- STL_bend: Computing Roughness Properties of Additive Manufacturing Surfaces and Creating Cylindrical Pipes for Roughness-Resolved CFD Meshing
During my stay at Lund University, I collaborated with Siemens Energy AB and other researchers on roughness-resolved, high-fidelity Large-Eddy Simulations (LES) of air flow over rough surfaces produced by Additive Manufacturing (AM) for turbine cooling applications. Siemens Energy AB provided us with Inconel 939 samples produced with AM, which creates surfaces with random roughness, impacting… Read more: STL_bend: Computing Roughness Properties of Additive Manufacturing Surfaces and Creating Cylindrical Pipes for Roughness-Resolved CFD Meshing - Neural Style Transfer with TensorFlow and Convolutional Neural Networks
Neural Style Transfer (NST) takes the semantic structure of one image (the content) and renders it in the texture and visual style of another. It is based on the pre-trained VGG-19 Convolutional Neural Network (CNN). This implementation is inspired by the official TensorFlow tutorial Style Transfer, and extends it with stronger diagnostics (effective learning rate,… Read more: Neural Style Transfer with TensorFlow and Convolutional Neural Networks