Failure strength prediction of aluminum spot-welded joints using kernel ridge regression

Maher Maalouf, Zuheir Barsoum

    Research output: Contribution to journalArticlepeer-review

    10 Scopus citations

    Abstract

    The current paper presents an alternative method for failure strength prediction of spot-welded joints in aluminum, based on nonlinear regression analysis, namely, the kernel ridge regression method. Welding parameters such as electrode force, welding current, and welding time are studied in the experimental investigation to measure their effects on the nugget size and failure strength of the resistance spot welds. Coupons are manufactured and tensile tested and the results show that the welding current and time have the largest effect on the nugget size and the failure strength. The results of this study are compared to those of the least squares method and they indicate that the truncated-regularized kernel ridge regression algorithm significantly improves the coefficient of determination and reduces the mean squared error.

    Original languageBritish English
    Pages (from-to)3717-3725
    Number of pages9
    JournalInternational Journal of Advanced Manufacturing Technology
    Volume91
    Issue number9-12
    DOIs
    StatePublished - 1 Aug 2017

    Keywords

    • Aluminum alloys
    • Failure strength
    • Kernel ridge regression
    • Resistance spot welding
    • Truncated Newton

    Fingerprint

    Dive into the research topics of 'Failure strength prediction of aluminum spot-welded joints using kernel ridge regression'. Together they form a unique fingerprint.

    Cite this