High gradient acceleration
Kyrre N. Sjøbæk*
FYS 4550 / FYS 9550 – Experimental high energy physicsUniversity of Oslo, 26/9/2013
* k.n.sjobak(at)fys.uio.noCERN & University of Oslo
Outline
Uses of high gradient acceleration Particle physics Other uses
Techniques for high gradient acceleration Superconducting Plasma-wakefield Normal-conducting RF
The next big particle physics machines
If the LHC discovers something new(apart from the Higgs) heavy & low cross section => high energy
Want a precision experiment Electron accelerator
Circular machines limited bysynchrotron radiation Use a linear accelerator
P∝ E4
m4R2
X-ra
ys
CLIC and ILC
Two projects to build the next big particle physics machine TeV-scale electron-
positron colliders ILC superconducting,
CLIC normal conducting Share detectors etc.
CLIC ILC
Energy [GeV] 500 – 3000 200 – 500
Luminocity [1034 cm-2 s-1]
Peak 1% :1.4 – 2.0
full spectrum:2.3 – 5.9
2
Accelerating gradient [MV/m]
80 – 100 31.5
Bunches/train 234 – 312 1000 – 5400
Particles/bunch [1010] 6.8 – 3.7 10 – 20
Bunch spacing [ns] 0.5 180 – 500
Wall-plug power [MW] 272 – 589 230
Site length [km] 13.2 – 48.3 31
Data from ILC Reference Design Reportand CLIC Conceptual Design Report
√s
High energy and high gradient
In a linear accelerator:Energy = L * q * E
z * f
f is the “fill factor”, fraction of L which are accelerating structures
Need to reach very high energies,O(energy) = 1 TeV Assuming f = 1, 1 TeV:
Ez
= 20 MV/m (SLC) => L = 50 kmE
z= 30 MV/m (ILC) => L = 33 km
Ez
= 100 MV/m (CLIC)=> L = 10 km
Gradient is critical!
CLIC – how compact is it?
CLIC gradient = 100 MV/mILC superconducting ~30 MV/m
“Laser straight” tunnel
Other usesfor high gradient
Free electron lasers:
Very bright UV/X-ray lasers with defined time structure Used for research into
materials Captures “snapshots” of
atomic structure with ~ 1nm and ~10fs resolution (LCLS)
Multiple proposals to build with “CLIC” technology:More compact & cheaper
Hadron therapy Smaller & cheaper
Superconducting high gradient
Superconducting cavities have extremely large Q-factors ~1010
Can store field for a long time Ideal for circular accelerators
Superconductivity breaks at high magnetic field Peak surface field proportional to
gradient Design cavities to minimize constant
of proportionality Even with optimal cavities, gradient
is limited to ~40 MV/m Need to keep at cryo temperatures,
liquid Helium necessary
Hc (T)
T
Tc
H0
Meissner state
(superconducting)
H
Normal state
Plasma-wakefield acceleration
Capable of extremely high gradients, ~100'000 MV/m
Drive beam or laser pushes away electrons in plasma Steep charge density
gradient => huge fields Still relatively unproven
technique
Normal-conductinghigh gradient structures
Able to reach ~100 MV/m Gradient limited by vacuum arc “breakdowns”
Breakdown probability determined bygradient, pulse length, material,and structure shape
Breakdown phenomenanot completely understood
Baseline for CLIC
Normal conducting breakdowns Breakdown = vacuum
arc Spontaneous formation
of plasma on the surface Breakdowns are bad:
Surface craters Deflects the beam Reflects RF power
Scanning E
lectron Microscope
image by M
arkus Aicheler
Measurem
ent byA
ndrea Palaia
BeamKicked
beam
Normal
“Standard model of vacuum arcs”Phases:
1.Field emission from tip
2.Ionization of neutrals
3.Creation of plasma sheath
=> Enhanced emission
4.Sputtering of neutrals
5.Growth
6.Saturation of energy supply
7.Extinction
Extremely high current densities on the order of106 A/cm^2 ≈ 1025 ions/cm^2/s
From pA to kA and Ångstrøm -> 100 µm in a few ns Creating plasma densities on the order of 1020 ions / cm2
Particle- and density plots
Potential & field
Scaling laws Tested large number of structures with different designs Approximate scaling law: Constant differs between structures...
E z30 t5
BDR=const
Electric field
Scaling law – predicting the constant
local complex power flow
Sc=ℜ( S⃗)+ 16
ℑ(S⃗)constC
P
global power flow
Different designs have different field patterns Field magnitude proportional to gradient Surface fields proportional to gradient Scale Esurf, sqrt(S
c), and sqrt(P/C) to same gradient and pulse
length
Find that sqrt(Sc) and sqrt(P/C)
clustered above some limitfor all structures
Structure design
Vary the shape of the structure
In each geometry, calculate field pattern
Minimize ratios of surface fields/gradient Avoid concentrating fields
Use scaling laws to predict which design yields the lowest BDR for any gradient
Summary
High gradient acceleration technology needed for next big particle physics machines Useful for other projects as well
Normal conducting technology can reach~100 MV/m accelerating gradient
Gradient limited by vacuum arcs Avoiding these is an ongoing research topic
Backup
TLEP
Circumference = 80 kmEnergy = 350 GeV(LEP: 209 GeV)
Normal conducting breakdowns
Simulation #1:Particle- and particle density plots
Phases visible: Emission Ignition Spreading
Also see powerful oscillations which some ions “surf” Electrostatic oscillations May be a numerical instability...
Simulation #1: Potential and field plots