Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Mauricio Lopes – FNAL
5
Stored Energy, Magnetic Forces
and Dynamic Effects
Outlook
US Particle Accelerator School – Austin, TX – Winter 2016 2
• Magnetic Forces
• Stored Energy and Inductance
• Fringe Fields
• End Chamfering
• Eddy Currents
Magnetic Forces
US Particle Accelerator School – Austin, TX – Winter 2016 3
• We use the figure to illustrate a simple example.
w a
x
B0
j = uniform current density
FF
Magnetic Force on a Conductor
US Particle Accelerator School – Austin, TX – Winter 2016 4
We use the expression, dvBjF ×= ∫Integrated over the volume of the conductor.
hLdxdvw
xBB
w
B
wh
NIj=
hBNI
==
==
0
0
0
0
0
µµL is the length into the paper.
0
20
2
02
20
002
20
00
0
22
µµµ
µhLBw
w
hLBxdx
w
hLB
hLdxw
xB
w
BB dvjF
w
===
==
∫
∫∫
0
20
2
µB
hL
F
Area
Forcepressurerepulsive ===
Magnetic Force on a Pole
US Particle Accelerator School – Austin, TX – Winter 2016 5
We use the expression, dvHF ⋅= ∫ ρ
h
Bρ
0=where the magnetic charge density ρ is given by,
The force is in the same direction as the H vector and is attractive.
dv=aLdyh
yBH
0
0
µ=
0
20
2
02
20
002
20
0
00
2=
2==
µµµµρ aLBh
h
aLBydy
h
aLBaLdy
h
yB
h
BdvHF
h
∫∫∫ ==
0
20
2
µB
aL
F
Area
Forcepressureattractive ===
Pressure
US Particle Accelerator School – Austin, TX – Winter 2016 6
The magnitudes of the repulsive pressure for the current and the attractive
pressure at the pole are identical. In general, the pressures parallel to the
field lines are attractive and the forces normal to the field lines are
repulsive. The pressure is proportional to the flux density squared.
0
20
2µB
pressure =
Example
US Particle Accelerator School – Austin, TX – Winter 2016 7
Let us perform a calculation for a flux density of 5 kG = 0.5T.
27
2
0
20 472,99
1042
5.0
2 m
NewtonsBpressure =
××== −πµ
atmosphere 143.14472,99 220.5T @ ≈==
in
lb
m
Newtonpressure f
Magnet Stored Energy
US Particle Accelerator School – Austin, TX – Winter 2016 8
• The magnet stored energy is given by; ∫= HBdvU2
1
the volume integral of the product of H and B.
Consider a window frame dipole field (illustrated earlier) with uniform field in
the space between the coil. If we ignore the field in the coil and in the iron,
volumemagneticB
U 2 0
20 ×=
µ
Magnet Inductance and Ramping Voltage
US Particle Accelerator School – Austin, TX – Winter 2016 9
Inductance is given by, 2
2
I
UL =
tI
URI
t
I
I
URI
dt
dILRIV
∆+=
∆×+=+= 22
2
In fast ramped magnets, the resistive term is small.
volumeB
tItI
UV ××
∆=
∆≈
0
20
2
22
µvolume=ahL
N
hBI
0
0
µ=
t
NaLBahL
B
tN
hBV
∆=××
∆≈ 0
0
20
0
0
1
µµ
Substituting,
Units and Design Options
US Particle Accelerator School – Austin, TX – Winter 2016 10
The units are,
VoltVoltm
m
WebersTm
t
NaLBV =−=×==
∆≈
sec
sec
secsec
2
2
20
t
NaLBV
∆≈ 0
Given the field = B0, pole width = a, Magnet Length = L and ramp time, ∆ t, the
only design option available for changing the voltage is the number of turns, N.
Other Magnet Geometries
US Particle Accelerator School – Austin, TX – Winter 2016 11
Normally, the stored energy in other magnets (ie. H dipoles, quadrupoles
and sextupoles) is not as easily computed. However, for more complex
geometries, two dimensional magnetostatic codes will compute the
stored energy per unit length of magnet.
Effective Length
US Particle Accelerator School – Austin, TX – Winter 2016 12
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
-300 -250 -200 -150 -100 -50 0 50 100 150 200 250 300
Bn
(a
u)
Z (mm)
Ideal
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
-300 -250 -200 -150 -100 -50 0 50 100 150 200 250 300
Bn
(a
u)
Z (mm)
Ideal Real
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
-300 -250 -200 -150 -100 -50 0 50 100 150 200 250 300
Bn
(a
u)
Z (mm)
Ideal
Real
Effective length
���� =� ��(�)��
����
��(� = 0)
z
Fringe Fields and Effective Lengths
• Often, canonical rules of thumb are adopted in order to estimate
the effective length of magnets.
