Abstract
A parameter optimization procedure for designing rivet dies is presented to improve the joining quality of an electromagnetic riveted joint. The optimal Latin hypercube sampling technique, radial basis function neural network model and Hooke–Jeeves algorithm were used to construct the optimization model. The results indicated that the rivet die angle was more sensitive to the expansion of the rivet shaft compared to the height. The optimal angle and height of the rivet die were found to be 20° and 2.68 mm, respectively. Relative to a joint with regular rivet die, the average relative interference of the optimized joint increased from 0.35% to 3.50%. This is because the designed rivet die restricted the deformation of the formed head and led to greater expansion of the rivet shaft in the restricted upsetting stage. The shear and pull-out strengths of the joint after optimization were increased by 5.20% and 11.41%, respectively.
Acknowledgements
This project is supported by the National Natural Science Foundation of China [no. 51975202], the Natural Science Foundation of Hunan Province [2019JJ30005] and the Key Research and Development Project of Hunan Province [2017GK2090].
Disclosure statement
No potential conflict of interest was reported by the authors.
Replication of results
The raw/processed data required to reproduce these findings cannot be shared at this time owing to technical or time limitations.
ORCID
Junjia Cui http://orcid.org/0000-0001-8004-6030