Jayakumar R. Particle Accelerators, Colliders, and the Story of High Energy Physics: Charming the Cosmic Snake

Подождите немного. Документ загружается.

with radiation resistant ﬁlm insulation and then with epoxy impregnated ﬁberglass

tapes. Since the strands become extremely brittle, the insulated ribbons are wound

with insulated wedges spacers in a winding tool and with end spacers, then held in a

special tool which puts the coil assembly under pressure and cures (with the setting

of the epoxy) at high temperatures. The tool dimensions and the process of winding

and curing are controlled to a great precision, in order to obtain the coil locations

within tens of microns of the code calculations. The coil is made in the form of race-

track-like saddle coils with long straight sections and short ends. Figure 8.6 shows a

fabricated coil and the curing tooling. Lower ﬁeld niobium–titanium coils are made

with ribbons of strands and wound into coils and used directly.

Typically, a dipole magnet is made by clamping a top and a bottom coil between

two metal collars. The collars are made from a number of laminates of stainless

steel (to avoid the so-called eddy currents which might be induced when the

magnetic ﬁeld changes during the current ramp up or down). The clamped coil

assembly is then placed between two halves of iron laminate yokes. The iron yoke

helps to obta in a higher magnetic ﬁeld and to prevent the leakage of magnetic ﬁelds

outside the magnet assembly. The whole assembly is then placed between two

aluminum or stainless steel semishells under compression and welded together.

Welding shrinkages and cool down to low temperatures cause the assembly be held

in compression. The art of mak ing the superconducting coil is intricate requiring

precise engineering modeling of both the design and the process. The fabrication

process has to be carried out with meticulously chosen materials which must

withstand high currents and magnetic ﬁelds, high potentials in case of magnet

trips, high mechanical pressures, and a large range of temperatures from 1,200



down to 270



C. The splice joints between coils, for example, between the upper

and lower coils are designed and developed with great care to withstand the large

mechanical forces exerted on them by the complex magnetic ﬁeld in that region,

Fig. 8.6 (Left) Saddle coil of a dipole magnet (Right) A niobium–tin coil being loaded into a curing

(reaction) oven. (Science@Berkeley Lab, Oct., 23, 2007, Courtesy: Lawrence Berkeley National

Laboratory)

122 8 The Next Generation: Supersynchrotrons

while the electrical resistance of the joint is minimized. (This is one of the most

sensitive and least reliable parts of the magnet which can, as it happened in the LHC

machine, cause major shutdowns.) All these parts of the magnet have to withstand

large doses of nuclear radiation over the operating life time. Special large tools

(with intimidating sizes) and jigs are fabricated to achieve the design parameters in

fabrication.

The iron yoke laminates have holes for the circulation of liquid helium which is

needed to cool down the coils to the operating temperature of 4



K or lower and to

remove any small heat that may be generated by mechanical movements or

somewhat resistive components like joints between the turns. During operation,

typically the magnets receive minimal amount of external heat load (for example,

the dipole magnet of the supercollider would have received about 2 watts per

magnet due to synchrotron radiation – radiation emitted by proton in moving in a

circular orbit and any additional heat caused by so-called hysteresis losses when

current is ramped in the magnet). Since material (even copper) thermal conductivity

becomes very small when cooled to very low temperatures, one cannot rely on

conduction through walls of the coil to cool the coil, and therefore the helium has to

ﬂow right near the coils. The allowed temperature rise in the cooling helium is

extremely small, of the order of 0.01



C since there is very little temperature margin

for the superconductor – meaning the superconductor would go normal with even a

small rise in temperature. Therefore, even the few watts of heat load can be onerous

and the cooling has to be exquisitely designed and conﬁrmed.

The whole cold mass is enclosed in a cryostat which is like a thermos ﬂask,

providing a vibration resistant, stable, and strong support while preventing any heat

leakage from the exte rnal environment. The cold mass support consists of multiple

steps of composite thermally insu lating materials arranged in such a way that the

path for heat leakage from outside is very long. The cold mass is draped in reﬂecting

foil to prevent receiving any heat radiation. There are two shells in the cryostat, an

outside iron shell at room temperature (steel conﬁnes all magnetic ﬁeld so that

external instruments may not be affected) and an inner alumi num shell which is

maintained at liquid nitrogen temperature (77



Kor196



C) so that the surface

“seen” by the cold mass is at low temperature and heat radiation is reduced. The use

of such a “shield” reduces the requirement for the more expensive liquid helium

(Fig. 8.7).

