Modular product design-sharing components across multiple products;is viewed as a convenient way to offer high product variety with low production cost. This study presents a novel method for the early identification of shareable components within a family of structures. Component sharing is first enforced as a constraint, and then posed as the outcome of the minimization of production costs considering the economies of scale. Based on the decomposition-based assembly synthesis method, the modular structural component design problem is posed as the optimal selection of joint locations and joint types within two structures, evaluating the geometric similarity of the resulting components as potential modules during the optimization process. For the given variant structures the designer defines the possible locations where a joint can be placed, and for each possible joint location also provides feasible types of joints, including a type for "no joint", among which the optimal selection can be made. The problem is solved to (1) minimize the reduction of structural strength due to the introduction of welds in each structure, (2) maximize the manufacturability/assemblability or equivalently minimize the manufacturing costs of two structures, considering the opportunities of cost reduction via component sharing under given production volumes. A multi-objective genetic algorithm is utilized for effective evaluation of trade-offs between different criteria. Real life case studies including automotive body-in-white models are presented to demonstrate the capabilities of the method.