Studies of the doubling of the Frascati Beam-Test Facility (BTF) line
B. Buonomo, C. Di Giulio, L. Foggetta
Istituto Nazionale di Fisica Nucleare, Laboratori Nazionali di Frascati
P. Valente
Istituto Nazionale di Fisica Nucleare, Sezione di Roma
Supported by the H2020 project AIDA-2020, GA no. 654168.
Motivation
The commissioning of the Beam-Test line [1] has been carried on in 2002, and in 2003 a first running with experimental groups was already possible, although with the target and selecting dipole DHSTB01 still on the main injection line and thus with a heavily reduced duty-cycle.
After the separation of the BTF transfer line in 2004, the facility has been providing access continuously to a large number of experimental teams, with an average number of beam-days close to 200/year. The full statistics for the last five years of BTF operation of the total number of beam-days and working hours (both scheduled and realized) is reported in Fig. 1.
Figure 1: Total number of beam-days assigned (left) and beam-hours, scheduled and realized (right), for the last five years of BTF operation.
The facility usually allocates slots of 1 week, Monday to Monday, and operates 24/7. More complex experimental setups of course required much longer beam-periods, up to the approximately five months of total allocation for the AGILE satellite pay-load calibration in 2005 (beam was not delivered 100% of this period, however).
The total number of groups for each of the last five years is shown in Fig. 2. Approximately 1/3 of the 150 average users/year comes from a not Italian institution.
Figure 2: Total number of user groups accessing the BTF in the last five years of operation.
The overbooking factor, i.e. the ratio between the required and assigned beam time, has ranged among about 120% up to 150%, depending on the year.
The large majority of runs performed at the BTF were related to the high-energy physics and astro-particles communities, mainly with the purpose of testing, characterizing and calibrating particle detectors. As a consequence, most of the runs are performed at low intensity (few particles regime), exploiting the full range of beam energy and spot parameters listed in Tab. 1.
Parameter / “Parasitic” operation / Dedicated operationWith target / Without target / With target / Without target
Particle species / e+ or e-
Selectable / e+ or e-
Same as DANE / e+ or e-
Selectable
Energy (MeV) / 25–500 / 510 / 25–700 (e-/e+) / 250–730 (e-)
250–530 (e+)
Energy spread / 1% at 500 MeV / 0.5% / 0.5%
Rep. rate (Hz) / Variable between 10 and 49, Depending on DANE status / 1–49
Selectable
Pulse duration (ns) / 10 / 1.5–40
(In 0.5 ns steps)
Intensity
(particles/bunch) / 1–105
Depending on energy / 103–1.5 1010 / 1–105
Depending on energy / 103–3 1010
Max. average flux / 3.125 1010 particles/s
Spot size (mm) / y)×x)=0.5–25×0.5–55
Divergence (mrad) / 1.5-2
Table 1: BTF beam parameters summary.
Practically all kind of detectors presently used in the HEP community and all major collaborations in this field have profited at some time of the BTF beam:calorimeters (tile, “spaghetti”, crystal, digital and double readout, etc.), scintillating detectors (segmented, timing detectors, fiber trackers), drift chambers, RPC, micro-pattern gas detectors (GEM, planar, mini-TPC and cylindrical, and MSGC), hybrid gas/Silicon detectors, diamond detectors, silicon pixels (also monolithic), silicon micro-strips, fluorescence detectors, nuclear emulsions, Cherenkov threshold and RICH, etc.
A smaller fraction of the beam shifts have been devoted to high intensity measurements or for small, dedicated experiments looking at specific electromagnetic processes, ranging from the measurement of the absolute fluorescence yield of air and other gases, to the detection of microwave emission in air showers, from the thermo-acoustic expansion of a super-conducting resonant antenna, to the characterization of innovative beam diagnostics (both position sensitive, like segmented diamonds, multi-channel plates, and proportional counters).
