Finite element (FE) analysis is being used extensively in aerospace industry to reduce component testing costs as well as to decrease design cycle time. One of the biggest challenges in FE analysis is to accurately model the joints between the components that constitute the machinery. Riveted joints are one of the various types of joints, that are widely used in aerospace industry. To accurately model riveted joints in an FE analysis, rivets need to be modeled with solid elements and contact elements need to be used. This would greatly increase the computational cost.
There are situations where the proximity of the riveted joints is neither very close nor very far from the region of interest. In such situations, the stiffness of the riveted joint needs to be captured accurately, though the stresses need not be very accurate. Various empirical as well as analytical solutions (like Huth, Swift, Tate & Rosenfeld, Grumman etc) are available in the literature to calculate the lateral stiffness of rivets. There are various assumptions that go into these models and stiffness obtained from the different analytical models does not match each other.
The first objective of the current study is to compare the stiffness obtained from the above mentioned analytical or empirical models with a 3D riveted joint FE model with contact elements over a range of geometric, material and load parameters. The second objective is to obtain rivet stiffness for a particular riveted joint in a machine assembly subject to certain type of loads. This is done in two steps. In the first step a full-fledged 3D submodel is used to obtain the rivet stiffness. The second step is to use the stiffness obtained and model the rivet with appropriate lumped springs in the global model, thereby reducing computation cost and time. This approach has not been widely studied in the literature.