Light Source Accelerator Physics, Lecture 1
Introduction to Accelerator Physics of Storage Rings
James Safranek Stanford Linear Accelerator Center
Outline
• Transverse optics • Longitudinal optics • Synchrotron Radiation
N.B. Lecture borrows heavily from David Robin’s USPAS’03 & USPAS’07 lectures (www…)
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 1 Concepts
Want to touch on a number of concepts including:
• Linear optics • Equation of motion • Transfer matrix • Twiss parameters and phase advance • Betatron tune • Dispersion • Momentum compaction • Chromaticity • Energy spread • Emittance
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 2 Particle Storage Rings
In a particle storage rings, charged particles circulate around the ring in bunches for a large number of turns.
Optics elements Particle bunches dipoles s R e F l o c p a u v r i d ty a u q Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 3 Coordinate System
Change dependent variable from time to longitudinal position, s
Coordinate system used to describe the motion is usually locally Cartesian or cylindrical y x s
Typically the coordinate system chosen is the one that allows the easiest field representation
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 4 Integrate
Integrate through the elements
Use the following coordinates* dx dy∆ p∆ L xx, '==== , yy , ' , δτ , ds ds p0 L
y x s
*Note sometimes one uses canonical momentum rather than x’ and y’
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 5 Equations of Motion in a Storage Ring
The motion of each charged particle is determined by the electric and magnetic forces that it encounters as it orbits the ring:
• Lorentz Force FmaeEvB==(), +× m is the relativistic mass of the particle, e is the charge of the particle, v is the velocity of the particle, ρ a is the acceleration of the particle, E is the electric field and, B is the magnetic field. ρ • Lorentz force in practical units (B-field only) dx' = ds/ρ → x''=1/ρ ρ 1/ [m] = 0.3 B(T)/E[GeV] or 1/ [m] = B(T)/(B ) where Bρ[Tm] = E[GeV]
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 6 Typical Magnet Types
There are several magnet types that are used in storage rings: Dipoles Æ used for guiding
Bx = 0 By = Bo Quadrupoles Æ used for focusing
Bx = Ky By = Kx Sextupoles Æ used for chromatic correction
Bx = 2Sxy 2 2 By = S(x –y)
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 7 Magnets that make up a storage ring
Dipole (bends e- beam) Quadrupole (focuses e- beam)
Assembled magnet lattice Sextupole (corrects aberrations) Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 8 Equations of motion – Linear fields
The equations of motion are:
Inhomogeneous equations with s-dependent coefficients Note that the term 1/ρ2 corresponds to the dipole weak focusing The term ∆p/(pρ) is present for off-momentum particles.
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 9 Hill’s equation
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 10 Two approaches
There are two approaches to introduce the motion of particles in a storage ring
1. The traditional way in which one begins with Hill’s equation, defines beta functions and dispersion, and how they are generated and propagate, …
2. The way that our computer models actually do it – transfer matrices/linear algebra, plus transfer maps.
It is worthwhile to become proficient in both approaches.
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 11 Transfer matrix of a drift space
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 12 Transfer matrix in a thin-lens quadrupole
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 13 Motion in finite-length quadrupole = harmonic oscillator
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 14 Transfer matrix of a finite-length quadrupole
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 15 Sector Dipole
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 16 Rectangular dipole
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 17 Transfer Matrix Formalism
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 18 4x4 matrices
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 19 Quadrupoles
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 20 Strong focusing
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 21 Alternate Gradient Focusing
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 22 Back to Hill’s equation …
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 23 Hills equation
The solution can be parameterized by a psuedo- harmonic oscillation of the form
xsβ ()=+εβ ϕ ()cos(() s ϕ s 0 )
' ββ xs()=−ε cos(()ϕϕ s +00 ) ϕϕ − sin(() s + ) β β ()αεss () where (s ) is the beta function, β '(s) α α(s) ≡ − (s ) is the alpha function, 2 ϕ γ (s ) is the betatron phase, and 1+α 2 (s) xy, (s) ≡ ε is an action variable β (s) s ds ϕ = ∫ & 2nd order differential equation for β(s). 0 β
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 24 Example of Twiss parameters and trajectories
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 25 Transfer matrices & β-functions
Transfer matrices and β-functions came from same equation They are related:
One-turn transfer map (i=f) π
β
Tune is defined as number β-oscillations/turn, ν = φone-turn/2π πν α 1 Tr(R) R R − R ν = cos−1 , = 12 , = 11 22 2 2 sin(2 ) 2sin(2πν )
This is how a computer calculates optics functions – using transfer matrices
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 26 Dispersion and momentum compaction
Assume that the energy is fixed Æ no cavity or damping • Find the closed orbit for a particle with slightly different energy than the nominal particle. The dispersion is the difference in closed orbit between them normalized by the relative momentum difference ∆p/p = 0 ∆p ∆p xD==, yD xyp p ∆p/p > 0 ∆p ∆p xD''==, yD '' xyp p
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 27 Dispersion
Dispersion, D, is the change in closed orbit as a function of energy ∆E ∆E/E = 0 xD= x E ∆E/E > 0
xCSDx x ' xCSDx''''= x δ fi 00 1 δ
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 28 Momentum Compaction
Momentum compaction, α, is the change in the closed orbit length as a function of energy. ∆∆LE ∆E/E = 0 = α LE ∆E/E > 0
L0 D α = ∫ x ds 0 ρ Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 29 Beam Ellipse
In an linear uncoupled machine the turn-by-turn positions and angles of the particle motion will lie on an ellipse
Area of the ellipse, ε :
εγ=+xxxx α22 β2 '' +
xsβ ()=+εβ ϕ ()cos(() s ϕ s 0 )
' ββ xs()=−ε cos(()ϕϕ s +00 ) ϕϕ − sin(() s + ) β ()αεss ()
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 30 Emittance Definition/Statistical
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 31 Transport of the beam ellipse
Beam ellipse matrix ε x β −α = ∑ beam x −α γ
Transformation of the beam ellipse matrix
xxT =RRxi, − f ∑∑beam,, f beam i xi, − f
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 32 Transport of the beam ellipse
Transport of the twiss parameters in terms of the transfer matrix elements β CCSS22−2 β α =−CC''1 + C S − SS ' α γ CCSS''22−2 '' γ fi
Transfer matrix can be expressed in terms of the twiss parameters and phase advances β f ()cosfi+ i sin fi f i sin fi βi ϕα ϕ ββ ϕ R fi = 1+−ifαα α α i f −+sin cosi cos − sin fi fi() fi f fi fiββ ββ ϕϕϕαϕ fi f
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 33 β β Chromatic Aberration
Focal length of the lens is dependent upon energy
Larger energy particles have longer focal lengths
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 34 Chromatic Aberration Correction
By including dispersion and sextupoles it is possible to compensate (to first order) for chromatic aberrations
The sextupole gives a position dependent Quadrupole
Bx = 2Sxy 2 2 By = S(x –y)
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 35 Chromatic Aberration Correction
Chromaticity, ν’, is the change in the tune with energy dν ν ' = dδ Sextupoles can change the chromaticity
1 ∆=∆νβxxx' 2π ( SD)
1 ∆=−∆νβyyx' 2 ()SD π where
2 ∂ By ∆=S 2 *glength ()2Bρ ∂x
Light Source Accelerator Physics, lecture 1 James Safranek, Indore, January 14-18, 2008 36