Five-axis hybrid robots are more extensively used in the machining of complex parts,
owing to their high dynamic performance and flexible attitude adjustment. Ball-end milling is
widely used in sculptured surface machining. Tool orientations can be adjusted for higher machining
efficiency and better cutting performance. However, hybrid robots’ complex kinematic structures
raise challenges to tool orientation planning. To generate a smooth toolpath that has a good cutting
performance and satisfies all the geometric and mechanical constraints, a tool orientation
optimization model is presented. The weighted sum of the joint kinematic performance index and
the effective tool diameter indexes are taken as the objective function. By constructing the robot
workspace in advance, the computation in the generation of feasible tool orientation regions is
reduced. The feasible orientation regions are further simplified as linear inequality constraints which
are updated during iteration. In each iteration, the initial nonconvex optimization problem is locally
approximated by a convex problem and solved by quadratic programming (QP). Simulations and
experiments are conducted for validation of the proposed method.