CERN Accelerating science

: : Caption not extracted

: (a) Cut-away view of the ATLAS calorimeter system. The TileCal consists of a barrel and two extended barrel sections. The sections of the ATLAS liquid-argon (LAr) calorimeters are also indicated. (b) An illustration of the mechanical assembly and optical readout of a single tile calorimeter module. A total of 256 such modules comprise the full tile calorimeter. Source tubes are used to circulate a $^{137}$Cs radioactive source contained in a capsule for calibration purposes. : Caption not extracted

The layout of the TileCal cells with $\eta \ge 0$, denoted by one or two letters (A to E) plus an integer number. The layer A is closest to the beamline.The naming convention is repeated for cells with negative $\eta$. The long barrel (extended barrel) cells are shown at the left (right).

The signal paths for each of the three calibration systems used by the TileCal. The signal produced by particles from collisions is denoted by the thick solid line, and the path taken by each of the calibration systems is shown with dashed lines.

The reference pulse shapes for high gain (dotted curve) and low gain (solid curve), shown in arbitrary units~\cite{TCAL-2010-01}.

:

: : (a) The reconstructed time of laser events as a function of the luminosity block. Data from six channels belonging to the same digitizer are superimposed. The timing jump lasted the entire duration of the run and all events are centered around $+\SI{15}{ns}$. The dashed line indicates the expected mean time value. (b) The 2D histogram shows the average channel time in physics events on a colour scale as a function of module number ($x$-axis) and channel number ($y$-axis).

:

: : (a) The reconstructed time of laser events as a function of the luminosity block. Data from three channels belonging to the same DMU are superimposed. The majority of events, centred around zero, are well timed in. The events affected by the the bunch-crossing offset are centred at $+\SI{25}{ns}$. (b) The reconstructed time in physics events in the same three channels with (corrected) and without (original) applying the algorithm mitigating the bunch-crossing offset events. The algorithm significantly reduces events centred around $+\SI{25}{ns}$.

:

: : The charge injection system constants ($C_{\mathrm{ADC}\to \mathrm{pC}}$) for the (a) high-gain and (b) low-gain ADCs, as a function of time, observed during the entire Run~2 (between CIS calibration runs taken on August 2015 and November 2018). Values for the average over all channels and for one typical channel are shown. The RMS values indicate the fluctuations present in calibrations. In addition, there is a 0.7\% systematic uncertainty present in individual calibrations, represented by the shaded error band. This uncertainty comes from the observed peak output amplitudes and is taken as characteristic of the channel-to-channel variation from this source, prior to any calibration. Only good channels not suffering from damaged components relevant to the charge injection calibration are included.

Variations in the detector-wide CIS constants (in per cent) between August 2015 and October 2018. The RMS variation is approximately 0.35$\%$. Only good channels not suffering from damaged components relevant to the charge injection calibration are included. The first and last bins contain overflow and underflow, respectively.

:

: : (a) The average response variation of the TileCal cells to the $^{137}$Cs source relative to the expected value, $\Delta R_{\mathrm{Cs}}$, as a function of time. The average runs over all cells in three radial layers. The increasing response corresponds to the periods without collisions. The LHC delivered integrated luminosity is shown by the shaded area. (b) The average response variation of the TileCal cells to the $^{137}$Cs source relative to the expected value, $\Delta R_{\mathrm{Cs}}$, after Run~2 data-taking, as a function of the cell position in $\eta$, for three different layers.

The variation of the average response to MB events $\Delta R_{\mathrm{MB}}$ for the cells in the gap/crack region of the extended barrel as a function of time. This quantity is derived relative to the luminosity measured using the inner tracker. The error bars correspond to the RMS of all the response distributions. The results are normalised to the values measured in the first run of Run 2 (16th July 2015). The integrated luminosity delivered by the LHC is shown by the shaded area.

The mean response variation $\Delta R_{\mathrm{Las}}$ in the PMTs for each cell type, in percent, averaged over $\phi$, observed during the entire $pp$ collisions data-taking period in 2018 (between standalone laser calibration runs taken on 18 April 2018 and 22 October 2018). For each cell type, the response variation is defined as the mean of a Gaussian function fit to the response variations in the channels associated with given cell type. A total of 64 modules in $\phi$ are used for each cell type, with the exclusion of known bad channels.

