Wuyang Zhao

Glassy polymers can transition from ductile to brittle under certain conditions such as aging or decreased temperatures, leading to sudden breakage with minimal energy absorption and potential accidents. Enhancing the toughness of glassy polymers without compromising stiffness is a significant scientific challenge, yet the mechanisms behind the brittle-to-ductile transition (BDT) remain not fully understood. In crystalline materials, the BDT is often attributed to the kinetics of dislocations, but this explanation cannot be directly applied to glassy materials due to the absence of well-defined microscopic structures of such plastic carriers like dislocations. Molecular dynamics (MD) simulations have shown that spatial fluctuations of local mechanical properties at the atomistic scale and geometric loading conditions are crucial in the BDT of glassy materials. However, addressing the effects of geometric loading conditions under non-uniform deformations is challenging in pure MD simulations due to computational constraints on system sizes. To overcome this limitation, this project employs a multiscale simulation method by embedding an MD domain into a continuum domain to conduct nonuniform deformation boundaries for the MD system. This approach enables a better understanding of the interactions between plastic carriers and the relationship between local structures and global mechanical properties in glassy polymers.

Nanocomposites have great potential for various applications since their properties may be tailored to particular needs. One of the most challenging fields of research is the investigation of mechanisms in nanocomposites which improve for instance the fracture toughness even at very low filler contents. Several failure processes may occur like crack pinning, bi-furcation, deflections, and separations. Since the nanofiller size is comparable to the typical dimensions of the monomers of the polymer chains, processes at the level of atoms and molecules have to be considered to model the material behaviour properly. In contrast, a pure particle-based description becomes computationally prohibitive for system sizes relevant in engineering. To overcome this, only e.g. the crack tip shall be resolved to the level of atoms or superatoms in a coarse-graining (CG) approach.

Thus, this project aims to extend the recently developed multiscale Capriccio method to adaptive particle-based regions moving within the continuum. With such a tool at hand, only the vicinity of a crack tip propagating through the material has to be described at CG resolution, whereas the remaining parts may be treated continuously with significantly less computational effort.

In engineering applications, plastics play an important role and offer new possibilities to achieve and to adjust a specific material behaviour. They consist of long-chained polymers and possess, together with additives, an enormous potential for tailored properties.

Recently, techniques have been established to produce and to disperse filler particles with typical dimensions in the range of nanometers. Even for low volume contents of filler particles, these socalled nanofillers may have significant impact on the properties of plastics. This can be most likely traced back to their very large volume-to-surface ratio. In this context, the polymer-particle interphase is of vital importance: as revealed by experiments, certain nanofillers may e.g. increase the fatigue lifetime of plastics by a factor of 15.

The effective design of such nanocomposites quite frequently requires elaborated mechanical testing, which might - if available - be substituted or supplemented by simulations. For this purpose, however, continuum mechanics together with the Finite Element Method (FE) as the usual tool for engineering applications is not well-suited since it is not able to capture processes at the molecular level. Therefore, particle-based techniques such as molecular dynamics (MD) have to be employed. However, these typically allow only for extremely small system sizes and simulation times. Thus, a multiscale technique that couples both approaches is required to enable the simulation of so-called representative volume elements (RVE) under consideration of atomistic effects.

The goal of this 4-year project is the development of a methodology which yields a continuum-based description of the material behaviour of the polymer-particle interphase of nanocomposites, whereby the required constitutive laws are derived from particle-based simulations. Due to their very small dimensions of some nanometers, the interphases cannot be accessed directly by experiments and particle-based simulations must substitute mechanical testing. The recently developed Capriccio method, designed as a simulation tool to couple MD and FE descriptions for amorphous systems, will be employed and refined accordingly in the course of the project.

In the first step, the mechanical properties of the polymer-particle interphase shall be determined by means of inverse parameter identification for small systems with one and two nanoparticles. In the second step, these properties shall be transferred to large RVEs. With this methodology at hand, various properties as e.g. the particles’ size and shape as well as grafting densities shall be mapped from pure particle-based considerations to continuum-based descriptions. Further consideration will then offer prospects to transfer the material description to applications relevant in engineering and eventually suited for the simulation of parts.