Prediction of ultrashort pulse laser ablation processing using machine learning

Toshiyuki Kusumoto; Kiyokazu Mori

doi:10.1117/12.2577447

5 March 2021 Prediction of ultrashort pulse laser ablation processing using machine learning

Toshiyuki Kusumoto, Kiyokazu Mori

Proceedings Volume 11673, Laser Applications in Microelectronic and Optoelectronic Manufacturing (LAMOM) XXVI; 1167303 (2021) https://doi.org/10.1117/12.2577447
Event: SPIE LASE, 2021, Online Only

Abstract

Several conditions—including optics, work material and the environment—can influence the results of laser processing. The physical model of laser-processing phenomena is represented as a complex function; thus, the development of a prediction model using machine learning (ML) may be an effective approach. In this study, the quantity of ablation under femtosecond ultrashort pulse laser processing was predicted using an ML model. For work materials, polycrystalline diamond was used as a composite material, cold rolled steel and aluminum were used as metals, silicon was used as a semiconductor and glass was used as an insulator. A total of 340 datasets were prepared for each material. The neuralnetwork algorithm was used to develop the prediction model. We also explored the challenges from the viewpoint of materials informatics. By applying the algorithm to each material dataset, prediction models with ±20 % precision were developed. When the algorithm was applied to all material datasets, the resulting prediction models had ±40 % precision for the single materials, whereas the prediction function was far from ideal in the case of the composite material. It is necessary to tune the input datasets (particularly the physical conditions of the composite material) for the ML model.

Conference Presentation

Citation Download Citation

Toshiyuki Kusumoto and Kiyokazu Mori "Prediction of ultrashort pulse laser ablation processing using machine learning", Proc. SPIE 11673, Laser Applications in Microelectronic and Optoelectronic Manufacturing (LAMOM) XXVI, 1167303 (5 March 2021); https://doi.org/10.1117/12.2577447

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available