Effect of temperature, physical aging and curing cycle on viscoelastic behavior of epoxy resin
Topic(s) : Material and Structural Behavior - Simulation & Testing
Co-authors :
Nunes STEPHANIE , Liva PUPURE (LATVIA), Janis VARNA (SWEDEN)Abstract :
Temperature and loading rate effects in glass or carbon fiber composites are caused mainly by the epoxy resin's viscoelastic (VE) and viscoplastic behavior. This is true for shear- and transverse loading in long fiber composites. For short fiber composites, it is correct for all loading cases.
The objective of the presented paper is to present the results of the experimental study and development of constitutive models that describe the stress-strain response of the epoxy at different material states (degree of cure, degree of physical aging etc.) at different temperatures.
DMTA has been considered a convenient tool for determining a shift factor and constructing master curves by performing frequency scans at different temperatures for linear VE. This study showed that it is in fact not reliable for predicting the behavior of macro-specimens, thus confirming the previously expressed doubts in other publications [1]. We used macro-specimens for master curve construction from a strain-controlled test consisting of 1 min uploading and 2h holding (L+H test) and using the recorded stress data to find coefficients in the Prony series. These coefficients are used to simulate an idealized relaxation test for the given temperature that is used for master curve construction. It is shown that fully cured specimens are thermo-rheologically simple (the temperature enters only the shift factor in the VE behavior).
The physical aging of the epoxy exposed to temperature below Tg was studied. It is shown that the epoxy behavior is thermo-aging-rheologically simple, and the shift factor can be expressed as a sum of two horizontal shifts in logt axis. Physical aging makes the epoxy stiffer, and the VE effects are smaller. As shown in Figure below 1000 min aging time, the aging shift factor is a linear function of aging time - the slope is the same for all temperatures used. After 1000 min, the aging shift factor increases faster for aging at higher temperatures.
Then, a cure cycle's effect on the epoxy's final properties was investigated using the knowledge of the phenomenon of physical aging with data-based fitting expression for the shift factor. The cure temperature and curing time were used as parameters, and a full cure was not always reached. The full cure was always completed by several hours of post-curing. The VE for fully cured epoxy with different cure cycles was tested at several temperatures to construct master curves and shift factors. The post-curing smoothens the differences between cases and the master curves and shift factors of fully cured specimens with different curing cycles are similar. The exception was when curing was done at 70°C for 20 min and the degree of cure before post-curing was the lowest among considered cases. The low sensitivity of the final behavior regarding the cure cycle can be used to minimize the cure time in the mold and perform the out-of-mold post-curing.
The objective of the presented paper is to present the results of the experimental study and development of constitutive models that describe the stress-strain response of the epoxy at different material states (degree of cure, degree of physical aging etc.) at different temperatures.
DMTA has been considered a convenient tool for determining a shift factor and constructing master curves by performing frequency scans at different temperatures for linear VE. This study showed that it is in fact not reliable for predicting the behavior of macro-specimens, thus confirming the previously expressed doubts in other publications [1]. We used macro-specimens for master curve construction from a strain-controlled test consisting of 1 min uploading and 2h holding (L+H test) and using the recorded stress data to find coefficients in the Prony series. These coefficients are used to simulate an idealized relaxation test for the given temperature that is used for master curve construction. It is shown that fully cured specimens are thermo-rheologically simple (the temperature enters only the shift factor in the VE behavior).
The physical aging of the epoxy exposed to temperature below Tg was studied. It is shown that the epoxy behavior is thermo-aging-rheologically simple, and the shift factor can be expressed as a sum of two horizontal shifts in logt axis. Physical aging makes the epoxy stiffer, and the VE effects are smaller. As shown in Figure below 1000 min aging time, the aging shift factor is a linear function of aging time - the slope is the same for all temperatures used. After 1000 min, the aging shift factor increases faster for aging at higher temperatures.
Then, a cure cycle's effect on the epoxy's final properties was investigated using the knowledge of the phenomenon of physical aging with data-based fitting expression for the shift factor. The cure temperature and curing time were used as parameters, and a full cure was not always reached. The full cure was always completed by several hours of post-curing. The VE for fully cured epoxy with different cure cycles was tested at several temperatures to construct master curves and shift factors. The post-curing smoothens the differences between cases and the master curves and shift factors of fully cured specimens with different curing cycles are similar. The exception was when curing was done at 70°C for 20 min and the degree of cure before post-curing was the lowest among considered cases. The low sensitivity of the final behavior regarding the cure cycle can be used to minimize the cure time in the mold and perform the out-of-mold post-curing.