DA! !NE DEVELOPMENTS FOR THE KLOE-2 EXPERIMENTAL RUN
C. Milardi for the DA!NE Commissioning Team*, INFN-LNF, Frascati, Italy.
Recently the peak luminosity achieved on the DA!NE
collider has been improved by almost a factor three by
implementing a novel collision scheme based on large
Piwinski angle and Crab-Waist. This encouraging result
opened new perspectives for physics research and a new
run with the KLOE-2 detector has been scheduled to start
by spring 2010. The KLOE-2 installation is a complex
operation requiring a careful design effort and a several
months long shutdown. The high luminosity interaction
region has been deeply revised in order to take into
account the effect on the beam caused by the solenoidal
field of the experimental detector and to ensure
background rejection. The shutdown has been also used to
implement several other modifications aimed at
improving beam dynamics: the wiggler poles have been
displaced from the magnet axis in order to cancel high
order terms in the field, the feedback systems have been
equipped with stronger power supplies and more efficient
kickers and electrodes have been inserted inside the
wiggler and the dipole vacuum chambers, in the positron
ring, to avoid the e-cloud formation. A low level RF
feedback has been added to the cavity control in both
DA!NE  the Frascati electron positron collider
began steady operations in 2001 and in the next seven
years provided high K meson rates, at the energy of the !
resonance, to three different experiments, which logged
~ 4.4 fb-1 total integrated luminosity in dedicated runs.
The KLOE detector collected the largest part of these
data, 3.0 fb-1. In the same period the collider
performances have been significantly improved by
several progressive upgrades and a wide program of
machine measurements and studies has been undertaken
aimed at pointing out the factors limiting the maximum
achievable luminosity. This activity largely contributed to
define a proposal for an inventive collision scheme based
on large Piwinski angle and Crab-Waist (CW)
compensation of the beam-beam interaction .
The novel approach to collision has been implemented
on DA!NE  during a six months shutdown already
planned to install a compact detector without solenoidal
field offering a simplified environment to test the new
BINP SB RAS, Novosibirsk; S. Bettoni, CERN, Geneva.
The results obtained  are remarkable: the luminosity
achieved a 4.53 ·1032 cm-2 s-1 peak value with a maximum
daily integrated luminosity of ~15 pb-1, eventually after
18 months of steady operation ~2.4 fb-1 have been
delivered to the experiment, see Fig. 1. At the same time
the effectiveness of the Crab-Waist collisions has been
unquestionably proved by several independent tests and
beam measurements .
Figure 1: Peak (dots) and integrated (line) luminosity
acquired on DA!NE by the four different experiments.
This relevant jump ahead has been attained despite the
collider performances were still affected by several
limiting factors. The maximum current storable in the
positron beam was 1.2 A due to e-cloud instability. The
beam lifetime at maximum current in collision was
~ 550 sec. Machine uptime was affected by hardware
reliability due to aged equipment.
MOTIVATION FOR THE KLOE-2 RUN
The interest in undertaking a new physics run with an
upgraded KLOE detector, KLOE-2 , is motivated by
physical considerations. These require delivering a high
statistics data sample to a detector having an improved
sensitivity; moreover compatibility of the CW collision
scheme with a large detector with a solenoidal field will
KLOE is a multipurpose experiment devoted, mainly,
to study decays of K mesons as well as several hadronic
physics and low energy quantum chromodynamics
phenomena. The detector consists of a large cylindrical
drift chamber, ~3.5 m long and 2 m in radius, surrounded
by a lead-scintillating fibre electromagnetic calorimeter.
A superconducting coil around the detector provides a
magnetic field of 0.52 T. With respect to the original
design, KLOE-2 has additional detector layers, including
new tracking and calorimeter devices. It also extends its
investigation capabilities to the study of gamma-gamma
reactions, by means of dedicated detectors  tagging the
scattered electron and positron, typical of those events.
