Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/81903
Title: Fiducial-aided calibration and positioning system for precision manufacturing of freeform surfaces
Authors: Wang, Shixiang
Advisors: Cheung, C. F. Benny (ISE)
Keywords: Machining
Measurement
Issue Date: 2019
Publisher: The Hong Kong Polytechnic University
Abstract: Freeform surfaces generally possessing non-rotational symmetry are gaining popularity with optical system designers. These freeform surfaces remarkably enhance the functional performance and efficiency of optical systems and conserve more material in the production process. Precision manufacturing of these surfaces with sub-micrometer form accuracy and good surface roughness require accurate positioning of the workpiece on different machining devices during the fabricating cycle. However, the geometric complexity of freeform surfaces creates significant difficulties in the precise localization of the surfaces for different processes, including the machining and measuring processes, which considerably limits further improvement of surface accuracy and machining efficiency. Motivated by the aforementioned limitations, this thesis presents a fiducial aided calibration and positioning method (FACP) to position freeform optics during the manufacturing cycle, including design, machining, measurement and evaluation. In order to achieve precise localization of the freeform surface on machine tools and the measurement instrument so as to achieve high alignment accuracy and compensation of the machining error in precision manufacturing of freeform surfaces, the FACP method utilizes fiducials to form a Fiducial-aided Computer Aided Design (FA-CAD) of the freeform surfaces. This not only establishes an inherent surface feature to link the coordinate frame among the machining and measuring machines but also calibrates and compensates the errors resulting from both machine tool and measurement instrument. To develop the fiducial aided calibration and positioning system, a fiducial aided positioning system was firstly developed by combining a square based fixture with standard spheres as fiducials. With the proposed method, this square based fixture with an adjustable number of fiducials was verified with high precision. The results show that the proposed method has the ability to position freeform surfaces with higher accuracy and is more robust than the traditional least square based method. In order to test the possible application of the proposed method in ultra-precision machine tool, a probe-based measurement system was developed on an ultra-precision raster milling machine tool. A compensation process carried out on several optical freeform surfaces successfully achieved sub-micrometer form accuracy.
For some complex freeform surfaces, which are usually measured by multi-sensor instruments, data fusion is particularly important for improving measuring accuracy. In this study, a fiducial-aided Gaussian process based data fusion method is developed. The method makes use of standard spherical fiducials as references to check the errors that influence measuring accuracy. The obtained uncertainty was used to determine the parameters used in the Gaussian process modeling and to determine the weights used in the weighted mean data fusion. The measured datasets from different probes in the coordinate measuring machine were fused by a weighted mean method with Gaussian process modeling. The errors lead to the on-machine measurement for fiducials being susceptible to geometrical volume error of the machine tool. To address this problem, a fiducial-based calibration system was established to provide information for error modeling of the machine where the workpiece is machined. This error modeling is intended to establish a relationship between the designed fiducial aided system and the machine tools used for machining and measuring the workpiece. A linear transforming model is firstly used for connecting the accuracy of the machine tool with that of the off-machine measurement instrument with high precision. A modified chi-square technique is then used to obtain all the uncertainties associated with the linear model. The main contributors of error are identified and considered in the model, including calibration error of the reference datum, error of the on-machine measurement system, error in determining the centers ofthe spheres, the defection of the fixture, and the volumetric error of the machine. Experiments have been conducted with results showing that the achievable accuracy of the fiducial-aided calibration and positioning system is sensitive to several factors, including accuracy of the on-machine probing system and distribution of the fiducials. The study not only provides an important means for significantly improving the efficiency and accuracy of the traditional manufacturing process of ultra-precision freeform surfaces, but also for enhancing the capability of the measuring instrument in the form characterization of these surfaces. The thesis contributes significantly to a novel precision metrology method and system for the precision manufacture of freeform surfaces.
Description: xx, 191 pages : color illustrations
PolyU Library Call No.: [THS] LG51 .H577P ISE 2019 Wang
URI: http://hdl.handle.net/10397/81903
Rights: All rights reserved.
Appears in Collections:Thesis

Files in This Item:
File Description SizeFormat 
991022289508103411_link.htmFor PolyU Users168 BHTMLView/Open
991022289508103411.pdfFor All Users5.21 MBAdobe PDFView/Open
Show full item record
PIRA download icon_1.1View/Download Contents

Page view(s)

14
Citations as of May 6, 2020

Download(s)

3
Citations as of May 6, 2020

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.