Fusion Engineering and Design 86 (2011) 2504–2507 Contents lists available at ScienceDirect Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes Development of a filled resin system for the TF coils of ITER Simon James Canfer a,∗ , Stephen J. Robertson a , Elwyn Baynham b , David Evans c , George E. Ellwood a , Stephanie H. Jones a , Juan Knaster d a STFC Rutherford Appleton Laboratory, Didcot, UK b Magnetech, Didcot, UK c Advanced Cryogenic Materials Ltd, Abingdon, UK d ITER Organisation, France article info Article history: Available online 28 May 2011 Keywords: ITER Toroidal Field coils TF Coil Casing Epoxy Filled epoxy abstract The final step in assembly of the ITER TF coils will be the insertion of the Winding Pack (WP) into the TF Coil Casing (TFCC). In order to facilitate mounting of the WP within the TFCC the design is made with clearance gaps of 7–10 mm to allow for manufacturing tolerances and WP adjustment during mounting. It will be essential to fill these gaps as the last step of the insertion operation. The fill process is a critical part of the operation; it completes the geometric location of the WP with respect to the casing and must provide for a uniform force transfer between the WP and casing. The properties and specification of the filler material are therefore driven by these two requirements. Two process options may be considered; direct fill by vacuum impregnation using a filled epoxy; or, pre-fill of the inter-space with dry particles (fibres or beads or powder) followed by vacuum impregnation with a low viscosity (unfilled) epoxy. This paper reports on experimental studies of both process options. The properties of a trifunctional epoxy resin system (TGPAP) filled with wollastonite, dolomite and milled glass fibres have been studied. Studies of flow and particle settling are reported. The same epoxy system has been used to impreg- nate samples pre-filled with dry particles, soda glass and alumina. Thermal contraction and mechanical strength measurements at 77 K are reported for both systems. © 2011 Elsevier B.V. All rights reserved. 1. Introduction The objective of this work was to compare concepts and materi- als for filling the space between the TF Coil Winding Pack (WP) and Case (TFCC). Manufacturing and assembly tolerances of the com- ponent parts will lead to a variable gap of 1–10 mm around the periphery of the Winding Pack, measuring approximately 800 mm 2 in section, and 10 m in height. Two different approaches were pro- posed by the ITER Organisation; firstly, using a filled epoxy resin where a finely divided filler is mixed into the resin prior to injec- tion into the TF coil gap, and secondly, a scheme where the gap is pre-filled with particles of up to 1 mm diameter before injection with an unfilled epoxy. Wollastonite, dolomite and milled glass fibres were investigated as epoxy fillers. Alumina and soda glass beads of 0.75–1.0 mm diam- eter were used to investigate the prefilled option. The epoxy resin chosen for this work is a trifunctional (three reactive groups on each molecule), radiation stable epoxy TGPAP (typically Huntsman MY0510) and an aromatic amine hardener ∗ Corresponding author. Tel.: +44 1235445370. E-mail address: simon.canfer@stfc.ac.uk (S.J. Canfer). DETDA (typically Huntsman HY5200). The mixture has a low vis- cosity and long useable life, so is suitable for vacuum impregnation applications. 2. Methods 2.1. Resin parameters Table 1 gives the formulations of the filled epoxies. All sys- tems used resin TGPAP and hardener DETDA, ratio 100:44 parts by weight. Firstly, the viscosity–time behaviour of the unfilled epoxy was investigated in order to determine a processing temperature that would allow a working time of 24 h at a suitably low viscosity (<300 mPa s). This was followed by experiments to determine a cure schedule compatible with TF coil gap filling by vacuum impregna- tion. The objective was to confirm that full cure could be achieved at relatively low temperature i.e. 90 ◦ C and realistic time (<24 h). A Brookfield DV II+ pro viscometer was used to investigate vis- cosity with time at temperatures controlled at 30, 40 and 50 ◦ C. A cure schedule was determined using a differential scanning calorimeter (DSC) Netzsch model 200F3. Typically, 15 mg of mixed resin and hardener were placed in vented aluminium pans. The 0920-3796/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.fusengdes.2011.04.040