space charge in high current lower energy electron bunches · space charge in high current lower...

Space Charge in High CurrentLower Energy Electron Bunches

Seminar DESY, 29.05.2007, S. Becker,T. Eichner, M. Fuchs, F. Grüner, U. Schramm,

R. Weingartner, D. Habs

● Experimental Status

● Self acceleration:

– Visual explanation

– Renormalization of the electron mass

● GPT: The applied calculation method

● Retardation

● Simulating 2 GeV, 1 nC

– Point 2 Point and Poisson-Solver still differ

– Convergence of Point 2 Point

● Simulation Benchmark: Energy conservation assuming mass renormalization

Experimental Status: Beam transport

Quadrupoles:

- Aperture 6 mm (5 mm effective)- Field gradient 503 T/m

Measurements at MaMi Mainz:

Experimental Status: Undulator

Test Undulator:

- 60 Periods- 5 mm period length- 0.9 T field on axis at a gap of 2.5 mm

1st Harmonic 3rd Harmonic

Measurements at MaMi Mainz:

laser pulse

~ 1 µm only!!~ nC charge

typical length scale = plasma wavelength

Experimental Status: Electron acceleration

Electron acceleration at MQP:

Experimental Status: Electron acceleration

● σx = σ

y = σ

z = 1 µm

● Charge: 1.25 nC -> I = 150 kA● Initial distribution: Gaussian

“Extreme” case:

The Regime:

Self Acceleration

Self Acceleration

Considering a beam expansion due to emittance > 0, neglecting space charge.Beam evolves “Pythagorean”Particles have uncorrelated energies

s= 2

2⋅L

Considering a beam propagation with no emittance, neglecting space charge.

Self Acceleration

z ' = ⋅ zDue to relativistic contraction, the longitudinal dimension becomes larger:

Considering a Coulomb explosion: Assuming x = y = z in the lab frame.

= 260

“Explosion” is transversal dominated.

Self Acceleration

Considering a beam expansion with no emittance, respecting space charge.

Electrons off axishave to gain speed,hence gain energyto “keep up” with theelectrons on axis.

Self Acceleration

Quantitative approach: Compressing electrons correspondsto a gain of potential energy,which can be expressed in terms of a mass:

E = m c2

me

The self field can be considered as a mass itself,which “dissolves” as the beam expands:

m f

E f = m f c2 =U '

Momentum before Expansion:

Momentum after Expansion:

Total energy in theelectrons' mean rest frame:

E ' = E0E f = m0c2E f = m0m f c

2

E ' = m' c2 = ' m0 c2

Potential energy leads to a mass of the field: E f = m f c2

Mean Lorentz factor due to Coulomb explosion:

' = 1m f

m0

Electron's rest frame:

E = 0m' c2 = 0m0c

2⋅1m f

m0

E = 0 ' m0c2 = m0c

2

= 0⋅ '

Laboratory frame: Electron bunch leaves the plasma with 0

Total energy:

after potential energy “released”Kinetic energy gainsby a factor of : '

after potential energy “released”

Mass Renormalization

(All energies E are normalized on one electron)

Potential energy in average is estimated to be of the order of 74 keV / electron

Self Acceleration: Lab Frame

Space charge (in the mean rest frame) acts transversally.

Where does the longitudinal momentum come from in the Laboratory Frame?

F E=d pdt

∥I

B =0I

2R

F res

GPT

● GPT adapts time steps dynamically -> energy conservation● Calculation precision is defined by relative momentum changes

stepwise● Space charge is considered point to point using Lorentz

transformation. Calculation time t ~ N²

For equation of motion:

Interaction in rest frame:

From Lorentz transformations:

Discussion of limitations due toRetardation Effects

"Event horizon"

Going to the electron bunch's mean rest frame:

~ 1

r2

The contribution of close neighbors is strongest

Assuming the bunch is “born” instantaneously- fields did not “spread” yet:

Discussion of limitations due toRetardation Effects

Possibilities of respecting retardation in calculations:

Interaction is described in terms of

Possible solution:Interaction only on case of " r' < Event Horizon "

