Data analysis of the strain gauges system of the W7-X superconducting coils

Abstract

Wendelstein 7-X (W7-X) is an advanced stellarator which is presently under development and construction at the Max-Planck-Institut für Plasmaphysik in Greifswald, Germany. The W7-X is a magnetic confinement device where the main magnetic configuration is produced by a set of 70 superconducting coils, arranged in five identical modules. A module can be further divided in two half-modules, each made up of five non-planar coils (Types 1–5) and two planar coils (Types A and B). To monitor the integrity of the coil system and its neighbouring components during the operational and accidental transients of the W7-X device, a strain gauges (SG) system was included in the machine design, distributed in those positions which will have, realistically and according to finite elements (FE) calculations, the most high stress levels. Specifically, each superconducting coil of the W7-X system is monitored via a set of four SG sensors which are firmly attached in specific locations of the coil steel casings. Each sensor consists of a bridge of two individual foil gauges, connected so to compensate the effects of the broad excursion from room to cryogenic temperature, and the four sensors are arranged in a double redundant array for increased reliability. To estimate the response of the SG system in terms of accuracy and precision, so actually in terms of degree of conformity to FE anticipated results and degree of repeatability of similar measurements, and also to explain some unpredictable results obtained at cryogenic temperature, a comprehensive data analysis was carried on, based on the SG data collected during the cold acceptance tests of the W7-X superconducting magnets. The cold testing phase was, indeed, a valuable tool to verify the response of the SG system in the framework of the overall acceptance process of the coils. Thanks to the consistent data pool, a meaningful statistical analysis was possible, and the behaviour of the SG system was observed from a new perspective. This paper describes the procedures adopted in the aforementioned data analysis, and provides a critical overview of the main results. The main issues faced and solved in the accomplishment of the analysis were mainly related to the handling of huge raw data files, and to the conversion of such data into meaningful information.