The maximum daily integrated luminosity measured
with Crab-Waist collision is a factor 50% higher than the
best achieved during the past KLOE run . This value is
*D. Alesini, M.E. Biagini, C. Biscari, R. Boni, M. Boscolo, F. Bossi,
B. Buonomo, A. Clozza, G. Delle Monache, T. Demma,
E. Di Pasquale, G. Di Pirro, A. Drago, M. Esposito, A. Gallo,
A. Ghigo, S. Guiducci, C. Ligi, F. Marcellini, G. Mazzitelli,
C. Milardi, L. Pellegrino, M. Preger, L. Quintieri, P. Raimondi,
R. Ricci, U. Rotundo, C. Sanelli, M. Serio, F. Sgamma, B. Spataro,
A. Stecchi, A. Stella, S. Tomassini, C. Vaccarezza, M. Zobov,
INFN-LNF, Frascati; E. Levichev, S. Nikitin, P. Piminov, D. Shatilov,
consistent with the average daily integrated luminosity
per month (~10 pb-1) delivered in the final part of the run
operating the collider in a moderate injection regime to
avoid excessive background in the detector. Tests carried
out adopting a continuous injection regime, compatible
with the KLOE-2 data-taking, yielded a hourly integrated
luminosity L!1 hour ~1.0 pb-1 . Scaling this best integrated
luminosity measured over two hours it is reasonable to
expect a daily integrated luminosity larger than 20 pb-1,
and assuming 80% collider uptime a monthly integrated
luminosity of "0.5 fb-1.
THE KLOE-2 INTERACTION REGION
Integrating the high luminosity collision scheme with
the KLOE-2 detector introduces new challenges in terms
of Interaction Region (IR) layout and optics, beam
acceptance and coupling correction. The IR magnetic
layout, see Fig. 2, has been designed in order to maximize
the beam stay clear letting the beam trajectory pass as
much as possible through the center of the magnetic
elements. The IR optics provides the prescribed low-#
parameters at the IP (#x = 0.265 m, #y = 0.0085 m, $x = $y
= 0.0, %x = %'x = 0.0), matching at the same time the ring
original layout in the arcs, as well as the phase advance
between the IP and the CW sextupoles. Transverse
coupling compensation is achieved by means of two anti-
solenoids for each beam, and by rotating some
quadrupoles in the IR.
All these aspects have been studied in details and the
results are presented a discussed in a dedicated
Figure 2: The KLOE-2 detector and its interaction region.
IMPROVING BEAM DYNAMICS
The limiting factors observed during the run for the
CW test required many developments related to ring
impedance, lattice non-linearities, instabilities mitigation
and e-cloud formation. These are meant to achieve higher
currents, especially for the positrons, with better lifetime
and dynamic aperture.
The wiggler magnets have been modified once again to
pursue the fight against non-linear terms in the magnetic
field (B) started in 2001 .
The longitudinally and horizontally shimmed plates
added on the poles in 2004 , which led to an
improvement by a factor 2 in the dynamic aperture and in
the energy acceptance , have been removed and a new
approach has been defined in order to:
reduce the higher order components in B,
increase the maximum B at a given current,
cancel the field integral on the beam trajectory.
It consists in changing the pole disposition in a such a
way to keep the beam trajectory as much as possible
centered with respect to the pole face . A
mathematical analysis shows how even terms in B, having
the same symmetry as the field, still approximately cancel
each other because the B sign changes from pole to pole.
On the opposite odd terms, having the same symmetry as
the B derivative tend to cancel inside each pole. The new
configuration with shifted poles has been tested on a spare
wiggler with a 37 mm pole gap, the same used for
shimmed pole wigglers. A 1.726 T peak field has been
measured at 550 A; reducing the current at 450 A the field
becomes 1.644 T, a value still higher than the one
achieved at 550 A, during the past DA!NE operations. A
further improvement has been obtained by short-
circuiting one out of the five windings in the terminal
poles coils; in fact the field integral of the whole wiggler
can be perfectly compensated powering in series the
central and terminal coils. This approach allows to get rid
of eight power supplies reducing the wall plug power.
The eight wigglers installed on the DA!NE collider have
been all removed, modified and measured. The residual
field integral has been experimentally compensated for
each device below 1 Gm by tuning the end pole clamps
Each DA!NE ring is equipped with three independent
sophisticated bunch-by-bunch feedbacks.
In the longitudinal plane a new system is going to
replace the 15 years old one. Its main characteristic
consists in being affected by a lower noise level in
detecting and damping the beam longitudinal oscillations.
The vertical feedback will be equipped with a 12-bit ADC
based hardware in order to reduce the impact of the
quantization noise on the system gain. Concerning the
horizontal feedback, the one installed on the positron ring
deserves special attention; in fact it must keep under
control a fast horizontal instability due to the e-cloud
limiting the maximum positron current and, in turn, the
peak luminosity. During the past run this effect has been
mitigated by adding a second horizontal feedback system
kicking the beam by using two out of the four injection
kickers striplines powered by spare hardware. In this way
~1.2 A of positron have been stored in a stable beam with
the design transverse beam dimension .
