Dear Yaping,
thank you very much for your offer of help. We appreciate it very much
There are indeed multiple things I would like to ask about with regard to the simulation of the scintillator wall and subsequent choices.
We are now in a stage of starting simulations. We can already identify
some of the issue that we think will be important.
At the low collision energies the high-charge fragment will play important role an we need to simulate them correctly. So I would be quite interested to learn how you did that and what you would recommend.
This question will be answered by my colleague, Dr. Hua Pei, who is take charge of the simulation.
As we gave a first though to it we are a little bit worried bout possible effects of saturation when using SiPMs for readout. I know that people who use standard photo multiplies sometimes modify its response for linearity (HADES did that).
In our project, we decide to use the standard PMT for readout, and readout two dynode signals (for example, we readout the 5th and 8th dynode signals at same time) for each PMT to cover the saturation effects you mentioned.
With this is connected the choice of the scintillator thickness. Signal from a thin scintillator is bound to be affected by Landau fluctuations, like in the STAR EPD. It seems that the common choice is around 5cm. As I understood from Nu you plan to be roughly in that range. So I'm wondering was the main reason for such a choice. Was it connected with the light collection time?
We used scintillators of thickness 4 cm in our prototype development. And currently we are testing the choice of thickness 2 cm for the scintillator. We will/need to compare the performance in simulation. As you mentioned, the thicker scintillator, the less Landau fluctuations.
Another issue of outmost importance is the precision of event plane and mainly centrality determination. I can foresee that we will be able to reconstruct reasonably well the event plane. For the centrality determination we will likely have to rely on charge summing similarly as HADES did in their proton fluctuation paper. I'm wondering how we could optimize that. I would expect that thicker detector may help or perhaps using neutron sensitive material. Any advice or experience in this direction is most appreciated. Since we are now at the star I'm sure there may be other issues that we are not aware of yet.
Our ZDC detector consists of 24 sectors covering azimuthal angle of 360, each sector is divided into 8 wedges with same length in r-phi direction. In total there are 192 modules (384 readouts). Then the geometry is divided into 8 rings. The current design has very good performance on the 1st order event plane reconstruction (high EP resolution). To the centrality determination, we apply the BDT method to determine the centrality offline. The input variables for the BDT training include ratios of the collected energy in each ring to the total collected energy by ZDC, number of fired modules, total energy collected by ZDC. Finally we can have a reasonable efficiency with high purity to separate central, semi-central and perpherial collisions.
Actually is there any paper describing you detector?
We do not have paper yet. But later, I will send you two talks from our project about the EP reconstruction and centrality determination.
Best regards, Yaping