Determination of Compressibility Parameters of Clay-Silt Soils of Lagoon Origin by means of Regression and Artificial Neural Network Methods


DİPOVA N., Cangir B.

TEKNIK DERGI, cilt.21, sa.3, ss.5069-5086, 2010 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 21 Sayı: 3
  • Basım Tarihi: 2010
  • Dergi Adı: TEKNIK DERGI
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.5069-5086
  • Anahtar Kelimeler: ANN, Antalya, CPT, Lagoon, Regression, SPT
  • Akdeniz Üniversitesi Adresli: Evet

Özet

In this research, compressibility properties of blue soft clays in the Hurma and Sansu regions (Antalya) were investigated by means of in situ and laboratory tests. Various depositional environments, sea level changes and drying effects have resulted in highly complex soil profiles throughout the area. Soil parameters can vary 5-10 times only in one borehole. In practice, oedometer tests are being carried out in limited number, and so, it was thought that the number of compressibility parameters can be increased by means of correlations with index parameters and in-situ tests. For this purpose, samples were taken in every 0.5 m from 7 boreholes of 10 in depth and were investigated by laboratory methods. Additionally, as in situ tests, standard penetration test (SPT) and cone penetration test (CPT) were conducted. Geotechnical properties were determined by in situ and laboratory test results. Comprehensive multi-statistical analysis was performed by SPSS and Data Fit software for establishing specific correlation between soil parameters. Artificial Neural Network was developed to simulate the mapping between index and compressibility parameters by Mat lab. As conclusion, to evaluate compressibility parameters, ANN models seem to estimate values more close to the measured values than empirical equations of regression analysis.