Femtosecond laser micro-nano structuring and prediction of aluminum bronze QAL9-4 driven by transformer

To overcome multi scale morphology manipulation challenges in aluminum bronze QAL9-4 (Al − 8 %, Fe − 2 %, Cu − balance) during femtosecond laser processing, we developed an architecture optimized Transformer model. Systematic multi pulse experiments revealed energy accumulation induced ablation threshold attenuation and quantitatively established the governing influence of laser fluence and scanning speed on surface topography modulation. In contrast to conventional sequence models, the proposed approach demonstrates markedly enhanced prediction accuracy for machined groove depth and width dimensions through decoder module elimination and encoder optimization. This work quantitatively confirms the predominant influence of laser fluence on feature width (ρ = 0.73) and uncovers nonlinear parameter interdependencies that inform process optimization strategies for QAL9-4 aluminum bronze in laser micro-nano fabrication. The developed process parameter geometry mapping framework enables controllable surface texturing for aerospace and microelectronics applications while expanding Transformer based multi physics modeling of laser material interactions.