SUMMARY OF THE FERMILAB PROTON DRIVER DESIGN STUDY*
W. Chou
†
and C. Ankenbrandt, FNAL, Batavia, IL 60510, USA
Abstract
This paper is a summary report of the Proton Driver
design study that has recently been completed at
Fermilab. It describes the design of a new 16 GeV high
intensity rapid cycling synchrotron as a replacement of
the present Booster. The major design goals are: (1) 1
MW beam power; (2) 1 ns (r.m.s.) bunch length. The
construction will be staged. It has also an upgrade path to
4 MW.
1 INTRODUCTION
In a 1997 summer study, a team led by Steve Holmes
formulated a development plan for the Fermilab proton
source and described the results in Ref. [1]. Subsequently,
at the end of 1998, a task group was formed to prepare a
detailed design of a high intensity facility called the
Proton Driver. The results of that effort are summarized in
Ref. [2]. The design includes a 16 GeV synchrotron, two
new beam transport lines (a 400 MeV injection line and a
12/16 GeV extraction line), and related improvements to
the present negative ion source and the 400 MeV Linac
The Proton Driver serves a number of purposes in the
Fermilab HEP program. In the near term, it replaces the
present Booster and increases the proton beam intensity in
the Main Injector by a factor of four, thereby providing an
upgrade path for NuMI and other 120 GeV fixed target
programs. It also opens the avenue for new physics
programs based on its stand-alone capabilities as a source
of intense proton beams. In the long term, it could serve
an intense muon source, a neutrino factory and a muon
collider by generating intense short muon bunches from a
target. The design also allows an upgrade path to a 4 MW
proton source by adding a 600 MeV linac and a 3 GeV
Pre-Booster.
At present, the Booster is the bottleneck that limits the
proton beam intensity in the Fermilab accelerator
complex. Its upstream machine, the Linac, is capable of
providing 3.4 × 10
13
particles per cycle at 15 Hz.
However, due to numerous problems, the Booster
intensity is limited to 5 × 10
12
particles per cycle. After
some modest upgrades, the downstream machine, the
Main Injector, is capable of accelerating four times more
protons than the Booster can provide. The Proton Driver,
as a complete functional replacement for the Booster,
removes this bottleneck and makes full use of the
capabilities of the Linac and Main Injector.
____________________________
*Work supported by the Universities Research Association, Inc.,
under contract DE-AC02-76CH03000 with the US Dept. of Energy.
†
chou@fnal.gov
2 MACHINE LAYOUT & PARAMETERS
The layout of the Proton Driver is shown in Figure 1.
The existing 400 MeV Linac will be reused. The H
-
beam
will be extracted from the Linac to the new 400 MeV
transport line and injected into the Proton Driver in the
same way as in the present Booster, namely, through a
charge exchange process. The H
+
(proton) beam will then
be accelerated to 16 GeV (or 12 GeV in Stage 1) in about
38 ms and extracted to the 12/16 GeV transport line. It is
then either injected into the MI-10 section of the Main
Injector or directed to a target for fixed target experiment.
The Proton Driver has a circumference of 711.3 m,
which is exactly 1.5 times the size of the present Booster
(474.2 m). It is of a triangular shape and has 3-fold
symmetry. It has three arcs and three long straight
sections. Each arc is about 173 m long and each straight
section about 64 m long. These straight sections are used
for injection, collimation, rf cavities and extraction.
In order to minimize any possible interruption to the
ongoing Fermilab HEP program, the site of the Proton
Driver is chosen at the west side of Kautz Road. In this
layout, the new 400 MeV beam line includes 150 m free
space for a future Linac upgrade. A future Pre-Booster can
also easily fit in. About two thirds of the new 12/16 beam
line will be in the existing enclosure. The elevation of the
Proton Driver is the same as that of the Main Injector.
This ensures appropriate radiation shielding.
The design presents a 2-stage implementation of the
Proton Driver. Stage 1 provides a maximum beam energy
of 12 GeV with a 53 MHz rf system, whereas Stage 2
increases the beam energy to 16 GeV with a new 7.5 MHz
rf system. The reasons for this staged implementation are:
(1) In Stage 1, one may reuse the rf system of the present
Booster and thus reduce the capital cost. (2) A 53 MHz rf
system matches that of the Main Injector. (3) In order to
match the acceptance of the Main Injector (40p mm-mrad
at 8 GeV) to the emittance of the Proton Driver beam (60p
mm-mrad, normalized), the Main Injector injection energy
needs to be raised to 12 GeV. (4) In Stage 2, it is
envisioned that a neutrino factory will be in place. It
requires a small number of proton bunches. Therefore a
low frequency (7.5 MHz) rf system replaces the 53 MHz
system. The maximum beam energy of the Proton Driver
is also increased to 16 GeV in order to generate enough
muons for a neutrino factory.
The major design parameters of the Proton Driver in
Stage 1 and 2 are listed in Table 1. As a comparison, the
parameters of the present Linac and Booster are also
listed.
0-7803-7191-7/01/$10.00 ©2001 IEEE.
Proceedings of the 2001 Particle Accelerator Conference, Chicago
594
3 TECHNICAL SYSTEMS
In order to achieve the demanding performance
specifications of the Proton Driver, a number of state-of-
the-art features are incorporated in its design.
