Grasping sheet metal objects for manufacturing operations requires custom-made robot-mounted end-effectors to grip the parts. Modern end-effectors use multi-type grasp where a combination of gripper types such as suction cups, magnets, and fingers may be used. This paper presents a genetic algorithm-based approach of grasp design automation. The algorithm first generates an option space of possible grasping locations by analyzing the geometry of the sheet metal part and then uses a genetic algorithm to optimize the grasp using up to five magnets and suction cups. The algorithm includes as fitness criteria the factor of safety of the total gripping force against part weight, the unbalanced moment created by the gripping forces and part weight, the cost of the grasp, and three combinations of these parameters. The GA features asexual reproduction, mutation, and elitism. The algorithm is implemented in the Siemens NX™ Knowledge Fusion language and on Microsoft VBA code. The paper presents detailed test results and sensitivity analyses that indicate that genetic algorithms can produce viable solutions for multi-type grasp configurations and that the algorithm behaves in response to varying its control parameters in ways that are logically anticipated.