Shape design and optimisation for additive manufacturing – COFFA

For these reasons, the COFFA project has chosen to provide designer assistance, using both modeling (and/or shape modification) tools adapted to complex shapes and analysis tools based on various relevant indicators such as performance (mass, stress, thermal, etc.) and manufacturability.
Initial work has been carried out within I2M on a new hybrid topology optimization strategy (SIMP + level-set) based on the use of NURBS hypersurfaces to describe the domain whose topology is to be optimized. This ongoing work is being carried out through the Nouvelle Aquitaine regional project Futurprod and its technological platform based on the SLM process. The work has not yet been compared with experimental results from actual manufactured parts, and therefore feedback from experienced users has not yet been effectively incorporated. Furthermore, this work focuses mainly on the digital topological optimization part, but the interaction between the designer and the produced part is not considered. Taking into account the designers' preferences in the choice of part shapes is impossible, only heuristics translated by rules compatible with the digital process can be taken into account as long as the calculation times remain reasonable. Also, business expertise cannot be valued. The major challenge in COFFA will be to be able to integrate this user/optimization tool interaction in an ergonomic and efficient way from a calculation time point of view by using the know-how and feedback from the G-SCOP laboratory which has significant expertise in the implementation of ALM processes.

The result of COFFA consists of a design methodology adapted to ALM (COFFA method) allowing to converge towards an optimal design in a short time and a general computer model that can be applied to cases of industrial interest. This approach takes place in 5 successive phases (see illustration). This approach reveals an important step of orientation and manufacturability analysis, this is explained by the fact that manufacturability is largely dependent on the orientation of the part in its manufacturing chamber. Consequently, the orientation is chosen to allow and optimize the manufacturability of the part. COFFA has given rise to publications in national and international conferences and journals concerning topological optimization, surface reconstruction, knowledge elicitation and formalization, part orientation, and the evaluation of the decision support tool in part orientation. Following the COFFA project, an ANR project involving the G-SCOP and I2M laboratories, and the company DPRI (acquired by Hexagon) was launched, the KAM4AM project (ANR-20-CE10-0012). A collaborative project with the company DPRI/Hexagon was launched during the project and funded 48 months of post-doctoral work on knowledge-based decision support in the context of additive manufacturing.

The COFFA project opened a lasting collaboration between G-SCOP and I2M which gave rise to the ANR KAM4AM project and a collaborative project within the framework of Carnot.

Bertolino, G.; et al. A general surface reconstruction method for post-processing of topology optimisation results. SIM AM. 2019. Pavie, Italy.

Bertolino, G.; et al. An Efficient Hybrid Optimization Strategy for Surface Reconstruction. Computer Graphics Forum. 2021, 1–37.

Bertolino, G.; Montemurro, M. Two-scale topology optimisation of cellular materials under mixed boundary conditions. International Journal of Mechanical Sciences. 2022, 216, 106961.

Mechekour, E.-H.; Vignat, F.; Grandvallet, C.; Pourroy, F.; René Marin, P.; et al. Evaluation of a decision-support tool for part orientation in EBM additive manufacturing. International Journal of Advanced Manufacturing Technology, In press. 2023.

Topology optimization is today experiencing a major boom in connection with additive manufacturing. The direction taken by the majority of players is advanced automation with integration of multi physic objectives and constraints and this for different stages of the product life cycle. The expression of these constraints is not obvious for the users of these solutions and they tend to develop complex methodologies to converge towards an acceptable solution.
The project COFFA has chosen the way of assistance to the designer, both by proposing complex shape modeling and editing tools and analysis tools based on indicators of performance and manufacturability. COFFA aims to offer designers a methodology and a set of tools to assist them in designing optimal shape for additive manufacturing.