Content Background I II Main research of my dissertation Papers - PDF document

2010/8/16 Seminar Presentation: Research on mechanism of HPM dielectric window breakdown and its application Dr. C. Chang Department of Engineering Physics, Department of Engineering Physics, Tsinghua University, Beijing 100084, China changc02@mails.tsinghua.edu.cn Content Background I II Main research of my dissertation Papers and patents III 1

2010/8/16 I. Background Breakdown at vacuum/dielectric interface of HPM window has been major limitation of power radiation [1]. Methods of improving breakdown threshold become key issues of HPM system [1]. y y [ ] vacuum SF 6 bag HPM Source Breakdown is triggered by multipactor and finally realized by plasma avalanche in ambient desorbed or evaporated gas layer above dielectric. [1] R. Barker, E. Schamiloglu, High power microwave source and technology, 2001. I. Background The principle courses of HPM window breakdown under vacuum: Seed electron production Seed electron production Multipactor, charge accumulation, Multipactor, charge accumulation, power deposition power deposition Saturation in Saturation in Gas desorption or Gas desorption or ambient gas ambient gas material vaporization material vaporization Saturation Saturation strengthening strengthening Plasma avalanche in Plasma avalanche in Space charge shielding Space charge shielding g g g g ambient gas ambient gas bi bi t t Interaction of plasma and Interaction of plasma and dielectric surface dielectric surface HPM transmission HPM transmission cutoff cutoff 2

2010/8/16 II. Main research of my dissertation HPM multipactor and breakdown mechanisms (A) 1. Influence of releasing gas on breakdown 2. Space charge field and potential model 3. Model of plasma discharge on a dielectric Theories and experiments of improving thresholds (B) 4. Periodic retangular surface 5. Periodic triangular surface 5 P i di t i l f 6. Resonant magnetic field Design feed horn and apply periodic surfaces (C) 1. Improved HPM multipactor model Propose model considering collision and ionization of multipactor electrons with ambient desorption or evaporation gases. The former models [1,2] assuming vacuum contrast to experiments showing desorption or evaporation gas forming local high p. [3,4] g p p g g g p Tangent and normal velocity: ⎛ ⎞ eE eE = − + − ν + ( ) exp( ) DC ⎜ DC ⎟ u t t u ν 0 ν z t z ⎝ ⎠ m m t t ( ) ( − ν exp eE t eE ) ( ) = 0 ω φ + ν φ − 0 ω ω + φ + ν ω + φ ( ) rf t sin( ) cos( ) rf sin( ) cos( ) u t t t ω ω 2 + + ν ν 2 ω ω 2 + + ν ν 2 y ( ( ) ) t ( ( ) ) t m m m m t t 2 2 1 ( ) e E ( ) ( ) 2 = − ν + − ω − ν Collision energy: exp 2 1 2cos( )exp mu t t t ω 2 + ν 2 t t 2 4 ( ) m t ⎛ ⎞ eE eE Transit time: ( ) − − ν τ + − τ = Firstly fit ν m and ν i 1 exp( ) ⎜ ⎟ 0 DC DC u 0 t z ν ⎝ ⎠ m m t 1, R. Kishek, et al. Phys. Rev. Lett. , 1998; 2,Valfells, et al. Phys. Plasmas, 2000. 3,A. Neuber, et al, J. Appl. Phys., 1999 ; 4,A. Neuber, et al. IEEE Trans. Plasma Sci. 2000. 3

2010/8/16 1. Improved HPM multipactor model fitting ionization and collision cross-section of N 2 from eV to keV ⎛ ⎛ ⎞ ⎞ ε ε ε − ε 4 ( ) a σ ε = 0 − − ⎜ 1 exp ⎜ ⎟ ⎟ b i ⎜ ⎟ ( ( ) ) ε − ε i 2 + + a ε ε ε ε 1 1 ⎝ ⎝ ⎝ ⎝ ⎠ ⎠ ⎠ ⎠ a 0 0 0 0 i i ( ) σ ε = ε 1.2 ε 2 1.2 1.5* /(1+0.008* ) m ε ε 2.5 +70* /(1+1.05* ) Assuming Maxwellian distribution, numerically integrating ν m and ν i ⎛ ⎞ 5 / 2 ∞ ε 8 1 v ⎠ ∫ ( ) = σ ε − ε ε ⎜ ⎟ exp( ) d ⎝ π p m kT kT 0 1. Improved HPM multipactor model ↑ Vaccuum [1] Kishek 98 Obtain variation of ε e and saturation boundaries with p. Extend the vacuum solutions [1] to that under different p. Calculation: the power deposited by multipactor leads to surface material melting, evaporating, to further improve local pressure 1, R. Kishek, et al. Phys. Rev. Lett. , 1998; 4

2010/8/16 1. HPM experiment research on releasing gas BWO, 1GW, X-band, 20ns, single and repetitive 50Hz Distinguished Δ P for different material Δ P : PMMA>PE>PTFE> QTZ E B : PMMA<PE<PTFE< QTZ E B : PMMA PE PTFE QTZ Close relation between E B with Δ P Deduced Δ P ~1-10Torr in 20ns Guide material selection and conditioning: Guide material selection and conditioning: Weak SEY, low gas absorption, high melting point; multiple conditioning C. Chang, et al., Phys. Plasmas15, 093508, 2008. C. Chang, et al., Phys. Plasmas17, 053301, 2010. 2. Space charge field and potential model Space charge influences multipactor, improves the deposited power. The former model [1, 2] only considered multipactor electrons. Plasma significantly affects field. Propose space charge model of multipactor electrons and plasma Solving Poisson Equations: ( ) 2 ψ 2 d n e ( ) ( ) ( ) − 0.5 = ψ + β αψ − + β − γψ exp exp (1 ) 1 esB 2 ε Co-decided potential d x T 0 es Extend former solution [1] to positive space potential showing β =1 curve [1] valfells 00 Analytical φ under vacuum ： -1 ⎛ ω ⎞ − φ 2 T e φ = + ln exp( ) p es ⎜ w ⎟ x 2 e ⎝ T v ⎠ es t ⎛ ⎞ β T m ( ) Analytical positive φ by plasma and Δ φ ≈ − + β ln ⎜ 1 ⎟ es e ⎜ ⎟ γ α e ⎝ M ⎠ multipactor electrons ： 1, A. Valfells, et al., IEEE Trans. Plasma Sci. 2000;2, A. Neuber, et al, J. Appl. Phys., 1999 5

