Seminar via Zoom By Lance Cooley, Applied Superconductivity Center, National MagLab

For Scientists
07/01/2020 11:00 am - 12:00 pm


Title: A case study of project lifecycle from research to construction: Conductor development and acquisition for the High Luminosity LHC Accelerator Upgrade Project

Abstract: The Large Hadron Collider (LHC) is arguably the largest machine that humankind has built. It is designed to collide protons and ions to produce new particles tied to the foundations of the universe, which are extremely rare even at 13 TeV energy. Before operation of the LHC began in 2008, it was already apparent that the physics reach of the machine would become constrained midway into planned operations, when increasingly rare events would be expected to dominate attention as the physics of less rare events became worked out. To ensure that future Nobel Prizes did not require more than the lifetime of a physicist, a 20-year roadmap of particle physics in 2003 gave the highest priority to starting a program to increase the luminosity of the LHC and increase the production rate of rare events by a factor of ~10. The High Luminosity LHC (HL-LHC) would require focusing quadrupole magnets beyond the state of the art, which up to 2007 was based on Nb-Ti technology. Catalyzed by the report, focused R&D programs in the US and Europe were begun to develop technology for an eventual HL-LHC, being called the LHC Accelerator Research Program (LARP) in the US Department of Energy (DOE) Office of High-Energy Physics (HEP).

Among the technologies developed under LARP, Nb3Sn superconducting quadrupole magnets were conceived, developed, and matured to a level sufficient to define a conceptual design and begin a formal construction project under DOE O.413.b controls by 2014, now called the HL-LHC Accelerator Upgrade Project (AUP). Development of the Nb3Sn conductor was a necessary underpinning of magnet development, since the conductor status at the start of LARP was strongly influenced by fusion and only capable of current density Jc of ~2000 A/mm² at 12 T, 4.2 K in a production configuration. Sustained support from HEP, including the Conductor Development Program (CDP), was essential to investigate a progression of conductor designs based on the Restacked Rod Process (RRP) at Oxford Superconducting Technology (OST, now Bruker-OST), leading to higher current density and eventually to the conductor specified for acquisition for HL-LHC magnets.

To merit initial consideration by LARP, RRP conductors in 10 kg billets with 350 m pieces using a 54/61 stack achieved a promising Jc of ~3000 A/mm² at 12 T, 4.2 K in 2005. Conductor maturity was gained by holding the Jc requirement constant while:

  1. requesting longer minimum pieces (350 m → 500 m → 750 m → 1000 m);
  2. requesting more complicated restacks (54/61 → 108/127 → 132/169 → 198/217);
  3. acquiring cost and yield data based on 100 kg (20 km) annual procurements;
  4. demanding multi-parameter optimizations (Jc → Jc & RRR → Jc & RRR for both round and 15% rolled strand).

The steady acquisition according to a draft specification also provided magnet technology research with a consistent conductor. By 2015, when the magnet conceptual design was fixed and the initial specification for AUP conductor was set by DOE and CERN, production was understood to the level where statistical process controls could be enforced with 3σ (99.7%) confidence level. In 2017, magnet production prototypes were begun, leading to the successful verification test of the first production magnet in January 2020. Notice here the 3-year lag between conductor specification and the time that the magnet design was validated by a successful test, which is a risk that the project managed by incorporating rather conservative conductor criteria, since the operating point corresponds to a Jc of 2450 A/mm2 at 12 T, 4.2 K. Now, in mid-2020, about 60% of the 10-ton conductor acquisition for the AUP has been received, and the production statistics show the fruit of this development story. I will exemplify some of the production data and problems that were prevented due to production controls, and I will include discussion about project cost estimating and earned value management.