Title: Ferrite Tests for Mu2e Beam-Line Extinction Uses
1Ferrite Tests for Mu2e Beam-Line Extinction Uses
- G. Velev
- Technical Division
- Magnet Systems Department
2Introduction
- Ferrite pulsed magnets are commonly used in the
accelerator applications - kickers injection, extraction, gap clearing
(recently MI) - Orbump beam orbit manipulation
- All of them have a low operational duty cycle,
practically lt1-3 - For Mu2e, an experiment which searches for a µ-e
conversion with an unprecedented sensitivity of
10-16 a new type AC dipoles are needed - These dipoles will be used to extinguish the
protons at the level of 10-6 -10-7 between the
bunches. They should work continuously at 300
kHz ( Bmax 160 G) and possibly at 5.1 MHz
(Bmax 10 G) at 100 duty cycle - In 2009, we started a RD to select suitable
ferrites for these dipoles - Collaboration with Japan, COMET experiment needs
similar technology.
3Beam cleaning
- The idea is to synchronize the beam bunches and
AC magnetic field - 100 ns bunches separated with 1.7 µs gap 600
kHz - The bunches are moving on the nodes - 300 kHz
- More information Eric Prebys note
http//mu2e-docdb.fnal.gov/cgi-bin/RetrieveFile?do
cid709
4Beam cleaning current version
5B-H curve
- B m H, in ferrites
- m m(H,T(C), )
- The losses in ferrite core
- area under the B-H curve
- At low frequency - hysteresis loss
- At high frequency - eddy current loss
B
6Ferrite samples
Ferrites
200x200x10 mm3 200x200x5 mm3
- Ferrite types MnZn, NiZn
- Frequency 300kHz, 5.1MHz
Material Properties MnZn NiZn Material Properties MnZn NiZn Material Properties MnZn NiZn Material Properties MnZn NiZn
Rel.Permeabillity _at_ 25 C 6500 625
Resistivity (W-m) 102 106
Thermal Conductivity (W/K/m) 4.5 4.3
7Test system
8ANSYS simulation
Magnetic flux density _at_ direction
Temperature Distribution
9B-H curves - MnZn
10Heating comparison1 plate vs 2 plates MnZn
High eddy currents
Low eddy currents
11MnZn_at_300 kHz
Current A-turns Bmax (G) Bbegin (G) Bend (G) T max (C)
MnZn , 300 kHz, 1 plate MnZn , 300 kHz, 1 plate MnZn , 300 kHz, 1 plate MnZn , 300 kHz, 1 plate MnZn , 300 kHz, 1 plate
0.5 60 58 60 22.7
1.2 168 154 166 31.6
1.8 243 230 196 42.6
3.0 356 341 201 65.0
4.1 401 297 207 78.4
MnZn, 300kHz, 2 plates MnZn, 300kHz, 2 plates MnZn, 300kHz, 2 plates MnZn, 300kHz, 2 plates MnZn, 300kHz, 2 plates
0.7 60 60 54 22.3
1.4 171 164 154 31.2
2.0 268 256 202 36.5
2.7 342 296 231 40.9
12MnZn_at_5.1MHz
Current A-turns Bmax (G) Bbegin (G) Bend (G) T max (C)
Current A-turns Bmax (G) Bbegin (G) Bend (G) T max (C)
MnZn , 5.1 MHz, 2 plates MnZn , 5.1 MHz, 2 plates MnZn , 5.1 MHz, 2 plates MnZn , 5.1 MHz, 2 plates MnZn , 5.1 MHz, 2 plates
3.2 5.8 4.2 5.3 23.3
7.2 10.4 9.2 9.8 32.9
10.6 15.0 13.5 14.0 46.9
17.9 21.1 20.6 18.7 77.4
13B-H curve - NiZn
14NiZn_at_5.1MHz
Current A-turns Bmax (G) Bbegin (G) Bend (G) T max (C)
NiZn , 300 kHz, 1 plate NiZn , 300 kHz, 1 plate NiZn , 300 kHz, 1 plate NiZn , 300 kHz, 1 plate NiZn , 300 kHz, 1 plate
3.6 64 61 64 23.5
5.7 102 101 95 28.3
7.8 146 143 111 40.1
9.3 167 160 116 73.2
NiZn , 5.1 MHz, 1 plate NiZn , 5.1 MHz, 1 plate NiZn , 5.1 MHz, 1 plate NiZn , 5.1 MHz, 1 plate NiZn , 5.1 MHz, 1 plate
3.3 6.0 4.9 5.5 21.7
9.6 9.9 8.5 9.4 28.0
12.3 11.7 10.6 11.2 37.2
13.7 12.3 11.4 10.7 42.4
25.1 18.9 18.7 12.4 65.2
15Ferrite selection
- Both materials satisfy the criteria for magnet
strength - Due to the low resistivity and large eddy current
effect, the thickness of MnZn ferrite plates
should be 5 mm. - At such high frequencies and power, for MnZn
plate we need good insulator between ferrites and
power bus - problem with insulation due to corona
discharge
16Magnet design - x-section
- Two designs were considered magnet with C and
H shape of the ferrite plates.
IEEE Applied Superconductivity, v. 20, p.
1642.
17Current Model Design
NiZn Ferrite plates
Beam direction
18Summary
- We measured MnZn and NiZn ferrite samples at 300
kHz and 5.1 MHz. Both materials will satisfy the
AC dipole requirements. - Building a magnet prototype based on the selected
NiZn ferrites simple design due to the high
resistivity of the material and no-insulation
between the ferrites and copper bus - This summer we plan to test the prototype,
including 5.1 MHz - Depending on the result an iteration may be
needed. - Contributions V. Kashikhin, S. Makarov, D.
Harding, E. Prebys and PARTI students I.
Iedemska and E. Bulushev.