Periodic Reporting for period 2 - NOVITAS (Nonlinear analysis for virtual design of composite deployable space booms and membranes)
Reporting period: 2025-03-15 to 2026-03-14
Summary of the context and overall objectives of the project
NOVITAS addressed the need for accurate, efficient and physically reliable numerical tools for the virtual design of ultra-thin deployable composite space structures, with particular focus on deployable booms, TRAC longerons, membranes and foldable structural assemblies. These systems are increasingly relevant in space engineering because they allow large functional surfaces, such as antennas, telescopes, solar arrays and membrane-supported platforms, to be compactly stowed during launch and reliably deployed in orbit. The project was motivated by the fact that future space missions require larger, lighter and more efficient structures, while launch volume and mass constraints remain severe. Deployable composite booms and membranes are therefore key enabling technologies for advanced space systems, but their design remains challenging because they operate in regimes dominated by large rotations, large displacements, local buckling, contact, cross-sectional deformation, composite anisotropy and dynamic effects. The original NOVITAS project summary explicitly identified these issues as central open challenges, including the need for mathematical models able to describe multiscale three-dimensional stress states, failure identification, material effects and deployment simulations.
The overall objective of the project was to advance the mathematical and computational modelling of thin and ultra-thin deployable composite structures, with the final goal of supporting their virtual design. The project pathway to impact was based on the development of modelling approaches able to reduce the computational burden of conventional two-dimensional shell or three-dimensional solid finite element models, while retaining the ability to describe local effects that are essential for structural reliability. In particular, the project focused on the simulation of the complete life cycle of deployable structures, from folding and stowage to deployment-related phenomena and dynamic behaviour. This required the development and validation of nonlinear numerical models capable of representing both global kinematics and local structural mechanisms, such as cross-sectional distortion, flange contact, local buckling, stiffness degradation and vibration attenuation.
The scientific motivation was also connected to the limitations of standard finite element approaches. Conventional shell and solid models are powerful but often require very fine meshes to capture the behaviour of ultra-thin composite longerons, especially when high aspect ratios, contact nonlinearities and local instabilities are involved. This leads to high computational costs, making extensive parametric design, optimization and uncertainty studies difficult. NOVITAS therefore aimed to provide a modelling framework that could preserve high-fidelity structural descriptions while making analyses more computationally efficient and suitable for engineering design loops.
The expected impact of the project is significant at both scientific and technological levels. Scientifically, NOVITAS contributes to the understanding and simulation of nonlinear behaviour in slender composite deployable structures, including folding, local buckling, contact and dynamic response. Technologically, the project supports the design of lighter, more compact and more reliable deployable space structures. The results may help reduce the number of expensive experimental campaigns, accelerate design iterations and provide engineers with virtual testing tools for complex space structures. This is consistent with the project’s original impact pathway, which identified the development of efficient modelling techniques as a contribution to space-based technological innovation and to a sustainable European economy.
The project was implemented through an interdisciplinary pathway combining theoretical modelling, numerical simulation and experimental validation. The DoA structured the action around management and training, dissemination, theoretical modelling of dynamic behaviours, experimental testing and numerical simulations. The technical core of the project was represented by WP3, WP4 and WP5, which addressed theoretical modelling, testing and numerical simulations for ultra-thin deployable structures.
No specific integration of social sciences and humanities was required by the topic, and SSH disciplines were not a technical component of the project. However, the project contributes indirectly to broader societal needs by enabling more reliable and efficient space structures, which can support future systems for Earth observation, telecommunications, scientific missions and space-based energy concepts.
The overall objective of the project was to advance the mathematical and computational modelling of thin and ultra-thin deployable composite structures, with the final goal of supporting their virtual design. The project pathway to impact was based on the development of modelling approaches able to reduce the computational burden of conventional two-dimensional shell or three-dimensional solid finite element models, while retaining the ability to describe local effects that are essential for structural reliability. In particular, the project focused on the simulation of the complete life cycle of deployable structures, from folding and stowage to deployment-related phenomena and dynamic behaviour. This required the development and validation of nonlinear numerical models capable of representing both global kinematics and local structural mechanisms, such as cross-sectional distortion, flange contact, local buckling, stiffness degradation and vibration attenuation.
