Methods of evaluation of the positioning capability of Cartesian and non-Cartesian machines
Manufacture and assembly of ever more precise components has been the driving force for many research projects. Error avoidance and error correction are used to improve the accuracy of the final output, but it is only by error evaluation that a manufacturer can quantify his production capability. The ability of a machine to perform the task designated to it is of critical importance. In particular it is essential to be able to determine the capability of a production machine to produce a part accurately, a measuring machine to dimension a part reliably or a handling machine to place a part in the appropriate position. In order to achieve this it is necessary to establish the positioning accuracy of the device throughout its working volume. In addition to the need for the assessment of machining performance, there is a strong desire within the machine-building community to allow for the post-assembly correction of errors inherent in the manufacture of machines. Such techniques are often a cost-effective complement to error avoidance. During this project, a new geometric model and supporting measurement methods are produced for the evaluation of errors in Cartesian-based machines. This work is an extension of that performed by Ford, et. al. Ell as discussed in chapter 3. The new work addresses the previously unresolved problem of determining the errors throughout the working volume of a machine with volumetric compensation and a tool or probe offset. This simulation method is in contrast to other techniques that quantify machine performance based upon a small subset of the machine volume. It is proposed that a figure for volumetric accuracy derived from these methods cannot stand on its own as a description of the manufacturing capability of a machine. The effects of measurement uncertainty on the synthesis technique have been examined and modelled. This has produced a method of quantifying machining capability based upon machine configuration, tool or head configuration, and supported by uncertainty based upon the test data input to the model. An alternative method of evaluating errors through the working volume is to measure directly using, for example, a tracking laser. One such system (LaserTrace) is based upon absolute position being resolved by trilateration from two tracking lasers. This system has been investigated for its applicability to the measurement process. Two methods have been produced to improve the accuracy of the system and reduce the time required for its calibration. One is based upon photogrammetry techniques, the other on a novel use of a -imachine checking gauge (MCG). The artefact is used for acquisition of data to perform parameter identification on a model of the system that has been found from first principles. This MCG-based calibration technique was successful, within the constraints of the resolution and repeatability of the control loop. Attempts were then made to apply this methodology to a second machine of the non-Cartesian type and configuration (UMD). Simulation shows this technique to be applicable, but the instability of the prototype precluded comprehensive on-machine testing. In the course of this research a thennal model of the UMD has also been produced to overcome the sensitivity of the prototype device to temperature changes. Such a model could be used to provide software correction of LTMD position values.