Comparative Analysis of Numerical Simulations and Experimental Tests for Acoustic Performance in Horizontal Partitions of Buildings

Acoustic forecasting models from computational simulation (SC) has been the tool used to evaluate the acoustic behavior of the building, even before the work is done, thus enabling the reduction of post-adaptation costs. This research aims to evaluate the reliability of computational simulation by analyzing data from simulations and field measurements (MC) of 14 case studies, in relation to horizontal partition systems (SVH). For the 228 SVH data, of which 116 were for SVH-RA and 102 for SVH-RI, the comparative study showed 38% acceptability between SVH-RA measurement and simulation; and 46% for SVH-RI. This acceptability refers to the values within the ±2 dB range of the aforementioned uncertainty. The data also showed variation between SC and MC, sometimes presenting larger SC, sometimes larger MC. The predicted hypothesis was that SC would always be higher due to the ideal conditions of the simulation, which do not occur in the field. However, where situations of MC greater than SC occurred, it was deduced that deciding on the simulation means choosing to work in favor of safety, since the minimum level of the standard is projected and better results are obtained in the field. It was also verified, through variation of the input data, that the fidelity with the technical specification of the designed systems reproduces reliable values, and the opposite, results in doubtful and discrepant data. In general, the results of the comparative analysis point to the use of computer simulation as positive, since part of the percentage that is not acceptable in this study represents doubtful measurements, and therefore, the simulation is an effective predictive tool and useful in the search for acoustic quality in residential buildings.