A look back at events leading up to first collisions in LEP, 20 years ago.
On 13 November 1989, heads of state, heads of government and ministers from the member states assembled at CERN together with more than a thousand invited guests for the inauguration of the Large Electron–Positron (LEP) collider (PS and LEP: a walk down memory lane.). Precisely one month earlier, on 13 October, large audiences had packed CERN’s auditorium and also taken advantage of every available closed-circuit TV to see the presentation of the first results from the four LEP experiments, ALEPH, DELPHI, L3 and OPAL – results that more or less closed the door on the possibility that a fourth type of neutrino could join those that were already known. This milestone came only two months after the first collisions on 13 August and three months after beam had circulated around LEP for the first time.
Champagne corks had already popped the previous summer, soon after 23.55 p.m. on 12 July 1988, when four bunches of positrons made the first successful journey between Point 1, close to CERN’s main site at Meyrin (Switzerland) and Point 2 in Sergy (France) – a distance of 2.5 km through much of the first of eight sectors of the 27-km LEP ring. It was a heady moment and the culmination of several weeks of final hardware commissioning. Elsewhere, the tunnel was still in various stages of completion, the last part of the difficult excavation under the Jura having been finished only five months earlier.
A year to do it all
Steve Myers led the first commissioning test and a week later he reported to the LEP Management Board, making the following conclusions: “It worked! We learnt a lot. It was an extremely useful (essential) exercise – exciting and fun to do. The octant behaved as predicted theoretically.” This led to the observation that, “LEP will be more interesting for higher-energy physics than for accelerator physics!”. However, he also warned, “We should not be smug or complacent because it worked so well! Crash testing took 4 months for about a tenth of LEP; at the same rate of testing the other nine tenths will require 36 months.” Yet the full start-up was already pencilled in for July 1989, in only 12 months’ time.
The following months saw a huge effort to install all of the equipment in the remaining 24 km of the tunnel – magnets, vacuum chambers, RF cavities, beam instrumentation, control systems, injection equipment, electrostatic separators, electrical cabling, water cooling, ventilation etc. This was followed by the individual testing of 800 power converters and connecting them to their corresponding magnets while carefully ensuring the correct polarity. In parallel, the vacuum chambers were baked out at high temperature and leak-tested. The RF units, which were located at interaction-regions 2 and 6, were commissioned and the cavities conditioned by powering them to the maximum of 16 MW. Much of this had to be co-ordinated carefully to avoid conflicts between testing and installation work in the final sector, sector 3-4. At the same time a great deal of effort – with limited manpower – went into preparing the software needed to operate the collider, in close collaboration with the accelerator physicists and the machine operators.
The goal for the first phase of LEP was to generate electron–positron collisions at a total energy of around 90 GeV, equivalent to the mass of the Z0, the neutral carrier of the weak force. It was to be a veritable Z0 factory, delivering Z0s galore to make precision tests of the Standard Model of particle physics – which it went to do with outstanding success.
To “mass produce” the Z0s required beams not only of high energy, but also of high intensity. To deliver such beams required four major steps. The first was the accumulation of the highest possible beam current at the injection energy of 20 GeV, from the injection chain. (This was itself a major operation involving the purpose-built LEP Injection Linac (LIL) and Electron–Positron Accumulator (EPA), the Proton Synchrotron (PS), the Super Proton Synchrotron (SPS) and, finally, transfer lines to inject electrons and positrons in opposite directions, which curved not only horizontally but also vertically as LEP and the SPS were at different heights). The second step was to ramp up the accumulated current to the energy of the Z0, with minimal losses. Then, to improve the collision rate at the interaction regions the beam had to be “squeezed”, by reducing the amplitude of the betatron oscillations (beam oscillations about the nominal orbit) to a minimum value. Finally the cross-section of the beam had to be reduced at the collision points.
The first turn
In June 1989 the LEP commissioning team began testing the accelerator components piece by piece, while the rest of CERN’s accelerator complex continued as normal. Indeed, the small team found themselves running the largest accelerator ever built in what was basically a back room of the SPS Control Room at Prévessin.
