Experiments based analysis of thermal decomposition kinetics model. Case of carbon fibers PolyPhenylene Sulfide composites. (April 2021)

• Record Type: Journal Article
• Authors: Carpier, Y.; Alia, A.; Vieille, B.; Barbe, F.
• Is Part Of: Polymer degradation and stability. Volume 186(2021)
• Publication Date: 2021-04
• Subjects: Thermal decomposition kinetics -- Thermoplastic composites -- Pyrolysis kinetics models -- Thermogravimetry -- Fire exposure; Polymers -- Deterioration -- Periodicals; Stabilizing agents -- Periodicals; Polymères -- Dégradation -- Périodiques; Stabilisants -- Périodiques; 668.9
• DOI: 10.1016/j.polymdegradstab.2021.109525
• Languages: English
• ISSNs: 0141-3910

Abstract

Highlights: This study allows a better understanding of the thermal decomposition of composite structures during critical events such as aircraft engine fires. The most physically consistent approach to model the isothermal decomposition kinetics of C/PPS composites was identified. The predictive capabilities of three types of approaches (modelistic, model free/isoconversional model, isoconversional model) have been discussed based on statistical analyses (Fisher's test). These approaches are virtually statistically equivalent. Model-fitting methods result in relatively accurate prediction of the TGA curves at a given temperature. The model-free approach is the most predictive, from a physical point of view, to account for the thermal decomposition of C/PPS composites under isothermal conditions. Abstract: Considering their increasing use in high-temperature aeronautical applications, thermoplastic-based laminates have to meet certification requirements in terms of critical service conditions (e.g. aircraft engine's fire). This work is aimed at identifying the most relevant approach to model the isothermal decomposition of C/PPS composites. The first step in the understanding of the thermal decomposition of these materials relies on TGA and MT-TGA tests. They can be conducted in isothermal and anisothermal conditions to characterize the thermal decomposition kinetics taking place in a fire scenario. The experimental database obtained from these tests is of major importance to identify input parameters of pyrolysis kinetics models. The predictive capabilities of three different approaches (model-fitting, isoconversional/model-fitting and model-free) is discussed based on statistical analyses (Fisher's test). For C/PPS laminated composites, these approaches are virtually statistically equivalent. Model-fitting methods result in relatively accurate predictions of the TGA curves at a given heating rate. However, this type of approaches is not fully satisfactory especially when it is used to predict the thermal decomposition of polymer-matrix composites under anisothermal conditions. Finally, the model-free approach is identified as the most relevant from a physical point of view to account for the thermal decomposition of C/PPS composites under isothermal conditions.