A probabilistic approach for the full response estimation of tall buildings with 3D modes using the HFFB

• Traditional HFFB methods are affected by shortcomings limiting their applicability.
• A novel method is proposed to overcome them within a probabilistic framework.
• Probabilistic models for the mean loads and load spectral structure are developed.
• The method is general and valid for any response parameter and any kind of building.
• The application to two representative case studies shows the efficacy of the method.

The high frequency force balance technique for the assessment of wind actions on tall buildings is widely adopted mainly due to its relatively fast setup and capacity for providing estimates of the resultant wind action independently of building geometry. Nevertheless, some shortcomings represent a limit to its adoption as an experimental methodology. In particular, the lack of knowledge on the pressure field makes it necessary to make arbitrary assumptions on the floor loads that can consequently generate important errors – usually more relevant for buildings with non-linear and coupled modes – in the response estimation. Also, the adoption of the method is often hindered by the limitation of considering only the first three fundamental modes in the response assessment.The novel methodology which is proposed in this paper has the aim of overcoming the aforementioned problems. The reliability framework in which the method is set allows the consideration of the uncertainties arising from the lack of knowledge on the pressure field through the definition of probabilistic models and provides the possibility of estimating any response parameter of interest, for any kind of building, considering exclusively the base measurements. The application to two representative case study buildings with coupled and non-linear modes shows, through comparison with reference results obtained from synchronous pressure measurements, the effectiveness of the method in providing conservative estimates of the response parameters.