Longitudinal structure optimization for the high density electromagnetic calorimeter

High density electromagnetic sandwich calorimeters with high readout granularity are considered for many future collider and fix-target experiments. Optimization of the calorimeter structure from the point of view of the electromagnetic shower energy, position and direction measurement is one of the key aspects of the design. However, mostly uniform sampling structures were considered so far. We developed a semi-analytical approach to study calorimeter performance based on the detailed Geant 4 simulation, which also allows to compare the expected performance for different non-uniform longitudinal readout structures. For multi-objective optimization, procedure based on the genetic algorithm is complemented with non dominated sorting algorithm. This methodology opens new prospects for calorimeter design optimization directly addressing specific measurement scenarios or optimization goals.