Periodic Reporting for period 4 - PRE-ECO (A new paradigm to re-engineering printed composites)
Reporting period: 2024-08-01 to 2025-10-31
Additive Manufacturing and Automated Fibre Placement (AFP) technologies have enabled a new class of fibre-reinforced materials known as Variable Angle Tow (VAT) composites. Unlike traditional laminates, VAT composites allow fibres (typically carbon or glass) to follow curved, spatially varying paths. This capability lets engineers tailor stiffness and strength exactly where needed, achieving lighter, stronger, and more efficient structures.
Yet this potential remains largely untapped. Steering brittle fibres along tight curves introduces gaps, overlaps, and kinks that degrade performance. These defects arise from complex interactions among process parameters, such as temperature, speed, pressure, and curvature, and their effects propagate from the microscopic fibre–matrix scale to the full structure. Since existing design tools cannot capture this multi-scale behaviour, VAT composites remain costly to design, difficult to manufacture, and uncertain in performance.
The PRE-ECO project (A new Paradigm to RE-Engineering printed COmposites) addressed this challenge through a radical rethinking of how such materials are conceived and engineered. Its goal was to build a new framework linking the manufacturing process, material behaviour, and structural performance of tow-steered composites. Specifically, PRE-ECO aimed to:
1. Develop multi-scale models simulating the behaviour of VAT composites from fibre-matrix to component level;
2. Create a hybrid metamodeling platform combining physics-based and machine-learning approaches to quantify and optimise manufacturing defects; and
3. Define new design-for-manufacturing principles, allowing engineers to include AFP and 3D-printing constraints at the design stage.
PRE-ECO achieved these objectives, establishing a cross-disciplinary methodology that unites structural mechanics, computational modelling, materials science, additive manufacturing, and artificial intelligence. It delivered advanced numerical tools, validated models, and experimental data that together form a new paradigm for the digital design of composites. The results pave the way for lighter, more reliable, and sustainable structures in aerospace, transport, and energy applications, contributing to lower fuel consumption and reduced environmental impact.
Yet this potential remains largely untapped. Steering brittle fibres along tight curves introduces gaps, overlaps, and kinks that degrade performance. These defects arise from complex interactions among process parameters, such as temperature, speed, pressure, and curvature, and their effects propagate from the microscopic fibre–matrix scale to the full structure. Since existing design tools cannot capture this multi-scale behaviour, VAT composites remain costly to design, difficult to manufacture, and uncertain in performance.
The PRE-ECO project (A new Paradigm to RE-Engineering printed COmposites) addressed this challenge through a radical rethinking of how such materials are conceived and engineered. Its goal was to build a new framework linking the manufacturing process, material behaviour, and structural performance of tow-steered composites. Specifically, PRE-ECO aimed to:
1. Develop multi-scale models simulating the behaviour of VAT composites from fibre-matrix to component level;
2. Create a hybrid metamodeling platform combining physics-based and machine-learning approaches to quantify and optimise manufacturing defects; and
3. Define new design-for-manufacturing principles, allowing engineers to include AFP and 3D-printing constraints at the design stage.
PRE-ECO achieved these objectives, establishing a cross-disciplinary methodology that unites structural mechanics, computational modelling, materials science, additive manufacturing, and artificial intelligence. It delivered advanced numerical tools, validated models, and experimental data that together form a new paradigm for the digital design of composites. The results pave the way for lighter, more reliable, and sustainable structures in aerospace, transport, and energy applications, contributing to lower fuel consumption and reduced environmental impact.
The project developed an integrated framework for the design, analysis, and optimization of Variable Stiffness Composite (VSC) structures, bridging material manufacturing and structural performance.
A suite of multi-scale models was created to predict the static and dynamic, linear and nonlinear responses of composite beams, plates, and shells with curvilinear fibres. These models capture local and global effects and were extended to cover multi-physics behaviour, including thermo-elastic, hygro-elastic, moisture diffusion, piezoelectric, and thermal conductivity responses, allowing realistic simulations under coupled environmental conditions.
The models were applied to analyse manufacture-induced defects through deterministic and stochastic approaches. Studies addressed Continuous Tow Shearing (CTS) thickness variations and Automated Fibre Placement (AFP) signatures such as gaps and overlaps. A stochastic-field methodology was coupled with the in-house high-order finite element solver to represent spatially varying imperfections across scales. Supported by process simulations and experiments, this approach enabled uncertainty quantification of micro- and meso-scale defects and their impact on structural integrity.
A parallelised hybrid optimisation platform combining genetic algorithms and gradient-based methods was developed to solve continuous and discrete design problems efficiently. By integrating AFP constraints and defect patterns into the optimisation loop, the tool ensures manufacturable and structurally optimal solutions. It has been applied to multi-objective problems involving mass, strain energy, frequency, and buckling, and demonstrated in wing-box optimisation for lightweight aerospace applications.
Progress was achieved in damage and failure modelling through advanced Node-Dependent Kinematics (NDK) formulations, capturing localised damage with reduced computational cost while retaining 3D accuracy. These models enable scalable simulations of progressive failure in large composite structures.
Additional research explored:
– Reduction of stress concentrations in open-hole laminates via localised 3D-printed reinforcements;
– Non-local elasticity and peridynamic models for crack propagation;
– AI-based algorithms for inverse problems such as damage and defect detection;
– A multi-fidelity framework with probabilistic machine learning to quantify variability and uncertainty;
– Applications to thermally loaded deployable space structures and thin-ply composites; and
– Nonlinear models for hyperelastic anisotropic materials with distributed fibres, enabling simulation of soft composite systems.
