Experimental Validation of a Vision-Based Technique to Estimate the Crowd Loading on Stadium Grandstands

Stadium grandstands may experience high vibration levels due to the crowd action. Existing codes and guidelines used for the design and the assessment of grandstands highlight the need of an already ongoing revision, since there is a lack of information about the real nature of the crowd excitation and its accurate evaluation. This paper describes a highly improved computer vision method, based on the use of Digital Image Correlation (DIC), able to provide reliable estimates of the crowd induced loads on stadium grandstands. The effectiveness and the applicability of DIC is validated through a series of experimental tests held at the G. Meazza stadium in Milan. A first study with 27 jumping volunteers is carried out to assess the performance of DIC in a real environment. Then, the procedure is extended to the case of a larger crowd made of 108 individuals. In both cases, the forces estimated through image processing have been used in conjunction with the impulse response function of the stand to determine the resulting structural vibrations. The comparison between measured and computed vibrations provides a good agreement both in the time and the frequency domain.