Research On Method Of Shape Error Separation And Evaluation For Large-scale Cylinder

At present the most effective means of shape error measurement is still using 3-D machine or roundness measurement instrument, but both high prices, high cost of measurement, and strictly demanding in the operating environment and condition, so they are usually limited to use in the room. Furthermore limited by the size of equipment, they are general used only for measuring small parts. For accuracy measuring geometry of the shape and position error of large work-pieces, the general means is on-line measurement. As the error inevitably exists in the benchmark machine guides, rotary centre and installation, it would include benchmarks measuring signal errors. How to reduce the uncertainty of measurement, improving the measurement accuracy has become one of the most eye-catching issues in testing technology in recent years.For the enormous size of the measured object, special materials and high precision of the profile measurements, as well as the limitations of the environment, conventional shaft sizes or shapes error measurement methods is difficult to be efficiency. It's important to develop new large work-pieces'shapes online measurement system.This paper puts forward to the multi-sensor measurement with data redundancy method, combining separation technology with multi-sensor data redundancy method, and puts forward to the mathematical model of the measurement of straightness profile information ,conduct a study on the proposed multi-sensor data to realize the concept and method of separating movement error of measuring trestle, and research on the method in overcoming the cumulative effect caused by the multi-sensor's zero error and the reading error on the horizontal direction. By comparison with traditional there-probe error separation technology, confirmed that the method can effectively reduce the accuracy impaction of the multi-sensor reading error and zero error on the horizontal direction. Such impaction will be eroded with increasing the number of the horizontal direction's sensor, thereby confirm the data redundant technology can effectively improving the accuracy of error separation, then effectively realize self-tuning and self-adaptive measurement of the straightness.Mathematical model of roundness profile information has been established from the circumference direction by multi-point measurement with data redundancy method. Error separation can be achieved by means of prior separating roundness (PSR), prior separating splindle radial motion (PSS), or synchronal separating both roundness and splindle radial motion (SSRS) of the time-domain three-point roundness error separation methods. Research on the relationship between the harmonic distortion with sensor's location,the number of circles and sampling sites .Discover the proposed method of PSS in time domain roundness error separation method can avoid an order harmonic distortion, and can reduce harmonic distortion to a minimum on the way to increasing the sample points on circumference.After separating the straightness error and roundness error, the final mathematical model of profile information of cylinder is established. The model is used the principle of least squares match, and the corresponding algorithm is established. To verify the correctness of the algorithm and obtain the redundant measurement data, a series of software are designed for simulating both the shaft-beating error and the movement of measuring trestle in the measurement process. By simulating experiments, the correctness of the algorithm is confirmed, the facts that the harmonic distortion on the circumference also impact on achievement of profile information of cylinder is revealed.