Estimation of mode I fracture toughness of rocks exposed to different environmental conditions using simple and linear multiple regression

Creative Commons License

Özdemir E., Eren Sarıcı D.

MECHANICS OF TIME-DEPENDENT MATERIALS, cilt.28, ss.1-17, 2024 (SCI-Expanded, Scopus)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 28
  • Basım Tarihi: 2024
  • Doi Numarası: 10.1007/s11043-024-09731-2
  • Dergi Adı: MECHANICS OF TIME-DEPENDENT MATERIALS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC
  • Sayfa Sayıları: ss.1-17
  • İnönü Üniversitesi Adresli: Evet

Özet

Mode I fracture toughness (Kıc) is a critical parameter in rock mechanics that is essential for

understanding how rocks behave under tensile loading and crucial for applications ranging

from safety assessments to structural design in geotechnical engineering. This study comprehensively

investigates the influence of various environmental conditions (dry, saturated,

frozen, thermal shock and thermal aging) on the physico-mechanical properties and Kıc of

rocks. The primary novelty of this study lies in its comprehensive modeling approach under

diverse environmental conditions, providing a nuanced understanding of factors influencing

rock fracture toughness. By extending analysis to less-studied conditions such as freezing

and thermal shock cycles, the study enhances the predictive capacity of fracture toughness

models in practical geotechnical applications. Physico-mechanical properties, including uniaxial

compressive strength, point load strength, Brazilian tensile strength (BT), Schmidt

hardness, and ultrasonic wave velocity were evaluated across different environmental scenarios.

Simple and linear multiple regression models were developed using these properties

to predict Kıc. Notably, BT emerged as a significant predictor in the simple regression analyzes.

Ten linear multiple regression models were formulated using SPSS 20, combining

mechanical tests (UCS, BT, PL) with non-destructive testing methods (Vp, Vs, SH), demonstrating

robust predictive capabilities with R2 values exceeding 0.95. Performance metrics

(mean absolute error, mean absolute percentage error, root mean square error) were used to

verify the accuracy of the model.