Characterising the Thermoviscoelastic Bending Behaviour of Carbon Fibre Reinforced PA6 Laminates via the Cantilever Bend Test
Topic(s) : Experimental techniques
Co-authors :
George Edward STREET (UNITED KINGDOM), Michael Sylvester JOHNSON (UNITED KINGDOM)Abstract :
1) Introduction
Accurate characterisation of fibre-reinforced composite bending behaviour is required in the pursuit of improved process modelling and control.
While bending characterisation is easily achievable for dry fabrics, the thermoviscoelastic behaviour of laminates with a thermoplastic matrix increases the variables involved with laminate bending. The majority of previous literature uses the Kawabata test for fibre-reinforced thermoplastic (FRTP) bending characterisation [1-4]. While this test yields accurate results, it requires a relatively complex experimental setup that is not accessible to the wider research community. The cantilever bending test provides a simple alternative that can be undertaken with relative ease and inexpensive equipment. This work therefore aimed to characterise the thermoviscoelastic bending behaviour of Carbon Fibre (CF) reinforced PA6 laminates via the cantilever test. The results were then validated in an ABAQUS numerical simulation routine.
2) Methodology
Fig. 1a illustrates the cantilever test devised for this study, in which an overhead infrared heater (equipped with a pyrometer for closed-loop thermal control) acted alongside a thermal contact pad such to heat the laminate up to a specified temperature. Once an equilibrium temperature was obtained, a support was removed, allowing the laminate to bend under the influence of its own weight. The specimen was tracked visually to monitor its profile over time.
Single-ply CF/PA6 sheets, 0.23 mm thick, were obtained for the purposes of the study from BÜFA Thermoplastic Composites GmbH & Co KG.
Post-processing was completed by fitting a polynomial to each laminate using a Levenberg–Marquardt algorithm. This was followed by the calculation of curvature and bending moment down the length of the specimen. This was repeated for different laminate temperatures and different bending times to obtain the thermoviscoelastic response.
3) Results
Fig. 1b illustrates the characterised bending profiles for laminates at two superimposed times and four tested temperatures.
Fig. 2a shows an example of the post-processing step: a bending moment versus curvature graph that can be implemented in a numerical simulation routine. Note: this figure is only for results in a static condition (after 5 mins).
Finally, after implementing the post-processed results in an ABAQUS numerical simulation, the bending behaviour was validated by comparing the bending profiles of the experimental and simulated tests (Fig. 2b).
4) Conclusion
In conclusion, this work illustrates that accurate FRTP bending characterisation can be conducted using the cantilever bend test, as opposed to costly alternatives. Due to the thermoviscoelastic matrix, increasing the temperature and decreasing the bending rate both lead to a reduced bending stiffness. The output for FRTP cantilever tests is valid for use within simulation routines due to the strong correlation between experimental and simulated results.
Accurate characterisation of fibre-reinforced composite bending behaviour is required in the pursuit of improved process modelling and control.
While bending characterisation is easily achievable for dry fabrics, the thermoviscoelastic behaviour of laminates with a thermoplastic matrix increases the variables involved with laminate bending. The majority of previous literature uses the Kawabata test for fibre-reinforced thermoplastic (FRTP) bending characterisation [1-4]. While this test yields accurate results, it requires a relatively complex experimental setup that is not accessible to the wider research community. The cantilever bending test provides a simple alternative that can be undertaken with relative ease and inexpensive equipment. This work therefore aimed to characterise the thermoviscoelastic bending behaviour of Carbon Fibre (CF) reinforced PA6 laminates via the cantilever test. The results were then validated in an ABAQUS numerical simulation routine.
2) Methodology
Fig. 1a illustrates the cantilever test devised for this study, in which an overhead infrared heater (equipped with a pyrometer for closed-loop thermal control) acted alongside a thermal contact pad such to heat the laminate up to a specified temperature. Once an equilibrium temperature was obtained, a support was removed, allowing the laminate to bend under the influence of its own weight. The specimen was tracked visually to monitor its profile over time.
Single-ply CF/PA6 sheets, 0.23 mm thick, were obtained for the purposes of the study from BÜFA Thermoplastic Composites GmbH & Co KG.
Post-processing was completed by fitting a polynomial to each laminate using a Levenberg–Marquardt algorithm. This was followed by the calculation of curvature and bending moment down the length of the specimen. This was repeated for different laminate temperatures and different bending times to obtain the thermoviscoelastic response.
3) Results
Fig. 1b illustrates the characterised bending profiles for laminates at two superimposed times and four tested temperatures.
Fig. 2a shows an example of the post-processing step: a bending moment versus curvature graph that can be implemented in a numerical simulation routine. Note: this figure is only for results in a static condition (after 5 mins).
Finally, after implementing the post-processed results in an ABAQUS numerical simulation, the bending behaviour was validated by comparing the bending profiles of the experimental and simulated tests (Fig. 2b).
4) Conclusion
In conclusion, this work illustrates that accurate FRTP bending characterisation can be conducted using the cantilever bend test, as opposed to costly alternatives. Due to the thermoviscoelastic matrix, increasing the temperature and decreasing the bending rate both lead to a reduced bending stiffness. The output for FRTP cantilever tests is valid for use within simulation routines due to the strong correlation between experimental and simulated results.