As an example, the Superconducting Supercollider, which was not built, had

dipole magnets of advanced design that produced 6.6 T (66,000 gauss) at a current

of about 6,500 A in a bore of 5 cm diameter and were 15 m long with an outer

diameter of 66 cm. There would have been over 8,000 such magnets in the collider.

The qu adrupole magnets produced a magnetic ﬁeld gradient of over 200 T per

meter radius and had a length of 5.2 m. All these numbers are spectacularly large,

requiring a large research and development effort. The Large Hadron Collider

Magnets (see later chapter) have even larger ﬁelds with larger bore and are two

magnets in one. The superconducting accelerator magnets are a marvel of techno-

logical achievement combining precision with cutting-edge materials based on

advanced physics discoveries. The design and development of materials, tools

Basic Design of an Accelerator Magnet 123

(tooling), and processes for the manufacture of such magnets is an amazing

achievement of the 1980s and 1990s. Over the years, such experience in

superconducting design and fabrication had been gained in the Tevatron, HERA,

NUKLOTRON, and RHIC accelerators. This experience and that gained in the

Supercollider project provided considerable learni ng experience for pushing the

limits of the magnet technology further such that very advanced magnets such as in

LHC are real today.

In addition to the dipole and quadrupole magnets, correction magnets, which

correct ﬁeld errors in the dipole and quadrupole magnets, are also made from

superconducting wires. Since the correction ﬁelds are usually small but have

complex winding patterns, the choice of superconductor is more to avoid losses

and make the magnets compact. The technology of the correction magnets has

evolved for achieving this purpose and often the coils are patterns placed on

insulation and rolled up to get the desired ﬁeld conﬁguration.

Superconducting Radiofrequency Accelerator Cavities

The Radiofrequency (RF) cavity is where the particles increase their energy by

being accelerated by an applied electric voltage. The RF voltage, as stated before, is

synchronous with particles to continuously accelerate the particle bunches. The

bunches (buckets) are separated in space such that when one bunch arrives, the other

Vaccum

Vessel

Coldmass

20K Thermal

Shield

80K Thermal

Shield

Cast

Cradle

FRP Support Post

External Feet

Fig. 8.7 Cross section of a superconducting quadrupole magnet, SSC Site Speciﬁc Conceptual

Design Report, July 1990

124 8 The Next Generation: Supersynchrotrons

leaves. This way, each bunch sees an accelerating voltage as it enters and a repelling

voltage as it leaves. In addition to accelerating the particles, the accelerator cavity

must also replace the energy lost by particles through synchrotron radiation which

the partic les emit when they do not move along a straight line. (Therefore, this is

generally an issue for synchrotrons and not for linear accelerators.) The RF cavity

design has come a long way since the days of Wideroe, aided by extraordinary

developments in the application of superconducting materials. These developments,

while important for circular accelerators, are the key to linear accelerators of the

future since so many of them would be required for them.

The challenge of the RF cavity is to generate high accelerator voltages so that

particles can be accelerated without going through too many turns. The demand on

the voltage is particularly severe for linear accelerators where multiple cavities are

used along the accelerator length, which can be made shorter, reducing size and

cost, if the accelerating voltage per unit length of the cavity can be increased. When

these voltages are large, considerable energy loss occurs in the RF structure and

electrical brea kdowns are a serious issue. The total power that must be fed into the

cavity is the product of the cavity voltage and the beam current (number of particles

crossing the cavity per second  electric charge). One way to limit the energy

losses is to make the cavity structure that conﬁnes the strong radiofrequency electric

ﬁelds, out of superconducting materials, reducing power requirement to near zero.

Superconducting cavities have zero D.C. resistance and therefore can achieve high

“Quality Factor,” ratio of inductive to resistive voltage.

When the development of superconducting cavities began, initial results were

limited to a few MV/m ele ctric ﬁelds. At these ﬁelds, considering the cost of

cryostat, etc., these would not have been competitive. But, recent developments

in design, material, and clean room fabrication, including surface deposition

techniques, have pushed the gradient to near 40 MV/m, a stupendous number.

