1. The Four-Layer Conventional Magnetic Shield Brad Plaster, Caltech May 25, 2006  “Current” design  Estimated shielding factors  Production and time-scale  Estimated costs $$$

2. Overview of current design LayerRadius [cm] 196.7 2101.8 3106.8 4111.9  Upper cryostat cylinders [ J. Boissevain, 13:17 EDT Today !!! ] ~7.5 m ~5.0 m LayerRadius [cm] 1106.9 2111.9 3117.0 4122.1 62-mil layer thickness → ~5.5 English Tons  Lower cryostat cylinders Δ ~ 5.1 cm ~2.5 m 9 panels around circumference μ-metal tubes around penetrations

3. Overview of current design  Lower cryostat inner shielding  Proposed upper cryostat inner shielding [ferromagnetic shield]  Improve field uniformity from 3 He spin-holding cos θ coil  Dimensions driven by size of 3 He spin-holding cos θ coil ferromagnetic shield superconducting shield ItemRadius [cm] Superconducting62.9 Ferromagnetic62.2 B 0 cos θ coil61.0 Lengths little less than 4.0 m 4K shield [ J. Boissevain, March 2006, NCSU ]

4. Estimated shielding factors: I  Transverse shielding factor, S T, for n-layer configuration of infinitely-long concentric cylinders  Sumner, Pendlebury, and Smith, J. Phys. D 20, 1095 (1987)  Proper evaluation requires a value of μ i for each layer  μ n = μ(H) for H = H ext  μ n-1 = μ(H) for H = H ext / [ shielding from n th layer alone ]  μ n-2 = μ(H) for H = H ext / [ shielding from n th + (n-1) th layers ]  … μ-metal Cryoperm at 4K

5. Estimated shielding factors: I  Results for lower cryostat shielding  Four-layer conventional shield + 4K ferromagnetic [ Cryoperm ] shield  Presence of superconducting shielding ignored Layerr [cm]μ(H)Accumulated S T 4122.115900Layer 4: 11.3 3117.0170000Layers 4+3: 221.2 2111.9100000Layers 4+3+2: 1663.7 1106.970000Layers 4+3+2+1: 10116.8 C62.270600Layers 4+3+2+1+C: 6.60 х 10 5 Residual Shielded Field = 7.6 х 10 -7 Gauss Ideal case: infinitely-long cylinders, no penetrations, no end-caps

6. Estimated shielding factors: I  Results for upper cryostat shielding  Four-layer conventional shield only Layerr [cm]μ(H)Accumulated S T 4111.915900Layer 4: 12.2 3106.8170000Layers 4+3: 261.8 2101.8100000Layers 4+3+2: 2282.4 196.770000Layers 4+3+2+1: 1.61 х 10 4 Residual Shielded Field = 3.1 х 10 -5 Gauss Ideal case: infinitely-long cylinders, no penetrations, no end-caps Proposed inner ferromagnetic shield will provide greater shielding factor, comparable to lower cryostat

7. Estimated shielding factors: II  Zeroth-order finite-element analysis calculation of shielding factors [TOSCA → ASU collaborators]  Finite-length concentric cylinders without end-caps  Lower and upper cryostats considered separately 4-layer conventional shield Cryoperm shield H = 0.5 Gauss transverse field

8. Estimated shielding factors: II  Results for lower cryostat shielding  Four-layer conventional shield + 4K ferromagnetic [ Cryoperm ] shield  Presence of superconducting shielding ignored  TOSCA result for residual field in the center of symmetry of the five concentric layers 1.77 х 10 -6 Gauss  Within factor of ~2 of estimation assuming infinitely-long cylinders TOSCA results as function of length of the four-layer conventional shield

9. Estimated shielding factors: II  Results for upper cryostat shielding  Four-layer conventional shield only  TOSCA result for residual field in the center of symmetry of the four concentric layers 2.10 х 10 -3 Gauss  Calculation probably not that realistic  Lengths of cylinders (~2.5 m) taken to be distance from top of upper cryostat to lower- cryostat, upper-cryostat “tee” TOSCA results as function of length of the four-layer conventional shield

10. Shielding factor optimization  Inspection of the formulas for the transverse shielding factor suggests that S T can be enhanced via a judicious choice of the radii If (r j /r k ) sufficiently small, S T ~ S i T S j T S k T...

