Author/Authors :
Vivio Francesco نويسنده Department of Enterprise Engineering, University of Rome Tor Vergata, Via del Politecnico 1, 00133 Rome, Italy , Fanelli Pierluigi نويسنده DEIM, University of Tuscia, Largo dell’Università, 01100 Viterbo, Italy , Ferracci Michele نويسنده Department of Enterprise Engineering, University of Rome Tor Vergata, Via del Politecnico 1, 00133 Rome, Italy
Abstract :
In aeronautical and automotive industries the use of rivets for applications requiring several joining points is now very common. In spite of a very simple shape, a riveted junction has many contact surfaces and stress concentrations that make the
local stiffness very difficult to be calculated. To overcome this difficulty, commonly finite element models with very dense
meshes are performed for single joint analysis because the accuracy is crucial for a correct structural analysis. Anyhow, when
several riveted joints are present, the simulation becomes computationally too heavy and usually significant restrictions to
joint modelling are introduced, sacrificing the accuracy of local stiffness evaluation. In this paper, we tested the accuracy of
a rivet finite element presented in previous works by the authors. The structural behavior of a lap joint specimen with a rivet
joining is simulated numerically and compared to experimental measurements. The Rivet Element, based on a closed-form
solution of a reference theoretical model of the rivet joint, simulates local and overall stiffness of the junction combining
high accuracy with low degrees of freedom contribution. In this paper the Rivet Element performances are compared to that
of a FE non-linear model of the rivet, built with solid elements and dense mesh, and to experimental data. The promising
results reported allow to consider the Rivet Element able to simulate, with a great accuracy, actual structures with several
rivet connections.
