What’s new Alessandro D’Elia 1. 12345 Top of the cryostat 631013162000 The line had been divided in 5 pieces. The dimensions are in mm We will propagate.

What’s new

Alessandro D’Elia

1

12345

Top of the cryostat

631013162000

The line had been divided in 5 pieces. The dimensions are in mmWe will propagate the reflection coefficient along the line

0x

12345L12345

gV

gZ

L1L2L3L4L5

2

cavit

y

Ohmic dissipation on the Power Coupler Line

We add more: N connectors, cables…. (see next slide)

OLD

Probably not yet the final configuration but very close to the final one

3

Ohmic dissipation on the Power Coupler Line

0x

123410L123410

gV

gZ

L1L2L3L4L10

cavit

y

Top of the cryostat

5

6

7

8

9

10

50

N connectors

900

25

25

50

1000

Thermal shielding

coupler

Thermal shield

Th

erm

al s

hie

ld

NEW

Transmission Line Input parameters

W)350(W230]W7for [

11

12

cavcav

f PP

P

985.01

1

L

Let us assume Q0=6.6x108, =130 (previous, 200) and a resonant frequency of 101.28MHz

)30 previous,(20111 0

00

HzHzQ

ff

QQf

f

loadload

4

As usual

xL ex 12

1 )(

)(1)(

)(1)(

111

111

1

1

xeIxI

xeVxVx

x

With

111 j

111 dc

)/ln(

)/1()/1(

21inout

outinsc RR

RRR

tan1 c

fd

r0

1

2

fL PZV 11 2

11 2

L

f

Z

PI

Propagation of the Reflection Coefficient

Voltage and Current in the line

Propagation coefficient

Attenuation Factor

Attenuation Factor of the conductors

Attenuation Factor of the dielectrics

Propagation constant

5

This is evaluated for coax having inner and outer conductors of the same material

The perturbation method

6

This is a standard and useful technique (e.g. R. E. Collin, ‘Foundations for Microwave Engineering’, p. 77) which avoids using L, C, R and G parameters and instead uses the fields of the lossless line with the assumption that they are not much different from lossless line fields

xePxP 20)(

Power loss per unit length

)(22)( 20 xPeP

x

PxP x

l

)(2

)(

xP

xPl

New alfa

7

The fields of a TEM mode propagating trough the coax in cylindrical coordinates are the following:

aB

iE ρ

ˆ2

ˆ)/(ln

0

0

0

xj

xj

eZ

V

eab

V

0

2

0*

2)(

2

1)(

Z

VdSxP

S

HE With the surface S defined as (ab)

and (0 2)

b

R

a

R

Z

VldSH

RdSH

RxP sosi

S

toso

S

tisi

l

oi

20

2

022

4/

22)(

b

R

a

R

ZxP

xP sosil

04

1

)(2

)(

Dissipated power in the “old” line

8

Old: surface resistance calculated by considering only stainless steel conductivity

New: surface resistance calculated by considering different conductivity for inner and outer conductors

Results new line

9

Some consideration

10

Dissipation has an exponential behavior depending on (in figure is small and we are still in the linear part). Therefore in principle it doesn’t matter where N connectors are. In practice is always better to put them in current node positions in case of losses due to bad contacts

Comments• Total power dissipation is compatible with the

one in TRIUMF• Final values for each part of the line may

change after thermal analysis• We know values of the dissipation factor of

the cables and connectors only at room temperature

• Possible cryogenic tests of cables and N connectors (in Liquid Nitrogen) to measure the behaviour of the attenuation at low temperatures... If really needed

11

Radiation Pressure

• We don’t care about microphonic excitation as we work in CW

• We don’t care about Lorentz detuning as the symmetry of our structure and the thickness of the copper layer exhibit a quite safety margin

• We care about additional forces on tuner plate (more expensive moving engine to buy!!)

12

Radiation Pressure

13

The general formula (Franceschetti, “Campi elettromagnetici”) for evaluating the radiation pressure is

)()( 00 nnn ihhieeiP W

Where it is e the electric field, h the magnetic field and in the unit vector orthogonal to the surface. The quantity W is the local energy density

2

0

2

02

1EH W

In case of oscillating fields the value to take into account is the effective value which means that we have to multiply for a factor 1/2

2

0

2

04

1EH W

In case of perfect conducting media we have on a planar surface only the component of e normal to the surface and tangential of h. The small surfaces defined in a meshing are always planar so that the equation (1) gets

(1)

2

0

2

02

4

1)/( nstss EHmNP

Radiation pressure on the tuning plate

14

P. Sekalski, S. Simrock, L. Lilje, C. Albrecht, “Lorentz force detuning compensation system for accelerating field gradients up to 35 MV/m for superconducting XFEL and TESLA nine-cell cavities” CARE-conf-04-001-SRF

2

0

2

0

22

4)/( nstss EH

gmNP

E-Field

H-Field

15

Results

Min=-140 N/m2

Total Force=-1.77N

Total Strength=180g

It seems we are safe

CLIC• Walter Wuensch would like to have from me as soon

as possible (less then 1 month) a document containing what we intend to realize by pointing out the first practical steps.

• He also agrees on the possibility of having someone (namely Vasim...) here at CERN for a while in order to make easier the exchange of information, but this is a future step at the moment.

• I’ve spoken with Vasim on Tuesday asking for some info and especially for any kind of material in order to have a more detailed overview of the structure

• I discussed also with Riccardo Zennaro and he would like to have the matrix of circuit model and also all the possible HFSS or whatever files of the structure

16