John Puentes,1 Alexander Chaloupka,2 Natalie Rudolph,1 Tim A. Osswald1
1Polymer Engineering Center, Department of Mechanical Engineering, University of Wisconsin-Madison Madison, Wisconsin 53706-1691
2NETZSCH-Ger€atebau GmbH, Selb/Bayern D-95100, Germany
Correspondence to: J. Puentes (E - mail: firstname.lastname@example.org)
Imagine you are a composite manufacturer of thermoset CFRP and you have to set up your process: How do you choose the optimal curing profile for your part? Do you take environmental changes influencing the material behavior into account, e.g., seasonal changes from summer to winter? What do you measure in order to check for a constant reactivity of your resin?
The solution to that is a so-called TTT diagram for your resin system that shows the phase transitions including curing and vitrification as a function of time and temperature. If you have this diagram available, you have the full information to predict the resin behavior during cure under process conditions. In order to get this diagram, you need DSC measurements and kinetic simulations to fit the DSC data with an appropriate reaction model. The paper shows the procedure to create such a TTT diagram.
However, the best kinetic simulation model for the TTT diagram can only be derived from DSC measurements exactly representing the temperature profile of your process. Here, the authors used the NETZSCH DSC 214 Polyma for the curing scans: It is the fastest regular heat-flux DSC available, therefore one can quickly heat up and investigate the entire curing reaction at isothermal conditions. This, in turn, is exactly what happens during the curing process in manufacturing: The resin heats up quickly as it enters the composite mold and is kept at a constant temperature for a while. So the take-home message of the paper by Puentes et al. is that you only get the best model of your process and the correct TTT diagram, if you measure the data in the right way.
The characterization of film adhesives is challenging because they required freezer storage, contain an inseparable filler— thermoplastic knit or fiber-reinforcement, and are heat activated systems with a pre-cure and unknown chemistry. A testing protocol that eliminates these sources of error is proposed. This study presents a method to generate time–temperature-transformation (TTT) diagrams of epoxy film adhesives via differential scanning calorimetry (DSC). Non-isothermal and isothermal DSC scans are used to capture the reaction and the glass transition temperature. The use of an initial fast ramp—up to 500 K/min—in the isothermal scans is explored for the first time. This technique shows the potential to produce a quasi-isothermal cycle, eliminating the loss of data in the initial stage of the reaction. The total heat released, the activation energy, and the fractional kinetic parameter, are estimated via model-free methods. The Kamal–Sourour model and the formal kinetic model are fit to model the rate of cure. The simplest model that accurately captures the reaction, a parallel two-step model, A ⇒ B, is outlined. The glass transition temperature is modeled via DiBenedetto’s equation to include the diffusion-controlled mechanism. The TTT-diagrams of two commercial adhesives, DA 408 and DA 409, are shown with an analysis of processing optimization. The use of quasi-isothermal scans with initial fast ramps combined with the correction for filler, moisture, and pre-curing history can be applied to characterize fast curing thermosets, complex B-stage resins, and thermosetting composites. The modeling results can also be used in numerical studies of residual stresses and dimensional stability in the manufacturing of thermosetting composites. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2017, 135, 45791.