Periodic Reporting for period 2 - TREAL (Thermoplastic material allowable generation using a reliability-based virtual modeling platform)
Période du rapport: 2021-04-01 au 2023-03-31
In the quest towards reduced carbon emissions and fuel consumption, aeronautical industries are moving towards the development and application of thermoplastic materials in the airframes. But the expensive and lengthy qualification test campaigns are a challenge to their application in structural designs. In this framework, TREAL aims to reduce the expensive physical testing through simulation. At the first step of the project, extensive experimental campaigns will be performed to obtain the material data of the thermoplastic materials. These will be used to implement reliable analysis models to predict the damage modes at different levels. Uncertainty quantification and management technologies are applied to obtain reliable material and design allowables using the developed analysis models.
Thermoplastics are a type of plastic that can be melted and recast almost indefinitely. This means they are amenable to reprocessing and recycling, enhancing their sustainability. Further, they do not require a curing step after consolidation and this reduces production times. TREAL plans to help the EU aerospace sector exploit the potential of thermoplastics through the development of an advanced thermoplastic modelling platform. It will ensure accurate determination of design allowables, the allowable materials’ property values derived from test data. More specifically, the reliable determination of the allowable limits of stress, strain, or stiffness for certain configurations and conditions will support the utilisation of thermoplastic composites in future aircraft designs.
More specifically, the overall objective of TREAL is ‘To develop an efficient and comprehensive framework to generate material allowables for thermoplastic based composite materials: from the material characterization to the virtual generation of design allowables’.
The following sub-objectives were defined in order to accomplish the global objective:
• To understand the mechanical behaviour of a thermoplastic composite material
• To develop analysis models for thermoplastic composite materials which are able to predict all damage and failure modes at both material and component levels
• To define the test and data reduction methods needed to define the material card of thermoplastic composite material
• To develop numerical methodologies to propagate material, process and testing uncertainties along the model chain up to the allowable computation
• To define a reduced test matrix for the validation of the virtually generated allowables
• To define a complete framework and implement it in a user-friendly platform for advanced and non-advanced users
• To assess accuracy, efficiency, user-friendliness, etc., of the virtual material allowables generation platform, and perform the necessary incremental models and methodology improvements.
Thermoplastics are a type of plastic that can be melted and recast almost indefinitely. This means they are amenable to reprocessing and recycling, enhancing their sustainability. Further, they do not require a curing step after consolidation and this reduces production times. TREAL plans to help the EU aerospace sector exploit the potential of thermoplastics through the development of an advanced thermoplastic modelling platform. It will ensure accurate determination of design allowables, the allowable materials’ property values derived from test data. More specifically, the reliable determination of the allowable limits of stress, strain, or stiffness for certain configurations and conditions will support the utilisation of thermoplastic composites in future aircraft designs.
More specifically, the overall objective of TREAL is ‘To develop an efficient and comprehensive framework to generate material allowables for thermoplastic based composite materials: from the material characterization to the virtual generation of design allowables’.
The following sub-objectives were defined in order to accomplish the global objective:
• To understand the mechanical behaviour of a thermoplastic composite material
• To develop analysis models for thermoplastic composite materials which are able to predict all damage and failure modes at both material and component levels
• To define the test and data reduction methods needed to define the material card of thermoplastic composite material
• To develop numerical methodologies to propagate material, process and testing uncertainties along the model chain up to the allowable computation
• To define a reduced test matrix for the validation of the virtually generated allowables
• To define a complete framework and implement it in a user-friendly platform for advanced and non-advanced users
• To assess accuracy, efficiency, user-friendliness, etc., of the virtual material allowables generation platform, and perform the necessary incremental models and methodology improvements.
TREAL was successfully completed by achieving all the objectives. A brief overview of the project results will be presented here:
(a) WP1 produced all the thermoplastic composite panels that were needed for the experimental testing. A manufacturing consolidation cycle was selected to obtain panels according to the material supplier expectation. All the panels were manufactured and inspections were performed to assess the quality.
(b) All the planned experimental tests were performed between UdG and UPorto in WP2. Damage inspections were performed to understand the damage morphology of thermoplastic composites. Further, design allowables were obtained from the physical tests
(c) In WP3, a novel thermoplastic constitutive damage model was developed by UPORTO to predict the mechanical performance of thermoplastic composites
(d) In WP4, MSC-xstream developed parametric models to generate finite element virtual coupons tests. Further, validation of the thermoplastic test results with the virtual results were performed using different constitutive models to estimate the accuracy in prediction
(e) In WP5, UdG developed a methodology to perform global sensitivity analysis and uncertainty quantification & management, where using the validated models defined in WP4, simulations were performed to generate the virtual design allowables.
(f) in WP6, MSC-xstream generated the platform in Digimat for performing the variability analysis to obtain design allowables.
Project dissemination activities are briefed below:
(a) Peer-reviewed scientific publications: A total of 5 articles were published; 2 are under review, and 1 under preparation
(b) International conferences: Participation in 12 conferences to disseminate the results.
