Periodic Reporting for period 2 - DADIYSO COMP (DESIGN AGAINST DISTORTION WITH LAYUP AND SHAPE OPTIMIZATION OF CARBON-EPOXY AEROSPACE COMPOSITES)
Période du rapport: 2018-03-01 au 2019-06-30
"In developing the next generation of lighter and greener air transport, polymer composite materials are promising to achieve weight savings and other possible advantages related to material design and manufacturability.
The phenomena known as ""spring-back"" or “cure induced distortion” has long been a challenge in manufacture using high temperatures and pressures (i.e. autoclave conditions).
Differences in thermal expansion of the resin, fibres, and tooling, combined with chemical shrinkage of the matrix can cause large and unwanted deformations to components. This leads to problems in production and assembly, requiring rework, repair, or even scrappage of expensive parts.
Addressing these aspects can enable a more widespread adoption of composite materials within the European aviation industry. These materials can help the European aviation industry remain competitive, and offer light weight, high performance products with minimal impact on the environment. The tools developed within the project can enable the use of simulation in the early stages of design, reducing development time and costs significantly. Maintaining a competitive industry and reducing time-to-market are two key aspects of the project which are of significant interest to European society, as they will have a direct impact on the European economy within the aviation sector.
Historically, design rules and tool compensation by manufacturing experts have been the dominant methods of controlling distortion. Within DADIYSO COMP, optimisation methods will be developed to instead alter the layup and shape of components in order to eliminate the negative effects of cure induced distortion. Changes to address distortion need also consider stiffness and strength requirements. This approach has not previously been attempted and achieving a set of viable tools for this task is the overall objective of the project.
"
The phenomena known as ""spring-back"" or “cure induced distortion” has long been a challenge in manufacture using high temperatures and pressures (i.e. autoclave conditions).
Differences in thermal expansion of the resin, fibres, and tooling, combined with chemical shrinkage of the matrix can cause large and unwanted deformations to components. This leads to problems in production and assembly, requiring rework, repair, or even scrappage of expensive parts.
Addressing these aspects can enable a more widespread adoption of composite materials within the European aviation industry. These materials can help the European aviation industry remain competitive, and offer light weight, high performance products with minimal impact on the environment. The tools developed within the project can enable the use of simulation in the early stages of design, reducing development time and costs significantly. Maintaining a competitive industry and reducing time-to-market are two key aspects of the project which are of significant interest to European society, as they will have a direct impact on the European economy within the aviation sector.
Historically, design rules and tool compensation by manufacturing experts have been the dominant methods of controlling distortion. Within DADIYSO COMP, optimisation methods will be developed to instead alter the layup and shape of components in order to eliminate the negative effects of cure induced distortion. Changes to address distortion need also consider stiffness and strength requirements. This approach has not previously been attempted and achieving a set of viable tools for this task is the overall objective of the project.
"
"The project work has included a characterization campaign for an epoxy system suitable for out of autoclave manufacture of aircraft composites. This was done to obtain material data to use in simulation subroutines developed during the project.
The first portion of the project refined and improved a high-fidelity simulation code for residual stress analysis previously developed in-house. The effect of the changing modulus during the rubbery stages of the matrix curing was included. These code used to develop a method for rapid cure distortion analysis based on linear analysis using equivalent properties. This ""rapid method"" has so far led to one conference publication (ECCM18, Athens), and one manuscript for submission to a scientific journal. In addition, portions of the work have been discussed directly with other research groups around the world working on similar problems, in addition to popular science presentations in Sweden.
Manufacturing was also performed to create small coupons, and large parts for validation. Distortions were measured with a high accuracy laser scanning system and compared to the simulation results. The accuracy of the two modelling approaches were examined in detail. In both cases, good agreement between simulated distortions and measured distortions were observed, however there was a large amount of scatter in the results from manufacturing. The materials and manufacturing methods used proved impossible to obtain the level of precision and manufacturing tolerance associated with prepreg manufacture in an autoclave. Evaluation of the distortion of the components also required quite detailed study of the laser scanning results, due to the tendency of e.g. local curvatures to confound typical measurements of angles etc.
In the second half of the project, a method for optimization of the shape and layup of the composite parts was developed. A new metric to measure the distortion severity was developed which accounts for displacement and local stiffness in the model. A novel stacking sequence generation algorithm was developed to create a composite stack based on ratios of allowed orientations, and some basic design rules from industry. These tools were implemented in python scripts to work with Abaqus CAE.
In addition, a toolbox to interface between CAD and CAE software was developed. This allowed a parameterized CAD model to be included directly in the optimization loop. This means the native geometry CAD model could be updated with a few clicks in the software post optimization. This reduces design time caused by iterations between CAD and FEA specialists.
The complete set of tools were finally demonstrated on a used case part. A large, stiffened panel assembly was optimized to minimize distortion while simultaneously maintaining minimum buckling load factor constraints.
The work has led to one conference presentation, and two manuscripts for submission to academic journals. In addition, publications in social media, popular science articles and presentations have been part of the dissemination. Further, discussions to incorporate and market the tools in future projects is ongoing."
