Simulation-based parameter identification for online condition monitoring of spindle nut drive

Abstract

In this article the development of a method for simulation-based condition monitoring of a spindle nut drive for machine tools will be presented. Thereby, parallel to the operation of the spindle nut drive, an automatic parameter identification of a corresponding simulation model is to be carried out with the aim to identify high-level information like stiffness and damping of the significant components based on the available drive signals. The underlying model for the identification consists of Finite Element (FE) component models and the corresponding component parameters like stiffness and damping of the bearings, spindle nut, etc. Beyond the parameter identification, the characteristics of the components (here stiffness) will be computed by the mentioned model. The identification and the calculation method in this paper is based on finding optimum stiffness parameters which are correspondent to the current state of the system and using a neural network to find the relation between the physical parameters of the system and measurable parameters of the system behavior. The results depict a new diagnostic process which could be also applicable for online condition monitoring of different components.