Central to this achievem ent are the material processing techniques, viz., chemical

cleaning of the surface of the cavities, removal of any specs and sharp burrs which

would locally increase the electric ﬁelds, causing ﬁeld emission and electrical

breakdown, annealing of the superconducting niobium material at 1,400



Cto

increase its thermal conductivity to maintain superconducting stability, and rinsing

methods to remove surface contaminants, etc. Since such cavities operate at very

low temperatures, these too need a cryostat to minimize external heat. The simpli-

ﬁcation of cryostat design has been an important factor in this feasibility. While the

normal cavities operate at a frequency of the order of 10 GHz, superconducting

cavities operate at a lower frequency (around 1 GHz), making the sources simpler

and mechanical tolerances easier to achieve. The lower frequency also allows long

particle bunches (Fig. 8.8).

While niobium–titanium and niobium–tin are better superconductors, the maxi-

mum gradients these could sustain are limited partly due to their granular structure

and grain-boundary effects and because these materials are affected more by high

frequency ﬁelds. Superconductors are nondissipative in DC conditions, but when

alternating ﬁelds are applied, the hysteresis involved in their persistent currents

causes dissipation, much like in iron and this dissipates heat, causing

Superconducting Radiofrequency Accelerator Cavities 125

superconducting instability. The energy dissipation is given by the equivalent AC

resistivity,

¼ðko

=TÞexpð1:76T

=TÞ;

where T

is the critical temperature (temperature above which the superconductor

would turn normal), T is the operating temperature, and o is the RF frequency. Due

to the exponential dependence on the temperature and the dependence of the

proportionality constant k on temperature, the AC resistance drops by nearly two

orders of magnitude, if the cavity can be operated at supercooled liquid He (LHe)

temperature of 2 K rather than the 4.2 K. The maximum sustainable RF ﬁeld is also

limited by the magnetic ﬁeld associated with it and even earth’s magnetic ﬁeld

affects the cavity and shielding is required to achieve high RF ﬁelds. Niobium is the

usual choice for the RF cavities, either in the form of thin sheet or sputtered onto

copper surface. Accelerators, such as the Cornell synchrotron and DESY, have used

sheets while the Large Electron Positron Collider in CERN (Fig. 8.8) used

sputtering technique. The cavity is designed carefully using specialized codes to

get mos t optimum contours of the surfaces while giving the correct resonant

structure avoiding undesirable modes of electric ﬁeld oscillation. The design and

fabrication is both a science and an art and the gleaming product is usually a

testament to the RF cavity scientist-sculptor (Fig. 8.9).

Supporting Cast-Facilities and Systems Engineeing

As stated earlier, the superconducting magnets and RF cavities are required to be

cooled to very low temperatures and typically this is accomplished by circulating

liquid helium through or around the devices. Therefore, the accelerator complex

Fig. 8.8 Example of a superconducting cavity inside a cryostat for the Large Electron Positron

Collider II (Courtesy: CERN, European Organization for Nuclear Research, Image Reference:

CERN-DI-9107011)

126 8 The Next Generation: Supersynchrotrons

would require large liquid helium (and liquid nitrogen – for cooling shields and

other components) plants and associated distribution systems. The accelerator(s)

are typically housed in underground tunnels and so these areas would have to be

serviced with elevators and people and material movers. A complex array of

electrical supplies, electrical busses, sensors, and sensor arrays are required. A

complex control, operation, and monitoring system is needed to operate the accel-

erator efﬁciently and safely. Finally, an enormous amount (hundreds to thousands

of gegabytes) of experimental and monitoring data would be gathered during all

phases of the operation and sophisticated data management, analysis, and visuali-

zation system with highly advanced information technology support (see Chap. 11).

Each of these support systems is large by any standards and require considerable

planning, design, fabrication, installation, and maintenance usually with impecca-

ble reliability.

One of the key concepts developed through years of managing such large project

is the Systems Engineering and Quality Assurance approach. This is well described

elsewhere in management books. But sufﬁce it to say, that like the space programs

requiring a reliability of 99.99999% or higher, the accelerator programs are an

example of human achievement in prec ision and reliability. This reliability is

achieved by ensuring that device and component requirements are threshed out to

meet the operation, the design is reviewed at every stage and judged with regard to

satisfying requirements. Risks involved in not meeting the requirement or in failure

are assessed; operational safety, feasibil ity of timely and on-budget fabrication are

assured; and a design description that ensures ease, quality, and accuracy in

fabrication. Once the design is ﬁnalized, the Systems Engineering approach freezes

the design and creates a conﬁguration management system consisting of all

Fig. 8.9 The ICHIRO5 superconducting RF cavity, designed by KEK, Japan, which achieved 33