The mean response variation $\Delta R_{\mathrm{Las}}$ in the PMTs for each layer, as a function of time, observed during the entire Run~2 (between standalone laser calibration runs taken on 17 July 2015 and 22 October 2018). For each layer, the response variation is defined as the mean of a Gaussian function fit to the response variations in the channels associated with given layer. Known bad channels are excluded. The LHC integrated delivered luminosity is shown by the shaded area.

Schematic of a partial longitudinal cut of the ATLAS detector showing the $\eta$ region covered by the tile muon trigger system. The azimuth angle coverage is $0 \le \phi <2\pi$. The sub-systems not used in the tile muon trigger are sketched for completeness and their drawings do not represent the accurate geometry.

The basic architecture of the tile muon trigger system.

The layout of one side of the MBTS sub-detector in the period 2015--2017. The energy deposited in each sensitive region is read by one readout channel. Due to the large reduction of the scintillator response, the number of sensitive regions in the outer layer was doubled in 2018.

The average response variation of A13 cells measured by the laser, caesium and minimum-bias integrator systems as a function of time during the entire Run~2. Known bad channels are excluded. As discussed in the text the response behaviours are connected to the LHC delivered luminosity shown by the shaded area.

The measured relative light yield $I/I_0$ (Eq.~(\ref{eq:I/I_0})) of the TileCal cells at the end of Run~2. The uncertainty is of the order of 1\%.

Simulated ionisation dose deposited in the scintillator tiles of the cells and in the gap/crack scintillators in $4\times 4$~cm$^2$ bins in $r\times z$. The study was performed using 50~000 inelastic $pp$ collisions at $\sqrt{s}=13$~\TeV\ generated with \textsc{Pythia~8}. The response of the detector was obtained using the simulation program \textsc{Geant4}. The results are normalised to a cross-section of $\sigma_{\mathrm{inel}}$ = 78.42~mb and an integrated luminosity of 1~fb$^{-1}$~\cite{Pedro:2019lgd,DoseSimulation}.

Average relative light yield ($I/I_0$) measurements based on the caesium system (dots) and integrated minimum-bias currents (triangles) for A13 cells as a function of average simulated dose $d$ and LHC integrated luminosity. The dashed curve corresponds to the fit to the function in Eq.~(\ref{eq: degradation}) to the data. The surrounding opaque band represents the total uncertainty in the fit including the RMS of the dose distribution within the cell and systematic uncertainties in $I/I_0$ due to the intrinsic precision of the caesium, MB and laser measurements. The solid curve represents the expected average $I/I_0$ of the A13 cells in the HL-LHC phase including dose rate effects (see the text). The surrounding semi-transparent band is the total uncertainty on this extrapolation, obtained by propagating the uncertainty sources of the study. Results from measurements of bare scintillators performed one month after irradiations made in the laboratory before the detector construction are also shown~\cite{Abdallah:2007cca}. An exponential function is fitted to the data obtained from irradiations with $\gamma$s (open squares) and hadrons (full squares). Dashed vertical lines represent the expected dose by the end of the LHC (450~fb$^{-1}$) and HL-LHC (4000~fb$^{-1}$)~\cite{ZurbanoFernandez:2020cco}.

The degradation rate parameter $p_1$ obtained from the simple exponential model ($I/I_0=p_0e^{-d/p_1}$) as a function of the average simulated dose rate $\dot d$ for the most exposed cells. Results from a similar study performed using the CMS Hadron Endcap Calorimeter measurements are also displayed (crosses)~\cite{CMSHCAL:2016dvd}. The vertical error bars on the TileCal data points represent the total uncertainty including the $I/I_0$ measurement uncertainty and the dose spread within the cell volume. The nominal points are fitted with a power law function (continuous curve). This function is extrapolated to the higher dose rate region (dashed curve) expected at the HL-LHC phase and populated by the CMS data. The dashed vertical line indicates the expected dose rate of the A13 cells in the HL-LHC.

