Effect of turbulent flow on wall pressure coefficients of block arrays within urban boundary layer

Publication date: 1 May 2016
Source:Building and Environment, Volume 100
Author(s): N. Ikegaya, C. Hirose, A. Hagishima, J. Tanimoto
Various numerical simulations have been developed to evaluate the mean ventilation rates of a target building; however, the manner in which turbulent flow generated by buildings and surrounding conditions affects the mean and fluctuating ventilation rates is not well understood. Therefore, we have performed large-eddy simulation of flow and pressure fields above two types of block arrays (lattice-type square and staggered pattern) to clarify the turbulent characteristics of estimated ventilation rates based on pressure coefficient distribution on the block faces. The concept of short-term ventilation rates, which are estimated from filtered pressure coefficients, is introduced to investigate the temporal variation in the estimated ventilation rates for different locations of the block faces and arrangements. First, mean and second-order turbulent statistics agree well with previous results obtained from both wind-tunnel experiments and numerical simulations, indicating that the turbulent characteristics of the flow above urban-like arrays are well reproduced. Second, short-term ventilation rates are found to vary temporally and spatially. Therefore, these values instantaneously become larger or smaller than the mean ventilation rates. In addition, complex fluctuation patterns of pressure coefficient distribution are found to be caused by air flow introduction from the block arrays into gaps between the blocks as well as by small-scale turbulence generated by surrounding buildings themselves. Lastly, the temporal statistics of filtered pressure coefficients show that the short-term ventilation rates can possibly become stronger or weaker than the mean ventilation rates, whereas the accumulated fluctuating ventilation rates are almost comparable to those estimated by mean pressure coefficients.