Final report for Pressurized Gas Beam Monitor for Extremely High Intensities

K. Yonehara

Fermilab

Abstract

We propose a novel RF beam detector for future intensity frontier neutrino facilities. We built the RF test detector and carried out the beam test at the MI abort line. The result shows the linearity of the detector as a function of beam intensity. We evaluated a systematic error of the detector based on the experiment. We found that the ionization in a wave guide would be a source of a noise. We propose to build a high quality factor cavity to mitigate the issue.

Introduction

Fermilab is the hub institution for future intense neutrino beam experiments by using a MW-class proton driver, those facilities are called Accelerator Improvement Plan (AIP) NuMI facility and Long Baseline Neutrino Facility (LBNF). A new type of rad-hard beam monitor is needed which is located downstream of a pion-production target to align the beam with target, horns, and DUNE Far detectors which is located 1300 km away from the target. The beam monitor also diagnoses deterioration of the target so that the monitor should have a good linearity. A radiation-hard gas-filled RF beam test detector was proposed, built and demonstrated to validate its concept by using a proton beam at the Main Injector (MI) abort line. This report shows the result of beam experiments during the phase II studies, evaluate the monitor performance, and discuss a possible improvement for a future neutrino facility.

All materials that have radiation sensitivity, such as ordinary cables and electronics, are placed outside of the abort room. The only materials in the high radiation areas is aluminum, which is used to make the cavity wall and the wave guides. Aluminum is the standard material to make a radiation-hard instrumentation. We tested various ionization gases. The phase II studies have demonstrated that we will not need a pressurized gas to control the plasma dynamics as was done in earlier MTA experiments. Atmospheric gas can be flown to keep fresh in the cavity. The basic RF monitor component is so simple that it promises to be much more radiation robust in comparison with ionization chambers. The MI beam test carried out in FY19 and the detector received integrated > 2e15 protons. Even the amount is small, we do not find any gain change in the cavity.

Other great feature of the RF detector is that the RF signal can be calibrated without the beam calibration run. If the shunt impedance is known, the RF power consumption is estimated with a simple formula. The concept was tested and experimentally validated in the present project.

Principle

A gas-filled RF-resonator beam monitor will be simple and, thus radiation-robust. When the charged beam passes through the gas-filled RF resonator it produces a large amount of electron-ion pairs by interacting with gas. The permittivity in the resonator is changed proportional to the number of incident charged particles. The imaginary part of permittivity is RF power consumption in the beam-induced plasma. The power consumption per single ion pair is given

, (1)

where and are the mobilities of electrons and positive (negative) ions in a gas, which are functions of , gas pressure, and gas temperature. The coefficient and are the relative populations of the electrons and negative ions, respectively (). The relative population is varied by and gas species with electron capture processes. The number of ions is also varied by , gas species, and the plasma density with charge recombination processes. Often we can find the cross section of those processes from the past experiment, or we measured them in the past MTA experiments because we could not find them at our interested , in which the gas pressure and RF gradient are extremely high.

On the other hand, the plasma consumption is given by solving the power equation with a RF equivalent circuit,

, (2)

where and are a shunt impedance and capacitance of the RF cavity, respectively. The first term is the power gain from the RF source while the second one is the stored RF power in the cavity. At the equilibrium condition (), the plasma consumption power is equal to the RF power gain from the RF source.

The total number of ion pairs in the cavity can be proportional to ,

, (3)

where is the correction factor for the electric field variation over the plasma distribution. For simplicity, we assume that the beam is a zero spot size and hits on the peak gradient, thus is unit in this analysis. The amount is proportional to the number of incident particles in the RF detector. Please note that we estimated the absolute value of in the past MTA experiments. To do the same accuracy, we need more beam diagnostics which we do not have at the MI abort line. To validate the concept of beam detection in the RF detector, our present setup is sufficient.

Experiment

We built a RF test detector (Figure 1).

· 2.4 GHz gas-filled RF resonator

· Pillbox cavity (TM011)

· Inner diameter 3.6850.001’’

· Length 3.000’’

· RF body and wave guides are made of aluminum

· 1-mm thick beam window

· Cavity is pressurized up to 2 atm

Figure 1: RF test detector

Figure 2: Electronics of RF detector for the beam test.

Figure 3: Observed RF gain. Blue is taken at the driving frequency 2.385 GHz (nominal), Orange and green are taken at 2.390 and 2.379 GHz, respectively.

