Neutral Currents in BEBC -
The experiment WA21 (part 3)

The Detectors


The Big European Bubble Chamber was operational between 1973 and 1984. Based on an idea of Charles Peyrou, BEBC was the successor of the Gargamelle bubble chamber, which became famous for the discovery of the Neutral Currents in 1973.

Essentially, each bubble chamber (and particularly BEBC) consists of

BEBC is inserted after mounting the IPF [CERN]

Certainly, BEBC can be considered as summit of bubble chamber technology (with the driving force of Horst Wenninger). The disadvantage of the bubble chamber due to the restricted reaction mass (about 1000 kg of H2 in WA21) has been tried to compensate for as far as possible by means of (almost) unsurpassed homogeneous experimental physical conditions in any aspect.

However, during BEBC' operation the focus in High Energy Physics (HEP) was already shifted towards (the SPS) collider experiments and the potential of BEBC was never ever really exploited.

BEBC control room with the DEC PDP-10 computer [CERN]

Scanning and Measurement

During every beam spill, three to five photographs were taken to allow a stereoscopic reconstruction of events. This took place at particular measuring tables: ERASME.

Every picture was at first scanned three times and eventually measured and digitized in case a reaction (of any kind) was visible. The probability of a detection depends on the number of seen tracks and on the location of the event in the bubble chamber. The later was compensated for introducing a fiducial volume by means of geometrical cuts.

The tracks (and the vertices) were measured by the systems and already now the first assumptions of the generated particles can be made: Stopping in the liquid, considerable energy loss, downstream produced neutral reactions. In particular V0 events are of interest, being produced by the decay of a π0, K0 or Λ0. The later are long-lived strange particles.

Unfortunately, the H2 target suffers from a substantial loss of neutral particles, since those leave the bubble chamber without a trace. On the other side, the absence of nuclear interactions may lead to completely described events, which allow a precise fit and a determination of the kinematical variables and even initial particle produced: Showroom events.

Event measurement with the ERASME system [CERN]

External Muon Identifier EMI

In order to detect muons in Charged Current (CC) events unambiguously, BEBC was equipped with an External Muon Identifier EMI partly constructed and initially operated by Hans Jürgen Hilke who died in spring 2017. To allow a coincidence measurement of outgoing tracks, two planes of Multiwire Proportional Chambers were installed downstream (outer and inner plane). For later runs an additional plane became operational upstream of BEBC, in particular to act as a veto-trigger for incoming muons (veto plane).

As can be seen from the right-side picture, the geometrical coverage of the EMI is far from complete. However, one has to consider, that any reaction inside BEBC takes place in the lab frame and is therefore Lorentz boosted while the produced particles are emitted in a cone: The higher the energy of a particular particle, the more it is aligned towards the beam axis.

Any measured leaving track for a BEBC event is at first extrapolated to the EMI and finally associated with a potential hit in one of the about 100 recorded timeslots lasting 250 ns each by means of the EMI program. Due to the still high background of residual muons, the likelihood for an accidental coincidence is quite high, thus often events were initially wrongly labeled CC and of course falsely associated to the wrong timeslot.

Due to this, the THIRA program required coincident hits in both the inner and the outer plane of the EMI, reducing the geometrical efficiency even more.

The EMI detector downstream from (the removed) BEBC vessel [CERN]

The EMI/IPF control desk
The folders tell the physists what to do -- in particular in terms of a Hydrogen Alarm.
The EMI/IPF electronics
The often published diagrams on cosmic muons penetrating the EMI were visualized on the Tektronix systems visible in the foreground (the physicists had to watch the apparatus).

Internal Picket Fence IPF

The IPF was installed in 1981 as a two-layer system of 1440 single wire tubes (diameter 1.5 cm each and length of 220 cm) surrounding the body of BEBC literally as a 'picket fence'. The read-out system re-used those for conventional EMI chambers, thus no particular change in the electronic data reconstruction was required. Not only that the IPF had a limit spatial and geometrical acceptance, a hit in the IPF just indicated which tube was affected, but the z-coordinate of the hit was not recorded.

