No abstract

No abstract

It is widely known that the accuracy of the finite element method has a direct relation with the type of elements and meshes. Another issue which has remained less treated is the impact of loading type on the accuracy of responses. Changing the applied forces from concentrated to distributed loading has a great effect on the accuracy of certain types of elements and this action can greatly reduce their accuracy. Particularly in the coarse meshes, it creates a critical situation. Some elements do not have the ability to provide the exact answers in stated conditions. For example, the well-known plane element, LST, demonstrates promising performance under concentrated shear and bending loading as well as surface traction. In the case of distributed loads and coarse meshes, its accuracy diminishes considerably. To remedy this defect, in this paper, a new higher-order triangular element is proposed by using natural assumed strain approximation. Various numerical examples demonstrate high accuracy and efficiency of the element in comparison with common well-known finite elements in analysis of structures under distributed loading.

Existing methods to predict the seismic demand of non-structural components in current seismic design guidelines do not generally consider the intensity of the design earthquake and the expected performance level of the lateral load bearing system. This limitation is especially important in performance-based design of buildings and industrial facilities in seismic regions. In this study, a novel multilevel approach is proposed to predict the seismic demand of acceleration-sensitive non-structural components using two new parameters obtained based on site seismicity and seismic capacity of the lateral load carrying system. The main advantage of the new method is to take into account the seismic hazard level and the expected performance level of structure in the calculation of the seismic demand of non-structural components. Based on the results of a comprehensive reliability study on 5 and 10-storey steel frame structures, the efficiency of the proposed approach is demonstrated compared to the conventional seismic design methods. The results, in general, indicate that the current standards may provide inaccurate predictions and lead to unsafe design solutions for acceleration-sensitive non-structural components, especially in the case of higher seismic intensity or medium performance levels. It is shown that the estimated accelerations by NIST and ASCE suggested equations are up to 50% and 80% lower than the minimum demand accelerations calculated for the studied structures, respectively, under the selected design condition. Based on the results of this study, a simple but efficient design equation is proposed to estimate the maximum acceleration applied to nonstructural components for different earthquake intensity levels and performance targets.

No abstract