– Dipole fringe field length = 1 half gap at each end
– Quadrupole fringe field length = 1/2 pole radius at each end.
– Sextupole fringe field length = 1/3 pole radius at each end.
US Particle Accelerator School – Austin, TX – Winter 2016 13
Three Dimensional Fringe Fields
The shape of the three dimensional fringe field
contributes to the integrated multipole error of a
magnet.
US Particle Accelerator School – Austin, TX – Winter 2016 14
• Dipole Fringe Field
– Typically, the fringe
field is longer at the
center of the magnet
and drops off near the
edges.
– This distribution is
approximately
quadratic and the
integrated multipole
field looks like a
sextupole field.
h
h = "canonical" fringe field length
edge of pole
Fringe field length is typically longer in the center for an "unchamfered" pole end.
z
B
z
x
US Particle Accelerator School – Austin, TX – Winter 2016 15
Virtual Field Boundary
US Particle Accelerator School – Austin, TX – Winter 2016 16
beam
Magnet
virtual field boundary
ρβtan1 −=
Fx
ββ
• The photograph shows a removable insert with a machined chamfer installed on the SPEAR3 prototype gradient magnet.
• The shape of the chamfer depth was determined empirically and was approximately parabolic. It was designed to reduce the integrated sextupole field.
• The chamfer shape was machined onto the end packs of subsequently manufactured production magnets.
US Particle Accelerator School – Austin, TX – Winter 2016 17
Quadrupole 3-Dimensional Fringe Field
• The quadrupole “gap” is largest at its center. The gap decreases as the distance from its center (g1>g2>g3>g4>g5>g6). Since the fringe field is roughly proportional to the magnet gap, it is longest near the magnet pole center. g1 g2
g3 g4 g5 g6x
y
US Particle Accelerator School – Austin, TX – Winter 2016 18
Quadrupole Chamfer
• The quadrupole pole chamfer is a straight angled cut, which shortens the pole at its center. The angle is cut such that the pole is longer near its edge.
• Again, this cut was determined empirically by trial an error, minimizing the n=6 integrated multipole error.
US Particle Accelerator School – Austin, TX – Winter 2016 19
-10
-8
-6
-4
-2
0
2
4
0 3 4 5 6
B6B10B14
end chamfer length (mm)
R = 25 mmB
n/B
2. 104
Harmonics as function of the end
chamfer length
US Particle Accelerator School – Austin, TX – Winter 2016 20
Ideal end chamfer length as function
of quadrupole length
US Particle Accelerator School – Austin, TX – Winter 2016 21
R=20mm
3.0
3.5
4.0
4.5
5.0
5.5
6.0
100 150 200 250 300 350 400 450 500 550 600
Quadrupole length (mm)
Idea
l en
d c
ham
fer
len
gth
(m
m)
Effective length as function of the end
chamfer length
US Particle Accelerator School – Austin, TX – Winter 2016 22
Q530
563.0563.5564.0564.5565.0565.5566.0566.5567.0
0 1 2 3 4 5 6
Chamfer length (mm)
Eff
ecti
ve l
eng
th (
mm
)
Multipoles as function of the coil
excitation
US Particle Accelerator School – Austin, TX – Winter 2016 23
Dynamic effects
US Particle Accelerator School – Austin, TX – Winter 2016 24
dt
BE
∂−=×∇
h
w
g th
w
g tk B
x
y
jB .̂. tieBo ω=
jBj ˆˆ &−=
∂∂−
∂∂
x
E
z
E zx
oz Bi
x
E ω=∂
∂
xBidxBiEz oo ωω == ∫EJ σ=
xBiEJ ozz ωσσ ==
Dynamic effects
US Particle Accelerator School – Austin, TX – Winter 2016 25
o
e
Feo
gBdl
Bdl
BI
dlHI
µµµ=+=
=
∫∫
∫
..
.
