The Institut De Recherche Technologique (IRT) Jules Verne is leading the CAELESTIS work package on “Digital manufacturing and defects prediction” (WP4) in addition to contributing to a project UQ&M strategy for product and manufacturing.
IRT JULES VERNE is an industrial and collaborative research center that focus on manufacturing challenges such as performance, flexibility and transition.
Research at IRT Jules Verne focuses on topics aimed at short- and medium-term market opportunities. Its core business is the development of generic and multi-sector technological bricks in three key areas:
Integrated product/process design: bringing to the fore innovative technological concepts that ensure the best balance between the functional performance criteria of products and the operational performance of processes allowing for their respective constraints and potential.
Process innovation: developing new technologies for the implementation or assembly of composite, metal and multi-material hybrids aimed at reducing costs or ramping up production. Such development is based on the modified use of processes employed in other sectors, on the study of as-yet relatively immature technologies such as additive manufacturing, on the hybridization of existing processes while taking into account function integration issues, and on the integration of digital approaches to the solutions being studied.
Flexible and intelligent production systems: developing or adapting autonomous intelligent tools in the areas of production and control capable of improving the productivity, flexibility, agility and performance of production systems while ensuring that people remain firmly at the core of such developments.
IRT Jules Verne works in partnership with the largest industrial and academic players in the field of manufacturing, with whom it co-builds and sustains the research projects conducted within it. Since 2012, this community has continued to grow and in 2024 there were 73 members and partners: industrial groups, SMEs, technical centers, academic partners, local authorities.
IRT In the WP4 of CAELESTIS project, and in the tasks T4.2 and T4.3, IRT JVN’s team developed a full product/process tolerancing assembly digital twin of the composite vane and of the fixing titanium parts of the GKN’s OGV, based on an automatic numerical workflow between specialized software. The design of the OGV and of the RTM tool and the process strategy have been realized by the GKN’s team.
The workflow of the models and the simulations offers, at each step of the process, a global numerical verification tool for the product / process Virtual Quality Control, in the development of the product and of its process:
from the composites parts manufacturing simulation in PAM-RTM/DISTORTION® to the virtual measurements of the tolerances on the virtual parts in Python® with a Virtual-Coordinate Measuring Machine and to the final assembly, in MECAmaster®, in compliance with the standard NF E04-008:2013.
The RTM process simulation of the composite parts is realized with a high-fidelity Multiphysics model in PAM-RTM/DISTORTION® and ensure the simulation of the filling (Figure 3), of the curing and of the demolding (Figure 4) of the RTM process.
The estimation of the distortions of the composite parts in the curing and in the demolding phases are realized with PAM-DISTORTION®. These distortions, resulting from residual stresses generated by the polymerization, have a strong impact on the final geometry of the composite vane, on the upper, lower and lateral surfaces (Figure 4).
Applying the state-of-the-art algorithms developed for 3D metrology for geometry inspection, a virtual coordinate measuring machine, V-CMM (Figure 5), has been developed, based on the small torsor of displacements, to automatically estimate the surface profile tolerances of the composite vane (Figure 6) from the distortions field, coming from the demolding phase of the RTM process (Figure 4)
Thus, with the tolerancing assembly digital twin, based on MECAmaster® (Figure 7: Tolerancing assembly digital twin in MECAmaster® softwaresoftware and stack-up linear surrogate models in Python®, the tolerancing analysis of the functional requirements defined on the OGV assembly ensure two main objectives:
–> To have aerodynamical surfaces with minimum geometrical defects to ensure an optimal air flow in the nacelle (Figure 6),
–> To ensure the assembly of the final OGV in the engine and the nacelle environment (Figure 8).
All the numerical thread is automatic between the different software, PAM-RTM/DISTORTION®, Python® and MECAmaster®, with physical and numerical data exchange between the software.
At the level of the IRT JVN’s team:
–> Yvan Denis, R&D Modelling & Simulation Engineer, modeled the RTM process of the composite vane, from the filling to the demolding phase, with PAM-RTM/DISTORTION®, in compliance with the RTM mold design and injection strategy defined by GKN’s team
–> Clément Freyssinet, R&D Modelling & Simulation Engineer, programmed the V-CMM in Python® to automatically extract the tolerances of the surfaces of the composite vane, in compliance with the datum systems and tolerances specifications done by GKN’s team. And he integrated all the linear surrogate models of the stack-ups of the functional requirements to Python® environment
–> Tanguy Moro, R&D Modelling & Simulation Expert, realized the tolerancing assembly model of the composite vane and of the fixing parts in the RTM tool, in MECAmaster® and generated and exported the linear surrogate models of the functional requirements to Python® environment. He had also defined the strategy for the complete workflow
–> Coline Fiquet pilots CAELESTIS project as Project Manager and as Head of the European Service of IRT JV.
As an input of this JVN’s numerical workflow, underlined in blue in the Figure 9, ADDcomposite and ITAINNOVA’s teams provided:
- The layup of the preform, with the stacking of the different UD carbon plies, with ability to integrate some AFP process defects as misalignments, gaps and overlaps, with simulation of the layup on ADDpath® software
- The permeabilities values of the preform, at each point of the preform, estimated by a surrogate model, based on Tensor Rank Decomposition and on the AFP process simulations.
As the output of this numerical workflow, IRT JVN provided the distortions of the geometry of the composite vane for the structural mechanical simulations of the OGV, realized by Universitat de Girona (Figure 9).
All the developments will be integrated in the HPC environment of Barcelona Supercomputing Center.
Find out more about IRT Jules Verne’s work with CAELESTIS
IRT Jules Verne is an industrial and collaborative research centre specializing in manufacturing leading the CAELESTIS work package on “Digital manufacturing and defects prediction” (WP4).
