As-built based method for virtual alignment and assembly of the ITER tokamak

Abstract

In the project development of large fusion machines, the alignment and assembly phases represent an open challenge. Each mechanical component that constitutes the whole machine assembly and fixtures is affected by an intrinsic geometrical variation due to manufacturing and assembly processes. The complexity of the assembly together with the unavoidable geometrical imperfection can affect the expected machine functionalities. In this context, Reverse Engineering activities, aiming at obtaining the as-built of the components, coupled with virtual fitting studies, defining optimum alignment using the real shape of the components, become powerful risk mitigation tools in the assembly phases. This work presents a systematic method, complementary to the methodology currently followed in the ITER project, for the as-built management and for the validation of alignment phases of such a complex assembly. The first phase is to identify the key assembly requirements which drive the alignment of the ITER assembly. Then, the database for the requirement collection is clearly organized following the proper nomenclature rules and its architecture for data collection and analysis is indeed set up. Four virtual fitting pilot studies are presented to show how the methodology is helpful to assess the assembly requirements in a defined alignment scenario.