Be
stands for the magnetic
field due to the eddy currents
w1
2
3g tk
31
21
321
2 III
II
IIII
+==
++=
∫∫= JdxdyI i
( )2
0
1 2
1wtBixdxBitI
w
kook ∫ == ωσωσ
( )23 2
1k
w
tw
oo twtgBixdxBigIk
−== ∫−
ωσωσ
−+=22
2 gtwgw
g
tBiB kkoo
e
σωµ
Be (x=0) (theoretical) 0.0129 (T)
Be (x=0) (FEMM) 0.0127 (T)
Bo(T) 0.145
frequency (Hz) 100
tk (mm) 3
g (mm) 25w (mm) 28σ (MS/m) 1.334
Real-case example
US Particle Accelerator School – Austin, TX – Winter 2016 26
Real-case example
US Particle Accelerator School – Austin, TX – Winter 2016 27
thickness = 1 mm
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0.00
-30 -20 -10 0 10 20 30x (mm)
By
(T)
400 Hz
350 Hz
300 Hz
250 Hz
200 Hz
150 Hz
100 Hz
90 Hz
80 Hz
70 Hz
60 Hz
50 Hz
40 Hz
30 Hz
20 Hz
10 Hz
thickness = 2 mm
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0.00
-30 -20 -10 0 10 20 30x (mm)
By
(T)
400 Hz
350 Hz
300 Hz
250 Hz
200 Hz
150 Hz
100 Hz
90 Hz
80 Hz
70 Hz
60 Hz
50 Hz
40 Hz
30 Hz
20 Hz
10 Hz
thickness = 3 mm
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0.00
-30 -20 -10 0 10 20 30x (mm)
By
(T)
400 Hz
350 Hz
300 Hz
250 Hz
200 Hz
150 Hz
100 Hz
90 Hz
80 Hz
70 Hz
60 Hz
50 Hz
40 Hz
30 Hz
20 Hz
10 Hz
thickness = 4 mm
-0.070000
-0.060000
-0.050000
-0.040000
-0.030000
-0.020000
-0.010000
0.000000
-30 -20 -10 0 10 20 30x (mm)
By
(T)
400 Hz
350 Hz
300 Hz
250 Hz
200 Hz
150 Hz
100 Hz
90 Hz
80 Hz
70 Hz
60 Hz
50 Hz
40 Hz
30 Hz
20 Hz
10 Hz
x = 20 mm
0.00
0.20
0.40
0.60
0.80
1.00
0 100 200 300 400
f (Hz)
By(
f)/B
y(0
Hz) 4 mm
3 mm
2 mm
1 mm
US Particle Accelerator School – Austin, TX – Winter 2016 28
x = 0 mm
0.00
0.20
0.40
0.60
0.80
1.00
0 100 200 300 400
f (Hz)
By(
f)/B
y(0
Hz) 4 mm
3 mm
2 mm
1 mm
x = -20 mm
0.00
0.20
0.40
0.60
0.80
1.00
0 100 200 300 400
f (Hz)
By(
f)/B
y(0
Hz) 4 mm
3 mm
2 mm
1 mm
Real-case example
Phase delay
US Particle Accelerator School – Austin, TX – Winter 2016 29
σωµµδ
or
2=
[ ]radians δ
φ kt=
Attenuation due to lamination on the
material
US Particle Accelerator School – Austin, TX – Winter 2016 30
)2/tan(2
kdkdo
=ψψ
ωµσik −=2
µ = 2200.µo
σ = 9.93 MS/m
Example
Bo
Bmin
Bmean
d2
minBBmean =
Real-case example summary
US Particle Accelerator School – Austin, TX – Winter 2016 31
Vacuum
chamber
thickness
2 mm 3 mm
Core
lamination
thickness
0.60 0.54
0.5 mm 0.67 0.40 0.36
1.0 mm 0.46 0.27 0.25
@400 Hz
Induced sextupolar field due to eddy
currents
US Particle Accelerator School – Austin, TX – Winter 2016 32
xBEJ zz&σσ ==
h
w
g th
w
g t B
x
y
2xBtI k&σ=o
BhI
µ=
2)( xt
B
h
txB k
o ∂∂= σµ
t
B
Bh
tFk k
o ∂∂=
ρσµ 1
22
( ) dxxw
gxF
2/11
0
22
2 12∫
−
+=
F=1 when g<<w, and F=2 for a circular chamber
)()()( tfBBBtB iei −+=
The sextupole strength (or moment) is defined by
( ))cos(12
1)( ttf ω−=
Example – Alba booster dipole
US Particle Accelerator School – Austin, TX – Winter 2016 33
ρ 11.4592 (m)
µo
4π10-7 (H/m)
σ 1.344 (MS/m)
Freq. 3 (Hz)
Be
0.029 (T)
Bi
0.873 (T)
20 mm
30 mm
17 mm
50 mm
VC1
VC2
F
VC1 1.5739
VC2 1.2141
Summary
US Particle Accelerator School – Austin, TX – Winter 2016 34
This chapter showed a collection of “loose-ends” calculations:
• Magnetic Forces
• Stored Energy and Inductance
• Fringe Fields
• End Chamfering
• Eddy Currents
Those effects have an impact in the magnet design but also need to be
taken into consideration into the magnet fabrication, power supply
design, vacuum chamber design and beam optics.
Next…
US Particle Accelerator School – Austin, TX – Winter 2016 35
Magnetic measurements