Nonparametric independent component analysis for detecting pressure fluctuation induced by roof corner vortex

The Nonparametric Independent Component Analysis (NICA) method is introduced to detect/infer the presence of roof corner vortex from experimentally measured wind pressure data. Different from the widely recognized Proper Orthogonal Decomposition (POD) method, which decomposes orthogonal pressure modes from the measurement, the NICA method is to separate the statistically independent pressure components/modes from the pressure data. Assuming that the measured wind pressures are caused by mixed wind flow phenomena, this paper attempts to decompose these independent flow induced pressure modes from the measured wind pressure data. Application of the NICA method to wind pressure data collected on the Texas Tech University (TTU) laboratory building justifies that the NICA method successfully separates/extracts from collected wind pressure data the pressure mode that is highly correlated with roof corner vortex. The corresponding time series coefficient of this mode is indicative of the presence of roof corner vortex.