MV/m electric ﬁeld at the Jefferson Laboratory (Courtesy: Thomas Jefferson National Accelerator

Laboratory, VA, USA)

Supporting Cast-Facilities and Systems Engineeing 127

documents pertaining to the design requirements and the design description. Any

change to the conﬁguration would then have to be reviewed by a group of scientists

for approval or denial. The fabrication and assembly of the devices follows this

conﬁguration and the process of manufacturing would incorporate all quality

assurance methods to ensure the required reliability and would also document all

processes and in-process testing results. The latter assists in tracing an offendi ng

device or component if there is a failure. Same approach is adopted in

commissioning and operation so that all steps are deliberated, reviewed, planned,

and controlled for conformance to the plan and safe and efﬁcient performance.

Tera on Terra: Tevatron to Large Hadron Collider

The ﬁrst of the supersynchrotrons was the Tevatron, the second highest in energy in

the world, which had all superconducting magnets, but a normal RF cavity. HERA

in DESY is the ﬁrst machine with both superconducting magnets and

superconducting RF cavity.

Enrico Fermi Would Be Proud

In 1972, the Fermi National Laboratory (FNAL) achieved the ﬁrst FODO cell based

synchrotron, to the great delight of the accelerator physicists around the world.

Following the Main Ring, it had nowhere to go because in 1976, the decision was

taken by the High Energy Physi cs Panel (HEPAP) and Department of Energy, in US

government, to build the next machine (a collider named ISABELLE) at the

Brookhaven National Laboratory (BNL). Fermilab Director Robert Wilson

resigned in protest. In 1978, Fermilab was merely “in drift.” But in 1978, the

ISABELLE project ran into trouble because the superconducting magnets had

serious technical problems. While BNL scrapped this project and proposed another,

Nobel Prize winner and the new director of FNAL, Leon Le derman, grabbed the

opportunity and on November 15, 1978, secured the funding for doubling the

energy of the Fermi Main Ring, using superconducting magnets. The new machine,

3.9 miles (about 6.1 km) in circumference has been continuously upgraded ever

since such that it has now achieved the 1 TeV energy in each of the proton,

antiproton beams (circulating in the same beam tube), deserving its new name

Tevatron. In early 1984, the Tevatron achieved the phenomenal 800 GeV energy

and in 1986, 900 GeV. In late 1986, proton antipr oton collisions were obtained at

1.8 TeV total energy. After several new results in part icle, FNAL achieved its claim

to fame by detecting the long predicted Top Quark (Fig. 8.10).

The program of developing superconducting magnets and an accelerator system

to accelerate particles at a rate of 50 GeV per second had been started in 1972. The

Tevatron has a beam tube with 7 cm diameter and a magnet bore of nearly 8 cm

diameter. There are 774 bending magnets, each 6.4 m long with a ﬁeld of 4.2 T, and

128 8 The Next Generation: Supersynchrotrons

216 quadrupole magnets around the ring, operating at about 4 k A. The total amount

of superconducting wire used in that magnet was 42,500 miles, an astounding

production achievement at that time. Twenty-four local refrigerating units were

employed initially and were upgraded later. The cold boxes are over 13 m tall, with

four 4,000 horsepower compressors pumping liquid helium at a rate of over 250 L/

min and correspondingly large liquid helium and nitrogen storage systems. The

other achievement of this facility was the production and storage of the antiproton

beam. Any one, who has watched SciFi episodes like Star Trek, knows that this is

the stuff of explosions, engine drives, and so on. Since any contact with matter

containing proton will cause annihilation of the antiproton, the particles have to be

produced, collected into a beam and cooled so that it does not bounce around in the

beam tube, and then steered (see Chap. 11).

In 1940, Enrico Fermi, perhaps on a whim, had proposed a 100 TeV machine

that would be placed around the earth, without even anticipating advances such as

the one in Tevatron, but he would be proud that the ﬁrst machine to reach the TeV

particle energy was built by a laboratory named after him. On September 30, 2011,

the Tevatron collided its last particle and the eminent physicist Helen Edwards

pressed the button to drain the particles from the machine was the last time.

