors.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have