Phacoemulsification cataract surgery, a minimally invasive technique to remove a cloudy lens from the eye, is one of the most commonly performed surgical procedures in the western world. Conventional training for this procedure involves didactic lectures and practice on pig and human cadaver eyes, none of which allow trainees to form an accurate predictive model of human tissue behavior during surgery. A virtual environment simulator for capsulorrhexis, one of the first steps in cataract surgery, has been developed that allows a trainee to use surgical instruments to excise a circle of tissue on the anterior side of the lens capsule through tearing. The simulator invokes a deformable mass-spring-damper mesh model of the tissue that can be grasped and torn via shearing. A novel algorithm for mesh division and maintenance enables realistic tearing behavior. The trainee controls tool motion using a 3-degree-of-freedom haptic device, and haptic feedback is provided from the virtual tissue. Although the haptic feedback in a real capsulorrhexis procedure is below the human threshold of haptic sensing, this simulator enables an experiment to determine the effectiveness of “haptic training wheels” – the idea of haptic training for a task without haptic feedback.