The research aims to assess acoustic comfort in office spaces, focusing on mutual disturbances between occupants and the contribution of HVAC systems for thermal comfort. A case study was conducted in an open office at Sapienza University of Rome. Acoustic descriptors, such as T60, STI, Das, and D2s, were evaluated through modeling and in situ measurements, considering various sound-absorbing materials like suspended ceilings, wall panels, and separation screens. The impact of HVAC power on these descriptors was also assessed. A validated model was used to create a training database for a neural network predicting acoustic descriptors based on room volume, absorption area, HVAC power, and occupancy. Post-intervention data were used to develop a predictive model to identify the optimal corrective solutions, aiming for ISO 22955 and UNI 11532-2 acoustic comfort levels while minimizing sound-absorbing material area. The model aids early-stage design by simplifying complex calculations.
Fiorini et al. (Tue,) studied this question.