Optimization of Mechanical Properties of BIO-sourced Epoxy Resins by ARTificial Intelligence – When Emerging Technologies and Sustainable Development Work Side by Side – BIO ART

The goal of BIO ART is to develop new bio-epoxy resins from renewable resources without bisphenol A, which is toxic to humans and the environment. BIO ART’s originality comes from the synergy between green chemistry and emerging technologies (multiscale modeling and artificial neural network) and sustainable application in industry. In contrast to purely experimental or exclusively numerical approaches, BIO ART integrates simulations and experiments at length and time scales ranging from the atomistic level to the engineering scale. The proposed project will contribute to close the four knowledge gaps: i) use of exclusively bio-sourced molecules from abundant resources and natural fillers with competitive mechanical properties, ii) multiscale modeling of epoxy including its macromolecular network topology, iii) optimization of the resin formulation by an artificial neural network framework linking the chemical nature of the molecules to the mechanical properties, and iv) advanced mechanical characterization and processing of fiber-reinforced bio-composites. BIO ART’s consortium consists of four complementary Franco-German partners with recognized skills in the synthesis of bio-polymers and physicochemical characterization (ICMPE/FR), in microstructure generation and surrogate models based on artificial neural networks (MSME/FR) as well as in multiscale modeling of polymers and discrete-to-continuum coupling methods (FAU/DE), and in composite processing and advanced mechanical characterization (UBT/DE). The scientific program is divided into 5 work packages: WP1: Synthesis of bio-sourced epoxy, WP2: Characterization of bio-sourced epoxy, WP3: Multiscale modeling, WP4: Optimization of bio-sourced epoxy formulation by artificial neural network, and WP5: Composite processing and mechanical characterization. The work packages are defined in a way that they can be completed in 3 years by 3 collaborating doctoral researchers, one for experimental part and two for the numerical part. A technician will support the experimental PhD candidate as regards the processing and characterization of the obtained materials. Beyond them, BIO ART’s consortium, which is a well-balanced composition of early career and senior scientists, will actively contribute to achieve the project’s milestones. BIO ART’s methods are up-to-date, are based on recently published results, and benefit from the strong synergies with current projects of the project partners. In particular, the experimental and numerical methods will range from the atomistic scale (molecular structure, synthesis of constituents, molecular dynamics simulations), to the mesoscale (curing process, network characterization, network model), and to the macroscale (fracture properties, continuum mechanical simulations). This methodology will focus on the investigation of the relationship between the structure and the multiscale properties of the obtained materials. This approach will synergistically combine modelling with experimental characterizations, which will allow to address the scientific issues of this project. This multidisciplinary scientific approach will allow BIO ART to respond to a current crucial societal issue, i.e. biosourced polymer materials from circular bio-economy, aimed for sustainable development applications