The scientific motivation was also connected to the limitations of standard finite element approaches. Conventional shell and solid models are powerful but often require very fine meshes to capture the behaviour of ultra-thin composite longerons, especially when high aspect ratios, contact nonlinearities and local instabilities are involved. This leads to high computational costs, making extensive parametric design, optimization and uncertainty studies difficult. NOVITAS therefore aimed to provide a modelling framework that could preserve high-fidelity structural descriptions while making analyses more computationally efficient and suitable for engineering design loops.
The expected impact of the project is significant at both scientific and technological levels. Scientifically, NOVITAS contributes to the understanding and simulation of nonlinear behaviour in slender composite deployable structures, including folding, local buckling, contact and dynamic response. Technologically, the project supports the design of lighter, more compact and more reliable deployable space structures. The results may help reduce the number of expensive experimental campaigns, accelerate design iterations and provide engineers with virtual testing tools for complex space structures. This is consistent with the project’s original impact pathway, which identified the development of efficient modelling techniques as a contribution to space-based technological innovation and to a sustainable European economy.
The project was implemented through an interdisciplinary pathway combining theoretical modelling, numerical simulation and experimental validation. The DoA structured the action around management and training, dissemination, theoretical modelling of dynamic behaviours, experimental testing and numerical simulations. The technical core of the project was represented by WP3, WP4 and WP5, which addressed theoretical modelling, testing and numerical simulations for ultra-thin deployable structures.
No specific integration of social sciences and humanities was required by the topic, and SSH disciplines were not a technical component of the project. However, the project contributes indirectly to broader societal needs by enabling more reliable and efficient space structures, which can support future systems for Earth observation, telecommunications, scientific missions and space-based energy concepts.
Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far
During the project, the work focused on the development, implementation and validation of numerical models for ultra-thin deployable composite structures. The activities covered both fundamental modelling aspects and application-oriented simulations, with particular attention to deployable booms, TRAC longerons, foldable strip assemblies and flat array structures. The work was aligned with the technical objectives of NOVITAS, especially the development of reliable nonlinear models for deployable composite structures and the prediction of their structural behaviour during folding, deployment-related configurations and dynamic excitation.
A first major activity concerned the formulation and implementation of refined one-dimensional finite element models with three-dimensional capabilities for thin-walled composite deployable structures. These models were used to represent the complex cross-sectional behaviour of TRAC longerons while keeping a beam-like description along the structural axis. This allowed the project to capture local deformation mechanisms without requiring fully detailed shell or solid discretizations along the entire structure. The developed modelling strategy was applied to TRAC longerons, in which the flanges and web undergo large rotations, cross-sectional deformation and contact during folding.
A second major activity was the development of an implicit nonlinear simulation strategy for folding. The folding of ultra-thin deployable structures is numerically challenging because it involves large displacements, rotations, local curvature concentration and contact between thin structural parts. The project implemented implicit quasi-static analyses based on Newton–Raphson iterations and displacement control, enabling the simulation of folding paths and equilibrium configurations. A dedicated contact strategy was also developed, based on spring-like interactions between predefined node pairs representing the possible contact between the inner surfaces of the flanges. The early version of this approach demonstrated that including contact avoids non-physical interpenetration between flanges during TRAC folding and allows the model to reproduce strongly nonlinear folding configurations.
This contact-based modelling was further improved through nonlinear spring formulations. The final approach introduced polynomial nonlinear springs activated within a small tolerance from interpenetration. The spring parameters were tuned to improve convergence, and a seventh-degree polynomial was found to reduce the number of iterations compared with a linear stiffness. This strategy was successfully applied to the folding of a strip made of two longerons connected by transverse battens, and the results were compared with Abaqus simulations using shell elements. The proposed implicit approach showed good agreement with the explicit Abaqus results, while using fewer degrees of freedom and requiring less computational time for the reference case. In the comparison, the Abaqus simulation used 77,298 degrees of freedom and required approximately 6 hours, while the proposed model used 38,994 degrees of freedom and completed in approximately 2.3 hours.
The work was then extended to longer deployable strips of 1 m and 5 m. These simulations demonstrated the ability of the proposed modelling approach to handle very long and slender deployable structures while keeping the computational cost manageable. This achievement is particularly important because conventional two-dimensional shell models are constrained by aspect-ratio requirements, which make the simulation of very long, ultra-thin structures increasingly expensive. The 1 m and 5 m analyses confirmed that the developed model can capture progressive folding and localized curvature in long structures, while maintaining a reduced computational burden.
