Usually in a combined material and geometric nonlinear analysis of a space frame, the larger or more complicated the analytical frame, the more difficult it is to find the consistent stiffness, whose assumptions of loading or unloading for all elements are consistent with the directions of later-calculated strain increments. Moreover, it is difficult to maintain the consistent stiffness strictly during a large space frame analysis with existing computers, because an overall analytical increment depends on the smallest increment in all elements requiring the consistent stiffness.
Therefore, the authors presented a combined nonlinear analytical method in the former paper. The method has a member model with six sub-elements along the member axis and fibers at each end of the sub-elements to trace the stress-strain relation, selects an appropriate stiffness among trials in case the consistent stiffness is not found in the member analysis, and uses larger analytical increments in the overall analysis and subdivided-capable increments in the member analyses. However, whole buckling-like phenomena were excluded from the analytical subjects.
In the paper, the analytical method is applied to large deformation analyses of the overall model tests of a large and rigidly jointed space truss. The conclusion is as follows:
Using the analytical method, the tested models of the large and complicated space truss were analyzed successfully. Both the analytical results for the asymmetrical vertical load test and those for the transverse directional test agreed very well with the experimental results up to a large deformed state including elastoplastic member-coupled buckling.
In the analysis of the asymmetrical vertical load test, coupled bucklings of continued chords as shown in the experiment were well simulated with a good correspondence of the buckling deformed states. Having investigated the coupled buckling through the simple partial model study, it was observed that the existence of adjacent lattices to allow different axial forces between both chords causes large differences in post-buckling behavior between the chords and makes the coupled bucking asymmetrical as shown in the experiment and the analysis.
In the analysis of the transverse directional test, the complicated collapse behavior just above the bases, in which the members were not determined experimentally to be buckled or bent with a small number of measurements in comparison with the large frame, was successfully analyzed.
The analytical behavior showed an exchange of a lighter buckled chord and heavier buckled one in a coupled buckling under the cyclic loading, though it depended largely on the material properties.