Multiscale Process Modeling of Semicrystalline PEEK for Tailored Thermomechanical Properties

Abstract

Polyether ether ketone (PEEK) is a semicrystalline thermoplastic that is used in high-performance composites for a wide range of applications. Because the crystalline phase has a higher mass density than that of the amorphous phase, the evolution of the crystalline phase during high-temperature annealing processing steps results in the formation of residual stresses and laminate deformations, which can adversely affect the composite laminate performance. Multiscale process modeling, utilizing molecular dynamics, micromechanics, and phenomenological PEEK crystal kinetic laws, is used to predict the evolution of volumetric shrinkage, elastic properties, and thermal properties, as a function of crystalline phase evolution, and thus annealing time, in the 306−328 °C temperature range. The results indicate that lower annealing temperatures in this range result in a faster evolution of thermomechanical properties and shrinkage toward the pure crystalline values. Therefore, from the perspective of composite processing, it may be more advantageous to choose the higher annealing rates in this range to slow the volumetric shrinkage and allow PEEK stress relaxation mechanisms more time to relax internal residual stresses in PEEK composite laminates and structures.

Optimal Molecular Dynamics System Size for Increased Precision and Efficiency for Epoxy Materials

“…The Interface Force Field (IFF) was implemented to describe atomic interactions. This force field has been previously shown to accurately predict physical, mechanical, and thermal properties of polymers. ,,,− For consistency in comparing the simulation times, all simulations were run in parallel with 16 Intel Xeon E5-2683 2.10 GHz processors on a single thread per processor, with the same LAMMPS compiler options. , The MD run times (on the nanosecond time scale) for model equilibration and property predictions were preselected and equal for each model replicate. The corresponding simulation times (on the order of thousands of seconds) correspond to the time it took for the processors to complete the MD runs.…”

“…This force field has been previously shown to accurately predict physical, mechanical, and thermal properties of polymers. 4 , 16 , 17 , 23 − 26 For consistency in comparing the simulation times, all simulations were run in parallel with 16 Intel Xeon E5-2683 2.10 GHz processors on a single thread per processor, with the same LAMMPS compiler options. 5 , 27 The MD run times (on the nanosecond time scale) for model equilibration and property predictions were preselected and equal for each model replicate.…”

“…Accurate MD modeling methods are particularly valuable for understanding the material response in situations where experiments are expensive or challenging, including studying the behavior of nanoscale samples, examining polymers under extreme temperature and pressure conditions, and optimizing processing parameters to yield composites with minimal residual stresses. When combined with multiscale continuum models and experimental data in a synergistic manner, MD models can effectively reduce the cost and time required for the design and implementation of new polymer–matrix composites for specific engineering applications. − Perhaps the largest issue with MD modeling is the computational rigor required to simulate realistic material systems, which limits the size and time scales to nanometers and nanoseconds, respectively.…”

LUNAR: Automated Input Generation and Analysis for Reactive LAMMPS Simulations

“…PCFF/PCFF-IFF uses the Lennard-Jones 9–6 potential to describe the van der Waals interactions, where the 9–6 Lennard-Jones potential is defined as E LJ = ϵ [ 2 true( σ r true) 9 − 3 true( σ r true) 6 ] where r is the distance between interacting particles, ϵ is the depth of the potential well, and σ is the size of the atom. PCFF/PCFF-IFF has been used in numerous other polymer studies ,− ,, for systems built using msi2lmp, with missing parameters assigned manually using autoequivalences. For certain polymer systems, it has been shown that the van der Waals sigma parameter for the hydrogen atom in a hydrogen–carbon configuration (i.e., the “hc” atom type) can be adjusted from 2.995 to 2.3 Å to correctly reproduce experimental densities.…”

“…Molecular dynamics (MD) simulation is a powerful tool for modeling polymeric and composite materials to establish trends in their thermomechanical properties, chemical composition and morphology, and the effects of processing conditions. − Improving the throughput, accuracy, and reproducibility of atomistic MD simulations of polymer and ceramic matrix composites, such as high-temperature carbon–carbon composites, will accelerate the development of the next generation of materials in high-performance applications. − For example, MD simulations can be integrated with multiscale process modeling − to optimize processing conditions and tune the properties of composite materials for specific applications. Establishing quantitative relationships between processing conditions and material properties is a primary objective of Integrated Computational Materials Engineering (ICME) and the Materials Genome Initiative (MGI) for novel materials design and discovery.…”