Abstract
Compressive performance of unidirectional (UD) fibre-reinforced composites is essential when designing large composite structures loaded in bending, such as wind turbine blades. This article looks at a study that deals with compression tests of 5 mm thick pultruded carbon fibre profiles and aims to develop a repeatable and representative method.
Based on experimental observations and results, for instance, from Ferguson et al. [1], constant width compression specimens exhibit an incorrect failure mode at the end of the gripping part. The waisted test sample is found to carry a higher load even though the final failure is still observed outside the gauge section. This failure now occurs in the necking region, under the tabs, or in the junction with the grips. The failure mechanism exhibits kink-band and splitting, leading to the final failure of the test specimen. As it is not in the gauge area of the sample, this still does not represent the full potential of the compressive load performance of the pultruded profiles.
The objective is to drive the failure into the gauge section by sizing and optimising specimen shape/geometry. A modelling-based approach is relevant to tackle this issue, as it constitutes an excellent alternative to long experimental campaigns and allows for a parametric study.
In this framework, a 3D Abaqus implicit Finite Element (FE) model is built to study and give detailed information on the loading of the specimen in compression and analyse the phenomenon at the origin of the early fracture. A parametric study is performed, and stress distribution compared between different iterations. A new specimen geometry is manufactured and tested from this first iteration step. Obtained experimental results are used as experimental/numerical dialogue and constitute new input data for the analysis and a thorough investigation of the compressive behaviour.
Based on experimental observations and results, for instance, from Ferguson et al. [1], constant width compression specimens exhibit an incorrect failure mode at the end of the gripping part. The waisted test sample is found to carry a higher load even though the final failure is still observed outside the gauge section. This failure now occurs in the necking region, under the tabs, or in the junction with the grips. The failure mechanism exhibits kink-band and splitting, leading to the final failure of the test specimen. As it is not in the gauge area of the sample, this still does not represent the full potential of the compressive load performance of the pultruded profiles.
The objective is to drive the failure into the gauge section by sizing and optimising specimen shape/geometry. A modelling-based approach is relevant to tackle this issue, as it constitutes an excellent alternative to long experimental campaigns and allows for a parametric study.
In this framework, a 3D Abaqus implicit Finite Element (FE) model is built to study and give detailed information on the loading of the specimen in compression and analyse the phenomenon at the origin of the early fracture. A parametric study is performed, and stress distribution compared between different iterations. A new specimen geometry is manufactured and tested from this first iteration step. Obtained experimental results are used as experimental/numerical dialogue and constitute new input data for the analysis and a thorough investigation of the compressive behaviour.
| Original language | English |
|---|---|
| Journal | BENCHMARK Magazine |
| Volume | January 2024 |
| Pages (from-to) | 44-50 |
| Publication status | Published - 2024 |