Neutron sensitive beam loss monitoring system for the European Spallation Source linac

The European Spallation Source (ESS), currently under construction in Lund, Sweden, will be a neutron source based on a partly superconducting linac. The ESS linac will be accelerating protons to 2 GeV with a peak current of 62.5 mA and ultimately delivering a 5 MW beam to a rotating tungsten target for neutron production. For a successful tuning and operation of a linac, a beam loss monitoring (BLM) system is required. BLM systems are designed to protect the machine from beam-induced damage and unnecessary activation of the components. This paper focuses on one of the BLM systems to be deployed at the ESS linac, namely the neutron sensitive BLM (nBLM). An overview of the ESS nBLM system design will be presented. In addition to this, results of the tests performed with the full nBLM data acquisition chain will be discussed. These tests represent the first evaluation of the system prototype in a realistic environment. They served as an input to initial study of the procedure for neutron detection algorithm configuration discussed in this contribution as well.

Figure 1

A block diagram of the ESS linac [1, 2]. The normal-conducting (NC) linac is composed of an ion source, radio frequency quadrupole (RFQ), drift tube linac (DTL), low (LEBT) and medium energy beam transport (MEBT) sections. Spoke resonators, medium beta and high beta elliptical cavities together with high energy beta transfer (HEBT) line constitute the superconducting (SC) linac. Red and blue color refer to NC and SC parts of the linac, respectively.

Figure 2

Detector and cabling layout for DTL tank 5 and initial 7 subsections of the Spoke section. Each Spoke subsection consists of two parts, a so called linac warm unit (LWU) with a pair of quadrupole magnets and a cryomodule housing two spoke cavities. Light and dark blue squares, respectively, indicate nBLM-S and nBLM-F detector locations along these linac sections. Which AMC, HV and LV card number as well as LV card channel number each detector connects to is represented with green or red numbers stated above the detector. Here green and red color respectively mark odd and even detector group type. For example, detectors F19, S19, F21, S21, S23, and F23 belong to one of the odd groups and connect to AMC card 7, HV card 1 and the fourth channel on LV card 1. Similarly, detectors F20, S20, F22, and S22 belong to one of the even groups and connect to AMC card 8, HV card 2, and fourth channel on LV card 2. Shade of green and red color indicates different BEE racks, namely AMC cards 9 and 11 sit in the same rack, while card 8 is located in a different rack. Gas loops for circulating the gas through a series of detectors are indicated with grey and black lines representing the long haul and flexible tubes, respectively. Here “IN” refers to a distribution and “OUT” to a return long haul gas pipe. For example, the fourth gas loop circulates the gas through detectors in the first 6 Spoke subsections.

Figure 3

Schematic view of the nBLM system architecture consisting of detector modules (detectors with front-end electronics—FEE), back-end electronics (BEE) for data acquisition (DAQ), gas system, power supply (PS) system providing high (HV) and low (LV) voltage to detector modules and EPICS-based [22] control system. The system connects to the ESS machine protection (MPS [20]) and ESS timing [21] systems.

Figure 4

Simplified block diagram of the data processing chain for one nBLM detector. DoD refers to the data that can only be retrieved on demand, for explanation see text at the end of Sec. 5.

Figure 5

Sketch of a neutron event with NDA settings and some of the pulse characteristics.

Figure 6

Distribution of number of IEs over amplitude collected at LINAC4 in 2018. Blue histogram represents reconstructed data, while black refers to the data as recorded.

Figure 7

Distribution of reconstructed IEs over amplitude and time inside the LINAC4 machine cycle period for an 8-hour long data run.

Figure 8

Example of signals recorded in raw data during rf pulse (top) and a magnified view of one of the $\gamma$ signals (bottom).

Figure 9

Distribution of reconstructed IEs over amplitude and time in LINAC4 machine cycle period for data collected during the controlled loss experiment in 2019 data-taking campaign.

Figure 10

Projections of the histogram from Fig. 9 on the time axis for different threshold cuts on the minimum event amplitude. Orange scale on the right side of the plot refers to the scale for the measurement marked with orange line. This line represents the cavity forward rf amplitude of the DTL tank 1 as measured by the LINAC4 operations team.

Figure 11

Noise distributions extracted from raw data stream collected at different times and with the two voltage configurations used during the 2019 data-taking campaign. Noise RMS values are indicated in the legend.

Figure 12

Amplitude distributions extracted during the controlled loss experiment (red) and during an overnight data acquisition where beam was not present (blue). Dark and light colors refer to distributions extracted with and without cut on event start time applied, respectively. The dashed lines in the $\gamma$ part indicate result of fitting an exponential shape to this region, with solid line marking range of the fit. The red dashed line in the neutron region is drawn to guide the eye.

Figure 13

Amplitude distributions extracted during the controlled loss experiment (red) and during an overnight data acquisition where beam was not present (blue). The distributions are normalized to give counts/s in each amplitude bin. Dark blue dashed line indicates estimated contribution of the $\gamma$ part. Orange histogram represents the red distribution with $\gamma$ correction applied.

Figure 14

Relative efficiency extracted from controlled loss (red) and no beam data (blue). Dashed lines indicate exponential functions fitted to the data in order to extract intersection point as the $S_{\mathrm{NDT}}$ value.

Figure 15

Distribution of IE $t_{\mathrm{TOT}}$ over event start time inside machine cycle for data collected during the controlled loss experiment.

Figure 16

Correlation between IE amplitude and $t_{\mathrm{TOT}}$ for the data collected during the controlled loss experiment.

Figure 17

Estimated neutron rates (green and blue) during controlled loss experiment together with the differential BCT signal (orange). See text for details.

Figure 18

Example of estimated neutron rates during a typical day with LINAC4 commissioning activities focused on sections further downstream of the nBLM detector.

Figure 19

Example of estimated neutron rates (top) during LINAC4 commissioning activities focused on phase scans of cavities upstream of the nBLM detector. See text for details.