sis 100 main magnets g. moritz, gsi darmstadt (for e. fischer, mt-20 4v07)) cryogenic expert...
TRANSCRIPT
SIS 100 main magnetsSIS 100 main magnets
G. Moritz, GSI Darmstadt
(for E. Fischer, MT-20 4V07))
Cryogenic Expert Meeting, GSI,
September 19/20 2007
3334 W
7623 W
533 W
2078 W
1361 W
static load @ 4K
beam induced loss
dynamic load @ 4K
liquefaction
static load @ 50-80K(4K equivalent)
Example: Cryogenic load distribution of the Synchrotrons SIS 100/200 (maximum load)
AC loss contributions (dynamic heat load)
• Magnet iron yoke (hysteresis loss)• Structural elements (hysteresis and eddy current loss) • Beam pipe (eddy current loss)• Conductor (hysteresis and eddy current loss )
Mechanical structure / lifetime of the magnets
• SIS100 : 200 millions cycles within 20 years
• SIS 300: 1 million cycles within 20 years
minimization of movement of any part
R&D on material fatigue, crack propagation
Main R&D Topics for rapidly-cycling magnets (Hz-range)
Eddy and persistent currents • affect field quality
• produce large steady-state AC-losses in coil, yoke, structural elements, beam pipe
minimization of these effects
good heat removal
SIS 100 superferric dipole
• iron-dominated superferric design (window frame type)
• cold iron• maximum field: 1.9 T• ramp rate: 4 T/s• 3 m long, 16 turns
based on Nuclotron dipoleCollaboration: JINR (Dubna)
1- cooling tube, 3 - Superconducting wire, 3 - NiCr wire, 4 - Kapton tape, 5 – adhesive Kapton tape
• hollow-tube superconducting cable
with low hydraulic resistance
• two-phase helium cooling
• strands indirectly cooled
compact, low cost design
example: calculation of hysteresis loss in the brackets
AC loss reduction (model magnets / FEM-calculations)
70% of the Nuclotron dipole losses are in the yoke!!!
Experimental studies on model magnets Theoretical ANSYS calculations
example: loss reduction by slitting the pole and reduce brackets material
Heat release in test dipoles
0
10
20
30
40
50
Nuclotron stainlesssteel (SS)end plates
SMP endblock
insertions(z=5cm)
no SMP but6 slits
(z=20cm),reducedbrackets
reducedcoil end
loop
improvedsc-wire
(EAS) forNuclotron
cable
bracketsand end
plates fromSS, no slits
additional 6slits
(z=20cm)modifications =>
Qcy
cle,
J
totalyokecoil
Modifications: removed brackets laser cut lamination slits minimized coil ends
R & D Results: AC loss reduction @ 4 K
assumption for 20years operation costs comparison study:
nc version: 26M€
sc version: 2.6 M€
Loss (J) per cycle (0-2T, 4T/s), 1.4 m long test dipoles
SIS100 dipole coil design
Goal:
• accurate positioning
• reduction of point load
Result: tube will survive 20 years of operation!
Status:
• mockups were produced
• mechanical properties tested at 77K
Coil support structure Fatigue /crack propagation of the Cu-Ni- tube
detailed ANSYS model of the coil / conductor (Courtesy of E. Bobrov)
3 full length dipoles / 1 quadrupole under construction
• 2.1 T dipole, straight at BNG• 2.1 T dipole, straight at JINR• 1.9 T dipole, curved at BINP• 27 T/m quad, 1. 3 m at JINR
Cooling
• magnet– coil and yoke in series, coil first, determines the
flow resistance
• vacuum chamber (reinforced by rips)– by insulated pipes soldered to the rips
from BINP
operating cycles / loads
magnet designed for cycle 2c: 800 msec Injection, 1 sec Pulse (0-2 T, 4 T/s)
design limited by hydraulic resistance!!
new requirement: triangular cycle, 1 sec → load almost doubled!!!
→ Consequences for the refrigerator
pressure drop
Reduction of hydraulic resistance
• increase tube diameter → filling of aperture, double layer
• reduce number of turns → high current, single layer