I am working as a research assistant in the CRC/TRR 129 Oxyflame in project A7 the atomistic multiscale simulation of char combustion.
Our goal is the development of a computational char combustion model with parameters that are directly related to verifiable physico-chemical processes and are predictable via simulations. With this we can improve the parameters of the simulations of the other projects within the CRC and as synergetic effect generate better predictions.
We have two connected but parallel lines of research: One is building a char model and generate data on for example adsorption and diffusion. The other one is the investigation of the combustion process to determine reaction rates. My focus is the char model, I have to stress that to the best of our knowledge there is NO method present in the literature to construct char models with input parameters such as mass percentages of elements and porosity. But for our purpose it is essential to be able to construct different kinds of char models in an controlled and easy fashion having these input parameters.
As starting point for this goal I devised a python code that starting from a random plane as input builds a char model. The code adds planes in the carbon-carbon interlayer distance, places carbon atoms onto these, optimizes the positions with a LAMMPS code and adds hydrogen and oxygen atoms to it. In a proof of principle study we have already compared adsorption values of different gases from our simulations and our calculations and experiments finding the same trends for the gases.
In the next year I want to generate a char structure construction tool which only requires minimal input as for example HRTEM images and chemical information like mass percentages of elements and porosity. That means we will be able to generate models of all kinds of char by just pressing a button. For these models we can determine adsorption and diffusion behaviour and use these to simulate the char burnout.