Numerical Analysis of Laser-assisted Ti to Polyimide Welding Using a Statistical Approach

M. Saleh Shaikh Mohammad Meiabadi, Afshin Kazerooni, Mahmoud Moradi

Research output: Contribution to journalArticlepeer-review

53 Downloads (Pure)

Abstract

Many efforts have been made at attaining direct joint between metals and polymers. Laser welding of dissimilar materials is a promising method to bond metals and polymers. In this study, we have developed a finite element (FE) model using ABAQUS software to investigate thermal phenomena of the laser welding of Ti to polyimide. Cylindricalinvolution-normal (CIN) heat source model is used to describe heat source power distribution. The Fortran software is used to programme the CIN heat source model and is linked to ABAQUS. The validation of the FE model is confirmed by attaining the analogous results to experimental data. The design of experiments (DOE) approach is utilized to statistical analysis of laser welding of Ti to polyimide. The effects of laser power
(0.5 to 1.5W) and laser scanning speed (50 to 150mm/min) on average bond width and delta bond width have been investigated via response surface methodology (RSM). The analysis is designed based on central composite design (CCD) full replication with two factors five levels. Results reveal that laser power is the most significant variable on average bond width and the laser scanning speed is the most influential variable on delta bond width.
Original languageEnglish
Pages (from-to)185–205
Number of pages21
JournalInternational Journal of Laser Science: Fundamental Theory and Analytical Methods
Volume1
Issue number2
Publication statusPublished - 2018
Externally publishedYes

Bibliographical note

Open Access

Keywords

  • Laser welding
  • titanium
  • Ti
  • polyimide
  • numerical analysis
  • ABAQUS
  • Fortran
  • finite element (FE)
  • central composite design (CCD)
  • response surface methodology (RSM)

Fingerprint

Dive into the research topics of 'Numerical Analysis of Laser-assisted Ti to Polyimide Welding Using a Statistical Approach'. Together they form a unique fingerprint.

Cite this