HERA: Remarkable Marriage of Unequals

Perhaps the blessings from Hera, the Greek Goddess of marriage, was a part of this

very successful and remarkable superconducting synchrotron Hadron Elektron

Ring Alnalge (HERA) located in DESY, Hamburg, Germany. The same way, the

ﬁrst ever roofed temple to a Greek divinity was built for Hera, the ﬁrst ever fully

superconducting synchrot ron was named HERA. The remarkable marriage is the

Fig. 8.10 Aerial picture showing the Tevatron at Batavia, Illinois, USA (Illustrated with red line).

The larger ring is the Tevatron and the smaller one the Main Injector (Courtesy: Fermi National

Accelerator Laboratory, IL, USA)

HERA: Remarkable Marriage of Unequals 129

fact that this is the only collider which collides electrons and positrons with protons,

so that one gets a combination of high energy as well as the relatively less messy

collisions that come from colliding electrons or positrons. The proton beam nominal

energy is about 920 GeV, close to a TeV, but the electron energy has to be kept low

~27 GeV to keep synchrotron radiation effects small. The proton ring (see Fig. 8.11)

uses 422 superconducting dipoles which produce 4.7 T and 244 superconducting

quadrupoles. A standard full 47 m long FODO cell consists of 4 dipoles, 4

quadrupoles, and 4 sextupole error correction magnets. The total circumference is

about 6.4 km similar to the Tevatron. The injector ring PETRA with 2 km circumfer-

ence has normal magnets. But the particles are privileged to be accelerated by a

superconducting RF cavity which produced 5 MV/m. The electrons and protons are

stored in separate bunches and then made to collide. In 1998, the machine switched to

positron–proton collisions.

The hard work of technological development of not only superconducting

accelerator magnets and RF cavities, but also of massive superconducting detector

magnets has paid off. The days of superconducting synchrotrons with

superconducting dipole, quadrupole, and ﬁeld correction magnets and

superconducting RF cavities have come and there is no turning back. The new

higher temperature ceramic superconductors are already ﬁnding their place in leads,

busbars, etc., and operation of the superconducting synchrotron with high stability

at higher temperature may be in the future. Appropriately, the word “Super” has

been co-opted and particle colliders are now “Super Colliders.”

Fig. 8.11 HERA’s intersecting electron/positron and proton rings. The detector in the straight

sections is also shown

130 8 The Next Generation: Supersynchrotrons

Chapter 9

Linear Accelerators, the Straight Story

When it concerns science, (a) No new serious idea, even when not accepted in the

context when it was proposed, is ever wasted (b) the scientists do not give up easily.

This is the case with the linear accelerators. The straight accelerator was

proposed and proven in the early 1930s after Rolf Wideroe encountered many

hardships in imple menting his idea of a linear (straight) accelerator (see Chap. 5)

with a series of tubes, electriﬁed alternately positive and negative to coincide with

the arrival of the particles at the tubes. As particle energies increased, the machines

got longer and expensive, less efﬁcient, and the beam dynamics became more

complicated. The cyclotr on then showed that one could achieve high energies in

a more compact machine and for a while, there was less interest in linear

accelerators (LINACs).

Sloan Linac

Even so, in parallel, David Sloan and Lawrence agreed that LINACs were simpler

and continued to reach for higher energies in them. Lawrence had spent two

summers in 1929 and 1930 working on ionization phenomena. He and Sloan used

a robust GE oscillator from those days, operating it aggressively at 11 kV where as

it was rated for only a little over 2 kV and built a 90-keV accelerator with nine

electrodes using Wideroe’s concept. A 200-keV machine was then built and

quoting this success, Lawrence, the king of circular accelerators, cham pioned the

cause of LINACs in his report to the National Academy of Sciences. In May 1931,

Sloan crossed the one million volt milestone when he achieved 1.26 MeV mercury

ions in a LINAC using 30 electrodes and 10 MHz RF source (Fig. 9.1). (It must be

noted that the choice was limited to heavy ions which are slow for a given target

energy, because high velocities meant high frequencies, high power, sources for

which had not been developed). This was a phenomenal 42 kV per electrode and is

impressive even by today’s standards. The LINAC was instrumental in not only

crossing the million volt barrier but also in providing exper ience with a great

R. Jayakumar, Particle Accelerators, Colliders, and the Story of High Energy Physics,

DOI 10.1007/978-3-642-22064-7_9,

Springer-Verlag Berlin Heidelberg 2012

131