Another small, but significant fraction of the beam time has been devoted to the production of tagged photons, especially for the calibration of astro-particles detectors, with the very important example of the pre-launch calibration with photons of the payload of the AGILE gamma ray astronomy satellite.
Since 2008, one or two shifts/year have been devoted to the characterization and test of the BTF photo-production neutron source.
Finally, in the last few years, there has been a growing interest in using the BTF beam, especially selecting positrons, for detecting and measuring effects related to channeling in crystals, both bent and straight ones, at energy significantly lower with respect to the other facility used by this community (in Europe typically the CERN SPS at 450 GeV), or for the detection of parametric radiation from crystals, characterization of crystal ondulators, etc.
At the end of 2015 INFN has formally approved a new experiment, PADME [2], exploiting the DANE LINAC and the BTF beam-line for producing a positron beam with well defined and easily tunable parameters, mainly with the objective of extending the exclusion – or even better, aiming at the discovery – of a dark photon in a relevant part of the mass and coupling range interesting for the muon (g-2) problem.
The experiment aims at collecting 1013 electrons on target, after a quick R&D and construction phase (approximately two years) of the setup, built around a large gap dipole magnet instrumented with segmented scintillator detectors, for measuring the momentum of positrons irradiating a background Bremsstrahlung; and a high-performance electromagnetic calorimeter, made of BGO crystals (reused from the former L3 experiment), for measuring the angle and energy of photons produced at the target, for the rejection of the e+e- and backgrounds.
Recently a detailed proposal [3] for a substantial upgrade of the BTF has been put forward, and part of the improvement is supported in the framework of the AIDA-2020 project.
This technical note addresses the optics and the optimization of the proposed new beam-lines, using simulation codes like G4-beamline [4] and MAD-X [5]. In order to validate the results, the simulation of the existing beam-line is compared with available experimental data.
Present BTF line layout and simulation
The BTF has been designed as a part of the DAΦNE complex: it is composed of a transfer line driven by a dipole magnet allowing the diversion of electrons or positrons, usually injected into to the damping ring, from the high intensity LINAC towards a dedicated experimental hall. The facility can provide runtime tuneable electron and positron beams in a defined range of different parameters, depending on the choice of one of the following two main operation modes:
-“Single particle” regime: in this operation mode, a step Copper target, allowing the selection of three different radiation lengths (1.7, 2 or 2.3 X0), is inserted in the initial portion of the BTF line for intercepting the beam (TGTTB01). This produces a secondary beam with a continuous full-span energy (from LINAC energy down to few MeV) and intensity, down to a regime in which the particle multiplicity per bunch follows a Poisson distribution.
-High-intensity beam extraction: the LINAC beam is directly steered in the BTF hall with a fixed energy (i.e. the final LINAC one) and with a reduced capability in multiplicity selection (typically from 1010 down to 104 particles/bunch) by means of collimating Tungsten slits.
A dipole magnet (DHSTB01) steers the beam towards the BTF experimental hall and also has the task of defining the momentum of the particles, either the primary LINAC beam or secondary particles emerging from the BTF target, with the momentum band defined by a downstream horizontal collimator. In the actual configuration (since 2004) a three-way switch-yard can alternatively drive the beam to:
-The straight line, connecting the end of the LINAC with the damping ring (by means of a 45° pulsed dipole);
-The BTF line, with the beam attenuation and selection system, bent by 3° by means of a small dipole, fed by a pulsed supply (DHPTB101);
-The spectrometer at 6°, for the precise measurement of the beam momentum at the end of the LINAC, by means of a 60° static dipole coupled to a metallic strip segmented detector.
Due to the fact that the DHSTB01 is a 45° sector magnet, in order to take into account the 3° of the BTF line with respect to the straight transfer-line, it is used as a 42° bending, towards the BTF experimental hall. For this purpose, the current setting is re-scaled and the magnet is rotated by 1.5°.