:

: : (a) The average response variation of the MBTS inner (down triangles) and outer (up triangles) counters as a function of time during Run~2. The circle (diamond) markers show the relative response of the PMTs of the inner (outer) counters. The LHC delivered luminosity is shown by the shaded area. (b) The average relative light yield ($I/I_0$) of scintillators and fibres of inner (down triangles) and outer (up triangles) counters as a function of total ionising dose during Run~2. The values are the averages of the corresponding determinations obtained for the inner and outer counters. The uncertainties correspond to the RMS of the counter response distributions. The function obtained by fitting the inner MBTS data points is also shown (solid curve).

The hierarchy of the TileCal DCS within the ATLAS DCS~\cite{Martins:2016hrt}.

The fraction of channels and cells removed from the reconstruction (masked) as a function of time during Run~1 and Run~2. The number of masked cells (channels) at the end of Run~2, 3 December 2018, is about 0.5\% (1\%). The hatched area represents the maintenance periods of the detector.

: Visualisation of the TileCal in the ($z$, $r$) plane showing the 2017 estimated values of the ratios $R_r$ of the cells of a given ring $r$ (see the text). The values are obtained by maximising the likelihood function in Eq.~(\ref{eq:likelihood_ phi}). The statistical uncertainty in each determination is about 0.5\%.

: The response ratio $R_l$ of the cells of different radial layers in the LB and EB obtained by analysing 2015--2016, 2017 and 2018 data. Statistical (first value) and systematic (second value) uncertainties are shown. :

: : Visualisation of the TileCal in the ($z$, $r$) plane showing the relative difference of the fitted response ratios (a) $\Delta_r(2015-2016 \to 2017)$, obtained in 2015--2016 and 2017, and $\Delta_r(2017 \to 2018)$, obtained in 2017 and 2018. The average statistical error in the determinations is about 0.8\%.

The distribution of the ratio of the energy of isolated hadrons measured by the calorimeters divided by the momentum of the track measured by the inner detector ($E/p$). The distribution obtained by analysing simulated data is also shown. The distributions are normalised to an integrated area of one. The ratios of the experimental values to the simulated ones are plotted in the lower panel. The MC statistical uncertainties are shown.

: : Caption not extracted

: : Caption not extracted

: Caption not extracted

The mean reconstructed cell time as a function of the cell energy. Results obtained in Run~2 (2015, 2016, 2017 and 2018), are shown. Statistical uncertainties are smaller than the size of the markers. The discontinuity close to 22~\GeV\ corresponds to the ADC high-/low-gain transition.

:

: : (a) The time resolution (full circles) and RMS of the time distribution (open squares) as a function of the cell energy. The results were obtained by analysing 2018 data. Statistical uncertainties are smaller than the size of the markers. The fits of the function in Eq.~(\ref{eq:resolution_fit_function}) are superimposed on high- and low-gain resolution data as indicated by solid and dashed curves, respectively. (b) A comparison of the time resolution fit functions obtained by analysing 2015, 2016, 2017, 2018 data samples~\cite{TD_proceedings} as a function of the cell energy. The 2017 high-gain fit is affected by a worse time resolution in the lower energy bin that leads to slightly different fit parameters $ p_0 $ and $ p_1 $ (see Eq.~(\ref{eq:resolution_fit_function})) compared to the results of the other years. The effect translates into an increase of the 2017 and 2015 resolution ratio near the high-gain end point. The lower panel shows the fitted time resolution obtained by analysing 2016, 2017 and 2018 data relative to the values obtained analysing 2015 data.

The cell time resolution ($\sigma$) determined in the energy range $\SI{8}{\GeV} < E_{\mathrm{cell}} < \SI{10}{\GeV}$ as a function of the mean number of reconstructed primary vertices per bunch crossing within a luminosity block, $N_{\text{PV}}$, in the run. Each point corresponds to one run and is obtained using all the cells of the detector. In the analysis, only runs with 25~ns bunch spacing and the number of bunches in the LHC larger than 1000 are used. The worse resolution in 2015 is compatible with the larger run-to-run differences provided the phase changes are not strictly coupled to individual run boundaries, as discussed in Section~\ref{sec:run_to_run}.

