1,721,043 research outputs found
A comparative review between genetic algorithm use in composite optimisation and the state-of-the-art in evolutionary computation
No full textThe task of providing optimal composite structures is increasingly difficult. New analysis approaches seek to model the material at the fibre/matrix scale and increasingly more control is sought over the material, for example optimising individual tows, and the structure, where new manufacturing techniques are proposed that will allow revolutionary new topologies. This additional complexity will stretch design engineers and as such it is important that state-of-the-art design methods are implemented to help take advantage of these exciting new opportunities, including computational optimisation methods. To determine best practice and the current limitations of the techniques a review of Genetic Algorithms in optimisation of composite materials and structures is performed over the last 10 years. This is compared to a technical review of the developments of Genetic Algorithms in the evolutionary computation literature. By better understanding how Genetic Algorithms are used in composite structures and comparing to evolutionary computational literature, recommendations are provided to help increase the use of Genetic Algorithms in solving composite optimisation problems in the future
Many-objective design optimisation of a plain weave fabric composite
No full textPlain weave fabrics provide low-cost composites used in many applications. Their mechanical properties are dependent on the weave and the yarn dimensions, which provides a complex design space to ensure optimal properties for a given application. Genetic Algorithms are commonly used in the literature to optimise the performance of composite materials but are currently limited to two or three objectives, where the optimisation may improve the specified properties but degrade others. In this paper 9 top performing Genetic Algorithms are benchmarked to find designs that respectively satisfy five-objective, three-objective and bi-objective formulations. The results show that the consideration of the five-objective problem is important, since the designs for the five-objective formulation give a wider range of results. These results do not include designs from the optimisation with the more limited objectives, meaning that these designs would need to be redesigned to be practical and demonstrating the benefits of optimisation with more objectives. cMLSGA is shown to be the strongest solver for these problems, contradicting the findings from the Evolutionary Computation literature. When compared with a current weave pattern, the five-objective optimisation provides 101 designs which improve all 5 material properties, with up to 76.61% improvements on the four mechanical properties and a maximum 37.73% reduction on areal density; there are weave patterns with designs that are specific to each of the properties individually
Comparative design of E-glass and flax structures based on reliability
No full textLongitudinal and transverse tensile properties, shear properties and Poisson's ratio of UD flax / epoxy and E-glass / epoxy composites from the literature </span
Benchmarking the performance of genetic algorithms on constrained dynamic problems
No full textThe growing interest in dynamic optimisation has accelerated the development of genetic algorithms with specific mechanisms for these problems. To ensure that these developed mechanisms are capable of solving a wide range of practical problems it is important to have a diverse set of benchmarking functions to ensure the selection of the most appropriate Genetic Algorithm. However, the currently available benchmarking sets are limited to unconstrained problems with predominantly continuous characteristics. In this paper, the existing range of dynamic problems is extended with15 novel constrained multi-objective functions. To determine how genetic algorithms perform on these constrained problems, and how this behaviour relates to unconstrained dynamic optimisation, 6 top-performing dynamic genetic algorithms are compared alongside 4 re-initialization strategies on the proposed test set, as well as the currently existing unconstrained cases. The results show that there are no differences between constrained/unconstrained optimisation, in contrast to the static problems. Therefore, dynamicity is the prevalent characteristic of these problems, which is shown to be more important than the discontinuous nature of the search and objective spaces. The best performing algorithm overall is MOEA/D, and VP is the best re-initialisation strategy. It is demonstrated that there is a need for more dynamic specific methodologies with high convergence, as it is more important to performance on dynamic problems than diversity.<br/
Dataset "Improving the applicability of genetic algorithms to real problems"
No full textDataset for the thesis for the degree of Doctorate of Philosophy, University of Southampton 2021. "Improving the applicability of genetic algorithms to real problems", by Przemyslaw Andrzej Grudniewski.
