| dc.description.abstracteng | Polymeric materials are ubiquitous in our modern lives.
Their many applications in complex materials are accompanied by potentially huge benefits for technological advancement.
These applications range from batteries, fuel cells, molecular sieves, tires, and microelectronic devices.
The ability to self-assemble into nanostructures in combination with their viscoelastic properties make polymers attractive for this wide range of applications.
I perform computer simulations gaining knowledge about their properties for applications and manufacturing, to improve the understanding of these materials.
The simulation of multicomponent polymer melts poses an extreme computational challenge.
The large spatial extent of defects in self-assembled structures or nonperiodic metastable phases, which are prone to finite size effects, require the study of large system sizes.
Hence, I use a soft, coarse-grained polymer model reducing the degrees of freedom to gain insights into long time and length scales.
Consistent implementations of these models that scale well on modern GPUs accelerated HPCs hardware enable investigations with up to billions of particles. Consequently, I can address challenges that were deemed intractable before.
Firstly, I analyze metastable network phases as a function of the volume fraction, f, of diblock copolymers for polymeric battery electrolytes.
One polymer block provides the mechanical stability while the other is ion conducting.
The focus lies on the structure of the conducting phase.
Due to the trapped metastable states, I investigate systems of extreme sizes with billions of particles circumventing finite size effects.
In fact, I identify fractal structures on significant length scales inside the network phase, which influence the transport properties locally.
As such, this work highlights the necessity of soft models and scaling implementations obtaining insights on engineering scales.
Secondly, I will investigate the simulation of viscoelastic properties of polymeric materials with soft, coarse-grained models.
It is particularly challenging to correctly capture the entangled dynamics.
The noncrossability of polymer backbones introduces topological constraints on the motion of the chains.
A soft, coarse-grained model does not capture this noncrossability automatically.
Hence, I utilize a SLSP model to mimic the entanglements via dynamic bonds.
With this model and a novel technique to average the stress auto-correlation function G(t), I perform a dynamic mechanical analysis of polymer melts and a cross-linked network.
The obtained storage modulus G'(w) and loss modulus G''(w) meet the expectations for a comparison with experimental studies.
A nonequilibrium study of diblock copolymers in shear flow completes this work.
Shear flow is a powerful method to macroscopically order a metastable microstructure.
In a symmetric diblock copolymer melt, the equilibrium microstructure is a lamellar phase.
The first step determines the perpendicular orientation of the lamellae in shear flow as stable at all stresses according to the concept of the Rayleighian, R.
Further, I study the transition between a grain in the unstable orientation next to a grain in the stable orientation.
I identify two different transition pathways.
At low applied stresses, the grain boundary of the stable grain grows into the unstable grain.
At higher stresses, the unstable orientation is destabilized and forms an intermediate microemulsion-like phase with no local orientation.
This intermediate phase turns subsequently into the stable orientation. Oscillatory shear at high frequencies delays the onset of this microemulsion pathway.
In a collaboration with Matthias Heck and Manfred Wilhelm at KIT, these transitions have been studied in LAOS experiments as well. | de |