Research On Temperature Effect Of Continuous Beams Based On Stochastic Subspace Identification

With the development of bridge construction technologies and related scientific theory, andnow the bridge health monitoring has become a hot area. Bridge health monitoring needs to referto dynamic parameters of the bridges, in order to carry out damage identification of bridgestructure, so as to determine the status of bridge health. Natural frequenc y as an importantindicator of bridge dynamic parameters, is vulnerable to environmental interference, especiallythe influence of environmental temperature changes.It leads to that the natural frequency dataobtained in the actual project includes the influence of temperature,.Therefore, to determine theinfluence of temperature on natural vibration frequency, and obtain the true value of the bridgefrequency is the foundation of the bridge health monitoring.Firstly, the three temperature fields,which are day-change temperature field, sudden-droptemperature field and the annual-change temperature field, have been discussed and thecharacteristics of the thermal stress of continuous beams are analyzed. The thermal stress in thestatically indeterminate structure (continuous beam) consists of self-restrained thermal stress andsecondary thermal stress.While there is only self-restrained thermal stress in staticallyindeterminate structure (simply supported beams). Self-restrained thermal stress is caused by thefiber constraints between each layer of the inside; and the secondary thermal stress is caused bythe constraints outside(such as bearings, etc.). Besides, the thermal stress of the continuousbeams are theoretically analyzed, and self-restrained thermal stress and secondary thermal stressare calculated, Three kinds of methods to calculate the secondary temperature forces are given,and practical formula of horizontal and vertical thermal stress are obtained. The thermal stresscaused in a three-span continuous box-girder bridge under the sunshine temperature field iscalculated as examples.Then, the present parameter identification methods based on environmental incentives aresummed up and discussed. Among them, the traditional Stochastic Subspace Identificationmethod can identify the system natural frequency easily. Just simply enter white noise into the system, and obtain the output data. And the output data are analyzed to determine the appropriatesystem order, you can obtain the identified system natural frequency. The method is optimized totake only several output signal (if the signal quality is good, only one output signal is needed) asthe "past" output data to take place of the full output data in the traditional random subspacemethod. It greatly improves the computational efficiency, and is also effective to avoid the errormode and modal omissions.By using the optimized SSI method, a actual continuous box-girder bridge on the Xinchengstreet is tested to identify the natural frequency under the sunshine temperature field. Point outthat the relationship between temperature and vibration frequency is negative. When temperatureincreases, the natural frequency identified by SSI method is reduced from the situation withoutconsidering temperature effect, and the largest reduce degree is6.32%, By using the traditionalpeak-picking method, the identified results are checked. What’s more, combined with theidentification of dynamic load test of the actual project, the results of SSI are verified, and have ahigh identification accuracy.Finally, the influence factors and their influence degree on natural frequency of continuousgirder bridge under temperature fields are discussed. Among them, the reduction of elasticmodulus caused by temperature changes, is one of the main factor of the changes of continuousbeam natural vibration frequency. The influence degree of elastic modulus reduction oncontinuous beam frequency is5.13%. The influence degree of boundary condition changescaused by thermal stress on the continuous beam first-order frequency is biggest, and the value is1.86%.