Simulating 2 GeV, 1 nC

time=2e-009

Avg(G) = 4014.32 Std(G) = 12.0781

0.599582 0.599584 0.599586 0.599588GPT z

3960

3970

3980

3990

4000

4010

4020

4030

4040

4050

4060

4070

4080

G

Result of Poisson-Solver:after propagation of 60cm:

=4014/=0.003

E=2 GeV=3880 /=0.001

Initial distribution:

time=2e-009

Avg(G) = 4026.51 Std(G) = 17.1934

0.599582 0.599584 0.599586 0.599588GPT z

3980

3990

4000

4010

4020

4030

4040

4050

4060

4070

4080

G

=4027 /=0.0044

Result of Point2Pointafter propagation of 60cm:

Simulating 2 GeV, 1 nC

Simulation of designedbeam line for E = 2 GeV

Triplet focusing-Waist at approx. 5m

0 . 0 0 . 5 1 . 0 1 . 5 2 . 0 2 . 5 3 . 0 3 . 5 4 . 0 4 . 5 5 . 0 5 . 5

G P T a v g z

0

2 0

4 0

6 0

8 0

stdG

z

3.000 macro particles1.8%

0 . 0 0 . 5 1 . 0 1 . 5 2 . 0 2 . 5 3 . 0 3 . 5 4 . 0 4 . 5 5 . 0 5 . 5

G P T a v g z

0

2 0

4 0

6 0

8 0

std

G

z

10.000 macro particles 1.1%

Due to adaptive time steps,scaling is worse than

t ~ n2

Did not reach convergence

-> End of Point 2 Point ?!

0 . 0 0 . 5 1 . 0 1 . 5 2 . 0 2 . 5 3 . 0 3 . 5 4 . 0 4 . 5 5 . 0 5 . 5

G P T a v g z

0

2 0

4 0

6 0

8 0

std

G

Simulating 2 GeV, 1 nC

GPT-Poisson solver:

Energy spread decreases!

TODO: Try other solvers:

ASTRA

Simulation Benchmark: Energy conservationassuming mass renormalization

"Benchmark test" for the "extreme case": 130 MeV 1.2 nC: σx = σ

y = σ

z = 1 µm

0 1 0 2 0 3 0 4 0 5 00 , 0 0

0 , 0 1

0 , 0 2

0 , 0 3

0 , 0 4

0 , 0 5

0 , 0 6

0 , 0 7

0 , 0 8

Ek i n

Up o t

Up o t

+ Ek i n

E/

Me

V

t / a . u .

R e s t - F r a m e P 2 P

Mean rest frame:

0 , 0 0 , 2 0 , 4 0 , 6 0 , 8 1 , 00

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

1 6 0

E/

Me

V

z / m

Ek i n

Up o t

Up o t

+ Ek i n

P 2 P

Laboratory frame:

Point 2 Point:

simulation Benchmark: Energy conservationassuming mass renormalization

"Benchmark test" for 2 GeV 1.0 nC: σx = σ

y = σ

z = 1 µm

0 , 0 0 , 5 1 , 0 1 , 5 2 , 02 0 6 0

2 0 6 5

2 0 7 0

2 0 7 5

2 0 8 0

E/

Me

V

z / m

P 2 P

Ek i n

+ Ep o t

Ek i n

Point 2 Point

0 1 2 3 4 5

2 0 6 2

2 0 6 4

2 0 6 6

2 0 6 8

2 0 7 0

2 0 7 2

2 0 7 4

2 0 7 6

2 0 7 8

2 0 8 0

2 0 8 2

Ek i n

+ Ep o t

Ek i n

E/

Me

V

z / m

P o i s s o n - S o l v e r

Poisson-Solver

U ' = 180

∑i≠ j

qi q j∣r ' i−r ' j∣

rx '=r xr y '=r yr z '=z r x

pz=zm0 cz '≈ pz /m0 c

simulation Benchmark: Energy conservationassuming mass renormalization

Calculation of potential energy:Lorentz Transformation into mean beam rest frame:

E pot = zU '

Summary

Thank you !!!

● Experimental Status

● Self acceleration:

– Visual explanation

– Renormalization of the electron mass

● GPT: The applied calculation method

● Retardation

● Simulating 2 GeV, 1 nC

– Point 2 Point and Poisson-Solver still differ

– Convergence of Point 2 Point

● Simulation Benchmark: Energy conservation assuming mass renormalization