New optical sensing system applied to taut wire based straightness measurement

In modern manufacturing industry, precision components are typically produced on Computer Numerical Controlled (CNC) machine tools which translate their accuracy onto machined parts. This accuracy is affected by a set of different motion errors caused by inherent imperfections in the design and build of the machine, variations in the local environment such as temperature, the cutting process itself and human factors. The reduction of these effects is achieved primarily through design improvements and error compensation techniques. The latter requires detailed knowledge about the existing errors in order to deal with them effectively.
This thesis describes a novel sensor system for measurement of errors caused by deviation in the straightness of Cartesian axes present in the structural loop of most machine tools. Currently there are very few methods available to measure straightness directly, each having advantages and disadvantages when considering simplicity, accuracy and affordability. The proposed system uses a taut wire reference with a novel sensor, a two-point technique for reference error cancellation and software to enable fast and accurate measurement of straightness between any two points of the measured machine’s working volume.
The standout features of the sensing system include ultra-low cost and high performance when compared with existing state-of-the-art systems. It is capable of measuring a straightness error as low as 3μm and takes only 2s of dwell time between readings, while laser interferometer requires 4s to perform averaging when measuring the same error. Existing taut wire microscopy is limited by 10-20μm of measured error depending on optics quality and manual reading takes at least 5s to minimise the human error. Setup time is also different – the new system saves 15 minutes time on 2m axis and more on longer lengths compared the laser due to simpler reference alignment procedure.
Theoretical analysis and practical implementation are followed by detailed performance evaluation experiments carried out under typical manufacturing conditions comprising different machine tools, different axes, measured errors, environmental effects and alternative measuring equipment. Tests cover aspects of accuracy, repeatability and overall system stability providing a complete picture of the system’s capability and the method’s potential which is also supported by uncertainty analysis. In addition to defining setup and measuring procedures, a user-friendly software interface is described and its main units are explained with respect to overall measurement efficiency and setup fault detection.