Prototype beam test

The final detector system will consist of optically isolated segments. Therefore, the performance characteristics of the chosen geometry can be studied with a detector system consisting of one single segment only. A prototype has been built with a 5 cm thick $46 \times 23$ cm$^2$ aerogel layer that consisted of the Matsushita SP-50 material. The diffusion box has been read out with two 5 inch PMTs, Photonis XP4572B/D1. All passive surfaces inside the diffusion box have been lined with Millipore paper as a diffuse reflector. The detector has been tested during the KEK PS-T500 beam test (see also section 5.2). Figure 51 shows the sum of both tubes calibrated in number of photoelectrons. An average of roughly 19 photoelectrons is observed for pions with 1.2 GeV/$c$ momentum. It is planned to operate the detector as a veto set by a coincidence between the two PMTs. Figure 52 shows the probability of a pion for not setting the veto as a function of the threshold placed on the individual PMTs. Note that in this figure, for example, a threshold of 1 p.e. means that at least a total of 2 p.e. has to be detected. The final detector system will consist of three layers. Figure 53 shows the expected performance, if a one-out-of-three coincidence is required to set the veto. A threshold of 1 p.e. on each of the two PMTs in a segment will result in a pion suppression of roughly $10^{-4}$ with a an accidental veto for kaons of roughly 5%.