Application of Fast Prototyping Method into the New, Real-Time Runout Elimination Algorithm

摘要

Technical condition of rotating machinery and type of possible present inefficiency are evaluated on the base of some selected symptoms. Magnitude and shape of shaft orbit in oil-film journal bearing are one of the most useful and frequently used diagnostic signals, so proper measurement of shaft vibrations is critical for the diagnosing process. Shaft vibrations are usually measured using eddy-current probes, what produces characteristic, hard to reject noise called runout. A new approach to a problem of runout elimination, utilising Kalman filter, gives consideration to the dynamics of the monitored machine and allows real-time execution of necessary calculations. For validation of the proposed runout filtering method a procedure of computer aided design and fast prototyping was developed . This procedure consist of: 1. computer aided modelling of the rotor-bearings system; 2. model tuning; 3. filter design; 4. simulation of filter operation; 5. fast prototyping, based on DSP; 6. implementation of developed filtering method on industrial-standard computer. Special attention was paid to the automatization of prototyping and implementation processes. For model tuning and validation, trajectories generated by MESWIR ware used. MESWIR is experimentally verified software package, based on FEM model of the rotor-bearings system, developed in Institute of,Fluid-Flow Machinery PAS . Experimental investigations ware conducted on constructed in IFFM PAS large-size rotating machine model, which gives the possibility to perform precise, off-line runout elimination for validation purposes. The paper describes tools (hardware and software) used in the project, reports the course of work and results obtained so far.