University of Crete (UOC)
Development of a multi-scale physics model and incorporation of predictive modeling techniques to describe laser-induced processes in irradiated metals
Abstract
dc:descriptionFemtosecond laser-matter interaction involves a sequence of coupled physical phenomena ranging from electromagnetic, thermal, to hydrodynamic regimes, each operating on distinct timescales while collectively determining the final material response. Understanding and predicting these processes is of fundamental importance for laser-based surface engineering, since they enable precise surface modification and the formation of micro- and nanoscale structures. This includes the formation of Laser-Induced Periodic Surface Structures (LIPSS), textures that enable a broad spectrum of functional surfaces with significant technological importance.While precise surface patterning necessitates the modulation of a diverse array of laser and material variables, the high dimensionality of the parameter space makes traditional trial-and-error techniques involving experiments and simulation both time-consuming and costly. This thesis addresses this challenge by developing a systematic predictive modeling framework. The research follows a hierarchical progression: first, gaining a fundamental understanding of the underlying physics; second, translating that knowledge into a robust multi-scale numerical model and utilizing this model as an engine to generate extensive datasets; and finally, employing these data sets to enable predictive machine learning.By analyzing a two-layered geometry, this study explores how energy dissipation is governed by the competition between two mechanisms, namely electron diffusion and electron-phonon coupling. The findings reveal a non-monotonic relationship between the laser pulse duration and the damage threshold, with a distinct minimum occurring near the electron-phonon relaxation time. This investigation demonstrates that electron confinement in thin films significantly alters relaxation dynamics compared to bulk targets, providing a validated physical criterion for the onset of material melting.Building upon these fundamental insights, the second stage introduces a comprehensive multi-scale physics framework to simulate the progressive evolution of surface morphology over multiple laser pulses. To transition from simple melting to the formation of self-organized structures, the model integrates electromagnetic wave scattering, non-equilibrium thermal dynamics, and hydrodynamic melt flow. The results illustrate how nanoscale surface irregularities act as seeds that excite surface electromagnetic waves, leading to spatially modulated energy deposition. This inhomogeneous heating drives the redistribution of molten material via surface tension and pressure forces. This unified model successfully captures the positive feedback loop between evolving topography and field enhancement, reproducing the morphological changes and periodicity shifts observed in experimental settings. In the final stage, to demonstrate the scalability of the framework and overcome the computational limitations of full-scale simulations, classification algorithms are trained on a combined database of experimental and simulated results. This data-driven methodology enables the rapid prediction of surface outcomes --such as ripples, grooves, or spikes-- directly from processing parameters. Furthermore, by employing other more complex metrics, the model identifies regions of high parametric uncertainty, providing guidelines for more informed and efficient future experimental designs. Collectively, this thesis demonstrates that the integration of fundamental physics with data-driven modeling constitutes a powerful strategy for the understanding, prediction, and optimization of laser-based surface engineering.
Degree
thesis:*- Grantor dc:publisher
- University of Crete (UOC)
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Velli, Maria-Christina
- Βέλλη, Μαρία-Χριστίνα
Subjects
dc:subject × 24- Λέιζερ φεμτοδευτερολέπτου
- Περιοδικές επιφανειακές δομές επαγόμενες από λέιζερ (LIPSS)
- Πολυφυσική μοντελοποίηση
- Κατώφλι καταστροφής
- Μηχανική μάθηση
- Λεπτά υμένια νικελίου
- Πολαριτόνια επιφανειακών πλασμονίων
- Αλληλεπίδραση λέιζερ-ύλης
- Υπολογιστική μοντελοποίηση
- Femtosecond lasers
- Laser-induced periodic surface structures (LIPSS)
- Multiphysics modeling
- Damage threshold
- Machine learning
- Nickel thin films
- Surface plasmon polaritons
- Laser-matter interaction
- Computational modeling
- Φυσική
- Φυσικές Επιστήμες
- Φυσική, διεπιστημονική προσέγγιση
- Physical Sciences
- Natural Sciences
- Physics, interdisciplinary
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/61674