Multipacting with space charge effect in PIP-II 650 MHz cavities and their power couplers

Gennady Romanov

January 26, 2016

Content

• On parameters of MP simulations without space charge effects.• PIP-II 650 MHz cavities models and workflow.• Simulations with space charge effects ON, parameters.• MP in the cell of low β cavity and comparison with simulations

without space charge effects.• MP in the cell of high β cavity.• MP in power couplers.• Suppressing multipacting in the cells.• Conclusion

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities2

Exponential growth rate

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities3

Three reasons to be very careful about exponential growth rate:1. MP is space charge limited and self-destructive process, and in fact there is no exponential growth of number of

particles.2. Speculations on MP intensity based on α value are misleading, because there is no correlation between α and total

MP current.3. In some cases growth of number of particles is not exponential even without space charge effect.

From CST webinar on MP

CST automatic detection of MP

te

CST criterion

Space charge effect ON

Effective SEY

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities4

In some cases particle number growth is not exponential even without space charge effect.

From CST webinar on MP

This s a real device designed, built and tested in aerospace industry.

; SEYcollision

emission

II

Effective <SEY> :• Detects multipacting regardless character of particle growth.• Being compared to SEY of material it indicates how “good” dynamic conditions in a cavity are for MP

. (Paolo) )( if , SEY

0t

T

eNtNe

<SEY> >1 - MP<1 – no MP

Models and workflow

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities5

00.5

11.5

22.5

3

0 200 400 600 800 1000

SEY

Incident energy, eV

True SEY vs incident energy

Nb wet treatedNb bakedNb discharge cleaned

PIC simulations with space charge ON requires more iterations per time step. But due to the limited number of particles they are surprisingly faster than the ones without space charge effect.

For each model the simulations are performed with every SEY. The energy of collisions Wcoll, power of collisions Pcoll, power of secondary emission Pemitand secondary emission current Iemit averaged over last RF period and total number of particles are calculated as the functions of effective voltage of a cavity (energy gain).

From A.Lunin&I.Gonin

β=0.61

β=0.92

EM solver

PIC solver

Furman model

Space charge is ONNew criterion:<I_emission> = const ≠ 0

Min size 0.2 mm

Min size 0.35 mm

Select to start movie

MP in low β cell, nominal energy gain 11.9 MeV

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities6

0

100

200

300

400

500

600

0 5 10 15 20

W, e

V

Energy gain, MeV

Low β, W_collision, wet treatment

Charge OFFCharge ON

0

0.5

1

1.5

2

2.5

0

2

4

6

8

10

0 5 10 15 20

<SE

Y>

I, A

Energy gain, MeV

Low β, I_emission and <SEY>, wet treatment

I emission<SEY>

0

2

4

6

8

10

0 5 10 15 20

I, A

Energy gain, MeV

Low β, I_emission

Wet treatmentBakedDischarge cleaned

0100200300400500600

0 5 10 15 20

W, e

V

Energy gain, MeV

Low β, W_collision

Wet treatmentBakedDischarge cleaned

Experimental (SRF2011, MOPO070) MP barriers in single cells: Eacc = 7-8 MV/m (4.9-5.6 MeV, the barriers are weak)=>

MP band

I_emission is notproportional to SEY (!)

MP in high β cell, nominal energy gain 19.9 MeV

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities7

00.511.522.53

02468

1012

0 5 10 15 20 25

<SE

Y>

I, A

Energy gain, MeV

I_emission and <SEY>, wet treatment

I emission<SEY>

0

100

200

300

400

500

0 5 10 15 20 25

W, e

V

Energy gain, MeV

W_collision, wet treatment

Charge ONCharge OFF

02468

1012

0 5 10 15 20 25

I, A

Energy gain, MeV

I_emission

Wet treatmentBakedDischarge cleaned

0

100

200

300

400

500

0 5 10 15 20 25

W, e

V

Energy gain, MeV

W_collision

Wet treatmentBakedDischarge cleaned

Non-resonant MP

MP in the power couplers

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities8

β = 0.92

β = 0.61

The low beta power coupler is practically free of MP – it’s very week and observed in the coaxial part . In the high beta power coupler more powerful MP develops also exclusively in the coaxial line. In both cases a potential on the central electrode is supposed to be applied to suppress this multipacting.

0.001

0.01

0.1

1

10

100

0 5 10 15 20 25

I_em

issi

on, A

Energy gain, MeV

I_emission, log scale

Low betaHigh beta

0

0.5

1

1.5

2

3 5 7 9 11 13 15

I, A

Energy gain, MeV

Impact of convexity

Normal

Convexity

Suppressing MP with convexity

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities9

S. Kazakov

Hφ

≈10 times

≈3 times

E

W_coll = 65 eV W_coll = 103 eV

Conclusion

• PIC simulations of MP with space charge ON require equal or even less CPU time compare to ones with space charge OFF.

• Beam dynamic of MP with space charge ON is very different.• From practical point of view it is important that low boundary

of MP is the same with and without space charge, while total MP band is more narrow with space charge ON.

• With proper surface preparation MP should not be a problem for both high beta and low beta cavities.

• MP found in the high beta power coupler, but planned electric potential on the coupler central electrode should suppress it.

• A convexity works with space charge ON, though not as effectively as predicted by simulation without charge effects.

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities10

Backup 1

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities11

SEY_max=2.8 (wet treatment)

SEY_max=1.25 (discharge cleaned)

For higher SEY multipactor occupies bigger area, also it is more distributed over phases ( in time). That’s why more particles (and higher I_emis) at the same current density.

Backup 2

26/I-2016G.Romanov | Multipacting in PIP-II 650 MHz cavities12