ISIS Upgrade Modelling

plantcitybusinessΠολεοδομικά Έργα

26 Νοε 2013 (πριν από 3 χρόνια και 7 μήνες)

86 εμφανίσεις

ISIS Upgrade Modelling

Dean Adams


On behalf of

STFC/ISIS

C Warsop, B Jones, B Pine, R Williamson, H Smith, M Hughes, A
McFarland, A Seville, I Gardner, R Mathieson, S Payne, A Pertica,
S Fisher, S Jago, J Thomason

and

Imperial College London

J
Pasternack


PASI , Friday 5
th

April 2012, RAL

A 0.5 MW ISIS


Replace old 70 MeV
Linac

with 180 MeV
version and upgrade injection beam lines
and ring injection region.



Synchrotron Space
charge limit scales as
β
2
γ
3

hence 80
to 180 MeV ≈ factor of
2.60
so output scales from 0.2 to 0.5 MW.



Presentation focuses on Ring
Studies/Modelling: Transverse and
longitudinal dynamics, injection, foils,
magnets, RF and beam loss control.

Injection Scheme

M1

M2

M3

M4

inner radius

stripping foil

h. & v.

sweeper

magnets

H
-

p
+

4


pulsed
ferrite, magnets

(
0.17 T,
45


55
mrad
, 26,000
A in ~500

s)

beam dump


H
-

charge exchange injection
over 500 turns on either falling
rising or symmetric point of
main magnet field.


Horizontal painting using
dynamic injection bump
(50
-
200
π

mm
mrad
)


Vertical painting via sweeper
magnets (50
-
200
π

mm
mrad
).


Longitudinal paint
±
0
-
1.3 MeV
using
Linac

injection energy
and Ring RF bucket frequency
errors. Chopped at
±

110
°

wrt

Ring RF phase.


1D studies




In house 1D code with longitudinal space charge.


Paint chopped beam (
±
110
°
) using injection
energy and ring RF bucket energy offset.


Use a dual harmonic volts system
𝑉
=
𝑉

=
2
sin
𝜑

𝑉

=
4
sin
2𝜑
+
𝜃



High bunching factor , transverse stability by
Keil
-
Schnell
-
Boussard

Criterion (KSB)
for
bunched
beams < 1






3D Studies

Centred around use of ORBIT code (
Fermilab
, SNS).


Version used here modified to include RF
Offsets and Acceleration.


Models: Injection/Acceleration with




Ramping Tunes and Harmonic Envelope Errors.




Machine apertures and collimators (Beam Loss).



3D space charge’ routine
.


Foil scattering.


Run in parallel environment using ~ 2M macro particles.


Produces: 6D phase space,
emittance

evolution, beam losses,




foil hits, beam moments
etc


ORBIT Injection Studies

3D Injection painting simulated.


Produce beam with maximum
emittance

300
π

mm
mrad

(un
-
normalised)


Centroid painting roughly
constant at 100
π

mm
mrad
.

6D phase space at end of injection

H and V 99%
emittance

evolution

Dynamic injection bump

Foil:3.3
σ

RMS width

Injected Beam

Re
-
circulating beam

Foils

p

H
0

H
-

ORBIT model simulates foil hits

In
-
house codes simulates
striping
efficiencies
and foil temperatures.

~ 3.5 re
-
circulations/injected proton, 1322
K on hottest point.

Temperature Per Pixel

ANSYS modelling agrees well.

Double
foils
studies in progress

Pixel temperatures reach steady
state after 10 pulses, 0.2s

(200 µg/cm
2
carbon
(as per JPARC)

>99.6% stripping
efficiency

Injection Magnets modelled using Opera

Injection dipole, peak field

0.165 T @ 26000 A

Blue zone 0.125% uniformity


Injection Straight Magnets

Particle tracking
through complex
fringe fields

Beam Losses and Activation

MARS modelling (below) indicates ~ 5x
increase in activation
between 70 and180 MeV

mSv/h

Kinetic energy, MeV

Cu

Fe

steel

graphite

concrete

Loss,
Horizontal
,
Vertical
, Total

ORBIT simulation (right) predicts < 1 %
beam loss mainly located on collimators.

ORBIT used to model incoherent tune
spread over injection and acceleration

F=1 KV, 2 WB




Tune Space

0
0.1
0.2
0.3
0.4
0.5
0.6
0
2
4
6
8
10
Horizontal Incoherent Tune
Shift

0
0.1
0.2
0.3
0.4
0.5
0.6
0
2
4
6
8
10
Vertical Incoherent Tune

Shift

ORBIT max
KV
Waterbag
ORBIT mode
max

mode

Working point studies

Other
working points
under investigation
to avoid
instabilities,
half integer, head
-
tail

SET code developed
in
-
house,
2D
particle tracker
with images
.

Raising Vertical tune leads to
loss of dynamic aperture
(right) and coupling
resonances

Lowering Vertical Tune
below half integer leads to
sextupole

resonance driven
by images.

3D version of SET (SET3D) in development to
complement ORBIT studies

Nominal design tune



Simplified 2D beam dynamics



Drive beam onto coherent resonance



Loss observations as expect










What causes growth?



Simulations and theory suggest
parametric halo



Measuring halo development
in new
experiments



Giving

a

deeper

understanding

of

main

loss

mechanism




Confirmation

of

codes

and

methods

used

in

new

designs

0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0.0
0.1
0.2
0.3
0.4
0.5
Intensity
x
1.0
E13
ppp
Loss
1.0
E13
Study of Loss Mechanisms

Predicted Resonance

Halo Experiment Transverse Profiles

Experiment

Simulation

Drive phase 1

Drive phase 2

Loss
vs

Intensity

(Y,Y

)

(Y,Y

)

(Y,Y

)

New Storage Ring Mode Experiments



High

intensity

“space

charge

limit”
:

half

integer

resonance



Diagnostics

-0.15
-0.05
0.05
0.15
0.25
0.35
9.952
9.953
9.954
9.955
9.956
Magnitude (arbitrary
units)

Acceleration Cycle Time (ms)

R5VMS Sum Position Monitor
R5 Electron Cloud Monitor
Stripline

(monitor/kicker)

Multi Channel Profile Monitor

Electron Clouds

ANSYS


HFSS

Software

CST

Summary


Installing a new 180 MeV
linac

could increase ISIS power to ~ 0.5 MW



Looks technically challenging but studies have shown no ‘show stoppers’.



A variety of modelling software for beams and hardware used: ORBIT,
in
-
house foil code, ANSYS, Opera, CST, SET (in
-
house) and HFSS



3D beam code SET3D in development to benchmark against ORBIT.



Feasibility study almost complete. Report finalised in ~ 3 months.