I implemented Node-To-Surface contact algorithm (Wriggers, Peter, Computational contact mechanics., Berlin: Springer (ISBN 3-540-32608-1/hbk). xii, 518 p. (2006). ZBL1104.74002.). The code is done by first forming the contact element, comprising of a slave node on the ball and a master line of the base. In the example of a ball penetrating into a soft block, the simulation could run until a point that the projection within a contact element falls outside the corresponding master in the Newton-Raphson iteration. Subsequently, the implicit NR loop does not converge anymore. Does anyone have experience dealing with this issue?
This is the issue, noted the black circle, the slave node (blue) located close to one of the master nodes (purple), leading to the issue above.