First engineering framework for the out-of-plane robotic shaping of thin rheological objects

Cocuzza, Silvio and Yan, X.-T. (2018) First engineering framework for the out-of-plane robotic shaping of thin rheological objects. Robotics and Computer Integrated Manufacturing, 53. pp. 108-121. ISSN 0736-5845

[img]
Preview
Text (Cocuzza-Yan-RCIM-2018-First-engineering-framework-for-the-out-of-plane-robotic)
Cocuzza_Yan_RCIM_2018_First_engineering_framework_for_the_out_of_plane_robotic.pdf
Accepted Author Manuscript
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (1MB)| Preview

    Abstract

    The automatic manipulation and shaping of soft rheological objects is a very challenging and hot topic of research. The deformation control of soft objects is eagerly required in economically important applications in the food industry, in surgical robotics, and in the assembly of flexible objects. The main challenges are the estimation of the object deformation properties and the dynamic control of its shape using the motion of a tool, usually actuated by a robot manipulator. The proposed framework addresses these challenges and represents an important breakthrough in the robotic shaping of thin rheological objects. Indeed, this is the first work in the literature in which the out-of-plane robotic shaping of a thin rheological object is modeled and experimentally validated. As a foundation for the proposed framework, a novel catenary-like Mass-Spring-Damper (MSD) model formed of lumped masses interconnected with three-element units is proposed to study and simulate the dynamics of the shaping of a thin rheological material over a moulding object. First, a three-step method for identifying the material properties from experimental tensile tests is proposed and validated. The method has allowed both to identify the material properties and to validate the three-element material model used, thanks to the close agreement between the material mathematical model and the experimental data. Then, a multibody model of the whole rheological object is proposed, which has been validated thanks to the close agreement between the simulated deformation profile and the experimental one. Finally, the validated model is used in order to compare different shaping movements and velocities of the tool by dynamic simulations. A two-step shaping movement, which has been experimentally validated in a robotic cell in industrial environment, results more advantageous with respect to simpler motions since significantly reduced internal forces are generated in the material while the quality of the shaping operation is still guaranteed. The proposed framework has then enabled the realization of the world’s first smart manufacturing plant for the high quality robotic shaping of thin rheological objects in the food industry.