The collimators are two couples of horizontal and vertical Tungsten slits:
-SLTTB01 (vertical), upstream of the attenuating target TGTTB01;
-SLTTB02 (horizontal), upstream of the energy-selecting dipole DHSTB01;
-SLTTB03 (vertical) and SLTTB04 (horizontal) downstream the dipole DHSTB01.
The yield of secondary electrons (or positrons) can be optimized by the FODO quadrupole doublet at 1 m from the BTF target (QUATB101 and QUATB102). Only electrons (or positrons) with energy around the selected value Esel will be actually transported to the downstream portion of the BTF line, depending on the current set on the 42° dipole DHSTB01.
The transfer line downstream the energy selection and the collimators, at 45° with respect to the LINAC line, drives the beam to the BTF experimental hall. In this last portion of the transfer line two FODO quadrupole doublets are present for the optimization of the beam optics: QUATB01-02 (in the LINAC tunnel) and QUAT03-04 (in the BTF hall). Finally, another 45° dipole (DHSTB02) is used to bend the beam along the main axis of the experimental hall. The beam-pipe has also a straight line exit (with DHSTB02 off), for diagnostics purposes.
The general layout including all the magnetic elements is shown in Fig. 3.
Figure 3: General layout of the present BTF line.
The simulation of the existing beam-line, starting from a realistic beam with the nominal emittance parameters at the output of the LINAC, i.e. 1 mm mrad for electrons, with a typical beam charge of 1 nC/pulse, includes the TGTTB01 attenuating target, the effect of the two collimator pairs, and all the magnetic elements (dipoles and quadrupoles), with the actual range of operational currents.
The G4-beamline simulated setup is shown in Fig. 4.
Figure 4: Present BTF line setup in G4-beamline.
Present BTF line layout and simulation
The present BTF layout was simulated with the G4-beamline software by using the Message Passing Interface (MPI). G4beamline is an open-source particle tracking simulation software based on Geant4, optimized for simulating beam-lines and hosted in the BTF VM cluster. The G4Beamline embedded MPI method uses the 24 virtual cores cooperating in a single computation unit: such environment allows managing up to 5·108 primary particles with our computing resources.
The simulation preparation consisted in few steps: the setup of the computing VM cluster and the collection of the BTF active elements and detectors data, in order to reproduce the actual transfer line performances.
The precise measurement of the magnets position was done during the annual alignment check of the BTF transfer line: the measurements were consistent with the CAD outline provided by the reviewed LNF technical division.
The DHPTB101 was simulated using the G4-beamline elements “genericbend”, the DHSTB001 and DHSTB002 using the “idealsectorbend” and the quadrupoles using the “genericquad”. Thanks to the detailed options describing each single element, we were able to implement the measured relationship between the current and the magnetic field gradient for dipoles and quadrupoles of the BTF transfer line. This feature was fundamental for the DHPTB101, a pulsed dipole, actually a quadrupole modified as a dipole obtained by connecting in series the upward e downward coils, respectively [6]. In order to speed up the simulation, we decided to kill the off-track particles when hitting the outer sections of pipes, the magnets poles or the shielding.
In order to validate the simulation, we have implemented the typical setup of the present BTF line, including both the actual settings of all active elements (i.e. magnets, LINAC primary beam), and the positions of passive ones (i.e. scrapers, shielding, pipes, detectors). We have validated the simulation in both the regimes; transporting the full LINAC primary beam, and the BTF secondary one produced on the target. In both configurations we have compared the simulation output with real data, acquired in our facility in dedicated runs, mainly using the beam transverse diagnostics obtained from the WIDEPIX® FitPix silicon detector and fluorescent flags, and the particle multiplicity measured by the BTF calorimeter. The simulation essentially confirms the expected beam parameters: less than 1 mrad divergence and sub-millimetric beam spot sizes measured in BTF, as soon as one includes the effect of the multiple scattering on the exit Beryllium window (0.5 mm thick). The previously described optimization, i.e. killing the outgoing particles, seems not relevant in the last part of the tracking, since we notice, also in the simulation, the scrapers background in the last FitPix-like detector.
