Pillbox Cavity Testing RF Strategy Meeting 11/15 to 11/16/10 Al

Pillbox Cavity Testing
RF Strategy Meeting
11/15 to 11/16/10
Al Moretti,
APC, Fermilab
1
Outline of the Talk








History of the Testing Arrangements
History of The LBL Pillbox Cavity With Flat Cu and Curve Be Windows
First Pillbox “Small” Button Cavity Test Arrangement
Testing Arrangements for the Pillbox cavity
Dimensional drawings of the Smaller and Larger Buttons
Simulation of the LBL Pillbox cavity without buttons and discussion of peak
fields
Current Status of the LBL Pillbox Cavity
Summary
2
History of the Testing Arrangements
(Opposing)
Red
45
Electric Gradient in MV/m
 The LBL Pillbox cavity was originally
designed with removable 16 cm OD
windows to test different materials
especially Be because of its very low
Muon cross-section.
The first testing, proof of principle
experiments, was performed with flat Cu
windows. The results of this first testing
demonstrated a large reduction in
achievable gradient at large magnetic
fields.
The next testing was on flat Be widows
and lastly on curved Be windows. The
same magnetic field results.
It was then decided to modify the cavity
for single button testing with buttons of
different materials.
Because the spark damage observed in
the button testing occurred mainly in Cu
body of the cavity, it was decided to
modify the cavity for 2 button testing.
This will concentrate the peak gradient on
the button tip away for the body of the
cavity.
Safe Operating Gradient Limit vs Magnetic
Field Level at Window for the three different
Coil modes
(Single Coil)
4040
37.66
Black Diamond
37
36
35.2
35
34
32.4
31.7
31
30
28.8
28.5
27.327
26.4
25.9
25.7526.74
23.25
22.5
22
21.5
20.9
40
35
30
25
20
16.5
15
(Solenoid)
Yellow
15
13.5
10
5
0
0
1
2
3
4
Peak Magnectic Field in T at the Window
5
Parallel E and B Gradient breakdown
limit versus the Magnetic field in T.
3
History of The LBL Pillbox Cavity With Flat Cu and Curve Be Windows
serious Spark
damage areas
The LBL Pillbox cavity after 250 E6 pulses
with Flat, Curved and Button testing
accumulated a great deal of spark damage
as shown here and also on the next slide.
Serious spark
damage around
12 and 6 O ’clock
 The serious spark damage observed in
the coupling iris was due to the high
gradients because of the very small radii.
Because of the spark damage shown and
the small coupling iris radii. It was decided
to refurbish the cavity at JLAB.
The refurbishing at JLAB in their
cleanroom and a special tool was designed
to machine a 3.175 mm radius on the
coupling iris.
4
First Pillbox “Small” Button Cavity Test Arrangement
Curved Be Window
facing Outward
Test Arrangement
had a curved Be
Window
At radii from 2.5 to 6 cm radius on the Be side the field
is 15 to 30 % higher than at its center. This corresponds
to damage shown on the Cu Button Window side.
Button: Few
sparks and TiN
Coating eroded
away by RF
Cu Button
Window
• Be side: No (visible) damage
• Cu side: Damage on the Button Holder
5
Testing Arrangements for the Pillbox cavity
B&E
LBL Pillbox cavity
With flat windows; the
First RF commissioning
Test arrangement.
LBL Pillbox cavity
with smaller single button
and curved Be Window
Testing arrangement.
This was the first button
Testing arrangement.
LBL Pillbox cavity
with 2 larger buttons
Testing arrangement.
6
Dimensional drawings of the Smaller and Larger Buttons
Dimensions of first smaller test button
It had a smaller penetration into the
cavity. Its ration of peak gradient on
the button to peak gradient on the
coupling iris was equal to 0.675.
Dimensions of second Palmer test
button
It had larger penetration into the
cavity. Its ration of peak gradient on
the button to peak gradient on the
coupling iris was equal to 0.396.
The above ratio’s are important when we consider the gradient
to RF commission the cavity first with flat copper windows.
7
Simulation of the LBL Pillbox
cavity without buttons
Ratio of the peal electric gradients on the cavity
surfaces with Flat windows.
1. The ratio of A/C = 1.47;
where C is the gradient on the axis of the cavity and
A is the gradient on the coupling iris with r=1/8”
2.
The ration of B/C =1.25;
A
where C is the gradient on the axis of the cavity and
B is the gradient on the coupling window iris.
B
3.
Also note that the peak gradient is shifted
about 1 cm toward the coupling aperture
instead of being on axis.
4. The ratio’s will change when we add buttons:
With the palmer style (larger) button
The ratio of A/C becomes equal to 0.396
for the smaller first test button equal to 0.675.
C
The point of the above is to pick a safe gradient for the
first RF commissioning of the cavity with Flat copper
windows. There seems to be agreement that cycle
stress and strain are the primary cause of breakdown.
So we should limit the gradient for the Palmer buttons
to (40MV/m) x 0.396 x1.1= 17.4 MV/m, and for the
smaller button = (40MV/m) x 0.675 x1.1= 29.7 MV/m
8
Current Status of the LBL Pillbox Cavity

