Parametric evaluation and predictive modelling of formability in μ-SPIF process
Single Point Micro Incremental Forming (µ-SPIF) is a versatile route for complicated, and customized micro-components. Localised deformation in terms of bending stretching and through thickness shears add the complexity in SPIF which intensifies in case of µ-SPIF due to size effect. Hence, detailed...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2025-01-01
|
| Series: | MATEC Web of Conferences |
| Subjects: | |
| Online Access: | https://www.matec-conferences.org/articles/matecconf/pdf/2025/02/matecconf_iddrg2025_01045.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Single Point Micro Incremental Forming (µ-SPIF) is a versatile route for complicated, and customized micro-components. Localised deformation in terms of bending stretching and through thickness shears add the complexity in SPIF which intensifies in case of µ-SPIF due to size effect. Hence, detailed parametric study on the µ-SPIF process along with data driven models is need of the time to address formability in case of µ-SPIF. In this study, the effect of wall angle, step depth, spindle speed and feed rate on the formability, forming height, process time and surface roughness have been studied in the micro-forming of Aluminium sheet of 50µm thickness with a hemispherical-tip microtool of 1 mm diameter. The dynamic behaviour of the material during forming was monitored through continuous measurement of forming force using a multi-component dynamometer. Machine learning regression models e.g. Tri-layered Neural Network, Quadratic Support Vector Machine, and Gaussian Process Regression are developed based on experimental data to predict the formed height and the surface roughness. The study found correlations of process parameters with forming time, surface roughness and height of the deformed parts. This study emphasizes the integration of experimental data, process analysis, and predictive modelling as a means of creating a digital twin framework for µ-SPIF. |
|---|---|
| ISSN: | 2261-236X |