1. High-Efficiency Klystron Design for the CLIC Project
Antoine Mollard1, Claude Marchand1, Franck Peauger1, Juliette Plouin1 Armel Beunas2, Rodolphe Marchesin2
1DRF/Irfu/SACM/Lisah
CEA Saclay Gif-sur-Yvette Cedex
2Thales Electron Devices 2 Rue Marcel Dassault Vélizy-Villacoublay
DDays 2016| Mollard Antoine
Antoine Mollard – DDays 2016

2. Journées des Doctorants
Antoine Mollard Education Engineer Arts et Métiers ParisTech (AM2010) Mechanical and industrial engineering MSc Université d’Aix-Marseille Fusion Science and Technologies PhD contact PhD project proposed by Juliette Plouin (Irfu/SACM) Motivation The diversity of the topics and activities: RF engineering, material science, electromagnetism, modelling and simulation, industrial process monitoring, tests. CEA and research surroundings. Partnership with Thales. Contribution of this project to a bigger one (CLIC).
Antoine Mollard – DDays 2016

3. Journées des Doctorants
High-Efficiency Klystron Design for the CLIC Project CLIC is a electron-positron collider project with a cavities beam line. These cavities need to be conditioned and tested with high power RF signals provided by 12GHz-klystrons. We are developing a high-efficiency klystron in order to save power. Collider : particle accelerator involving directed beams of particles; collisions byproducts give evidences of subatomic structures and high-energy particles. Klystron : 100MHz – 100GHz radio-frequency amplifier used for radar, telecommunication, radiation therapy, particle accelerator or nuclear fusion power plant. Efficiency : klystron’s efficiency is the ratio between its RF output power and its electrical input power.
Antoine Mollard – DDays 2016

4. High-Efficiency Klystron Design for the CLIC Project
Klystron principle
CLIC and the kladistron project
How to make a kladistron?
Our work so far
Antoine Mollard – DDays 2016

5. Klystron principle
Antoine Mollard – DDays 2016

6. Working at 100MHz- 100GHz with a narrow bandwidth.
What is a klystron?
An amplifier
Working at 100MHz- 100GHz with a narrow bandwidth.
Used in:
Radar
Telecommunication
Radiation therapy
Plasma heating
Particle accelerator
Antoine Mollard – DDays 2016

7. Klystron principle: how does it work?
Cold water
Hot water
Linear-beam vacuum device which is used as a narrow bandwidth amplifier for high radio frequencies (100 MHz- 100GHz)
Creation of a stream of electrons by the electron gun (thermionic emission)
Confinement by an external magnetic field
Triggering of the beam modulation by the wave in the input cavity
Modulation of the beam into a set of electron bunches in the gain cavity/ies
Drifting of the electron bunches
Energy transfer between the bunches and the output antenna
Absorption of the lower energy electron beam by the collector
Collector
Ionic pump 7
RF output 6 Ps
Middle cavity 5
Solenoid 2 4
RF input 3 Pe
Control anode
Electron beam 1
Vk Ib
Klystron efficiency is the ratio of the output power to the electron gun supply power.
The microperveance characterizes the space-charge effect on the electron beam motion.
𝜂= 𝑃 𝑜𝑢𝑡 𝑉 𝑘 ∗ 𝐼 𝑏 µ𝑃= 𝐼 𝑏 ∗ 𝑉 𝑘 3 2
Antoine Mollard – DDays 2016

8. CLIC and the Kladistron project
Antoine Mollard – DDays 2016

9. CLIC accelerator RF cell Decelerated Drive Beam 100 A
The CLIC project
CLIC : Compact LInear Collider 3TeV 50km-long electron/positon collider
CLIC accelerator is designed with several cavities (about )
These cavities need conditioning on specific benches with high-power RF source (klystron). D R I V E B A M L O P S N J C T Y U e - G + F W Z H ( n o t s c a l ) d g r v i m 1 k
CLIC accelerator RF cell
RF Power (120 MW)
Decelerated Drive Beam 100 A
A high-efficiency klystron would lead us to a cheaper solution.
Accelerated Main Beam 1 A
Antoine Mollard – DDays 2016

10. Kladistron
𝑷 𝒊𝒏
𝑷 𝒐𝒖𝒕
A Kladistron (Kl-adi(adiabatic)-stron) is a high-efficiency klystron with a large number of cavities (at least twice as many as in a classical klystron).
𝑽 𝒌
N cavities ↗ 𝑹 𝑸 ↘ => Efficiency (η= 𝑷 𝒐𝒖𝒕 𝑽 𝒌 ∗ 𝑰 𝒃𝒆𝒂𝒎 ) ↗
𝑷 𝒊𝒏
𝑷′ 𝒐𝒖𝒕
𝑽 𝒌
Antoine Mollard – DDays 2016

11. AJDisk 1D-simulations of a 12GHz-Kladistron
10 cavities – 12GHz
µP = 1.5 A.V^-3/2
Efficiency 67.2 %
Length 197 mm
20 cavities – 12GHz
µP = 1.5 A.V^-3/2 Efficiency 78 %
Length 285 mm
Our preliminary AJDisk results show a higher efficiency than what we can expect from a klystron with such a high microperveance.

