Popular materials for lightweight applications, like helicopters, aircrafts and automobiles, are glass and carbon fiber-reinforced plastics. Traditionally, reactive resins, like epoxy, unsaturated polyester and polyurethane, are used for impregnation. The important cross-linked network is achieved by a chemical reaction. “During crosslinking at sufficiently high temperatures, the material changes from a liquid via a gel into a glass-like solid” . Therefore, the properties of the composite material are determined by the manufacturing process conditions and not only the properties of the basic components.
Thus, in technical processes and to predefine optimal manufacturing conditions, it is crucial to know the curing state reached during manufacturing as well as the correlation between the glass transition temperature (Tg) and degree of cure. Especially, knowledge about full cure (Tg∞) is important as the manufacturing temperature needs to approach or exceed Tg∞ to complete the reaction within a reasonable cure time. Otherwise, vitrification prevents or delays full curing.
The scientific paper “Cure state detection for pre-cured carbon-fibre reinforced epoxy prepreg (CFC) using Temperature-Modulated Differential Scanning Calorimetry (TMDSC)” by W. Stark, M. Jaunich and J. McHugh was published in the Journal Polymer Testing. It aims to “determine the correlation between the actual glass transition temperature, degree of cure and curing time at 180°C for carbon-fiber (CFR) prepreg […] using the TMDSC method” .
What is Temperature-Modulated Differential Scanning Calorimetry (TM-DSC)?
Traditional Differential Scanning Calorimetry (DSC) is used to investigate the cure state of pre-cured samples for different lengths of time in non-isothermal experiments. This way, it is possible to determine the correlation between Tg and degree of cure in only one measurement. “These experiments work well when the reaction temperature is higher than the maximal glass transition temperature. […] The situation is more complex when the actual glass transition temperature is in the same temperature range as the post-curing reaction. The term actual glass transition temperature (Tgact) will be used for the value achieved by partial curing, which is situated between Tg0 of the neat resin and Tg∞. In many cases, vitrification occurs during partial curing, as the cure temperature is lower than Tg∞” .
Temperature-modulated DSC allows for the separation of the glass transition and cross-linking reaction phenomena. The sample is subjected not only to a linear heating rate, but also to sinusoidal temperature variations. This method leads to the separation of the so-called reversing and non-reversing part of the heat flow. The reversing effects are, for example, the glass transition as well as melting and crystallization. The change in specific heat at the glass transition becomes apparent. The non-reversing processes are a function of time and cannot be repeated like curing and tempering effects. They are calculated as the difference between the total heat flow and the reversing heat flow. From this, the exothermic curing reaction is deductible.
High-level information from conventional DSC measurement
In order to gain first information on a higher level, the non-cured prepreg material was analyzed with a standard DSC measurement at heating rates of 2, 10 and 20 K/min. “The higher the heating rate, the more pronounced the step in the heat flow at Tg0. This is the reason that a high heating rate of 20 K/min is recommended for the glass transition detection using DSC” . The start of the exothermic cross-linking reaction was detected from approximately 140°C. In addition, two clear exothermic peaks were observed indication two-step or multiple-step reaction. Tgact was not recognizable in the curves.
Using TM-DSC on non-cured carbon-fiber prepreg
Based on previously published results, the parameter of modulation period was selected and was at 60 s. The highest possible heating rate is advantageous for determining Tg. Therefore, 10 K/min was selected as the highest possible underlying heating rate.
Figure 1 displays a typical behavior of a temperature-modulated DSC measurement. The heat flow shows the effect of overlaid modulation. Figure 2 displays the reversing and non-reversing signal as well as the total signal. It is observable that Tg0 from the reversing and the total signal are in good agreement. As expected, this shows that the use of this advanced method has no particular advantage for this material. Only when measuring partially cured samples where glass transition and reaction temperatures are close together, is the temperature modulation method needed to observe these effects.
TM-DSC measurement of pre-cured samples and determination of vitrification
Therefore, further analyses with samples cured at 180°C for 30 minutes were carried out. Different temperature modulations were applied, while the other measurement parameters stayed the same.
At the end of each measurement, a discrepancy in the reversing signal can be observed, which was further analyzed. The authors of the paper found that “at the end of the reaction the change in heat flow is too fast for the modulation period. Therefore, the symmetric modulation is disturbed” .
The results show that the start temperature of the remaining reaction significantly increases with pre-curing. Only in the reversing signal generated by TMDSC, the glass transition temperature Tgact is clearly detectable. A close correlation between the temperature at reaction start and Tgact was observed, which could indicate vitrification. To verify this, the degree of cure was calculated using the reaction enthalpy in the post-reaction:
Where α is the degree of cure (0 to 1), ΔHr is the residual heat and ΔHt is the total heat.
The authors found a degree of cure of approximately 72%.
Correlation between the degree of cure and curing time
In order to determine the relation between the degree of cure and curing time, pre-cured samples were measured between 10 min and 5 h simulating curing times in the temperature-modulated DSC (other parameters were kept constant: underlying heating rate: 10 K/min, modulation amplitude: 1.6 K, modulation period: 60 s).
“With increasing reaction time, the actual glass transition temperature increases. Also, the starting temperature of the post-curing reaction increases and the amount of heat released is reduced” .
After calculating the degree of cure, the analyses show that “the main part of the reaction proceeds during the first 60 min” . After that, the degree of curing and Tgact grow nearly linear.
Finding the correlation between curing conditions with TM-DSC
The scientific research by W. Stark et al. highlights that temperature-modulated DSC (TM-DSC) analysis allows for the detection of cure state of pre-cured carbon-fiber epoxy prepreg (CFC). The thermoanalytical method was used to find correlations between curing conditions, degree of cure and glass transition temperature as TMDSC “enables better determination of the glass transition temperature, which is often accompanied by an exothermic curing reaction and thus overshadowed”  in standard DSC measurements.
Knowledge about the glass transition temperature as a function of degree of cure is vital in predefining optimal manufacturing conditions and avoid vitrification.
 Stark, W., Jaunich, M. , McHugh, J. (2013): Cure state detection for pre-cured carbon-fibre reinforced epoxy prepreg (CFC) using Temperature-Modulated Differential Scanning Calorimetry (TMDSC), Polymer Testing, http://dx.doi.org/10.1016/j.polymertesting.2013.07.007