The comparison of measured dynamic characteristics or response of large structures with that of an appropriate finite element model with all its underlying assumptions often reveals discrepancies. This may be due to improperly determined parameters, such as interstory stiffness, mass of different stories, and the modulus of elasticity of the concrete, as well as the inadequacies of the model.The measured dynamic response generally occurs in one of three forms: time response, frequency response, and modal data. For time response data, either in free vibration or for a known input, parameters are estimated by proper adjustments to match more closely the measured motion. For steady-state frequency response, a sinusoidal load (or synchronized loads) is input mechanically and the response, both in amplitude and in phase, is measured for different frequencies of excitation. Damped resonant frequencies, the associated modal damping ratios, and the corresponding mode shapes are the measured quantities for modal data.The finite element models used for civil engineering structures often incorporate a large number of degrees of freedom. Measured response is sparse and usually limited to the lower frequency range. A procedure for estimating these parameters must be able to allow for the small amount of data and must utilize efficient numerical algorithms to determine the best parameters. Nonlinear least squares, within a Bayesian framework, is such a method. It can be applied to time-history data, steady-state response, and modal characteristics. This method is used to determine aerodynamic coefficients of a scale model of a suspension bridge deck from free response data in a wind tunnel, stiffness parameters from frequency measurements of a 5-story steel building frame loaded by mechanical exciters on the roof, and stiffness parameters from modal data of a 12-story reinforced concrete frame, as obtained from transient wind observation of lateral accelerations.
The response of the bovine ovary to pregnant mares’ serum and horse pituitary extract