12. Kladistron and classical klystrons
12GHz
« KLADISTRON »
4.9GHz « KLADISTRON » (technical demonstration)
Antoine Mollard – DDays 2016

13. How to make a kladistron?
Antoine Mollard – DDays 2016

14. Kladistron and classical klystrons
12GHz
« KLADISTRON »
4.9GHz « KLADISTRON » (technical demonstration)
Antoine Mollard – DDays 2016

15. TH2166 Klystron
TH2166 front view
TH2166 klystron was designed by Thales electron devices (TED) for Mainz Microtron.
Features
Frequency
4.9 GHz
Output power
56 kW
Efficiency
50%
Ibeam
4.3 A Vk
26 kV µP
1.066
Number of cavities 6
Interaction line length
233mm
233mm
This klystron will be modified to verify the kladistron principle.
Antoine Mollard – DDays 2016

16. klystron TH2166 enhancement Cavities preliminary design
The design we are looking for would let us :
Use the TH2166 klystron test and conditioning bench  Total interaction line length of 233mm, same input and output cavities, same solenoid
Use the TH2166 klystron electron gun and collector  Same microperveance of 1µA.V^(-3/2)
Check the kladistron principle  More than 6 cavities
Avoid cavities coupling  Drift space between cavities larger than 9mm
Avoid gain peaks  Low coupling cavities (low R/Q and Q0 values)
Thales-provided elements
Antoine Mollard – DDays 2016

17. Our work so far
Antoine Mollard – DDays 2016

18. Thales-provided elements
Steel rods and Ionic pump
Wave guide,
Output cavity and Output antenna
Collector
These elements have already been delivered.
Antoine Mollard – DDays 2016

19. Th2166 Klystron and Kladistron Comparison MAGIC2D Simulations
Our kladistron simulation results reach an efficiency of six points above TH2166 simulation results.
The electron bunching and the beam current growth are also smoother.
TH2166 beam current along the interaction line
TH2166 electron beam
Kladistron beam current along the interaction line
Kladistron electron beam
Antoine Mollard – DDays 2016

20. klystron TH2166 enhancement Cavities preliminary design
Type 1
Type 2
Type 3
Type 4
16 cavities interaction line
TH2166 cavity shape
Low-coupling cavity shapes
Type 1
Type 2
Type 3
Type 4
According to our COMSOL simulations, these low-coupling cavities are fit for smooth electron bunching.
Antoine Mollard – DDays 2016

21. Frequency shift : a reliable tuning system is required
Zgap
Zgap
Type 2
Type 3
Δf/Δzgap ≈ 2.5MHz/µm
Antoine Mollard – DDays 2016

22. Frequency shift : a reliable tuning system is required
→Δzgap ≈ -4µm
For each computation we shift the frequency of one of the cavities.
Kladistron efficiency is sensitive to its cavities frequency shifts, especially at the end of the interaction line.
Antoine Mollard – DDays 2016

23. Cavities and tuning system preliminary design
Cooling system channels (4)
Beam drift tube
Copper
This preliminary design is under study and it is considering the surrounding of the cavities (cooling system, solenoid,…).
The tuning system is inspired by CLIC accelerating cavities design; a thigh copper membrane is misshaped to adjust cavities frequencies. This strain is controlled by an accurate screw thread.
Thread steel pin brazed on the membrane
Cavity
Bronze bearing
Beam drift tube
Misshaped membrane
Screw thread
Tuning system (x4 at 60°)
Antoine Mollard – DDays 2016

24. Tuning system preliminary design COMSOL simulations
Displacement (mm)
Δf (for each tuning system) (MHz)
Δf (for 4 tuning systems) (MHz)
0.26
4.02
16.03
Antoine Mollard – DDays 2016

25. Prototype cavities to test
Prototype cavities to be tested to validate tuning system (and tuning range & resolution) and brazing process.
Prototype test bench
Antoine Mollard – DDays 2016

26. The multipactor phenomenon
Antoine Mollard – DDays 2016

27. Problem: the multipactor phenomenon
Electric field
Electric field
Copper
Copper
ETC…
Antoine Mollard – DDays 2016

28. Solution: titanium layer
Electric field
Titanium
Copper
Copper
Antoine Mollard – DDays 2016

29. Solution: titanium Titanium layer:
Titanium hydride deposition with a brush
Titanium sheet brazing
These two technologies will be tried out in a second testing phase.
Antoine Mollard – DDays 2016

30. Conclusion and outlook
TH2166 klystron study and codes validation
TH2166 kladistron preliminary design
TH2166 kladistron’s cavities design finished We will soon test two first prototype cavities (tuning system)
Invitation to tender
Second testing phase (titanium)
TH2166 Kladistron assembled and tested by the end of the year.
Antoine Mollard – DDays 2016

31. Thank you for your attention
Antoine Mollard – DDays 2016