DESY PETRA retires undefeated Theoryfest

The traditional annual Theory On 3 November, the PETRA elec­ nuclear force, delicate effects due Workshop at the German DESY tron-positron storage ring at the to these particles were neverthe­ Laboratory in concen­ DESY Laboratory in Hamburg less observed. This boosted con­ trated this time on quantum chro- ended its remarkable eight-year fidence in the underlying theory modynamics (QCD) — the unchal­ career as a machine for high ener­ and helped pave the way for the lenged but yet to be convincingly gy physics experiments. By cun­ discovery of the W and Z at confirmed theory of and ningly switching off ten days be­ CERN in 1983. forces in strong nuclear inter­ fore the new TRISTAN ring at PETRA experiments also made actions. the Japanese KEK Laboratory pioneer contributions in the field In principle, QCD describes a collided 25 GeV beams (50 GeV of two- physics — the broad range of phenomena, but it total energy, see page 1), PETRA scattering of light by light. is a long way to get from the ap­ managed to retire with its record Despite the big push for higher parently simple basic equations of of 23.35 GeV colliding energy, the sixth ('top') quark the theory to the observed exper­ and positron beams still intact. was not reached, but this pro­ imental data. Tests of QCD are Built with a design goal of vided important new limits. good only to within about 20 per 19 GeV per beam, additional As well as physics successes, cent, and in many cases a lot of PETRA accelerating power was the groups working at PETRA work is required even to attain this gradually wheeled in, eventually also made notable contributions goal. leading to the record figure to instrumentation progress — However the workshop showed early in 1984. As well as pushing the JADE jet chamber, precision that theorists are far from discour­ up the energy of the collisions, vertex chambers, etc. aged by technical difficulties and PETRA supplied lots of them, On the machine side, PETRA try to make progress with refined with good luminosity being en­ was to give the first example of perturbation methods and with joyed by five experiments over the 'mini-beta' scheme in action. numerical simulations. the years - CELLO, MARK-J, This technique had been pro­ A number of invited speakers JADE, PLUTO and TASSO, bring­ posed at DESY to squeeze the discussed the production of 'jets' ing together a total of about 400 colliding beams even tighter to­ of hadrons in electron-positron physicists from 12 countries. gether and improve the luminos­ and -antiproton annihilation. A milestone discovery in 1979 ity. Good polarization (spin align­ W. Scott of the UA1 experiment assured PETRA of a place in phy­ ment) levels were obtained at the CERN proton-antiproton Col­ sics history. Clear two-jet struc­ thanks to careful machine tuning, lider summarized new data, report­ tures, coming from quark pairs while new materials, such as alu­ ing broad agreement with QCD. formed in the electron-positron minium for vacuum chambers, But there are large uncertainties, annihilations, were seen almost were exploited. PETRA also in particular when measuring the as soon as PETRA was switched served as a test-bed for proto­ jet energy, making it difficult to on. The sighting of a third jet type superconducting radiofre- give accurate production rates. showed that were also quency accelerating cavities, pro­ The strength of the QCD coupling being released. As well as pro­ viding valuable experience in this quoted by Scott was higher than viding the first direct evidence new technology, now planned that from previous electron-posi­ for the carriers of the inter-quark for several future machines. tron measurements at DESY's PE- force, this new PETRA physics Over the next twelve months, TRA ring. It could be that higher enabled the strength of the quark the PETRA ring will be modified orders of QCD perturbation theory forces to be measured. for its new role as an electron, are required or that the formation Although pf=jRA 's energy was positron and proton injector for of hadrons from and gluons insufficient to isolate the W and the HERA storage ring now being ('fragmentation') does not follow Z boson carriers of the weak built at DESY. the assumed pattern.

CERN Courier, January/February 1987 23 PETRA at the DESY Laboratory in Hamburg, retired undefeated champion of the electron-positron ring.

(Photo DESY)

