Research On Aligning The Measured Data Of Complex Blank Parts To The Nominal CAD Models Under The Machining Allowance Constraint

The3D profile measuring technique is being used in industrial production more and morepopular and improving the traditional manufacturing method day by day. There are many structuralparts in industrial production. These structural parts are usually obtained by means of first casting theprefabricated parts whose size is very close to the design size and then finishing the prefabricatedparts to meet the design requirements. Towards this small allowance finishing processing of thesecomplex structural parts, the3D measurement can improve the processing stage on the following twoaspects.①The terribly deformed blank parts can be rejected before machining so they will not bemachined and the unnecessary waste is avoided.②Make sure the blank parts are positionedefficiently and reliably in the CNC machine tool so as to improve the efficiency while avoiding lackof allowance resulting from poor localization during machining. For these two aims, the measureddata of blank parts must be aligned to the nominal CAD models under the given machining allowanceconstraint. The constraint alignment technique has been fully studied in this thesis and an effectivealgorithm is proposed to solve this alignment problem. In addition, a software system is designed andmade with the key algorithms. The thesis’s main contributions are as follows.⑴The Half-Edge Data Structure is proposed to store triangle meshes after analyzing the featureand topological relation of triangle meshes. Also the algorithm flow for reconstructing meshes’topology from the file of obj format is presented.⑵A constrained nonlinear least squares mathematical mode is established for the aligning themeasured data of blank parts to the nominal CAD models under the given machining allowanceconstraint. And an efficient method for computing the closest point of the measured data of blankparts to the nominal CAD models is analyzed.⑶A2-step algorithm is proposed after analyzing the experiments and the optimization theory.Firstly, registration without considering any constraints is performed by the ICP algorithm. If theregistration result meets the machining allowance requirement, the algorithm terminates; otherwise,the further registration considering allowance constraint is performed to make sure the final result canmeet the given allowance requirement.⑷A software system is designed and implemented for solving the registration problems andseveral specific registration problems are settled by the software system. The encouraging experimentshows that the related theories of this thesis are correct and the software system is practical.