- The Dataset is separated according to chapters of the Thesis
- In each chapter the data is separated into 3 main categories: a) Figures in folder "Figures"; b) Table data in folder "Processed Table Data"; and c) raw data files in folder "Raw Data"
- Additionally, the code for the algorithm used to conduct all experiments is provided in "Additional_Data_Code_ReadMe.7z"</span
Effects of extensible modelling on composite riser mechanical responses
No full textThe change from steel risers to composites comes with uncertainties that led to large safety factors. One area of uncertainty is the predicted response and stresses derived from commercial packages that are based on formulations that assume in-extensible riser. However, composite pipes exhibit a lower axial stiffness and therefore the velocity of the axial waves will change with a corresponding change in dynamic response. To determine the effect of this assumption, this paper assesses the effect of extensibility on the time-domain response. It is found that the in-extensible model predicts 3 times the number of high frequency tension cycles in the 20kN tension range. To determine the impact of this change on the stress, the accuracy of available composite pipe models is benchmarked using shell, continuumshell and solid elements. The quadratic and continuum-shell elements provide a maximum percentage difference of 4% compared to solid elements but the continuum-shell is selected as it has a lower computational cost. The response from the extensible and in-extensible models are input into the pipe model, they provide similar Tsai-Wu failure factors, alleviating concerns when modelling the strength. However, the change in dynamics remains a concern for other applications such as machine-learning or digital-twins
Behaviour of multi-level selection genetic algorithm (MLSGA) using different individual-level selection mechanisms
No full textThe Multi-Level Selection Genetic Algorithm (MLSGA) is shown to increase the performance of a simple Genetic Algorithm. It is unique among evolutionary algorithms as its sub-populations use separate selection and reproduction mechanisms to generate offspring sub-populations, called collectives in this approach, to increase the selection pressure, and uses a split in the fitness function to maintain the diversity of the search. Currently how these novel mechanisms interact with different reproduction mechanisms, except for the one originally tested at the individual level is not known. This paper therefore creates three different variants of MLSGA and explores their behaviour, to see if the diversity and selection pressure benefits are retained with more complex individual selection mechanisms. These hybrid methods are tested using the CEC’09 competition, as it is the widest current benchmark of bi-objective problems, which is updated to reflect the current state-of-the-art. Guidance is given on the new mechanisms that are required to link MLSGA with the different individual level mechanisms and the hyperparameter tuning which results in optimal performance. The results show that the hybrid approach increases the performance of the proposed algorithms across all the problems except for MOEA/D on unconstrained problems. This shows the generality of the mechanisms across a range of Genetic Algorithms, which leads to a performance increase from the MLSGA collective level mechanism and split in the fitness function. It is shown that the collective level mechanism changes the behaviour from the methods selected at the individual level, promoting diversity first instead of convergence, and focuses the search on different regions, making it a particularly strong choice for problems with discontinuous Pareto fronts. This results in the best general solver for the updated bi-objective CEC’09 problem sets
Sustainable Sandwich Panels for Use in Ship Superstructures
No full textShip superstructures are commonly manufactured from steel but composite sandwich structures could be an alternative leading to significant weight savings. In addition to these weight savings using sustainable materials could reduce the environmental impact of ship production, operation and recyclability in shipping. However, these materials must be capable of equal performance to those that are currently used. Comparing sandwich panels is complex as there are many objectives, stress to strength ratio, sustainability, cost and mass, and variables such as the skin and core materials and thicknesses. Due to this complexity Genetic Algorithms are used to compare potential designs, providing different material selections for different combinations of objectives. The comparison between different Genetic Algorithms demonstrates that HEIA is the most effective algorithm but with all of the algorithms having equivalent performance on these problems. The optimisation provides a set of 716 feasible designs, with balsa being the most popular core but with feasible solutions split between the flax, carbon and glass skins
Concurrent engineering in the context of the composite leisure boatbuilding industry
No full textLeisure boatbuilding is an industry that has tight profit margins and growing competition due to the global nature of the industry. It is a growth market with the number of high-earning potential customers increasing worldwide. For British boatbuilding to retain and increase its high standing within these global markets investment is required to develop larger profits and market share. Concurrent engineering is a method of design that has given large benefits to a multitude of industries but is ill-defined within leisure boatbuilding.This thesis investigates the nature of British boatbuilding and develops concurrent engineering within this context. To develop faster design while increasing quality this thesis concentrates on automated communication. A number of tools are developed focusing on structures and production. These include a mass and cost multi-objective optimisation tool further developing first principles rules using a Genetic Algorithm, a reliability tool to increase the speed of iterative design and a design history tool focusing on data mining using neural networks within a grid computing structure. Furthermore, a concurrent engineering methodology specific to leisure boatbuilding has been developed leading to a design environment for use within this sector. The resulting work develops techniques that increase the knowledge available to engineers in an intuitive, quantitative, manner
Fatigue approaches for mooring chains subjected to wear degradation
No full textThere are currently 365 FPSOs in service around the world. These vessels all use mooring lines to maintain position and provide stability, keeping the vessel and cargo safe. However, more than 21 failures have occurred between 2001 and 2011 and approximately 50% of the reported failures occurred in the first 3 years of 20-year design life. Each mooring line failure represents the potential for serious environmental and economic consequences. Based on industry surveys, the most common failure mode is fatigue failure. In the current offshore standards, the surface degradation due to wear and corrosion is modelled as a diameter loss at a standards rate. To assess whether the uniform reduction in chain diameter suggested in the offshore standards is able to explain the early chain failures seen in service, this paper incorporates two wear rates into a fatigue life calculation; one wear rate is taken from DNV-OS-E301 and is compared against one taken from NORSOK M-001. Three fatigue life estimation approaches: tension, nominal stress and hotspot, are used to compare the differences in fatigue method. The stress in the chain is calculated using an analytical model, which is verified against an FE model. The effect of wear degradation on the ultimate strength of the chain is calculated based on the minimum breaking load. The results show that the diameter loss rates suggested in the offshore standards are not able to explain the early mooring chain failures seen in the past and that the reduction of diameter cannot solely explain the early failures seen in service. The hotspot approach, not often used in mooring line predictions, is best able to predict these shorter lives, as it offers more accurate fatigue predictions by considering high peak stresses compared to standard methods such as tension and nominal stress approaches
- …