The first portion of the project refined and improved a high-fidelity simulation code for residual stress analysis previously developed in-house. The effect of the changing modulus during the rubbery stages of the matrix curing was included. These code used to develop a method for rapid cure distortion analysis based on linear analysis using equivalent properties. This ""rapid method"" has so far led to one conference publication (ECCM18, Athens), and one manuscript for submission to a scientific journal. In addition, portions of the work have been discussed directly with other research groups around the world working on similar problems, in addition to popular science presentations in Sweden.
Manufacturing was also performed to create small coupons, and large parts for validation. Distortions were measured with a high accuracy laser scanning system and compared to the simulation results. The accuracy of the two modelling approaches were examined in detail. In both cases, good agreement between simulated distortions and measured distortions were observed, however there was a large amount of scatter in the results from manufacturing. The materials and manufacturing methods used proved impossible to obtain the level of precision and manufacturing tolerance associated with prepreg manufacture in an autoclave. Evaluation of the distortion of the components also required quite detailed study of the laser scanning results, due to the tendency of e.g. local curvatures to confound typical measurements of angles etc.
In the second half of the project, a method for optimization of the shape and layup of the composite parts was developed. A new metric to measure the distortion severity was developed which accounts for displacement and local stiffness in the model. A novel stacking sequence generation algorithm was developed to create a composite stack based on ratios of allowed orientations, and some basic design rules from industry. These tools were implemented in python scripts to work with Abaqus CAE.
In addition, a toolbox to interface between CAD and CAE software was developed. This allowed a parameterized CAD model to be included directly in the optimization loop. This means the native geometry CAD model could be updated with a few clicks in the software post optimization. This reduces design time caused by iterations between CAD and FEA specialists.
The complete set of tools were finally demonstrated on a used case part. A large, stiffened panel assembly was optimized to minimize distortion while simultaneously maintaining minimum buckling load factor constraints.
The work has led to one conference presentation, and two manuscripts for submission to academic journals. In addition, publications in social media, popular science articles and presentations have been part of the dissemination. Further, discussions to incorporate and market the tools in future projects is ongoing."
The advances beyond state-of-the art are as follows.
On the topic of cure distortion simulation, some important incremental improvements in the most advanced forms of modelling residual stresses have been achieved, such as accounting for the changing rubbery modulus during cure. The development of the rapid simulation methodology is however much more important. By using the high fidelity model to tune material properties in the rapid method, complex geometry can be simulated in a fraction of the time, a reduction 1000x was observed. This has large implications for the simulation of large complex structures with multiple components. The difficulty and time consumed in simulating such structures means that optimization has been impossible. With the rapid method in place, parameter studies and optimization can be done early on to minimize the effects of cure induced distortion in large structures.
The methods developed for optimization against cure distortion are novel. A new metric for evaluating the severity of distortion has been developed. A novel algorithm for stack generation has also been created. Together they form a framework for stack optimization which shows significant promise in its simple, and manufacturing centric approach.
These tools can be of great help to the aerospace industry. The ability to rapidly predict and perform pro-active design optimization against distortion of parts prior to commissioning tooling represents a significant step forward in composite manufacturing. This can lead to reduced costs and lower risks of using composites, as late-phase reworks of components and tooling can be avoided. This can lead to a wider acceptance of composite materials, and in the long term, more structurally weight efficient aircraft that will in turn produce fewer emissions.
The potential applications of the tools are not limited to the aeronautical industry. Composite materials have become of significant interest to the automotive industry where dimensional tolerances are important, and the ability to simulate every aspect of the manufacturing process is critical. In this sense, these tools could be of significant value to the automotive industry. This will also lead to lower weight and less polluting vehicles which can have significant societal benefits.
On the topic of cure distortion simulation, some important incremental improvements in the most advanced forms of modelling residual stresses have been achieved, such as accounting for the changing rubbery modulus during cure. The development of the rapid simulation methodology is however much more important. By using the high fidelity model to tune material properties in the rapid method, complex geometry can be simulated in a fraction of the time, a reduction 1000x was observed. This has large implications for the simulation of large complex structures with multiple components. The difficulty and time consumed in simulating such structures means that optimization has been impossible. With the rapid method in place, parameter studies and optimization can be done early on to minimize the effects of cure induced distortion in large structures.
The methods developed for optimization against cure distortion are novel. A new metric for evaluating the severity of distortion has been developed. A novel algorithm for stack generation has also been created. Together they form a framework for stack optimization which shows significant promise in its simple, and manufacturing centric approach.
These tools can be of great help to the aerospace industry. The ability to rapidly predict and perform pro-active design optimization against distortion of parts prior to commissioning tooling represents a significant step forward in composite manufacturing. This can lead to reduced costs and lower risks of using composites, as late-phase reworks of components and tooling can be avoided. This can lead to a wider acceptance of composite materials, and in the long term, more structurally weight efficient aircraft that will in turn produce fewer emissions.
The potential applications of the tools are not limited to the aeronautical industry. Composite materials have become of significant interest to the automotive industry where dimensional tolerances are important, and the ability to simulate every aspect of the manufacturing process is critical. In this sense, these tools could be of significant value to the automotive industry. This will also lead to lower weight and less polluting vehicles which can have significant societal benefits.