• Lattice: This is a transition-free FMC (flexible
momentum compaction) lattice. It uses 270°/270°
FODO modules in the arcs. It has large momentum
acceptance and dynamic aperture. [3]
• Magnets: The magnets employ external vacuum
skins like those in the Booster, have large apertures
like those in the Fermilab Accumulator, and use
stranded conductors for the coil in order to reduce
eddy current losses. [4]
• Power supplies: The power supply uses a dual-
harmonic resonant system (15 Hz plus 12.5% of 30
Hz component), thereby lowering the peak rf power
requirement by 25%. The trim coils in the main
quadrupoles use finite number of current harmonics
(up to the 7th) for tracking error correction. [5]
• Vacuum and beam pipe: Because the magnets are in
vacuum, the beam pipe is made of metallic stripes or
perforated liners (like that in the LHC) to provide a
low-impedance environment for the beam. [6]
• RF: In Stage 2, the 7.5 MHz rf cavities employ a new
type of alloy called Finemet for their magnetic cores.
The main advantages of the Finemet cores are high
accelerating gradient and wide bandwidth. [7]
• Injection: To keep space charge tune spread under
control, the transverse charge distribution of the
injected beam will be made as uniform as possible by
the painting technique. [8]
• Collimation: A sophisticated 2-stage beam collimator
system collects about 99% of the lost particles in a
small area, thereby allowing hands-on maintenance
in the rest of the enclosure. [9]
• H
-
source: A noiseless Dudnikov-type-source will
increase the beam intensity by a factor of two (to 115
mA) from that of the present source. [10]
• Linac front-end: This system consists of two RFQ
sections and a double-alpha transport line in
between. The latter provides a perfect match
between the two RFQ's. [11]
• RF chopper: A new type of chopper, which is similar
to a beam transformer, has been designed,
manufactured and tested in a collaboration between
Fermilab and the KEK. [12,13]
• Inductive inserts: These are ferrite modules for
compensating space charge effects. There has been a
successful beam experiment at the LANL. These
modules help increase the beam intensity in the PSR
without any adverse effect. [14]
4 R&D PROGRAM
A complete list of the R&D program can be found in
Chapter 19 of Reference [2]. An important category of the
program are those items that are not only needed by the
Proton Driver but will also be useful for improving the
performance of the present proton source. Therefore, they
have the highest priority. These include:
• High brightness H
-
source development
• Linac front-end improvement
• Booster 53 MHz rf cavity modification
• Finemet 7.5 MHz rf cavity development
• Beam loading compensation system development
• Inductive inserts study in the present Booster
• Booster magnet ac field and impedance measurement
Several other items are also currently underway:
• Material outgassing rate test
• Stranded conductor coil study
• Chopper development
• High gradient, low frequency rf system
A number of other R&D items have to wait until more
resources can be made available.
5 ACKNOWLEDGEMENTS
In the past two years the design effort has attracted
more than fifty participants, mostly from Fermilab.
Physicists from the IIT, Stanford Univ., Univ. of Hawaii,
CERN, RAL and IHEP (Russia) also contributed heavily.
The authors are deeply indebted to all of them. The
support from Fermilab management to this study is
greatly appreciated.
6 REFERENCES
[1] S. Holmes, editor, FERMILAB-TM-2021 (1997).
[2] W. Chou, C. Ankenbrandt and E. Malamud, editors,
FERMILAB-TM-2136 (2000).
[3] W. Chou et al., "An Imaginary ?
t
Lattice for the
Fermilab Proton Driver," this conference.
[4] F. Ostiguy and F. Mills, "Large Aperture Magnets for
a Future High Power Proton Synchrotron," this
conference.
[5] C. Jach and D. Wolff, "Proton Driver Power Supply
System," this conference.
[6] T. Anderson et al., "FNAL Proton Driver Canned
Vacuum System Design," this conference.
[7] D. Wildman et al., "A Prototype 7.5 MHz Finemet
Loaded RF Cavity and 200 kW Amplifier for the
Fermilab Proton Driver," this conference.
[8] C. Prior, " H
-
Injector Studies for the Fermilab Proton
Driver," this conference.
[9] A. Drozhdin et al., "Beam Loss and Collimation in the
Fermilab 16 GeV Proton Driver," this conference.
[10] V. Dudnikov, "30 Years of High-Intensity Negative
Ion Sources for Accelerators," this conference.
[11] D. Young et al., "Low Energy Improvements to the
Fermilab 400 MeV Linear Accelerator," this
conference.
[12] W. Chou et al., KEK Report 98-10 (1998).
[13] Y. Mori et al., "Fast Beam chopper with MA Cores,"
Proc. 2000 EPAC (Vienna).
[14] J.E. Griffin et al., FERMILAB-FN-661 (1997)
Proceedings of the 2001 Particle Accelerator Conference, Chicago
595
Table 1: Proton Driver Parameters of Present , Stage 1 and Stage 2
Parameters Present Stage 1
(MI)
Stage 2
(MI + ν-fact)
Linac (operating at 15 Hz)
Kinetic energy (MeV) 400 400 400
Peak current (mA) 40 60 60
Pulse length (µs) 25 90 90
H
-
per pulse
6.3 × 10
12
3.4 × 10
13
3.4 × 10
13
Average beam current (µA) 15 81 81
Beam power (kW) 6 32 32
Booster (operating at 15 Hz)
Extraction kinetic energy (GeV) 8 12 16
Protons per bunch
6 × 10
10
2.4 × 10
11
1.7 × 10
12
Number of bunches 84 126 18
Total number of protons
5 × 10
12
3 × 10
13
3 × 10
13
Normalized transverse emittance (mm-mrad) 15π 60π 60π
Longitudinal emittance (eV-s) 0.1 0.1 0.4
RF frequency (MHz) 53 53 7.5
Extracted bunch length σ
t
(ns) 0.2 1 1
Average beam current (µA) 12 72 72
Target beam power (MW) 0.1 0.9 1.2
Figure 1: Proton Driver Layout.
Proceedings of the 2001 Particle Accelerator Conference, Chicago
596