Another important achievement was the extension of the modelling approach from individual longerons and strips to more complex foldable array structures. The project investigated flat deployable arrays composed of pairs of TRAC longerons, transverse battens and hinges. The formulation allowed the full assembly to be represented through an equivalent one-dimensional model while still retaining higher-order cross-sectional kinematics. The full quadrant structure was also investigated, with the model reaching 267,156 degrees of freedom while preserving an efficient description of global rotation and local deformation mechanisms. The results showed that the internal energy remained essentially constant during the rotation step, indicating that the motion was close to a rigid-body-like rotation, while folding introduced stored elastic energy that can drive subsequent deployment.
A further scientific activity addressed the nonlinear dynamic behaviour and vibration damping of TRAC longerons. Experimental sine-sweep tests were performed under different boundary conditions, and damping ratios were evaluated using the Half-Power Bandwidth Method. The experiments showed that local buckling can appear under sufficiently high excitation levels and can be associated with an increase in damping. In one configuration, beyond 180 mV excitation, the damping ratio increased from approximately 0.01 to 0.019 up to 340 mV, and this transition coincided with the emergence of local buckling observed through high-speed camera measurements.
The experimental dynamic results were complemented by nonlinear numerical simulations. The numerical model reproduced the experimental setup and was used to compare static, modal and frequency-response behaviour. The comparison showed strong agreement in the prediction of the natural frequency of the deformed configuration: the experimental value at 10 mV was 88.85 Hz, while the numerical model predicted 89.16 Hz, corresponding to a difference of only 0.35%. The final outcome was the identification of local buckling not only as a potential instability or failure mechanism, but also as a phenomenon that can influence and, in specific conditions, enhance vibration attenuation in ultra-thin deployable structures.
In addition to deployable space structures, the project generated results related to advanced modelling of composite and architected structures. A component-wise modelling strategy was applied to laminated composite beams to study strain energy contributions at different scales, including laminate, layer, fibre and matrix levels. This work showed that refined models can reproduce stress distributions comparable to those of solid finite element models while using a much smaller number of degrees of freedom. In particular, one model achieved the same in-plane and out-of-plane stress distribution as the commercial solid model using only 12% of its degrees of freedom. This result supports one of the broader technical needs of NOVITAS: obtaining detailed information on local stress and energy distributions in composite structures without the prohibitive cost of full three-dimensional simulations.
Finally, the project also produced results on architected and metamaterial-like structures inspired by the Kagome lattice. The developed numerical approach was used to capture buckling, post-buckling and rigid-body rotations of lattice triangles, as well as the effect of local stiffness modifications. These results demonstrated that the modelling framework can be extended beyond deployable booms to more general architected structures requiring accurate prediction of local and global nonlinear behaviour.
Overall, the main outcomes of the action were: the development of high-fidelity and computationally efficient nonlinear models for ultra-thin deployable composite structures; the implementation of contact-based implicit folding simulations; the validation of the approach against commercial finite element models and experimental data; the identification of local buckling as a mechanism affecting vibration damping; and the extension of the modelling philosophy to multiscale composite analysis and architected metamaterial structures.
A first major activity concerned the formulation and implementation of refined one-dimensional finite element models with three-dimensional capabilities for thin-walled composite deployable structures. These models were used to represent the complex cross-sectional behaviour of TRAC longerons while keeping a beam-like description along the structural axis. This allowed the project to capture local deformation mechanisms without requiring fully detailed shell or solid discretizations along the entire structure. The developed modelling strategy was applied to TRAC longerons, in which the flanges and web undergo large rotations, cross-sectional deformation and contact during folding.
A second major activity was the development of an implicit nonlinear simulation strategy for folding. The folding of ultra-thin deployable structures is numerically challenging because it involves large displacements, rotations, local curvature concentration and contact between thin structural parts. The project implemented implicit quasi-static analyses based on Newton–Raphson iterations and displacement control, enabling the simulation of folding paths and equilibrium configurations. A dedicated contact strategy was also developed, based on spring-like interactions between predefined node pairs representing the possible contact between the inner surfaces of the flanges. The early version of this approach demonstrated that including contact avoids non-physical interpenetration between flanges during TRAC folding and allows the model to reproduce strongly nonlinear folding configurations.
