The paper presents a novel technique for optimizing the design of parts intended for 3D printing, with a focus on enhancing structural integrity and material utilization. The study addresses the challenges of lightening of 3D parts bounded by cylindrical, spherical surfaces, and planes. A smart technique of generating spherical voids of varying diameters within 3D parts of complex geometry is proposed. The optimization procedure is aimed to maximizing the total volume of spherical voids while considering 3D printing standards. A mathematical model of the nonstandard packing problem is constructed, using normalized and adjusted phi-functions for describing distance constraints. A solution algorithm is developed reducing the problem to a sequence of subproblems with smooth functions. An illustrative example is provided for a 3D part with complex geometry. The findings emphasize the significant benefits of utilizing the proposed technique in designing 3D components, with a particular focus on achieving substantial weight reduction.