Analysis of changes in coordinate measuring machines accuracy made by different nodes density in geometrical errors correction matrix. (May 2015)

• Record Type: Journal Article
• Authors: Gąska, A.; Sładek, J.; Ostrowska, K.; Kupiec, R.; Krawczyk, M.; Harmatys, W.; Gąska, P.; Gruza, M.; Owczarek, D.; Knapik, R.; Kmita, A.
• Is Part Of: Measurement. Volume 68(2015:May)
• Page Start: 155
• Page End: 163
• Publication Date: 2015-05
• Subjects: Geometric compensation -- CAA -- Geometric error -- Machine mapping -- CMM; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2015.02.056
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: Changes of CAA matrix nodes density between 40 to 100 mm does not influence accuracy of machine strongly. LaserTracer can map standard CMMs almost two times faster than classical interferometer. The resulting accuracy of machine correction is comparable for these two systems. LaserTracer is insufficient to determine geometric errors of high accuracy CMMs. Abstract: Advances in modern manufacturing techniques increase production efficiency but, at the same time, present new tasks and challenges for coordinate metrology and the manufacturers of Coordinate Measuring Machines (CMMs). The main goal of current research efforts is improving measurement accuracy. Seeing as many of the possible solutions regarding CMM construction had already been explored, there seems to be little left for improvement in that field. Further efforts at accuracy improvement rely mostly on using sophisticated mathematical algorithms designed to correct relevant errors. Many types of errors could be compensated using this approach, including: probe head errors, machine dynamics errors and, most importantly, machine geometrical errors. Almost all coordinate measuring machines produced nowadays are equipped with geometrical errors compensation matrix known as the CAA matrix (Computer Aided Accuracy). CAA matrices are based on a grid of reference points (nodes) in which certain values of the components of geometrical errors are determined experimentally. The error values between the nodes are estimated using simple interpolation methods. Theoretically, a higher density of reference points on the grid describing the CAA matrix should improve the accuracy of the machine utilizing the matrix. On the other hand, increasing the number of nodes simultaneously increases the amount of workload, time and money spent on constructing the CAA matrix. This paper presents a number of experiments aimed at creating CAA matrices with different number of matrix nodes using the LaserTracer system. The relations between maximum permissible errors obtained on a machine using matrices with different densities of nodes are also discussed. Additionally, the authors attempt to tackle the question of determining the most optimal density of nodes with regards to the ratio of time spent on matrix creation and the effect on accuracy.