This contact-based modelling was further improved through nonlinear spring formulations. The final approach introduced polynomial nonlinear springs activated within a small tolerance from interpenetration. The spring parameters were tuned to improve convergence, and a seventh-degree polynomial was found to reduce the number of iterations compared with a linear stiffness. This strategy was successfully applied to the folding of a strip made of two longerons connected by transverse battens, and the results were compared with Abaqus simulations using shell elements. The proposed implicit approach showed good agreement with the explicit Abaqus results, while using fewer degrees of freedom and requiring less computational time for the reference case. In the comparison, the Abaqus simulation used 77,298 degrees of freedom and required approximately 6 hours, while the proposed model used 38,994 degrees of freedom and completed in approximately 2.3 hours.
The work was then extended to longer deployable strips of 1 m and 5 m. These simulations demonstrated the ability of the proposed modelling approach to handle very long and slender deployable structures while keeping the computational cost manageable. This achievement is particularly important because conventional two-dimensional shell models are constrained by aspect-ratio requirements, which make the simulation of very long, ultra-thin structures increasingly expensive. The 1 m and 5 m analyses confirmed that the developed model can capture progressive folding and localized curvature in long structures, while maintaining a reduced computational burden.
Another important achievement was the extension of the modelling approach from individual longerons and strips to more complex foldable array structures. The project investigated flat deployable arrays composed of pairs of TRAC longerons, transverse battens and hinges. The formulation allowed the full assembly to be represented through an equivalent one-dimensional model while still retaining higher-order cross-sectional kinematics. The full quadrant structure was also investigated, with the model reaching 267,156 degrees of freedom while preserving an efficient description of global rotation and local deformation mechanisms. The results showed that the internal energy remained essentially constant during the rotation step, indicating that the motion was close to a rigid-body-like rotation, while folding introduced stored elastic energy that can drive subsequent deployment.
A further scientific activity addressed the nonlinear dynamic behaviour and vibration damping of TRAC longerons. Experimental sine-sweep tests were performed under different boundary conditions, and damping ratios were evaluated using the Half-Power Bandwidth Method. The experiments showed that local buckling can appear under sufficiently high excitation levels and can be associated with an increase in damping. In one configuration, beyond 180 mV excitation, the damping ratio increased from approximately 0.01 to 0.019 up to 340 mV, and this transition coincided with the emergence of local buckling observed through high-speed camera measurements.
The experimental dynamic results were complemented by nonlinear numerical simulations. The numerical model reproduced the experimental setup and was used to compare static, modal and frequency-response behaviour. The comparison showed strong agreement in the prediction of the natural frequency of the deformed configuration: the experimental value at 10 mV was 88.85 Hz, while the numerical model predicted 89.16 Hz, corresponding to a difference of only 0.35%. The final outcome was the identification of local buckling not only as a potential instability or failure mechanism, but also as a phenomenon that can influence and, in specific conditions, enhance vibration attenuation in ultra-thin deployable structures.
In addition to deployable space structures, the project generated results related to advanced modelling of composite and architected structures. A component-wise modelling strategy was applied to laminated composite beams to study strain energy contributions at different scales, including laminate, layer, fibre and matrix levels. This work showed that refined models can reproduce stress distributions comparable to those of solid finite element models while using a much smaller number of degrees of freedom. In particular, one model achieved the same in-plane and out-of-plane stress distribution as the commercial solid model using only 12% of its degrees of freedom. This result supports one of the broader technical needs of NOVITAS: obtaining detailed information on local stress and energy distributions in composite structures without the prohibitive cost of full three-dimensional simulations.
Finally, the project also produced results on architected and metamaterial-like structures inspired by the Kagome lattice. The developed numerical approach was used to capture buckling, post-buckling and rigid-body rotations of lattice triangles, as well as the effect of local stiffness modifications. These results demonstrated that the modelling framework can be extended beyond deployable booms to more general architected structures requiring accurate prediction of local and global nonlinear behaviour.
Overall, the main outcomes of the action were: the development of high-fidelity and computationally efficient nonlinear models for ultra-thin deployable composite structures; the implementation of contact-based implicit folding simulations; the validation of the approach against commercial finite element models and experimental data; the identification of local buckling as a mechanism affecting vibration damping; and the extension of the modelling philosophy to multiscale composite analysis and architected metamaterial structures.
Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)
The project advanced the state of the art in several complementary directions.
First, NOVITAS went beyond conventional finite element modelling by developing refined one-dimensional models capable of capturing three-dimensional cross-sectional deformation in ultra-thin composite deployable structures. The state of the art for such structures often relies on shell or solid finite element models, which are accurate but computationally expensive and strongly affected by mesh density and aspect-ratio constraints. The project demonstrated that a reduced-dimensional model can still capture essential local effects such as flange deformation, local curvature, contact, warping and buckling. This provides a substantial advance because it enables analyses that are both physically rich and computationally manageable.
Second, the project advanced the simulation of folding in ultra-thin deployable booms and strips. Folding is a particularly difficult problem because it combines large rotations, concentrated curvature, contact, geometric nonlinearities and composite structural behaviour. The developed implicit framework reproduced folding paths of TRAC longerons and strip assemblies and showed good agreement with explicit Abaqus simulations, while reducing the number of degrees of freedom and computational time. This is beyond the state of the art because implicit approaches are generally more difficult to apply to such strongly nonlinear contact problems, while explicit commercial approaches often require high computational effort and fine meshes.
Third, NOVITAS introduced and validated improved contact strategies for deployable structures. The project showed that neglecting contact leads to non-physical interpenetration between the TRAC flanges, while the inclusion of contact through node-pair spring formulations produces physically meaningful folded configurations. The later introduction of nonlinear spring laws further improved convergence, reducing the number of iterations and making the method more robust for practical simulations. This is an important methodological result because contact in ultra-thin flexible structures is often one of the main bottlenecks in nonlinear simulation.
Fourth, the project demonstrated scalability to longer structures. The modelling strategy was not limited to small academic examples but was extended to 1 m and 5 m foldable strips. These simulations showed that long, thin deployable structures can be analysed while keeping computational cost manageable, opening the door to realistic large-scale deployable space structures. This is particularly relevant for future applications involving large antennas, solar arrays, telescopes and membrane-supported platforms.
Fifth, the project produced new scientific evidence on the dynamic behaviour of ultra-thin deployable structures. The experimental and numerical study on vibration damping showed that local buckling may contribute to vibration attenuation. This result is significant because local buckling is usually treated only as an undesirable instability or possible failure precursor. NOVITAS showed that, in specific configurations and excitation ranges, the onset of local buckling can coincide with a marked increase in damping ratio. This suggests a more nuanced understanding of local instabilities in ultra-light deployable structures and opens the possibility of exploiting controlled nonlinearities for vibration mitigation.
Sixth, the project contributed to the multiscale analysis of composite materials through component-wise energy evaluation. The ability to isolate the strain energy stored in different structural constituents, such as fibres, matrix, laminae and laminate regions, is relevant for failure assessment and structural health interpretation. The project showed that the component-wise approach can reproduce solid-model-level stress distributions with a drastic reduction in computational cost and can identify regions where high strain energy accumulation may indicate potential failure initiation.
Seventh, the project extended the developed modelling concepts to architected and metamaterial-like structures, including Kagome-based lattices. The model captured buckling, post-buckling and rigid-body rotations of triangular lattice units, which are central mechanisms in the mechanical response of Kagome structures. The study also demonstrated the possibility of tailoring the mechanical response through local stiffness modifications, with implications for load-carrying capacity, controlled deformation and energy absorption.
The potential impacts of these results are substantial. At scientific level, the project provides new methods and evidence for the nonlinear analysis of deployable, composite and architected structures. At technological level, the results support more efficient virtual prototyping, allowing engineers to explore design alternatives, contact conditions, folding paths and dynamic responses before manufacturing. At economic level, the reduction of computational and experimental costs can accelerate design loops and reduce development risks. At societal level, the improved design of deployable space structures can contribute indirectly to future space systems supporting Earth observation, telecommunications, scientific exploration and potentially space-based energy technologies.
Further uptake would require additional research and demonstration activities. In particular, the next steps should include broader experimental validation under deployment conditions, integration of viscoelastic and temperature-dependent material behaviour, systematic treatment of imperfections and uncertainties, implementation in user-friendly simulation tools, and application to complete spacecraft-scale deployable architectures. Additional work may also be needed to establish standardized benchmark problems for ultra-thin deployable composite structures, since the field currently lacks widely accepted reference cases for comparing numerical methods.