of Liverpool), but the correlation of the two approaches is not straightforward. No convincing evidence was presented for glueballs (particles composed of gluons rather than quarks) although there are candi­ dates such as the eta 1440. M. Pennington of Durham cov­ ered a careful analysis of pion and kaon pair production in proton- proton collisions (see page 17) which suggests several reson­ ances, of which at least one could be a glueball. QCD sum rules are one method for avoiding perturbation theory difficulties, and were described by L. J. Reinders of Bonn. However the most promising attempts to solve QCD at laboratory energies New calculations in quark dy­ A traditional test-bed for QCD are coming from numerical calcu­ namics have already begun (re­ is the quark structure (structure lations based on a theoretical lat­ ported by R. K. Ellis of Fermilab). functions) of nucléons measured tice. While new predictions are Results for jets are not yet avail­ with penetrating beams. beginning to emerge, the method able but a study of single hadron C. Geweniger of Heidelberg con­ is still in its infancy (A. Kronfeld production shows large effects, cluded that the detailed kinematic of DESY - 'Solving QCD on a particularly in the angular depend­ dependence from QCD is applicable computer'). This is because realis­ ence. over a narrower range of ex­ tic numerical QCD simulations us­ A fundamental problem in all changed momentum than had been ing today's algorithms need 1020 this work is the appearance of thought (Q2 above about 10 GeV2). arithmetical operations, not easily terms which increase rapidly with For electron-positron physics in digestible by today's computers. energy and must therefore be sum­ the upsilon region, D. Wegener of But the algorithms are being med. D. Soper of Oregon explained Dortmund had many results to improved and computing time can how to deal with such soft gluons. report, including improved limits be saved by improving the simula­ When using only a few orders on certain lepton decays which, tion procedures and overcoming of QCD perturbation, it is important if seen, would upset conventional the effects of the lattice, holding to choose the coupling constant dogma. out new hope. More powerful com­ correctly to minimize contributions The upsilon particles are under­ puters would be an asset, and from higher orders. M. Fontannaz stood as bound states of beauty A. Terrano of Columbia related of Orsay gave a phenomenological quarks and antiquarks. Full QCD how some theoreticians have justification for this optimization treatment is difficult and physicists turned their attention to the design in the case of photoproduced habitually resort to non-relativistic of special computers with highly hadrons. potential models, covered by parallel architectures. During the fragmentation of high J. Kuhn of MPI Munich. This ap­ Several completed or nearly energy quarks into observable par­ proach describes the upsilons quite completed machines attain speeds ticles, cascades of gluons are ra­ well, but there are still a few miss­ of between 0.1 and 10 Giga-oper- diated away. B. Webber of Cam­ ing states. The emerging form of ations per second, and are signifi­ bridge showed that the coherence this potential is also suggested by cantly cheaper than commercial of this radiation has important numerical studies of lattice gauge machines with comparable perfor­ implications. theories (described by C. Michael mance. However the potential user

24 CERN Courier, January/February 1987 (left) and Martin Luscher at the traditional Theory Workshop organized by the DESY Laboratory in Hamburg.

(Photo DESY)

thermodynamic approaches can pay dividends. At low temperatures, H. Leut- wyler of Bern showed how the situation is well understood in terms of a pion gas whose proper­ ties can be calculated elegantly and accurately. At higher temper­ atures lattice techniques have to be brought in, and the phase tran­ sition to a quark-gluon plasma comes at an energy density of some 2.5 GeV/fm3. With 20- 50 GeV/fm3 expected from heavy ion collisions, the creation of the plasma looks on the cards. How could it be recognized? One pos­ sibility would be a dilepton spec­ trum without or with a reduced J/psi signal, as plasma quarks are not expected to form bound meson states. The members of the DESY Theory Workshop Organizing Com­ mittee were R. Baier, H. Fritzsch, M. Lûscher, R. D. Peccei and K. Schilling.

From M. Luscher has to confront reduced flexibility block spin transformation) gives and difficult software. reason to be optimistic. Even if a full QCD simulation on There is growing interest in the BROOKHAVEN a lattice is still a dream, some in­ link between QCD and nuclear phy­ More from spin teresting physics follows from bold sics as pictures of the strong nu­ approximations. G. Martinelli of clear force under different condi­ CERN told of promising results for tions. F. Lenz of SIN covered a For more than a decade it has been the weak decays of kaons into simple non-relativistic model for known that hyperons produced by pions. the interaction between many are polarized to an unex­ A crucial question underlying the quarks and antiquarks, opening pected degree. Most models try whole lattice approach is whether up, for example, the role of quark to relate these observations to the continuum limit of the (artificial) exchange in pion scattering and polarization effects at the underly­ lattice really exists, and whether the states of the helium-4 nucleus. ing quark level: polarization of the a perturbation approach is correct. The long-awaited quark-gluon quark-antiquark pairs produced Hints can come from numerical plasma, where the quarks and when colour-flux strings break, calculations, but a full analytic gluons break loose from their con­ and/or the Thomas precession proof would be more acceptable. ventional nucléon confinement, effect have been suggested as G. Mack of Hamburg spoke on the may be transiently produced in the mechanisms responsible for the continuum limit of lattice field the­ collision of high energy ions, as polarization. This type of model ories and showed how the 'phase indicated by H. Satz of Bielefeld. has had several successes and a space cell expansion' (essentially With no detailed understanding of few failures (see September 1986 a variant of the Wilson-Kadanoff the underlying scattering process. issue, page 15).

CERN Courier, January/February 1987 25