Abstract
Abstract
In silicon detectors, the self-shielding behavior of the dense electron–hole plasma generated along the track of an energetic heavy ion plays a critical role in governing the dynamics of signal formation. Although general-purpose TCAD tools are theoretically capable of modeling this phenomenon, the need for an extremely fine mesh can result in prohibitively high computational resource requirements for simulating full-scale detectors. Conversely, detector simulation approaches such as the AvalancheMC implementation in Garfield++ do not account for this plasma effect, which can lead to inaccurate waveforms. Previously, we introduced an adaptive-mesh framework that accurately modeled plasma effects for 5.5 MeV alpha particles with substantially lower computational resource requirements than equivalent TCAD simulations. In this work, we systematically evaluate the same framework across a broad range of ions (from alpha to titanium), energies (5.5–910 MeV), and detector conditions (five different planar silicon detectors, front and rear irradiation). The code is integrated with GEANT4 for energy deposition and uses a nested mesh that dynamically refines around the ion track. Predicted charge-signal rise times are compared against published experimental data from independent studies across 16 test cases. The absolute relative difference between simulation and experiment is ≤ 11% for 13 out of 16 cases, and ≤ 17% for all cases. These results demonstrate that the adaptive-mesh approach generalizes robustly from light to heavy ions without case-specific tuning, offering a computationally efficient alternative to TCAD for plasma-effect simulations in depleted silicon detectors.