TY - THES ID - 135671329 TI - Predicting the Structure and Thermal Conductivity of Silica Aerogels by Molecular Dynamics Simulations AU - Dierckx, Vincent AU - Seveno, David AU - KU Leuven. Faculteit Ingenieurswetenschappen. Opleiding Master in de ingenieurswetenschappen. Materiaalkunde (Leuven) PY - 2020 PB - Leuven KU Leuven. Faculteit Ingenieurswetenschappen DB - UniCat UR - https://www.unicat.be/uniCat?func=search&query=sysid:135671329 AB - Silica aerogels are nanoporous glasses that exhibit extremely low thermal conductivities due to their intricate three-dimensional porous network and low densities. They are promising candidates for isolating applications. However, their high prices, brittleness and strongly deviating thermal properties in commercially available aerogels hold back their breakthrough in mass-market. The last problem is strongly linked with their nanostructure. In this study, Molecular Dynamics simulations were used to generate silica aerogels, structurally characterize them and compute their thermal conductivity coefficients. Aerogels were created with densities in the range of 0.15−0.29 g/cm^3. These values are lower than ever thermally tested before using Molecular Dynamics and are close to optimal performing experimental aerogels. The Wolf-BKS force field was used due to its ability to generate realistic aerogel structures and its computational efficiency compared with other force fields. Starting from a -cristobalite crystal of 1.536 million atoms, a custom negative pressure rupturing technique with two-stage intermediate relaxation was implemented to reach low densities resulting in meta-stable aerogels with system sizes ranging from 55.5 nm to 67.6 nm. Besides pore size distribution, pore surface area density, degree of anisotropy, fractal dimension number, a novel structural characteristic was computed: the skeletal (silica backbone) thickness distribution. For this, voxilized image slices of the structure were generated and analyzed with CT-scan software; CTAn. The average skeletal size and fractal dimensions were directly proportional to the aerogels density and ranged from 3.0 nm to 4.1 nm and 2.025 to 2.125 respectively. The average pore size was found to be inversely proportional with the density and ranged from 12.0 nm to 14.0 nm. Compared to optimal performing experimental aerogels, the pore sizes and skeletal thicknesses of the modelled structures were slightly higher. Reverse Non-Equilibrium Molecular Dynamics simulations were performed to compute thermal conductivities. Due to the large system size, the heat transport through the system takes millions of timesteps to reach steady state. For this reason, the simulated temperature profiles during transient state were compared with numerically calculated (solving the first order heat transfer differential equation) temperature profiles in order to compute the thermal conductivity coefficients. This resulted in a power law of λ = 0.20ρ^1.27 with λ the thermal conductivity coefficients ranging from 0.0177 − 0.0427 W/mK and ρ the densities of the aerogel structures. ER -