(c) Project communication at different international events including JEC world at Paris.
(a) WP1 produced all the thermoplastic composite panels that were needed for the experimental testing. A manufacturing consolidation cycle was selected to obtain panels according to the material supplier expectation. All the panels were manufactured and inspections were performed to assess the quality.
(b) All the planned experimental tests were performed between UdG and UPorto in WP2. Damage inspections were performed to understand the damage morphology of thermoplastic composites. Further, design allowables were obtained from the physical tests
(c) In WP3, a novel thermoplastic constitutive damage model was developed by UPORTO to predict the mechanical performance of thermoplastic composites
(d) In WP4, MSC-xstream developed parametric models to generate finite element virtual coupons tests. Further, validation of the thermoplastic test results with the virtual results were performed using different constitutive models to estimate the accuracy in prediction
(e) In WP5, UdG developed a methodology to perform global sensitivity analysis and uncertainty quantification & management, where using the validated models defined in WP4, simulations were performed to generate the virtual design allowables.
(f) in WP6, MSC-xstream generated the platform in Digimat for performing the variability analysis to obtain design allowables.
Project dissemination activities are briefed below:
(a) Peer-reviewed scientific publications: A total of 5 articles were published; 2 are under review, and 1 under preparation
(b) International conferences: Participation in 12 conferences to disseminate the results.
(c) Project communication at different international events including JEC world at Paris.
TREAL has provided a beyond the state-of art methodology to estimate design allowables of composite structures. The project opens doors towards a modified methodology to design structures, where extensive and expensive experimental campaigns can be partly replaced by numerical simulations. This provides a positive impact to the whole approach of design and certification to the aircraft industries, where the current approaches are expensive, time consuming and provides very conservative design allowables. Such a change in the design and analysis paradigm is crucial for the timely entry of new generation aircrafts by 2035.
More specifically, the advances with respect to the current state of the art have been:
- Development and implementation of advanced constitutive models for the prediction of the mechanical response of novel thermoplastic materials. The predictive accuracy has been within ±10% from the mean experimental results (limits defined by the Topic Manager)
- A complete characterization of a novel thermoplastic material has been performed at coupon level, with more than 500 experimental tests performed.
- Material cards for the two advanced constitutive models. Statistical distribution and material allowable (B-basis) for a total of 47 properties following the CMH-17 approach.
- Design allowables for the novel thermoplastic material for different 9 design drivers (like open hole or bearing strength) following the CMH-17 approach.
- A methodology to identify the critical uncertainties, to propagate them using analysis models and quantify on the quantity of interest. Three different workflow appraoches (including machine learning models) have been implemented for UQ&M.
- Generation of virtual design allowables for 9 different design drivers and compared with the experimental data, providing a good level of correlation.
- The methods and tools implemented in the DIGIMAT-FE platform, allowing the calculation of the virtual design allowables using a user-friendly platform.
The activities and developments from the TREAL project resulted in open access publications in top journals (5 published, 3 more pending) and up to 12 presentations in international conferences. Moreover, TREAL results are currently being further developed (exploited) in 2 ongoing HE projects, a Clean Aviation project, a recent project proposal for a HE call related to "Aviation safety - Uncertainty quantification for safety and risk management". Further, TREAL demonstrates efficient methodologies to account for material uncertainties and to implement it in a software platform which can be uptaken in the short term to encourage software developers to incorporate UQ&M modules in their softwares, which directly impacts the creation of value added jobs to the aerospace society.
More specifically, the advances with respect to the current state of the art have been:
- Development and implementation of advanced constitutive models for the prediction of the mechanical response of novel thermoplastic materials. The predictive accuracy has been within ±10% from the mean experimental results (limits defined by the Topic Manager)
- A complete characterization of a novel thermoplastic material has been performed at coupon level, with more than 500 experimental tests performed.
- Material cards for the two advanced constitutive models. Statistical distribution and material allowable (B-basis) for a total of 47 properties following the CMH-17 approach.
- Design allowables for the novel thermoplastic material for different 9 design drivers (like open hole or bearing strength) following the CMH-17 approach.
- A methodology to identify the critical uncertainties, to propagate them using analysis models and quantify on the quantity of interest. Three different workflow appraoches (including machine learning models) have been implemented for UQ&M.
- Generation of virtual design allowables for 9 different design drivers and compared with the experimental data, providing a good level of correlation.
- The methods and tools implemented in the DIGIMAT-FE platform, allowing the calculation of the virtual design allowables using a user-friendly platform.
The activities and developments from the TREAL project resulted in open access publications in top journals (5 published, 3 more pending) and up to 12 presentations in international conferences. Moreover, TREAL results are currently being further developed (exploited) in 2 ongoing HE projects, a Clean Aviation project, a recent project proposal for a HE call related to "Aviation safety - Uncertainty quantification for safety and risk management". Further, TREAL demonstrates efficient methodologies to account for material uncertainties and to implement it in a software platform which can be uptaken in the short term to encourage software developers to incorporate UQ&M modules in their softwares, which directly impacts the creation of value added jobs to the aerospace society.