The plan was to make two “cold check-outs” – without beam – on 7 and 14 July, with the target of 15 July for the first beam test. The cold check-out involved operating all of the accelerator components under the control of the available software, which proved important for debugging the complete system of hardware and software for energy ramping in particular. On 14 July, however, positrons were already available from the final link in injection chain – the SPS – and so the second series of tests turned into a “hot check-out”. Over a period of 50 minutes, under the massed gaze of a packed control room, the commissioning team coaxed the first beam round a complete circuit of the machine – one day ahead of schedule.
In the days that followed, the team began to commission the RF, essential for eventual acceleration in LEP. The next month proved crucial but exciting as it saw the transition from a single turn round the machine to a collider with beams stored ready for physics.
By 18 July the first RF unit was in operation, with the RF timed in correctly to “capture” the beam for 100 turns round the machine. Two days later, the Beam Orbit Monitoring system was put into action, which allowed the team to measure and correct the beam’s trajectory. Measurements showed that the revolution frequency was correct to around 100 Hz in 352 MHz, or equivalently, that LEP’s 27 km circumference was good to around 8 mm. Work then continued on measuring and correcting the “tune” of the betatron oscillations, so that by 23 July a positron beam was able to circulate with a measured lifetime – derived from the observed decay of the beam current – of 25 minutes. Then, following a day of commissioning yet more RF units, the first electrons were successfully injected to travel the opposite way round the machine on 25 July.
Now it was time to try to accumulate more injected beam in the LEP bunches and to see how this affected the vacuum pressure in the beam pipe. By 1 August the team was observing good accumulation rates and measured a record current of 500 μA for one beam. This was the first critical step towards turning LEP into a useful collider. The next would be to ramp up the energy of the beam.
The late evening of 3 August saw the first ramp from the injection energy of 20 GeV, step by step up to 42.5 GeV, when two RF units tripped. On the third attempt – at 3.30 a.m. on 4 August – the beam reached 47.5 GeV with a measured lifetime of 1 hour. Three days later, both electrons and positrons had separately reached 45.5 GeV. Then 10 August saw the next important step towards a good luminosity in the machine – an energy ramp to 47.5 GeV followed by a squeeze of the betatron oscillations.
In business
On 12 August LEP finally accumulated both electrons and positrons. The next day the beams were ramped and squeezed to 32 cm, yielding stable beams of 270 μA per beam. It was time to turn off the electrostatic separators that allowed the two beams to coast without colliding. The minutes passed and then, just after 11 p.m., Aldo Michelini, the spokesperson of the OPAL experiment, reported seeing the first collision. LEP was in business for physics.
So began a five-day pilot-physics run that lasted until 18 August. During this time various technical problems arose and the four experiments collected physics data for a total of only 15 hours. Nevertheless, the maximum luminosity achieved of 5 × 1028 cm–2s–1 was important for “debugging” the detector systems and allowed for the detection of around 20 Z0 particles at each interaction region.
A period of machine studies followed, allowing big improvements to be made in the collider’s performance and resulting in a maximum total beam current of 1.6 mA at 45.5 GeV with a squeeze to 20 cm. Then, on 20 September, the first physics run began, with LEP’s total energy tuned for five days to the mass peak for the Z0 and sufficient luminosity to generate a total of some 1400 Z0s in each experiment. A second period followed, this time with the energy scanned through the width of the Z0 at five different beam energies – at the peak and at ±1 GeV and ±2 GeV from the peak. This allowed the four experiments to measure the width of the Z0 and so announce the first physics results, on 13 October, only three months after the final testing of the accelerator’s components.
By the end of the year LEP had achieved a top luminosity of around 5 × 1030 cm–2s–1 – about a third of the design value – and the four experiments had bagged more than 30,000 Z0s each. The Z0 factory was ready to gear up for much more to come.
• Based on several reports by Steve Myers, including his paper at the second EPAC meeting, in Nice on 12–16 June 1990.
