Simulation approaches to support composite laminate analysis from early stage design

Abstract

Composite laminates have enabled the creation of components that are lighter and stronger than their metallic equivalents, making them an enticing material for the development of novel components. However, the cost to build accurate and efficient structural analysis models of these components can be a limiting factor in the development process. This issue stems from the composite laminate being composed of a selection of orthotropic layers forming an anisotropic material that behaves more akin to a structure than a traditional material. The use of this laminate structure introduces increased material data that needs to be represented within the analysis.

This composite representation is usually generated using one of several common composite analysis methods, with the selection of the most suitable one being guided by the requirements of the analysis. Reviewing the common methods identified a gap in the state-of-the-art for 3D analysis of composite laminates at early design stages, i.e. when design data is limited. To address this a novel analysis method, referred to here as the internally smeared method, was developed and tested. The internally smeared method models the plies on the outer surfaces in full, with the remaining plies modelled using a smeared material property. The method was first tested using classical laminate theory (CLT) on a data set of lay-ups to establish the trade-off between the number of plies modelled in full, i.e. the computational expense of the analysis, and accuracy of results. Following this trade-off study, finite element testing was conducted on benchmark problems. These tests showed that the number of plies needing to be modelled in full to achieve a desired accuracy is up to 97% less than the complete ply lay-up, with the remaining region modelled using the computationally cheaper smeared material property.

The second aim of this work was to capture the analysis decisions that are made when generating the composite analysis using the Simulation Intent framework. These composite specific decisions have been identified by investigating the common simulation attributes for the previously mentioned analysis methods and identifying the high-level decisions that the analyst makes when using each method. Capturing these analysis decisions in a high-level and structured manner allows the creation and dissemination of neutral methods for an analysis. In this work the captured information has been utilised as inputs for a semi-automated workflow, showcasing a reduction in pre-processing time for the generation and updating of composite laminate analysis models.

Overall, this leads to a more cost efficient, robust and repeatable workflow for analysing composite laminates, enabling the consideration of composite laminates earlier in the design stage and the redefinition of the analysis as the design progresses.

Date of Award	Jul 2022
Original language	English
Awarding Institution	Queen's University Belfast
Sponsors	Northern Ireland Department for the Economy
Supervisor	Declan Nolan (Supervisor) & Trevor T Robinson (Supervisor)