Large Magnetic Volumes for Neutrino Factory Detectors A.Bross ISS Detector Phone Meeting June 22, 2006
Options We have begun looking into the engineering realities of trying to magnetize very large (>30k m 3 ) volumes What we are considering is something much much larger than what has been done to date Technologies u Room temperature Cu or Al conductor - NO s Power dissipation is MUCH too high u High T c superconductor – NO * s At this point in time for the same Ampere-Turns: 200X more expensive than convention SC s * However, development progress in recent years has been rapid so the situation could change in the near (5 yr) future. u Conventional SC s Lots of experience, but this size is new. s Technically – certainly doable s BUT WHAT IS THE COST?
15 m X 15 m X 15m modules B = 0.5T Magnet Steel Magnetic Tunnel Magnetic Cavern Multiple Solenoids - Conceptual Layouts
Cost Modeling Green and Lorant is a good starting point u “Estimating the Cost of Large Superconducting Thin Solenoid Magnets” – 1993 u C(M$) = 0.5(E s (MJ)) C(M$) = 0.4(B(T)V) We can also take the CMS Coil as-built cost ( $55M) as a more recent reference point u B = 4T u V = 340 m 3 u Stored Energy – 2.7 GJ For the NF case take a 15 X 15 X 15 m 3 volume with B=0.5T u Don’t worry now about whether this is a cylindrical solenoid or a box. s This will of course be very important mechanically
Cost Extrapolations for Baseline Magnet Cost via stored energy Stored energy 340 MJ u From Green and Lorant C(M$) 0.5(340) 24M$ Cost via Magnetic Volume u From Green and Lorant C(M$) 0.4(.5 X 3400) 45M$ Reference Point – CMS Solenoid C(M$) 0.5(2700) 93M$ (Stored energy) C(M$) 0.4(4 X 370) 41M$ (Magnetic volume) Most Optimistic Extrapolation u Use stored energy and conclude formula overestimates by factor of 1.7 (93/54) based on CMS case s Then NF magnet extrapolated cost – 14M$ Most Pessimistic Extrapolation u Use magnetic volume and conclude formula underestimates by a factor of 1.3 (54/41) based on CMS case s Then NF magnet extrapolated cost – 60M$
Magnet Costs So it is clear that there is a large uncertainty in the cost just based on extrapolations Conclusions: u For low-field case (B<.5T) scaling formulae may not be accurate due to the large size of magnets being considered s Vacuum loading (vacuum vessel) will be a major consideration and will strongly impact cost s Superconductor itself is not a cost driver –Based on recent MICE order, cost for baseline NF magnet discussed here is <0.5M$ s On-site fabrication required u Magnets of this size can certainly be built, but better cost estimates will only come after some real engineering analysis s 3-6 month effort s We have started looking at the vacuum vessel here at Fermilab