2010/8/16 2. Space charge field and potential model No shielding No shielding ↑ multipactor electrons ↑ [1] plasma and multipactor electrons shielding Plasma leads to further shielding E x , improving P ↑ 。 Analytical curves agree with PIC simulation [1] Positive space potential was experimentally found [2] C. Chang, et al., Phys. Plasmas16, 053506, 2009. 1,H. Kim, et al. Phys. Plasmas, 2006 2, L. Schiesko, et al., J. Nucl. Mater., 2007. 3. Model of plasma discharge on a dielectric Strong interaction between HPM dielectric breakdown and plasma discharge in local gas, no corresponding analytical model. Density, momentum, energy equation: � Space charge dn = ν − − δ δ ν ( (1 ) ) field and e n Secondary i w t l e dt � electron diffusion loss � ( ) � d n u = − ν compensation e m en E n m u e rf e m dt � (3 / 2) � d n T = − < ⋅ > − − = n ν ε = + + ( ) 2 e e en u E P P P P eV T n v T n v e ie cL cL e i c ie QW i e l e e l dt ν e E 2 2 ( ) ⎛ 2 ⎞ ⎛ 2 ⎞ ( ) ( d T ) 0 = rf t − ε + ν − + + − − δ δ ν e 2 1 ⎜ T ⎟ ⎜ eV T T T ⎟ ( ) ω 2 + ν 2 ⎝ 3 c e ⎠ i ⎝ 3 QW e i W t e ⎠ l 3m dt t ⎛ ⎛ ⎞ ⎞ ( ) ( ) 1 3 T M ( ) Potential drop = ⎜ + − δ δ + γ γ + γ δ δ 2 − γ + δ δ ⎟ ln ⎜ ⎟ ln 1 ln ln e V ⎜ ⎟ ⎜ ⎟ π QW 2 w t 2 w t 2 w t e ⎝ m ⎠ ⎝ ⎠ e 6

2010/8/16 3. Model of plasma discharge on a dielectric ↑ ↑ τ by PIC simulation [1] Considered interaction of plasma and dielectric surface, τ↓ Experiment [2] found, τ diel < τ space C. Chang, et al., Phys. Plasmas16, 033505, 2009. 1.Y. Lau, et al., Appl. Phys. Lett., 2006 2.D. Hemmert, et al, SPIE, 2000 II. Main research of my dissertation HPM multipactor and breakdown mechanisms (A) 1. Influence of releasing gas on breakdown 2. Space charge field and potential model 3. Model of plasma discharge on a dielectric Theories and experiments of improving thresholds (B) 4. Periodic retangular surface 5 P i di t i 5. Periodic triangular surface l f 6. Resonant magnetic field Design multi-mode feed horn (C) 7

2010/8/16 4. Multipactor on periodic rectangular surfaces Plasma avalanche Multipactor In ambient gas Periodic dielectric surfaces to alter the trajectories of electrons to decrease ε e < ε p1 and τ << T/2 and τ << T/2 Suppress multipactor in developmental stage Isolate local high pressure and plasma avalanche 4. Multipactor on periodic rectangular surfaces π 2 + 2 2 ⎛ 2 ⎞ 2.86-9.38GHz ， E rf ~30kV/cm eE τ = = 0 ϕ − / 2 rf sin arctan( ) S ( ) T ⎜ ⎟ S m (T/2) ~0.56-6.1mm y ω 2 π ⎝ ⎠ m ↑ E rf ↑ , S y ↑ ; E dc ↑ , S y ↓ , S x ↓ ; p ↑ , S y ↓ , S x ↓ , E dc and p significant influence Width d decided by E rf , ω , p, E dc . d<< λ , not to influence HPM transmission 8

2010/8/16 4. Multipactor 2D-PIC simulation Y(mm) Y(mm) X(mm) X(mm) X(mm) f=2.86GHz, E rf =30kV/cm, d=1mm, h=1mm 4. Multipactor 2D-PIC simulation ↑ F rf τ 1 < T/2 , F rf strong restoring force. SE: τ <<T/2, ε e << ε p1 , within the duration T/2- τ 1 9

2010/8/16 4. Proof-of-principle experimental verification Four port circulator Experiment platform ： Klystron ， S-band(2.86GHz) ， μ s width ↑ Incident vacuum SF 6 Reflected Transmitted Intense reflection+transmission cutoff 4. Proof-of-principle experimental verification PTFE, H=1mm P=2mm H/mm P/mm Capacity/MW E B /kV/cm 0.5 1 >16 >28 increase E B 1 2 >16 >28 of about 2 1 3 5-6 16 Fl t Flat surface f 4 4 14 14 When P in >16MW, the alumina windows limit power further increasing. C. Chang, et al., J. Appl. Phys. 105, 123305(2009) 10