First, NOVITAS went beyond conventional finite element modelling by developing refined one-dimensional models capable of capturing three-dimensional cross-sectional deformation in ultra-thin composite deployable structures. The state of the art for such structures often relies on shell or solid finite element models, which are accurate but computationally expensive and strongly affected by mesh density and aspect-ratio constraints. The project demonstrated that a reduced-dimensional model can still capture essential local effects such as flange deformation, local curvature, contact, warping and buckling. This provides a substantial advance because it enables analyses that are both physically rich and computationally manageable.
Second, the project advanced the simulation of folding in ultra-thin deployable booms and strips. Folding is a particularly difficult problem because it combines large rotations, concentrated curvature, contact, geometric nonlinearities and composite structural behaviour. The developed implicit framework reproduced folding paths of TRAC longerons and strip assemblies and showed good agreement with explicit Abaqus simulations, while reducing the number of degrees of freedom and computational time. This is beyond the state of the art because implicit approaches are generally more difficult to apply to such strongly nonlinear contact problems, while explicit commercial approaches often require high computational effort and fine meshes.
Third, NOVITAS introduced and validated improved contact strategies for deployable structures. The project showed that neglecting contact leads to non-physical interpenetration between the TRAC flanges, while the inclusion of contact through node-pair spring formulations produces physically meaningful folded configurations. The later introduction of nonlinear spring laws further improved convergence, reducing the number of iterations and making the method more robust for practical simulations. This is an important methodological result because contact in ultra-thin flexible structures is often one of the main bottlenecks in nonlinear simulation.
Fourth, the project demonstrated scalability to longer structures. The modelling strategy was not limited to small academic examples but was extended to 1 m and 5 m foldable strips. These simulations showed that long, thin deployable structures can be analysed while keeping computational cost manageable, opening the door to realistic large-scale deployable space structures. This is particularly relevant for future applications involving large antennas, solar arrays, telescopes and membrane-supported platforms.
Fifth, the project produced new scientific evidence on the dynamic behaviour of ultra-thin deployable structures. The experimental and numerical study on vibration damping showed that local buckling may contribute to vibration attenuation. This result is significant because local buckling is usually treated only as an undesirable instability or possible failure precursor. NOVITAS showed that, in specific configurations and excitation ranges, the onset of local buckling can coincide with a marked increase in damping ratio. This suggests a more nuanced understanding of local instabilities in ultra-light deployable structures and opens the possibility of exploiting controlled nonlinearities for vibration mitigation.
Sixth, the project contributed to the multiscale analysis of composite materials through component-wise energy evaluation. The ability to isolate the strain energy stored in different structural constituents, such as fibres, matrix, laminae and laminate regions, is relevant for failure assessment and structural health interpretation. The project showed that the component-wise approach can reproduce solid-model-level stress distributions with a drastic reduction in computational cost and can identify regions where high strain energy accumulation may indicate potential failure initiation.
Seventh, the project extended the developed modelling concepts to architected and metamaterial-like structures, including Kagome-based lattices. The model captured buckling, post-buckling and rigid-body rotations of triangular lattice units, which are central mechanisms in the mechanical response of Kagome structures. The study also demonstrated the possibility of tailoring the mechanical response through local stiffness modifications, with implications for load-carrying capacity, controlled deformation and energy absorption.
The potential impacts of these results are substantial. At scientific level, the project provides new methods and evidence for the nonlinear analysis of deployable, composite and architected structures. At technological level, the results support more efficient virtual prototyping, allowing engineers to explore design alternatives, contact conditions, folding paths and dynamic responses before manufacturing. At economic level, the reduction of computational and experimental costs can accelerate design loops and reduce development risks. At societal level, the improved design of deployable space structures can contribute indirectly to future space systems supporting Earth observation, telecommunications, scientific exploration and potentially space-based energy technologies.
Further uptake would require additional research and demonstration activities. In particular, the next steps should include broader experimental validation under deployment conditions, integration of viscoelastic and temperature-dependent material behaviour, systematic treatment of imperfections and uncertainties, implementation in user-friendly simulation tools, and application to complete spacecraft-scale deployable architectures. Additional work may also be needed to establish standardized benchmark problems for ultra-thin deployable composite structures, since the field currently lacks widely accepted reference cases for comparing numerical methods.