The historic story of the superconducting (SC) magnet for MHD ETL Mark V Facility (cf. Fig. 1) is presented. This SC magnet was developed in the MHD Project (1966-75), which was one of the first MITI/AIST Large-Scale R & D Projects and the first national project for superconductor applications in Japan. This SC magnet had been the largest in Japan through 1982 when the Japanese LCT coil was made by JAERI. It was completed in 1973 after many difficulties, some fatal and some trivial, because of a lack of knowledge, before it could generate maximum magnetic field 7T with stored energy 65MJ. Because technological problems had piled up and because no management know-how of national projects on technology development had been accumulated before then in the forefront of worldwide technological advances, “step by step” advances and “trial and error” attempts in the progress of the project had to be done over again. The paper is divided into three parts. Part I describes the process of determining the specifications of the magnet and the conductor tests before the magnet was manufactured.
Hybrid composite pipes reinforced with high-strength polyethylene fiber (DF) and alumina fiber (AlF) were prepared to develop the coil bobbin for stable superconducting coils. The bobbin in which the circumferential thermal strain expands with cooling and in which the circumferential Young's modulus is large would be effective for stable coils. The unidirectional hybrid composite (ADFRP) showed 0 thermal expansion coefficient when the ratio between DF and AIF volume was 5/5 in fiber direction, and its Young's modulus was larger than that of DF reinforced plastic (DFRP) both in parallel and perpendicular to fiber direction. The circumferential and longitudinal Young's moduli of ADFRP pipe were larger than those of DFRP pipe. The average value of inner and outer circumferential thermal strains with cooling down showed 0 with a filament winding (FW) angle of 90deg when the ratio between DF and AIF volume (D/A) was 5/5. When D/A equals 5/5, the calculated thermal strain with cooling of the pipe showed good agreement with the average of observed inner and outer thermal strains. The circumferential thermal strain showed an expansion with a FW angle of 50-90deg, and an absolute value was smaller than those of DFRP. The inner and outer circumferential thermal strains were different. The difference decreased with an increase of the ratio inner diameter/thickness, and the differences were smaller than those of DFRP with a decrease of the degree of anisotropy of thermal expansion coefficients in UD-FRP. The experimental data were obtained to make it possible to devise a coil bobbin with negative thermal expansion coefficient by ADFRP.