Physical Parameter Identification of an Rc Frame Structure on Elastic Foundation

In this paper, the simple genetic algorithm (SGA) is improved by combining with the simulated annealing algorithm (SAA) for the parameter identification of a reinforced concrete (RC) frame on elastic foundation. SGA adopts parallel search strategy, which is based on the concept of "survival of the fittest" in optimization while SAA adopts a serial form and the process is endowed with time-variety probable jumping property so that local optimization could be prevented. The global searching ability is developed by combining the two methods and the new algorithm is named genetic annealing hybrid algorithm (GAHA). Modal experiments were carried out on a four-storey RC frame structural model with isolated embedded footings in laboratory. The measured natural frequencies and mode shapes have been utilized to identify the physical parameters of the frame by the proposed method. Four cases of concrete elastic modulus and foundation dynamic shear modulus are identified, and the results are compared with the usual sensitivity methods (SM). By model updating, the results show that the elastic modulus of concrete increases with respect to the storey. The identified elastic modulus of the concrete is generally larger than that found by compressive testing because the dynamic modulus of concrete is larger than the static modulus of concrete. The identified soil dynamic shear modulus also increases with the storey since the soil property depends on the pressure exerted on the soil. It is also shown that the identified results by GAHA are better than that of SM.