These developments form a comprehensive digital design chain integrating multi-scale modelling, uncertainty quantification, AI-enhanced optimisation, and process-aware design rules. This marks a major advance in composite engineering, enabling predictive and optimised design of printed and AFP structures under realistic manufacturing conditions.
The results, widely disseminated through peer-reviewed publications, establish new theoretical, computational, and practical foundations for industrial use of variable-stiffness composites in aerospace, transport, and energy sectors.
A suite of multi-scale models was created to predict the static and dynamic, linear and nonlinear responses of composite beams, plates, and shells with curvilinear fibres. These models capture local and global effects and were extended to cover multi-physics behaviour, including thermo-elastic, hygro-elastic, moisture diffusion, piezoelectric, and thermal conductivity responses, allowing realistic simulations under coupled environmental conditions.
The models were applied to analyse manufacture-induced defects through deterministic and stochastic approaches. Studies addressed Continuous Tow Shearing (CTS) thickness variations and Automated Fibre Placement (AFP) signatures such as gaps and overlaps. A stochastic-field methodology was coupled with the in-house high-order finite element solver to represent spatially varying imperfections across scales. Supported by process simulations and experiments, this approach enabled uncertainty quantification of micro- and meso-scale defects and their impact on structural integrity.
A parallelised hybrid optimisation platform combining genetic algorithms and gradient-based methods was developed to solve continuous and discrete design problems efficiently. By integrating AFP constraints and defect patterns into the optimisation loop, the tool ensures manufacturable and structurally optimal solutions. It has been applied to multi-objective problems involving mass, strain energy, frequency, and buckling, and demonstrated in wing-box optimisation for lightweight aerospace applications.
Progress was achieved in damage and failure modelling through advanced Node-Dependent Kinematics (NDK) formulations, capturing localised damage with reduced computational cost while retaining 3D accuracy. These models enable scalable simulations of progressive failure in large composite structures.
Additional research explored:
– Reduction of stress concentrations in open-hole laminates via localised 3D-printed reinforcements;
– Non-local elasticity and peridynamic models for crack propagation;
– AI-based algorithms for inverse problems such as damage and defect detection;
– A multi-fidelity framework with probabilistic machine learning to quantify variability and uncertainty;
– Applications to thermally loaded deployable space structures and thin-ply composites; and
– Nonlinear models for hyperelastic anisotropic materials with distributed fibres, enabling simulation of soft composite systems.
These developments form a comprehensive digital design chain integrating multi-scale modelling, uncertainty quantification, AI-enhanced optimisation, and process-aware design rules. This marks a major advance in composite engineering, enabling predictive and optimised design of printed and AFP structures under realistic manufacturing conditions.
The results, widely disseminated through peer-reviewed publications, establish new theoretical, computational, and practical foundations for industrial use of variable-stiffness composites in aerospace, transport, and energy sectors.
The project has advanced far beyond the state of the art in modelling, analysis, and optimization of Variable Stiffness Composite (VSC) structures. Traditional design methods treat material behaviour, manufacturing, and performance as separate processes, relying on simplified models and empirical assumptions. PRE-ECO overcame these limitations by creating a unified, hybrid simulation and data-driven framework that connects all relevant scales and embeds manufacturing variability directly into design.
The project established a robust, flexible, and intelligent simulation technology for printed and AFP composite materials, composed of:
1. A high-fidelity multi-scale modelling framework for micro-, meso- and macro-mechanical characterisation. The models account for process-induced defects, environmental coupling, and nonlinear effects with high precision, supporting deterministic and stochastic multi-physics analyses.
2. Innovative optimisation algorithms and multi-fidelity strategies combining high-resolution finite-element models with reduced-order and surrogate representations. Using probabilistic machine learning, the framework efficiently explores large design spaces while quantifying uncertainty, enabling manufacturing-aware, reliability-based optimisation.
3. An integrated hybrid platform that merges physics-based simulation, virtual testing, and data intelligence, enabling real-time prediction, sensitivity analysis, and the creation of digital twins for composite components.
By merging computational mechanics, data science, and advanced manufacturing, PRE-ECO has defined a new paradigm for digital composite design. The methodology allows engineers to create lighter, stronger, and more reliable structures while fully accounting for the technological signature of the manufacturing process.
The tools and models developed are expected to accelerate industrial adoption of tow-steered composites.
The project established a robust, flexible, and intelligent simulation technology for printed and AFP composite materials, composed of:
1. A high-fidelity multi-scale modelling framework for micro-, meso- and macro-mechanical characterisation. The models account for process-induced defects, environmental coupling, and nonlinear effects with high precision, supporting deterministic and stochastic multi-physics analyses.
2. Innovative optimisation algorithms and multi-fidelity strategies combining high-resolution finite-element models with reduced-order and surrogate representations. Using probabilistic machine learning, the framework efficiently explores large design spaces while quantifying uncertainty, enabling manufacturing-aware, reliability-based optimisation.
3. An integrated hybrid platform that merges physics-based simulation, virtual testing, and data intelligence, enabling real-time prediction, sensitivity analysis, and the creation of digital twins for composite components.
By merging computational mechanics, data science, and advanced manufacturing, PRE-ECO has defined a new paradigm for digital composite design. The methodology allows engineers to create lighter, stronger, and more reliable structures while fully accounting for the technological signature of the manufacturing process.
The tools and models developed are expected to accelerate industrial adoption of tow-steered composites.