Operational performance of the CERN injector complex with transversely split beams

With the progress made in 2015, the beams produced by the CERN Proton Synchrotron using multiturn extraction (MTE) have been delivered to the Super Proton Synchrotron (SPS) for the fixed-target physics run. Operation successfully started in the second half of September 2015 and continued until the end of the proton physics program by mid November. In this paper the overall performance and beam quality is discussed in detail considering the complete chain of accelerators, from the PS-Booster to the SPS. Moreover, a thorough comparison of the global performance of the MTE scheme against the previously used technique, the so-called continuous transfer (CT), is also carried out.

Figure 1

Calculated zero-intensity working point evolution along the PSB cycle for the MTE beam. Injection takes place in the upper right end of the working point lines. Resonance lines up to 4th order are shown.

Figure 2

Measured transverse normalized emittances in the PSB as a function of bunch intensity. The situation for the standard setup (“Initial”) as well as the special one (“After optimization”) is shown and proves the reduced dependence on intensity of the vertical emittances for the special configuration. The emittance is expressed as $1\sigma$ normalized value.

Figure 3

Upper: Time evolution of the strength of the key families of sextupoles, octupoles, and of rf voltage during the resonance crossing process. Bottom: The variation of the corresponding main observables of beam dynamics is also shown. ${\delta }_b$ is the bucket height and is equal to $3{\sigma }_{\delta }$, with ${\sigma }_{\delta }$ the rms beam momentum spread. The location of the main elements in the PS ring is shown in Fig. 5. The elements with name starting with X or O are sextupoles, or octupoles, respectively.

Figure 4

Normalized $1\sigma$ vertical emittance in the PS ring at injection and after beam splitting as a function of total beam intensity. The relative emittance growth is also shown together with the measured emittance evolution vs intensity out of the PSB (see Fig. 2).

Figure 5

Upper: Evolution along the PS circumference of the lattice functions, showing also the location of the essential elements for MTE. Bottom: Horizontal ring aperture together with beamlets and core envelope just prior to the pulsing of the kickers. The extraction region indicates the section of the ring where orbit bumps are present. The element name includes the number of the straight section in which the device is installed.

Figure 6

Sketch of the PS main magnet (upper) and of the five circuits used to control tunes and chromaticities (lower), from Ref. [35]. The additional coils enable generating a transverse variation of the magnetic field, which is used to produce nonlinear field components essential for the control of the transverse splitting.

Figure 7

Left: Measured horizontal beam distribution for different efficiency of the beam splitting. Right: Corresponding intensity profile measured by a beam current transformer in the transfer line between the PS and the SPS.

Figure 8

Left: Wire scanner vertical profile measurement, showing the profile tails, cut by the aperture restrictions, leading to beam scraping and losses. The extent of the profile is hence a measure of the global available aperture. Right: Expected losses at the SPS flat bottom as a function of the vertical normalised emittance. The global aperture bottleneck is the internal dump, TIDVG.

Figure 9

Evolution of the measured SPS vertical aperture from 1995 to 2015. The extent of the transverse particle distribution of a blown up beam is evaluated with wire scanners as a measure of the global aperture. The measurements with two wire scanners are compared and translated to the location of the known SPS aperture bottleneck TIDVG.

Figure 10

Upper: Evolution of the PS beam intensity and extraction efficiency during the 2015 MTE run (from Ref. [25]). The three colored regions towards the end of the run correspond approximately to the selected time frames used in the analysis reported here. Bottom: PS intensity over time, shown together with the cuts in intensity, and ${\eta }_{\mathrm{MTE}}$.

Figure 11

Example spills for both CT and MTE.

Figure 12

Upper: Histogram showing the distribution of the MTE efficiency. Bottom: Histogram showing the distribution of the DC efficiency including CT data. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 13

Upper: Histogram showing the distribution of the PS extraction efficiency. Bottom: Histogram showing the distribution of the transfer line transmission efficiency. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 14

Example of beam intensity in the SPS as a function of time.

Figure 15

Global picture of the distribution of the beam transmission through the SPS cycle at the key times used for the performance evaluation. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 16

Histogram showing the distribution of ${\eta }_{\text{Overall}}$. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 17

Intensity losses in % as a function of momentum for 2015 data, both CT and MTE. The highest intensity losses (for all four time frames) come around $22\mathrm{GeV}/\mathrm{c}$, corresponding to transition crossing, reaching as high as 2.5%. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 18

Correlation heat maps for the figures-of-merit used in the analysis. The three plots refer to the different MTE beam intensities, namely $N_p=1500$, 1800, 2000 ppp for the left, center, and right panel, respectively. It is worth noting that the general tendency of correlations was found to be positive.

Figure 19

Example of correlation plots between PS and SPS efficiency figures of merit. The progressively lower correlation between ${\eta }_{\mathrm{MTE}}$ and transmission efficiencies for later stages in the SPS cycle is clearly visible. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 20

Transmission efficiency ${\eta }_{\mathrm{FB}}$ (upper row) and ${\eta }_{\text{Overall}}$ (bottom row) for MTE (left column) and CT (right column) for the various types of SPS super-cycles (Standard for the normal super-cycle; Special for super-cycle for beam physics development studies; Filling for LHC filling super-cycle). The difference is clear and mainly stems from the different efficiency in the low-energy part of the SPS cycle.

Figure 21

Correlation heat map for the figures of merit used in the analysis of CT performance.

Figure 22

BLM readings for MTE and CT in the PS (upper) and SPS (lower) rings. The bars represent the average, while the error bars the rms values of the BLMs. For the PS case the readings are given in ADC counts, while for the SPS, the readings have been normalized to the beam intensity. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 23

Bar plot of losses for MTE and CT beams. The height of each bar represents the total losses, from before ejection in the PS, to the start of Flat Top in the SPS. The typical intensity values are in units of $10^{10}\mathrm{ppp}$ at extraction from the PS.

Figure 24

Amplitudes of frequencies from FFT for CT (left column) and MTE (right column) extracted spills as a function of time. The upper plot shows all frequencies present in the spectrum, while in the center plot a zoom in the region around the 200 MHz component is shown. The bottom plot shows the evolution over time of the 200 MHz and of the 192.5 MHz components only.

Figure 25

Investigating relationships between $|c_n|$ and ${\eta }_{\mathrm{MTE}},{\eta }_{\mathrm{DC}}$.