11. Shielding factor optimization  Inspection of the formulas for the transverse shielding factor suggests that S T can be enhanced via a judicious choice of the radii If (r j /r k ) sufficiently small, S T ~ S i T S j T S k T...  Carried out a grid search over radii of four-layer shield  4K ferromagnetic shield and inner-most layer of four-layer shield radii fixed  Baseline configuration  ~220 m 2 for lower cryostat  ~ 67 m 2 for upper cryostat  Computed S T for the radii of each grid combination, then estimated cost by scaling material costs to surface areas → ~70% of the cost for the shield will be material cost (more later about costs…)

12. Shielding factor optimization Lower cryostat Upper cryostat

13. Calculations: future TOSCA  First approximation  Finite-length cylinders, no end-caps  Ignore lower-cryostat and upper-cryostat connecting “tee”  Transverse and axial shielding factors  Verify results of layer spacing optimization  Second approximation  Add end-caps with geometry similar to that in Jan’s “latest reference design”  Optimize the geometry of the end-caps  Full model  Join the lower-cryostat and upper-cryostat cylinders  Model penetrations through the structure

14. Production of the shield  Vendor selected is Amuneal  Located in Philadelphia (in the north-side ghetto…)  Magnetic shields we have procured from Amuneal for our R&D activities have all been delivered on-time  Recently, have provided a free Cryoperm shield (valued at ~$3000) as part of joint R&D efforts between Caltech and Amuneal  Delivered a ~1/4-scale multi-layer shield to J. Kirschvink at Caltech (geophysics)  Provided magnetic shielding for JLab, BNL, SNS (Beam Line 15), … ; now in discussions of magnetic shielding requirements for the ILC  Preliminary design review conducted on-site at Amuneal on March 6, 2006  B. Filippone, J. Boissevain, W. Sondheim, S. Currie, and BP

15. Production of the shield  Driving constraint for production is the size of their furnace  5’ diameter х 12’ long  Assembly will consist of many “panels”  For magnetic continuity, Amuneal has proposed using 4”-wide battens along the circumferential and horizontal seams to be tightened with aluminum clamps  Envisage an Al support structure for the shield  Tied into the main support structure for the lower and upper cryostats, without transferring any of the load from the cryostats onto the μ-metal  Implicit that Amuneal will provide the Al spacers between the layers  Also implicit that Amuneal will construct the Al support structure, and “protective covering” for outer-most layer

16. Assembly issues  Modular assembly  Lower cryostat shielding around the measurement cells → permits B-field tests, etc. in lower cryostat  Region near “rear” end-caps easily removable  Upper cryostat shielding, after ready to drop in upper cryostat “guts”  “Top” end-caps  “Rail system” for removing “rear” end-caps for access to B-field coils and measurement cells  Upper cryostat “cut in half”, for access to interior after installation Serious integration issues !!

17. Time scale  Now – April 2008: Design for shield, support structure, degaussing system, integration issues, etc.  April 2008 – December 2008: Procurement period  Cost fixed at time of PO  Shielding shipped via semi-truck to Caltech for test assembly/installation starting January 2009 in high-bay area  Measure residual fields  Shielding used for tests of all B-field coils to be fabricated at Caltech/ASU  Installation of shielding structure and magnets at ORNL beginning November 2009

18. Expected cost $$$  Prior to February 2005 review, quote obtained from Amuneal

19. Expected cost $$$  Prior to February 2005 review, quote obtained from Amuneal  Design to be delivered to Amuneal in early-June, quote to be generated for CD-1 Review $800 to $900 k

20. Expected cost $$$ Quote for Feb. 2005 Review

21. Saturation ?

22. Long-Standing Problem Motivation to investigate impact of weld beads/seams on N=20 cos θ coil + shield uniformity came from March 6 meeting at Amuneal μ-metal Cryoperm

23. Long-Standing Problem TOSCA results –Modeled cos θ coil + 62-mil thick μ-metal shield WITH 102-mil thick weld seam μ-metal