The same description used for the present line has been implemented in the simulation of the new BTF lines, of course including the new elements and new positions.
The two new lines in the proposed new configuration of the BTF lines are shown in the schematic 2D and 3D layouts of Figg. 5 and 6.
The expected performances of the new BTF line configurations are shown in Fig. 8, where the performance obtained in the simulation are shown. The minimization procedure was implemented in the G4-Beamline simulation, in order to obtain the currents value (I) optimizing the beam performances for the quadrupole FODO cell.
The obtained currents are within the specifications of the existing quadruples and relative power supplies.
Figure 5: Layout of the new BTF lines: in red the new elements.
Figure 6: 3D layout of the new BTF lines.
The G4-beamline simulated layout is shown in Fig. 7.
Figure 7: New BTF lines setup in G4-beamline, showing the new elements and the detecting volumes (DIAG1 to DIAG5). The line coincides with the present line up to the DIAG3 position. The two final detecting volumes are OUT1 (BTF1 line, in the present area) and OUT2 (BTF2 line, in the new hall).
Figure 8.a: Example of transverse beam spots for 510 MeV electrons in the first part of the BTF1 and BTF2 line, at the first three diagnostics positions: (from left to right) immediately after the TGTTB01 target (DIAG1) with a widespread and fully symmetric x-y distribution, immediately downstream of the first horizontal collimator SLTTB02 at the entrance of the energy selecting dipole DHSTB01 (DIAG2), and downstream of the second horizontal collimator SLTTB04 (DIAG3), showing the dispersive effect of the bending.
Figure 8.b: Beam spot for 510 MeV electrons immediately upstream of the splitting of the two BTF lines and downstream of the new quadrupole triplet (DIAG4), using the same dataset of Fig. 3.21.a.
Figure 8.c: Example of transverse beam spots for 510 MeV electrons at the BTF1 line exit (OUT1 position), and beam divergence in x (=0.7 mrad) and y coordinates (y=1.1 mrad), with the same dataset as Figg. 3.21.a and 3.21.b.
Figure 8.d: Example of transverse beam spots for 510 MeV electrons at the BTF2 line exit (OUT2 position), and beam divergence in x (=2 mrad) and y coordinates (y=0.8 mrad), with the same dataset as Figg. 3.21.a and 3.21.b.
The simulation of the present BTF line and the new line was implemented also in the MAD-X simulation tool to simulate the beam dynamics and to optimize the beam optics.
The value of currents obtained for the dipoles and for the quadrupoles with the MAD-X minimization are consistent with the value of the actual data-set for the BTF present line and are well within the specifications of the actual quadrupoles for the new BTF line (dubbed BTF1 and BTF2, respectively, following the convention shown in Fig. 7).
The beam parameters with those currents settings are shown in Fig. 9.a and 9.b, where in particular the Twiss parameters are shown.
Figure 9.a: From left to right: MAD-X simulation results for the function and dispersion present BTF, new lines “BTF1” and “BTF2” respectively.
Figure 9.b: From left to right: MAD-X simulation results for the envelope (x and y) of present BTF, new lines “BTF1” and “BTF2” respectively.
Acknowledgements
We are especially grateful to Riccardo Gargana, whose contribution in the VM environment setup was crucial for our simulations, and Giancarlo Sensolini for the very detailed information about elements actual positions, as well as on the approximations of the existing layout of the real beam-line.
References
[1] G. Mazzitelli et al., Nucl. Instrum. Meth. A 515, 516 (2003).
[2] M. Raggi, V. Kozhuharov and P. Valente, EPJ Web Conf. 96, 01025 (2015).
[3] P. Valente et al., Linear Accelerator Test Facility at LNF Conceptual Design Report, INFN-16-04/LNF.
[4] G4-beamline –
[5] MAD-X –
[6] G. Mazzitelli, M.A. Preger, C. Sanelli, F. Sgamma, P. Valente, Technical noteBFT-1, 18/02/2003