Carbon fibre textile composite for automotive applications
Lead Research Organisation:
University of Bristol
Department Name: Aerospace Engineering
Abstract
A JLR body structure must perform a significant number of functions including all of those
required for crash, NVH & durability attributes. JLR aspiration is that all of these attributes
are assessed using virtual methods e.g. finite element models. The accuracy of these models
depends on 2 main inputs 1. Geometry (a true representation of the design) & 2. Material
models (how the design responds when loaded) - get either wrong & significant risk (timing,
cost etc.) is added into a programme.
Carbon Fibre Composites are a highly anisotropic material where effectively a new material
property is created every time a fibre angle, fibre size or fibre/resin ratio is changed, even
within the same nominal material grade. In order to optimise the usage of these textile
materials, their material properties must be understood.
Rather than adopting a traditional test & measure approach to materials characterisation
(which typically costs thousands of pounds per material, takes weeks to complete, and needs
to be repeated if there is any change to any of the vast number of variables in the format of
the composite material), we propose developing innovative virtual models of the material at
the micro-scale and then using these models to extract the macro-scale material properties
required for vehicle attribute model inputs.
required for crash, NVH & durability attributes. JLR aspiration is that all of these attributes
are assessed using virtual methods e.g. finite element models. The accuracy of these models
depends on 2 main inputs 1. Geometry (a true representation of the design) & 2. Material
models (how the design responds when loaded) - get either wrong & significant risk (timing,
cost etc.) is added into a programme.
Carbon Fibre Composites are a highly anisotropic material where effectively a new material
property is created every time a fibre angle, fibre size or fibre/resin ratio is changed, even
within the same nominal material grade. In order to optimise the usage of these textile
materials, their material properties must be understood.
Rather than adopting a traditional test & measure approach to materials characterisation
(which typically costs thousands of pounds per material, takes weeks to complete, and needs
to be repeated if there is any change to any of the vast number of variables in the format of
the composite material), we propose developing innovative virtual models of the material at
the micro-scale and then using these models to extract the macro-scale material properties
required for vehicle attribute model inputs.
People |
ORCID iD |
| Stephen Hallett (Primary Supervisor) | |
| Meng Yi Song (Student) |
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/P510427/1 | 01/10/2016 | 31/12/2021 | |||
| 1834948 | Studentship | EP/P510427/1 | 01/07/2016 | 30/11/2020 | Meng Yi Song |
| Description | The tows in a 2D woven carbon fibre material rotates away from its orthogonal position as a 2D fabric conforms to a 3D shape. The effect of localised shear in the final composite part is commonly ignored in macro-scale mechanical simulations. My current research allows sheared unit cell properties to be implemented in macro-scale FE simulation to improve accuracy. Update 2020 March 3rd: Damage models are recently implemented into both meso scale and macro scale model. At meso scale, the Young's modulus and damage initation could be predicted, which in turn is used as an input for macro scale model to predict local damage of the final composite part. Experimental work for validating the meso scale model is completed. The material used is SolvaLite710-1, a rapid cure resin system. The findings are: 1. Increased shear angle reduce the void content in the finished part. 2. For unsheared prepreg, as it comes off the roll, have an average shear angle of 4 degrees. 3. Shear angle is very difficult to control when plys must be removed from the picture frame, the higher the shear angle the higher the retraction. While making the sheared cured plates, it was found that for the 10 degrees shear, the actual shear angle is at 7 degrees, for 20 and 30 degree shear, the actual shear angle is 14 degrees. There appears to be an "ideal resting shear angle" which is based on the desired shear angle and shear stiffness. 4. The shear angles for unit cells at towards the edge of the picture frame varies with the unit cell shear angle at the centre of the sheared fabric. The unit cell at centre tends to be sheared less than edge if harderend with picture frame constraint. And tends to retain the shear angle when fabric is not constrained. 5. When making 6ply coupons, it was found that that nominal thickness per ply cannot be achieved when forming just one ply, due to the nesting effect between plys. For example, a single ply can be compressed to 0.52mm, however for a 6ply coupon, the nominal thickness can be compressed to 0.47mm. The nesting regions are identified visually using CT scan. 6. Unsheared have a little more composite shrinkage compared to the high shear coupons, although this change is so small that it can be negligible. 7. If the shear allows the fibres to skew towards the tensile loading direction, then the Young's modulus behaves exponentially as the shear angle increases. However, if the fibres are skewing away from the tensile direction, the Young's modulus appears to be insensitive. 8. Although the Young's modulus increases exponentially at higher shear angles (when fibres skewing towards the tensile direction), strength seems to be capped up to a certain point. 9. The different in temperature do have an effect on shear stiffness and bending stiffness. at 50 and 80 degrees, it is observed that both parameters for the prepreg behaves similarly to the dry fabric. Whereas the prepreg is very stiff at room temperature. No clear differences are found between 50 and 80 degrees, this could due to the difficulty in controlling the temperature for such a rapid cure system. Update 2022 February 15th: 1. Current FE model is capable of predicting the stiffness of 2D woven carbon fibre composite with good accuracy, with shear taken into account. 2. Proposed a method where the local changes in material's properties due to shear can be applied onto the final, component level, macro scale model. 2. Damage initiation is hard to predict mainly due to the lack of debonding mechanism in the FE model. |
| Exploitation Route | My outcome can be used to improve accuracy in macro-scale FE model. Therefore, any applications where the user need to perform a macro-scale mechanical FE model of a balanced 2D woven material (not restricted to carbon fibre) , may use my work to improve their accuracy. The amount of improvement should be dependent on the complexity of the part, as more complex geometries will create more regions of local shears. This is currently in process so the actual amount of improvement need to be researched. This can also used as a base work to add more types of fabric weaves such as UD, NCF, unbalanced 2D or 3D woven materials. Update 10/03/2020 An FE work flow has established from the material creation (weave, unit cell size, tow properties, resin properties), to material property prediction, to forming simulation to obtain fabric deformation, till macro-scale component level simulation to predict component's performance. This allows any designer to quickly obtain an realistic component behaviour from simply construct the woven composite. A large amount of experimental data are generated for validating my FE models, these data can also be used for future researches of similar field that involves fabric reinforced composites. For example, the effect of shear void content, the shear stiffness of the fabric (prepreg under different temperature and compare that to the dry), and the effect on the Young's modulus and strength as a result of the shear angles locked up in the composite. Since the rapid cure system is a new technology, and speed is a major factor to be considered for mass production, it is expected there will be more research projects coming up in the future for these rapid cure systems. My experimental result provides first hand data for shear behaviour and the effect of the shear of this novel rapid cure system. |
| Sectors | Aerospace, Defence and Marine,Chemicals,Construction,Digital/Communication/Information Technologies (including Software),Manufacturing, including Industrial Biotechology,Transport |