JLAB has completed the refurbishment
of the Pillbox Cavity under cleanroom
conditions.

It has been shipped to Fermilab with Flat
CU windows attached. It was
successfully vacuum leak checked
before shipment.

It was checked to be vacuum leak tight
after it arrived at Fermilab.

The waveguide high power RF drive
coupler has been attached and has been
successfully vacuum leak checked.

It has been mounted on its transport
dolly and is ready to be installed in the
MTA magnet.
Cavity on Dolly
LBL Pillbox cavity
9
Plans for the Pillbox Cavity








Install the Pillbox cavity in the MTA magnet. Connect
the cooling water, vacuum and calibrate the pickup
signals (2).
Connect the cavity to the 12 MW klystron waveguide
line.
Connect the optical fiber diagnostic, photo-tube and
spectrum analyzer.
RF commission the Pillbox cavity with the flat Cu
windows to 18 MV/m for 3 E6 or until the sparking rate
is less than 1/100000.
Remove the cavity and have the Lab photographer take
pictures of the inside of the cavity especially in the
area where there maybe damage.
Install the Be buttons into the cavity. Connect the
cooling water, vacuum, Fiber optics and calibrate the
pickup signals (2).
RF commission, in view of chart, and cycle stress and
strain causing breakdown to 25 MV/m without
magnetic field for 3 E6 or until the sparking rate is less
than 1/100000. Should we skit the above and go
straight to 3 T running?
Do the same as above with Cu buttons.
10
Summary of Pillbox cavity
• The history of the Pillbox cavity has been presented.
•The serious spark damage observed was presented
•The damage noted required refurbishment of the cavity
•The refurbishment was done at JLAB in their cleanroom
•The cavity has been refurbished and is has been returned to
Fermilab.
•The cavity is vacuum tight and has been mounted to its
transport dolly
•The cavity is now ready to be installed in the MTA magnet and
begin the 2 Be button testing after first testing with flat Cu
windows
• The first testing with flat windows should only be RF
commissioned to 18 MV/m because of the peak field levels at
the coupling iris even with a radius of 3.175 mm.
11
Discussion of the Magnetically Insulated Cavity
The orthogonal Box Cavity
Palmer and D. Stratakis developed as theory that a large
magnetic field at right angle to the RF electric would magnetically
insulate a vacuum RF cavity from electric breakdown.
 It was then decided to construct a build cavity to test this
theory.
A low cost design was chosen in the shape of a box
Two cavities of this type have been built and only one has been
tested.
We need to decide what to do with the second cavity. It has
been suggested that this cavity be coated with TiN and tested in
the MTA.
There is in storage material to built 2 more cavities. It has been
suggested that the material be built for the parallel B and E
confirguration.
12
Orthogonal E and H Cavity Coupled to the Waveguide
Rectangular Coupling
aperture with rounded
edges
E
Sapphire
Viewing
Port
B
Original LBL
Waveguide RF power
Coupler section; ½
height Standard WG
section
¼ height Coupling Cell.
This is built in two
sections with WG;
flanges not shown.
Pickup
Ports
HFSS Model RF cavity and WG coupler
13
Calculated Parameters of the cavity and cavity dimensions
HFSS normalizes all parameters to 1 W of
input power to the waveguide coupler and
solves for the frequency, gradient, coupling
factor, Qo (in cavity mode) etc.
1W produces a gradient of 25 kV/m by
scaling:
25 MV/m would take 1 MW ideal.
Aperture Fields
about a factor 4
lower than
central max.
field.
250
mm
276.50
mm
E
The Impedance across the center of the cavity
is
B
Imp =9.5 MΩ.
This is the resistance across the center of the
cavity given by
Imp = (gap Voltage)²/1 W.
123.82
mm
This uses the peak voltage and is in agreement
with SuperFish and most published accelerator
designs.
Qo = 27,400.
14
