Predicting the effect of adherend dimensions on the strength of adhesively bonded joints using M5P and M5 classifiers


Ayaz Y.

Journal Of The Brazilian Society Of Mechanical Sciences And Engineering, cilt.42, ss.465, 2020 (SCI Expanded İndekslerine Giren Dergi)

  • Cilt numarası: 42 Konu: 9
  • Basım Tarihi: 2020
  • Doi Numarası: 10.1007/s40430-020-02547-4
  • Dergi Adı: Journal Of The Brazilian Society Of Mechanical Sciences And Engineering
  • Sayfa Sayıları: ss.465

Özet

The configuration of adhesive bonding, especially adhesive–adherent interface, is very important in adhesive bonding. Adherend thickness, overlap length and adherend width have a great effect on failure load. It is important to know the effect of these parameters in the estimation of failure load. In this study, two models based on data mining techniques were generated for the estimation of the effect of adherend thickness, width and overlap length on failure load in adhesively bonding joints. One of the models is based on the M5P model tree and the other is based on the M5 rule algorithm. Experimental data were used to evaluate the performance of the models. The correlation coefficient parameter was used to compare the models’ performance. The evaluated correlation coefficient is 93% for the M5P model tree and 96% for the M5 rule.