Master assignment
Background
Ultrashort pulsed laser is an excellent tool for material removal with little or no thermal damage that enables plethora of three-dimensional surface textures at nano and micro meter scales. High energy deposition at extremely short time interval makes the laser ablation process a subject of intense debate in terms of governing physical phenomena. Additionally, the vast parameter space of laser processing makes it a time and effort demanding step to identify the optimum processing conditions through trial-and-error basis. With the advancement of machine learning technology, the potential of extracting hidden features from large datasets, or the proficiency of rigorous interpolation across multidimensional space with limited data points provides an alternative approach to planning and optimization of laser ablation process for different materials. Such an approach can enable us not only to gain insight into the physical phenomena occurring at small length and time scale, but also to bolster surface engineering and to explore advanced materials.
Assignment
In this assignment, neural-network based approaches will be investigated in order to achieve realistic, predictive three-dimensional visualization of laser ablated surface topology on polished Zinc surfaces after single and multi-pulse irradiation on same location with arbitrary laser pulse energies.
Tasks (preliminary)
- Literature review ,
- Simulate complex system directly from a small experimental dataset,
- Provide methodology for identification of the optimum laser processing parameters,
- Writing MSc thesis,
- Co-authoring a paper (if relevant)
Reporting is an ongoing task in this MSc-project
Start & duration
Start: As soon as possible
Duration: About 9 months (40 EC)
Contact persons
