The KARMEN Upgrade

Wenn Sie die deutsche Version bevorzugen, klicken Sie hier!

In the year 1996 the KARMEN experiment has been upgraded substantially. The aim of this so called KARMEN upgrade was to reduce the background for the search for neutrino oscillations dramatically. This aim has been achieved completely as will be shown below.

Extensive simulations and the identification of the background in special measurements preceeded the KARMEN Upgrade. The simulations were also used to optimize the design. Moreover these simulations can be used to visualize the background and to explain how the upgrade works. The background for the search for oscillations in the channel `nm ® `ne is caused by cosmic muons wich produce in rare cases high energy neutrons in the iron of the KARMEN bunker. The picture below (taken from a simulation performed with the GEANT simulation package from CERN) shows the situation for the so called deep-inelastic muon-nucleon scattering at the nuclei of the iron atoms. These reactions are mediated by the exchange of virtual photons with energies of up to several GeV destroying the nucleus in a spallation reaction. During this spallation a lot of elementary particles like pions, kaons but also neutrons are produced having energies of up to several GeV. The picture below shows a simulated spallation together with the tracks of the particles produced.


This spallation reaction was caused by a positive muon m+ with an energy of 41.2 GeV. The spallation produced charged particles, neutrons and gammas g but only the neutrons have a large range and thus the biggest chance to reach the central detector undetected by any anti counters. The picture below illustrates this. Here a vertical cut through the KARMEN detector, the anti counter systems and the surrounding iron shielding is shown. Several spallation reactions are drawn into the picture, which could not be detected with the anti counters existing before 1996. The neutrons penetrate into the central detector and cause events which nearly look like the detection of `ne-neutrinos and thus result in background for the search for `nm ® `ne oscillations.


The only way to reduce this background is to detect the cosmic muons which produce the neutrons. The detection has to happen far enough from the central detector in order to prevent neutrons produced by undetected muons from penetrating into the detector. This can be achieved if about 1 metre of steel sits between the veto counters detecting the muons and the central detector. Thus the only solution was to put large scale veto counters into the the walls and the roof of the iron bunker roughly in the middle of the shielding. Those veto counters are drawn in red in the above picture already. All muon tracks shown cross the new veto counters and are detected. Thus potential successive events in the central detector can be precisely identified and discarded. Detailed simulation calculations showed that a reduction of the background by a factor of 40 could be expected. Such a reduction of the background results in an enhancement of the sensitvity for the search for neutrino oscillations by roughly a factor of 10 assuming the same amount of measuring time.

The actions performed during the upgrade covered a partial dismantling of the iron bunker (about 4000 tons of steal were moved), the production of 136 large scale scintillation counters, the attachment of these counters to the remaining walls of the bunker and after that the reconstruction of the bunker to regain the previous strength of the shielding. Within the 2-3 m thick walls of the 7000 ton heavy iron bunker a total of 300 square metres of plastic scintillators were installed to detect penetrating muons. The picture below shows a sketch of the iron bunker, partially cut away to show the new veto counters in red.

The picture below shows a single plastic scintillator module. It is 4 m long, 65 cm wide and 5 cm thick and consists of a material which emits light in the visible range when a charged particle (e.g. a muon) crosses. To demonstrate the light tansport the track of a single photon is shown in the picture which is guided inside the scintillator via total internal reflection.

At both ends of the modules there is a 180 degree light bending system, guiding the light to 4 so called photomultipliers at each end. Those devices are shown in red. They are extremely sensitive to light and are used for the detection of single photons. The light bending is supported by aluminium mirrors shown in blue at the ends of the modules. Because of the light bending the modules can be put close to each other providing the smallest possible gaps between the modules and minimizing the number of particles flying through the gaps without being detected.

The success of the KARMEN upgrade can by measured by a comparison of the background rates before and after the upgrade if one normalizes the number of background events to the measuring time used in each case. Both the background with a single prong signature (i.e. only one isolated event is measured without looking for successive events) as well as background with a sequential structure contributes. The last is especially important because one looks for sequential events if one searches for `nm ® `ne oscillations. The picture below compares the event rates measured before and after the upgrade vs. the visible energy of the events in the detector for both cases.


The black lines show the background rate before the upgrade, the yellow areas the one after the upgrade. Above 20 MeV, i.e. in the energy range which is most important for the search for neutrino oscillations, in the case of a single prong signature a reduction by a faktor of 30 has been achieved, in the case of a sequential signature the background is reduced even more by a factor of 40 like it was expected from simlation calculations. This reduction improves the background situation so much, that one can speak about a new quality in detecting neutrinos with KARMEN. E.g. neutrino reactions with a single prong signature can now be detected with a signal to background ratio of 50:1. Moreover the background for the search for `nm ® `ne oscillations induced by cosmic muons could be essentially eliminated.

Enough theory ... have a look at our KARMEN photo gallery, including a collection of photographs of the new VETO frames in England.

Home TO Latest Results