Determination of nonlinear behavior of a ball joint system by model updating

Behavior of a ball joint system when evaluated discretely and disconnected from structure, does not represent its true behavior in the structure. In the present work, to study the actual behavior of a typical ball joint system, available experimental responses of a full scale double layer grid have been used as reference data. Using inverse problem method and specifically finite element model updating technique, an optimization problem is defined. The objective function that must be minimized is the difference between existing experimental deflections and corresponding analytical deflections obtained from an appropriate finite element model which has been prepared for the double layer grid. With modeling a general nonlinear behavior for joints in the finite element model of the double layer grid and considering it as updating parameter (optimization variable), the optimization problem is solved through the genetic algorithm. Therefore, by updating the finite element model of the double layer grid, the behavior of the ball joint system can be estimated in real conditions. The obtained results show that the updated model can predict with good accuracy the true deflections of the double layer grid. Also, the attained force-displacement relationships of the ball joint system demonstrate that its behavior in the initial stages of loading is nonlinear with a relatively low stiffness. However, after that it becomes stiffer then it behaves linearly such as in the ideal conditions.

► The actual behavior of a ball joint system was studied through inverse modeling.
► Nearly accurate updated model of a grid was obtained in terms of its deflections.
► Axial force-displacement behavior of the joint is nonlinear at low force values.
► The force range of joint nonlinear behavior depends on support condition of the grid.
► When applied force increase, the joint behaves linearly and show more stiffness.