Looking at the IPF from the bottom before the BEBC vessel was inserted [CERN]
In the middle one can see the five cameras, the beam is traversing the chamber system from left to right.

Apart from an additional veto plane (the upstream IPF) and the downstream IPF simply was used (by EMIANA) as extra coincidence layer for charged particles leaving BEBC and perhaps improving the timeslot selection for events.

At that stage, CC events were considered only and Maurice Retter (and Simon Towers for his Phd thesis) from Oxford facilitated a careful analysis of the IPF activity. While introducing a Bayesian weight used to associate hits, they improved the timeslot decision significantly for all events. In addition, a Veto- and Exit Picket function was introduced to measure any un-associated activity in the IPF as part of the PICKET program. Aim was to discriminate NC from so-called N* (hadronic induced) events by means of the Veto-Picket Function.

Another remaining problem regarding the muon detection in CC events was the accidental coincidence of potentially outgoing particle tracks from the BEBC events with hits in the EMI of so-called thrugoing muons. Not only beam related muons have to be considered (discussed above), but CC (and NC) interactions happen also upstream BEBC in the material (in particular the coils) surrounding the bubble chamber. The interaction rates are essential proportional to the mass of the potential target system, which of course disfavors the BEBC' H2 filling (about 1500 kg) by an order of magnitude.

As Diploma student from the University Bonn for WA21 in Klaus Böckmann's group, I was delegated during that time as Associate to CERN solving this problem to some extend, while creating the TMRECO thrugoing muon reconstruction program. Further, this proofed that the Veto Picket activity is related to events produced upstream of BEBC and which are accompanied (in the very same timeslot) by a thrugoing muon: A CC event created outside BEBC punching thru and partially materializing in the bubble chamber as N*.

By the very same token, it became feasible to detect hadronic induced events with good efficiency by the detector and not relying on kinematical cuts on the event distributions (typically, the seen transverse momentum of the hadronic system, as carried out in Madjjid Mobayyen's analysis) improving significantly any analysis of Neutral Current events.

Data Taking

The data taking thus involves three different detectors:

  1. The bubble chamber BEBC with data from the reconstructed tracks given by the GEOMETRY progam.
    Given these data, one can already perform physics data on the tracks (given their curvature in the magnetic field, seen interactions, and decays) providing a hypothesis for the particle (KINEMATICS).
    Taking these hypothesis one can extrapolate the track to the circumfencing detectors, which was subject of the THIRA program.
  2. The recorded hits in the EMI detector available by the EMI data were taken care of the EMIANA program, giving hit coordinates (x,y,z) together with the event time.
  3. Unfortunately, the IPF was not treated accordingly and just the 'tube' hitted was recorded together with the time frame, thus neither a 'z' coordinate nor a signal strength was availalbe. Our PICKET program suffered thus from an uncomplete coverage, even though the data were available but never recorded

These information needed to be merged and irrelevant information to be filtered. The resulting data set was written to tape a DST (Data Summary Tape) and shipped to the involved colloborating labs. Initially, a 'reel' tape was used and written with 6250 BPI covering at most 175 Mbyte of data.
It was my duty to produce those tapes for the last runs of BEBC and WA 21. Unfortunately, the 'tape cutting' process - taking place on CERN's CDC - was in particular limited by the available memory (and no virtual memory in place) which required to skip data with a huge amount of input data. Thus, we might have lost some interesting and high-energetic data.

For my own NC analysis I rewrote most of those programs resulting in my SEVEN program. The name SEVEN originates from the fact, that the initial data structure was enhanced at the virtual address q(lkt-7) of the kinematic 'bank'.

Currently, all CERN data are available here, the CERN Tape Archive.

Continue with Events in BEBC