In an embodiment of the electronics illustrated in Figure 2 for the beam tests, we used a network analyzer (NWA) as a RF power source which generates RF power in the range from 0 to 20 dBm. The RF signal is sent through a long ½’’ heliax cable (labeled Orange). There is an amplifier unit in the electric line to compensate the insertion loss of the cable. The unit consists of two amplifiers (labeled 1 W and 5 W) and two attenuators (labeled 20 dB and 10 dB), which are used to filter out RF noise and to minimize the non-linear behavior of a single amplifier. Figure 3 shows the RF gain in the RF electronics. The amplified RF signal then goes through two directional couplers to measure the forward and reflected signals (with output labeled White and Blue, respectively). The RF signal is sent through another ½’’ heliax cable (labeled red) then fed by the 3’-long wave guide to the input (port 1) of the 2.45 GHz gas filled RF resonator that is used to measure the charged-particle flux. The output RF signal from the 2.45 GHz resonator is sent through another 3’-long wave guide and taken from the coupler (labeled port 2). In order to tune the quality factor (QF) of the resonator (i.e. to have the correct loaded QF), the port 2 coupler is slightly under-coupled. The output RF signal from the port 2 is then sent through a heliax cable (labeled Green) and filtered (labeled 1.6 dB) then fed into a real-time RF power meter. The measurement of the power lost in the RF resonator (or cavity) by comparing the power sent by the NWA to that received by the RF power meter and the knowledge of the gas composition and pressure are all that are needed to determine the intensity of the beam passing through the resonator.

The equation (2) shows that the RF power consumption due to a beam-induced plasma can be evaluated by using a shunt impedance at the equilibrium condition. The shunt impedance can be measured from the non-beam RF signal. We measured the quality factor of the RF test detector with two different methods. Figure 4 shows the spectrum of the RF test detector. The cavity Q is determined from the band width, where and are the resonant frequency and the bandwidth, respectively. While figure 5 shows the observed decay constant in the test detector. The cavity Q is determined, , where is the time constant. Discrepancy of both measurements is less than 10 %.

Figure 4: Orange is the spectrum of RF test detector. . Green and cyan are the observed forward and reflected RF power, respectively.

Figure 5: The observed decay time constant. =247.

Figure 6 shows the RF detector signal of a six-batches 120 GeV/c MI proton beam. A red line shows the initial RF voltage () and a blue line is the voltage at equilibrium condition (). The RF is loaded by beam when the beam is turned on at 0 sec. The RF is recovered when the beam is turned off at 9.6 sec. The sampling time of the detector is 10 nsec. We observed the batch gap in the RF test detector. From the signal, the gap can be 40-50 ns. This is consistent with the MI beam structure. The RF voltage is fluctuated at the equilibrium condition. This may be a beating between the beam bunch timing and the RF sampling. Further analysis will be made.

Figure 6: Observed six-batch MI beam signal in the RF test detector.

Figure 7 shows the typical beam event in the RF test detector for the demonstration test. The beam is a single batched beam. We observed that the beam batch intensity went down by 20 % in the single spill. The MI experts confirmed this is real. Unfortunately, we do not have a fast beam intensity monitor in this run. The quality factor of the RF test detector can be tuned by applying the off resonant driving frequency. The shun impedance is changed accordingly. This is a calibration to convert the RF signal to the RF power consumption.

Figure 7: (Top left) Normal RF signal, (Top right) The beam intensity wend down by 20 % in one spill, (Bottom left) The test detector is over coupled with the RF power source when the driving frequency is +2 MHz off from the resonance, (Bottom right) The test detector is under coupled with the power source when the driving frequency is -2 MHz off from the resonance.

Figure 8 shows the estimated RF power consumption in the RF test detector at the equilibrium voltage by using equation. The plot suggests that the atmospheric N2 gas (blue point) is the most sensitive to the beam intensity. The deviation of point from the fitting curve is at most 10 % level. The error can be improved by adding a correction. We will discuss it in the next section.

Figure 8: Estimated RF power consumption in the RF test detector. Blue is taken with atmospheric N2 gas (), orange is taken with 2-atm N2 gas (, green is tanek with 2-atm dry air (), and red is taken with 1-atm He gas (

Discussion

The measurement shows a good linearity of RF signal with respect to the beam intensity. We tested N2, He, and dry air and the atmospheric N2 gas shows the best sensitivity in the range of 1e12 to 4e12 protons per spill. However, we see the deviation of data points, at most 10 % level. There are several sources of the fraction for each group. The beam intensity is varied in the beam spill. Figure 7 shows the example. We estimated the equilibrium potential from the first 1 sec beam. This makes 5-10 % fraction of the estimated beam signal. This event suggests that the RF detector is sensitive to the time structure of incident beam.

The beam position and size on the RF detector are other sources of the fraction. Because the MI abort line is not designed to precisely control the beam, the beam position is drifted and beam size is varied during the experiment. We have a SWIC in front of the RF test detector. The beam profile correction () will be taken into account in the next higher level of analysis.

