| Hybrid simulation test is an anti-seismic test method that combines numerical simulation with the real physical test.It can simplify the real test model,thus saving the test cost and enhancing the operability of the test.In traditional hybrid simulation experiments,numerical models often only take out uncomplicated parts of the overall structure for analysis,and it is difficult to perform accurate simulations for new or complex structures.Therefore,it is necessary to calculate the parameters of the substructure model in real time.This paper first establishes a numerical simulation program for the hybrid simulation model update test.The uniform design method is used to establish the constitutive parameters of the steel and the physical geometric parameter sample space.A verification operator structure model is established for each parameter to be identified in the space,and then the parameters of the model closest to the test substructure response are selected from the verification operator structure group and transmitted to the model of the numerical substructure according to the observation value of the test substructure.Thereby the purpose of updating the parameters of the numerical substructure model in real time has been achieved.Firstly,Numerical verification of indicator is selected,then parameter saliency is analysised,and the model update was performed.The two parameters and three parameters of the square steel tube and H-section steel were identified separately,and the feasibility of the method was verified by numerical simulation.In this paper,it is proved by experiments that the calculation of substructure parameter identification and updating has a good effect.In this paper,the single column of the single-row frame structure is used as the test sub-structure to complete the experimental design of the model update hybrid simulation test.The test process takes the pseudo-dynamic test of the overall model as the real value,and compares the relative cumulative error of the displacement and force of the traditional sub-structure overall model and the model update hybrid simulation test model,indicating the validity of the model update.The elastic phase model update result is 1/3~1/5 of the relative cumulative error of the traditional hybrid simulation test.The elastic phase compares the results of different initial elastic modulus parameters,indicating that the initial parameters have little effect on the recognition results,and the initial parameter identification results are between 0.86 and 0.88.Since the experimental substructure is consistent with the calculated substructure assumed model,the relative cumulative error of the experimental substructure and the calculated substructure feedback force in the hybrid simulation test is updated by the comparison model,and the feedback force results under the same displacement are analyzed.Under the same displacement,the relative cumulative error of the feedback force of the substructure members is stable within 5%.In the first working condition of the plastic phase,the two-parameter model update hybrid simulation test is completed,and the obtained model update test is 1/4～1/6 of the tr aditional hybrid simulation test error.Then,the three parameters of plastic constitutive are identified.The error of parameter identification update when the plastic development is more complete can still be smaller than the traditional hybrid simulation test. |
