Design and optimisation of composite structures
Lead Research Organisation:
Imperial College London
Department Name: Aeronautics
Abstract
Numerical analysis and design of structures is a field with ever growing potential. Advancements in technology and hardware enable designers to rely on numerical methods for the design of complex structures. Aeronautics and aerospace are examples of areas in which these complex structures exist. These complex structures are often made of composite materials, which aside from the numerous advantages are very complex in nature; thus, they require advanced methods and models for their design and mechanical analysis, especially when considering the failure of the material.
Optimisation is a design tool that enables the engineer to achieve a better design configuration of the structure through the aid of numerical methods. Topology optimisation is one method with the remit of optimising the mass distribution of a structure to satisfy some requirements and to provide better properties. Evolving or moving boundaries are an intrinsic part of topology optimisation. Traditionally, the efforts to model evolving boundaries rely on implicit schemes, which provide the means to efficiently model the evolution of a boundary but lack the ability to transmit information at the boundary. Explicit alternatives are often computationally expensive and inherently complex to implement. In this PhD, one objective is to investigate a solution to this dichotomy: a novel finite element method (FEM) based formulation, using the floating node method (FNM), capable of explicitly representing moving boundaries in a domain. Such methodology has the potential to bridge the gap between analysis and design by providing an accurate translation between the numerical domain and the design domain.
The simultaneous analysis of different time and length scales is designated multiscale analysis. Multiscale analysis of composite materials is paramount when one aims to investigate the failure mechanisms of composite materials. These mechanisms are varied and happen at very different time and length scales. Being able to capture all these scales within one numerical analysis leads to being able to capture the complex material behaviour in real structures. This project has the aim of investigating new multiscale techniques, using the floating node method, that enable the accurate and detailed analysis of the failure of real-world composite structures.
In summary, the overall aim of this PhD is to develop numerical methods that have the potential of being a step-change in the analysis and design of composite structures. The more detailed objectives include the development of methods that bridge the design and analysis domains through optimisation methods; the development of new methods capable of handling physical phenomena at different time and length scales simultaneously (multiscale); and the development of methods that join optimisation and multiscale analysis to tackle large-scale, non-linear failure of composite structures.
Optimisation is a design tool that enables the engineer to achieve a better design configuration of the structure through the aid of numerical methods. Topology optimisation is one method with the remit of optimising the mass distribution of a structure to satisfy some requirements and to provide better properties. Evolving or moving boundaries are an intrinsic part of topology optimisation. Traditionally, the efforts to model evolving boundaries rely on implicit schemes, which provide the means to efficiently model the evolution of a boundary but lack the ability to transmit information at the boundary. Explicit alternatives are often computationally expensive and inherently complex to implement. In this PhD, one objective is to investigate a solution to this dichotomy: a novel finite element method (FEM) based formulation, using the floating node method (FNM), capable of explicitly representing moving boundaries in a domain. Such methodology has the potential to bridge the gap between analysis and design by providing an accurate translation between the numerical domain and the design domain.
The simultaneous analysis of different time and length scales is designated multiscale analysis. Multiscale analysis of composite materials is paramount when one aims to investigate the failure mechanisms of composite materials. These mechanisms are varied and happen at very different time and length scales. Being able to capture all these scales within one numerical analysis leads to being able to capture the complex material behaviour in real structures. This project has the aim of investigating new multiscale techniques, using the floating node method, that enable the accurate and detailed analysis of the failure of real-world composite structures.
In summary, the overall aim of this PhD is to develop numerical methods that have the potential of being a step-change in the analysis and design of composite structures. The more detailed objectives include the development of methods that bridge the design and analysis domains through optimisation methods; the development of new methods capable of handling physical phenomena at different time and length scales simultaneously (multiscale); and the development of methods that join optimisation and multiscale analysis to tackle large-scale, non-linear failure of composite structures.
Organisations
Publications
Da Costa R
(2020)
A novel formulation for the explicit discretisation of evolving boundaries with application to topology optimisation
in Computer Methods in Applied Mechanics and Engineering
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/N509486/1 | 30/09/2016 | 30/03/2022 | |||
2296924 | Studentship | EP/N509486/1 | 30/09/2018 | 30/03/2022 | Rui Silva Sampaio Da Costa |
EP/R513052/1 | 30/09/2018 | 29/09/2023 | |||
2296924 | Studentship | EP/R513052/1 | 30/09/2018 | 30/03/2022 | Rui Silva Sampaio Da Costa |
Description | So far, the work carried out through this award led to a publication and a conference presentation which outlines the findings for part of the objectives of the award. The publication focuses on the optimisation aspect and the development of the methodology that enables the optimisation in a seamless fashion, which was the first key objective. At the moment the ongoing work is linking those findings with other key objectives through problems with industrial relevance and of engineering significance; more specifically answering the question of 'how can we use this seamless optimisation methodology on a large scale composite structure?'. |
Exploitation Route | The published worked is highly relevant to different fields of research and is generalisable to different applications. All details of the methodology developed as well as complete source code for the work is available online to promote collaboration and further development. As an example, the methodology can be readily applied to fields such as fluid-structure interaction and multi-phase flow. Also, on the topic of this award the methodology can be expanded to address different problems relating to composite materials and highly complex composite structures such as a wing-box. |
Sectors | Aerospace Defence and Marine |
URL | https://doi.org/10.1016/j.cma.2020.113077 |
Title | A novel formulation for the explicit discretisation of evolving boundaries with application to topology optimisation |
Description | A methodology based on finite element that couples floating nodes and the level set method to allow for a seamless discretisation of evolving boundaries |
Type Of Technology | New/Improved Technique/Technology |
Year Produced | 2020 |
Open Source License? | Yes |
Impact | It allows to seamlessly optimise the shape of a structure within the finite element framework; something that would previously require fundamental changes to the inner workings of a finite element solver. |
URL | https://doi.org/10.1016/j.cma.2020.113077 |
Description | Presentation at the 14th WCCM ECCOMMAS 2020 conference |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Postgraduate students |
Results and Impact | This was a presentation in the largest international conference on computational mechanics and numerical methods for which I was granted a scholarship to attend. It allowed for me to contact with different research and visions as well as expose my own. |
Year(s) Of Engagement Activity | 2020 |
URL | https://slideslive.com/38943906 |