Research On Modal Property And Parameter Identification Of Simply Supported Bridge With Crack

Bridge damage most directly reflected in the cracks of the bridge.As a concrete bridge in its service period,cracks cannot be avoided.Therefore,the analysis of the dynamic characteristics of beams which have cracks and recognize the damage is significant.Currently,a variety of identification technologies have been proposed in the field of structural damage identification,and vibration-based damage identification technology has been promoted as an extremely important part of the bridge health monitoring system.In the analysis of cracked beams,there are usually two kinds of crack model: one assumed the beam cracks in the whole vibration process is always in the open state,called open crack model.And the other one assumed beam cracks will change its opening and closing state over time,called breathing crack model.Most domestic scholars use open crack model when they analysis the vibration of beams with cracks,and the actual state,since the beam loads and other reasons,alternately opening and closing of cracks phenomenon occurs during vibration,breathing crack model is more reasonable.Following researches were conducted in this thesis.Firstly,this paper choose the open crack model to analysis the impact of the presence of cracks on the dynamic characteristics of simply supported beam preliminary.The analysis result showed that the dynamic characteristics of simply supported beam with cracks is different from a beam having a complete fracture.For a complete beam,natural frequencies are constant during the whole vibration.While cracks occur in simply supported beam,its vibration frequencies decreases.The depth and location of the crack can affect the dynamic characteristics of the beam,but when the crack is located in a modal nodes,it will have no effect on the natural frequency.Damage location and severity identification were realized based on neural networks using frequency as input variables.Moreover,simply supported beam with open cracks were simulated through numerical simulation.Modal curvature,modal curvature difference,modal flexibility difference and modal flexibility difference curvature were used as damage identification indicators and their identification effects were verified.Based on neural networks,damage identification of simply supported beam without baseline model was demonstrated.And then apply the breathing crack model,and compare the time-frequency analysis methods of breathing cracks.Finally,choose the Hilbert-Huang transform technology to analysis cracked beams was a detailed study and acquire the simply supported beam dynamic response of Hilbert spectroscopy.The results show that the vibration of a simply supported beam with a breathing crack appeared higher harmonic components in integer multiples of the fundamental frequency of the position in the spectrum,showed that the appearance of breathing cracks make the vibration of beams showing a significant nonlinear characteristics.The fundamental frequency of simply supported beam decreases with the increase of relative depth of the crack.Conversely,the order and the amplitude of higher harmonic components are increased with the increase of injury degree.Once cracks occur in simply supported beam,the instantaneous frequency of beam is no longer a constant,but fluctuate within a certain range,and the range of ?IF(the mean variation of the IF)increasing with the increase of degree damage.Finally,extracting the instantaneous frequency by EMD and proposing to use ?IF(the relative difference of instantaneous frequency)?2IF?(variance of instantaneous frequency)??(peak ratio)and?(ratio of frequency and harmonic concepts)to identify the damage of a simply supported beam.This paper improved that the mean variation of the IF ?2IF? ?? and? increase with ?(relative depth of the crack)following a polynomial law,and they have the initial preliminary ability to identify the breath cracks.Besides,? is sensitive to crack's location.