Finite element modelling of composite materials under impact loading, challenges and potential ways forward.
Topic(s) : Material and Structural Behavior - Simulation & Testing
Co-authors :
Sakineh FOTOUHI (FRANCE), Amin FARROKHABADI , Mohammad FOTOUHI (NETHERLANDS)Abstract :
Despite the high performance of composite materials, they have an Achilles' heel in their susceptibility to impact damage, causing barely visible impact damage (BVID) and catastrophic failure at low loads. BVID significantly affects the performance and safety of composite materials. Optimizing the design of composite materials is costly due to the need for numerous experiments, and modelling these materials under impact loading is essential but challenging.
The focus of this paper is to fundamentally understand composite materials under impact loading using finite element modelling (FEM). The methodology comprises an in-depth review of current FEMs and the challenges that need to be addressed. The suggested solutions are investigated for a carbon epoxy laminate, and the results are compared with experimental findings. The main interactive factors that need to match the experiment with FEM include real stiffness, BVID initiation points in terms of displacement and time, damage pattern, and damage progression.
BVID comprises matrix cracking and delamination damages, with the latter being the main concern with BVIDs. Delamination initiation and propagation patterns are not well-predicted using the ultimate failure criteria with the available materials' library of the existing FEM software. This issue is addressed by employing cohesive elements for both modelling matrix cracking within plies and for interface delamination, as delamination initiates due to matrix cracks, according to the literature review. This approach ensures that the damage pattern and progression match much more closely with experimental observations.
Special consideration is required to account for the enhancement effect of through-thickness compression on shear behaviour, identified as the primary influencing factor of delamination. This particularly affects the modelling of the outermost layers underneath the impactor, where damage initiates at a very early stage, not aligning with experimental results. This phenomenon cannot be modelled with the current materials library for cohesive elements. In this work, user-defined materials are implemented to facilitate the through-thickness compression effect. This has been investigated by defining different factors, including shear strength and Mode II critical energy release rate, to linearly increase by a material-dependent enhancement factor. There is a considerable improvement in damage initiation and its progression with the use of user-defined materials. The conclusion also addresses the dependency of the enhancement factor on the impact energy level.
The experimental thickness slightly differs from the FEM due to additional cohesive interfaces between each composite ply. This discrepancy is accounted for by applying a scaling factor to the FEM results, improving stiffness matching with the experiments.
The focus of this paper is to fundamentally understand composite materials under impact loading using finite element modelling (FEM). The methodology comprises an in-depth review of current FEMs and the challenges that need to be addressed. The suggested solutions are investigated for a carbon epoxy laminate, and the results are compared with experimental findings. The main interactive factors that need to match the experiment with FEM include real stiffness, BVID initiation points in terms of displacement and time, damage pattern, and damage progression.
BVID comprises matrix cracking and delamination damages, with the latter being the main concern with BVIDs. Delamination initiation and propagation patterns are not well-predicted using the ultimate failure criteria with the available materials' library of the existing FEM software. This issue is addressed by employing cohesive elements for both modelling matrix cracking within plies and for interface delamination, as delamination initiates due to matrix cracks, according to the literature review. This approach ensures that the damage pattern and progression match much more closely with experimental observations.
Special consideration is required to account for the enhancement effect of through-thickness compression on shear behaviour, identified as the primary influencing factor of delamination. This particularly affects the modelling of the outermost layers underneath the impactor, where damage initiates at a very early stage, not aligning with experimental results. This phenomenon cannot be modelled with the current materials library for cohesive elements. In this work, user-defined materials are implemented to facilitate the through-thickness compression effect. This has been investigated by defining different factors, including shear strength and Mode II critical energy release rate, to linearly increase by a material-dependent enhancement factor. There is a considerable improvement in damage initiation and its progression with the use of user-defined materials. The conclusion also addresses the dependency of the enhancement factor on the impact energy level.
The experimental thickness slightly differs from the FEM due to additional cohesive interfaces between each composite ply. This discrepancy is accounted for by applying a scaling factor to the FEM results, improving stiffness matching with the experiments.