Another concern is the quality factor change. In the present data acquisition process, the quality factor measurement is taken typically before and after whole RF beam measurements. We occasionally took the quality factor measurement during the beam test and found that the quality factor varied by 10 % during the measurement. That directly changes the estimated RF power consumption. Ideally, the quality factor measurement must be done at each beam spill. We will modify the data acquisition procedure to take the quality factor measurement at each beam spill.

We should discuss about a possible issue on the RF detector for neutrino applications. Ionization in the waveguide can be taken place because the beam size in a neutrino beam line can be a few orders of magnitude larger than the RF detector. In case of the test detector, the field gradient in the wave guide is about 1/6 of the gradient in the cavity. It means that the 1/36 of RF power will be consumed by a plasma in the waveguide per the volume of RF cavity. Because the volume of waveguide is much bigger than the cavity, the RF power consumption in the waveguide becomes comparable, or bigger than that in the RF detector. One possible solution is using a high quality factor cavity. The high Q factor cavity can store the RF power with low RF input power. Therefore, the field potential in the waveguide will be significantly lower than the RF cavity.

Conclusion

We carried out the beam demonstration test of the novel RF detector and found a good linearity of the RF response with respect to the beam intensity. We validated the concept of RF beam detector.

We evaluated a possible systematics of the RF detector from the experimental results. Especially, the ionization in the waveguide can be a potential source of background noise. A high-Q cavity is a potential solution to mitigate the noise. We plan to study the waveguide issue by using the RF test detector. The detector will be moved to directly send a beam into the waveguide during the Summer shutdown in 2019. The beam result will tell us how the RF circuit is disturbed by the plasma in the wave guide.

0.000000 5.×10-6 0.000010 0.000015020

40

60

80

100

120

Time (s)

Voltageincavity

(V)

0.000000

5.×10

-6

0.000010 0.000015

0

20

40

60

80

100

120

Time(s)

V

o

l

t

a

g

e

i

n

c

a

v

i

t

y

(

V

)

0 1.×10-6 2.×10-6 3.×10-6 4.×10-60.000.05

0.10

0.15

0.20

Time(second)

RFsignal

(V)

0

1.×10

-6

2.×10

-6

3.×10

-6

4.×10

-6

0.00

0.05

0.10

0.15

0.20

Time(second)

R

F

s

i

g

n

a

l

(

V

)

0 1.×10-6 2.×10-6 3.×10-6 4.×10-60.000.05

0.10

0.15

0.20

Time(second)

RFsignal

(V)

0

1.×10

-6

2.×10

-6

3.×10

-6

4.×10

-6

0.00

0.05

0.10

0.15

0.20

Time(second)

R

F

s

i

g

n

a

l

(

V

)

0 1.×10-6 2.×10-6 3.×10-6 4.×10-60.000.05

0.10

0.15

0.20

Time(second)

RFsignal

(V)

0

1.×10

-6

2.×10

-6

3.×10

-6

4.×10

-6

0.00

0.05

0.10

0.15

0.20

Time(second)

R

F

s

i

g

n

a

l

(

V

)

0 1.×10-6 2.×10-6 3.×10-6 4.×10-60.000.05

0.10

0.15

0.20

Time(second)

RFsignal

(V)

0

1.×10

-6

2.×10

-6

3.×10

-6

4.×10

-6

0.00

0.05

0.10

0.15

0.20

Time(second)

R

F

s

i

g

n

a

l

(

V

)

1 2 3 40.000

0.005

0.010

0.015

0.020

Beam Intensity (1e12 protons/spill)

RFpowerconsumption(J)

1 2 3 4

0.000

0.005

0.010

0.015

0.020

BeamIntensity(1e12protons/spill)

R

F

p

o

w

e

r

c

o

n

s

u

m

p

t

i

o

n

(

J

)

-10 -5 0 5 10 15 20-20-15-10-50

5

10

Readout on NWA (dBm)

ReadoutonRTP

(dBm)

-10 -5 0 5 10 15 20

-20

-15

-10

-5

0

5

10

ReadoutonNWA(dBm)

R

e

a

d

o

u

t

o

n

R

T

P

(

d

B

m

)

2.375 2.380 2.385 2.390 2.395-50-40-30-20-10

Driving frequency (GHz)

ReadoutonRTP

(dBm)

2.375 2.380 2.385 2.390 2.395

-50

-40

-30

-20

-10

Drivingfrequency (GHz)

R

e

a

d

o

u

t

o

n

R

T

P

(

d

B

m

)

-2.×10-7 -1.×10-7 0 1.×10-7 2.×10-70.020.05

0.10

0.20

0.50

Time (s)

VoltageatRTP

(V)

-2.×10

-7

-1.×10

-7

0

1.×10

-7

2.×10

-7

0.02

0.05

0.10

0.20

0.50

Time(s)

V

o

l

t

a

g

e

a

t

R

T

P

(

V

)