Orthogonal Cavity Electric Field Simulation Measurement Results
Note 1:
Ratio Eps/ Epa=
3.5 for orthogonal
case.
Note 2:
Ratio Eps/ Epa= 2
for parallel case.
This should be
considered when
determining the pre
RF commissioning
of the cavities (LBL
or
Box cavity) with flat
windows without
buttons and without
the presence of the
magnetic field.
Peak
Surface field
on-axis Eps
Network analyzer measurements:
 Fo = 805.33 MHz test lab;
Simulation Fo=806.2 MHz
 Qo= 27,400;
Simulation Qo= 27,400
β= 0.96 coupling factor;
Simulation β= 1.06.
These values are
preliminary and may change
when attached to the LBL
stepped WG coupler and
mounted in the magnet. Even
with a coupling factor of β=
0.96, 99.95 % of the
transmitted RF power go into
the cavity.
Peak Surface
Field on
Aperture Epa
15
Box Cavity in A0 Test Lab
Pickup ports
Compound Tin
Seal Location
Sapphire
Viewing port
location
Water
cooling
tubes
End ¼ height
Waveguide
Coupling
Piece
16
Box Cavity in the MTA Magnet on its Rail Support
Water cooling
tubes
Pick-up and Sapphire
ports
Y. Torun
17
Calibration Setup befor connection to 12 MW Supply WG
WG Flex Section
for Angle
Changing
Ceramic window for
Vacuum
Calibration Adapter
WG to type N
Y. Torun
Vacuum Pump
18
Spark Gradient vs. number of pulses at
0 deg and 3 T
45
40
Gradient in MV/m
35
30
25
20
Large YOYO Effect
starts here
15
10
5
0
0.00E+005.00E+051.00E+061.50E+062.00E+062.50E+063.00E+063.50E+064.00E+06
Number of RF pulses
19
Spark Gradient vs. Angle between E and B at 3 T
45
40
35
Gradient in MV/m
30
Series1
= 90 deg
Series2
= 89 deg
20
Series3
=87 deg
15
Series4
=91 deg
Series5
=86 deg
25
10
5
0
0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
3.50E+06
4.00E+06
Number of Pulses
20
40
35
30
Gradient in MV/m
The cavity was also
tested at magnetic field
levels from 0.5 to 3.5 T at 0
Deg. There was very little
drop off observed in the
break down levels.
There was no X-Ray’s
observed with the Magnetic
field on
There may have been
single surface multpactiring observed in the
early phase of the RF
commissioning with the
magnetic field
The vacuum burst were
10 to 100 times larger
during a spark than without
the magnetic field.
Box Cavity Gradient vs Angle
Between E &B at 3 T
25
Gradient Line for
Very Low Sparking
Rate.
20
15
10
5
0
85
86
87
88
89
90
91
92
RF field to DC Magnetic Field Angle
21
Pictures of Damaged Area in the Box Cavity
Blistered Areas
The inside of the cavity was very clean with not many spark pits visible if any.
 There were blistered areas in appearance in the top and bottom corners along
the axis of the cavity both the upstream and downstream ends.
These are pictures of the cavity at about ½ the running time. At the end of the
experiment, the areas appeared darker and spread out further and this was also
after the 50 MV/m RF commissioning at 0 T.
22
The SPDT RF Switch has been installed in the MTA
This gives us 2 RF test location in the MTA Hall.
23
Parallel E and H field Design:
Optical View
port
E Field pick-ups
E&B
Original LBL
Waveguide RF power
Coupler section; ½
height Standard WG
section
¼ height Coupling Cell.
This is built in two
sections with WG
flanges not shown.
24
Summary Of the Box cavity
•
•
•
•
•
We have completed the Box cavity Study ( magnetic field
insulation effect) at 91,90, 89, 87 and 86 Degrees at 3. Tesla.
Before each run we commissioned the cavity upto 33 MV/m at 0.
T except for the first 0 Degree run commissioned to 23 MV/m. In
later runs, we commissioned the cavity to 50 MV/m at 0 T.
The results have been shown
The plans are then to install the LBL Pill box cavity and begin the
study with new Be button shape. It is smaller in diameter and
more peaked to give a enhancement factor of 3.
We need to decide if the second orthogonal box cavity should be
tested with a TiN coating.
We also need to decide if we need to build a parallel Box for
testing in the MTA.
25