In view of the KLOE-2 run a more systematic approach
has been undertaken to cope with the e-cloud driven
instability. The transverse horizontal feedback power has
been doubled (500 W now) providing ~40% increase in
the kick strength which scales as the square root of the Download full-text
power itself. The horizontal feedback kicker has been
replaced with a device with a double stripline length and
reduced plate separation, providing larger shunt
impedance at the low frequencies typical of the unstable
modes. Moreover the kicker has been moved in a lattice
position having a higher #x value.
However the best way to overcome the threshold in the
positron current consists in avoiding the e-cloud
formation; for this reason stripline electrodes have been
designed and inserted in wiggler and dipole vacuum
chambers of the positron ring relying on beam studies
 showing a clear dependence of the e-cloud instability
grow-rate on the beam orbit in those magnets. The
e-cloud electrodes  consist of Cu strips, equipped with
dielectric (shapal) contacts, having 2 mm total thickness,
50 mm width and 1.5 m length. An almost complete
neutralization of the emitted photo-electrons  is
expected to be achieved by applying to the electrodes a
moderate dc voltage of the order of 0.5 kV.
Figure 3: The e-cloud clearing electrode installed inside a
dipole beam pipe, the strip shape is bent to follow the
A direct RF feedback system in the low level RF is
being developed. This allows to reduce the cavity
detuning angle, increasing the overall efficiency and
limiting the reduction of the coherent ‘0-mode’
synchrotron frequency with beam current.
The leftover old-style bellows relying on shields
implemented by contiguous mini bellows have been
replaced with new ones having lower impedance and
providing long lasting shield contour uniformity when
The ion clearing electrodes still present in the electron
ring and no longer used have been removed.
The Collimator rectangular vacuum chambers, (20 mm
high and 90 mm wide), have been replaced by square
ones (55 mm) to reduce their contribution to the ring
impedance. In this way it becomes also possible to move
the blades closer to the beam improving their
effectiveness in intercepting the background otherwise
hitting the experimental detector.
OTHER HARDWARE DEVELOPMENTS
Several other systems have been improved to fulfil the
requirements set by the KLOE-2 detectors and to achieve
more reliable operating conditions.
The cryogenic plant has been maintained and provided
with new transfer lines to cool the 4 superconducting
anti-solenoids installed in the IR.
The Linac gun operating since 15 years has been
replaced with a new one. A new accelerating section is
going to be added at the end of the machine to make
operation less critical during the positron beam injection.
The new kicker developed for the transverse horizontal
positron feedback has been also used as a beam dumper.
It has been installed in the opposite section with respect to
the IR and will allow to dump the beam in a controlled
way reducing the radiation level in the area and avoiding
dangerous detector trips.
The Control System functionality has been extended to
all new collider elements, and the system itself is
undergoing a deep upgrade; in fact the low level control
boards — based on 68040 processors and MacOS
operating system — are going to be progressively
replaced by Intel boards running under Linux.
The developments on the DA!NE collider have been
completed in a six months shutdown. The KLOE-2
experiment is in place as well as its new IR. The detector
has been already cooled down to 4.4
accelerator components and subsystems have been
modified looking for higher and more stable currents
having longer lifetime and for more reliable operations.
The DA!NE and KLOE-2 commissioning will start in
the next few weeks.
The activities on DA!NE and KLOE-2 have been
successfully finished on schedule thanks to the
commitment of the Accelerator and Technical Divisions
 G. Vignola et al., Frascati Phys. Ser. 4:19-30,1996;
 P. Raimondi et al., physics/0702033;
 C. Milardi et al., arXiv:0803.1450v1.
 C. Milardi et al., Int.J.Mod.Phys.A24:360-368, 2009.
C. Milardi et al., ICFA Newslett.48:23-33, 2009.
M. Zobov et al., ICFA Newslett.48:34-44, 2009.
M. Zobov et al., Phys.Rev.Lett.104:174801, 2010.
 A. Drago et al., THOBRA01, this conference.
 F. Bossi, Journal of Physics: 171 (2009) 012099.
 KLOE-2 Collab., LNF - 10/14(P) April 13, 2010.
 A. Gallo et al., EPAC06, p. 604-606.
 C. Milardi et al., PAC09, MO4RAI01.
 C. Milardi et al., TUPEB005, this conference.
 C. Milardi et al., PAC2001, p. 1720.
 M. Preger et al., DA!NE Technical note, MM-34, 04.
 C. Milardi et al., e-Print: physics/0408073.
 S. Bettoni et al., THPE065, this conference.
 A. Drago, PAC09, TH6REP072.
 A. Drago, ICFA Newslett.48:50-58, 2009.
 D. Alesini et al., TUPEB002, this conference.
 T. Demma et al., TUPD036, this conference.