The $\phi$-averaged electronic noise as a function of $\eta$ of the cell, with both contributing read-out channels in high-gain mode. For each cell type the average value over all modules is taken. The statistical uncertainties are smaller than the marker size. Values are extracted using a single representative pedestal calibration run taken in 2017. The different cell types are shown separately for each layer. The transition between the long and extended barrels can be seen in the range of $0.7 < |\eta| < 1.0$.

Normalised energy distributions in cells A12 ($1.1<|\eta|<1.2$) observed in $pp$ collision data with 25~ns bunch spacing at $\sqrt{s}=13$~\TeV\ collected in 2016 in the zero-bias stream and in the \textsc{Pythia 8} MC simulation with the A3 minimum-bias tune. An integration over all modules is performed. The depicted distributions correspond to two different \pileup conditions with $\langle\mu\rangle=20$ (squares) and $\langle\mu\rangle=30$ (circles). The ratio of the data to the MC simulation is shown in the lower panel.

The total noise in TileCal cells, as a function of $\eta$, observed in $pp$ collision data with 25~ns bunch spacing at $\sqrt{s}=13$~\TeV\ collected in 2016 in the zero-bias stream with an average number of interactions $\langle\mu\rangle=30$ per bunch crossing and in the \textsc{Pythia 8} MC simulation with the A3 minimum-bias tune. The noise is estimated as the standard deviation (RMS) of the measured cell energy distribution. The data (MC simulation) are plotted with closed (open) markers. The cells of different layers are shown with different colours.

:

: : The total noise in TileCal cells, as a function of the average number of interactions per bunch crossing $\langle\mu\rangle$, observed in $pp$ collision data with 25~ns bunch spacing at $\sqrt{s}=13$~\TeV\ collected in 2016 in the zero-bias stream with an average number of $\langle\mu\rangle$ equal to 30, and in \textsc{Pythia 8} MC simulation with the A3 minimum-bias tune. The noise is estimated as the standard deviation of the energy distribution per cell. The data (MC simulation) are plotted with closed (open) markers. The noise is shown for cells located in the region (a) $0.4<|\eta|<0.5$ in the LB and (b) $1.2<|\eta|<1.3$ in the EB. Due to statistical limitations, the total noise measured in data is shown only for $\langle\mu\rangle>15$. The fit functions from Eq.~(\ref{eq:total_noise_fit}) are overlayed on the experimental and simulated data points (dashed curves) in the figure.

The $\eta$ distribution of particles with transverse momentum larger than 20~\GeV\ measured in events selected using the L1 muon trigger alone (triangles) and the tile muon system coincidence (line). The ratios of the bin contents are shown in the lower panel. The coincidence regions, $1.0<|\eta|<1.3$, are indicated by the vertical lines. The $\eta$ of the tracks is reconstructed using online information. The asymmetry of the distribution is due to the different acceptance of the muon spectrometer within the toroidal magnetic field.

: : Caption not extracted

: (a) The efficiency of the MBTS trigger during $pp$ collisions at $\sqrt{s}=13~\TeV$ as a function of the multiplicity of charged particles compatible with the beam line, $n^{\mathrm{BL}}_{\mathrm{sel}}$, with $\pt>500~\MeV$ and $|\eta| < 2.5$ reconstructed in the inner detector. The data were collected in 2015. The statistical uncertainties are shown as vertical bars, the sum in quadrature of statistical and systematic uncertainties is shown as the height of the shaded rectangles~\cite{STDM-2015-02}. Systematic uncertainties dominate. (b) The efficiency of the MBTS trigger during $pp$ collisions at $\sqrt{s}$=13~\TeV\ as a function of the multiplicity of charged particles compatible with the beam line, $n^{\mathrm{BL}}_{\mathrm{sel}}$, with $\pt>500~\MeV$ and $|\eta| < 2.5$ reconstructed in the inner detector. Results from the analysis of data collected in 2017 are compared with those from 2015 presented in (a). The statistical uncertainties are shown as vertical bars, the sum in quadrature of statistical and systematic uncertainties of the 2017 data are shown by the shaded rectangles. For 2017 data, statistical uncertainties dominate, while the uncertainties for 2015 data in (b) are smaller than the size of the symbols because of the much larger range of the vertical axis than in (a). : Caption not extracted