Application of numerical modeling and genetic programming to estimate rock mass modulus of deformation

Purpose The shear strength of reinforced concrete (RC) columns under cyclic lateral loading is a crucial concern, particularly, in the seismic design of RC structures. Considering the costly procedure of testing methods for measuring the real value of the shear strength factor and the existence of several parameters impacting the system behavior, numerical modeling techniques have been very much appreciated by engineers and researchers. This study aims to propose a new model for estimation of the shear strength of cyclically loaded circular RC columns through a robust computational intelligence approach, namely, linear genetic programming (LGP). Design/methodology/approach LGP is a data-driven self-adaptive algorithm recently used for classification, pattern recognition and numerical modeling of engineering problems. A reliable database consisting of 64 experimental data is collected for the development of shear strength LGP models here. The obtained models are evaluated from both engineering and accuracy perspectives by means of several indicators and supplementary studies and the optimal model is presented for further purposes. Additionally, the capability of LGP is examined to be used as an alternative approach for the numerical analysis of engineering problems. Findings A new predictive model is proposed for the estimation of the shear strength of cyclically loaded circular RC columns using the LGP approach. To demonstrate the capability of the proposed model, the analysis results are compared to those obtained by some well-known models recommended in the existing literature. The results confirm the potential of the LGP approach for numerical analysis of engineering problems in addition to the fact that the obtained LGP model outperforms existing models in estimation and predictability. Originality/value This paper mainly represents the capability of the LGP approach as a robust alternative approach among existing analytical and numerical methods for modeling and analysis of relevant engineering approximation and estimation problems. The authors are confident that the shear strength model proposed can be used for design and pre-design aims. The authors also declare that they have no conflict of interest.

Download Full-text

Rockslides on limestone cliffs with subhorizontal bedding in the southwestern calcareous area of China

Natural Hazards and Earth System Science ◽

10.5194/nhess-14-2627-2014 ◽

2014 ◽

Vol 14 (9) ◽

pp. 2627-2635 ◽

Cited By ~ 7

Author(s):

Z. Feng ◽

B. Li ◽

Y. P. Yin ◽

K. He

Keyword(s):

Numerical Modeling ◽

Rock Mass ◽

Hard Rock ◽

Compression Fracture ◽

Field Observations ◽

Mountainous Areas ◽

Hard Rocks ◽

Rupture Plane ◽

Rock Collapse ◽

Cliff Retreat

Abstract. Calcareous mountainous areas are highly prone to geohazards, and rockslides play an important role in cliff retreat. This study presents three examples of failures of limestone cliffs with subhorizontal bedding in the southwestern calcareous area of China. Field observations and numerical modeling of Yudong Escarpment, Zengzi Cliff, and Wangxia Cliff showed that pre-existing vertical joints passing through thick limestone and the alternation of competent and incompetent layers are the most significant features for rockslides. A "hard-on-soft" cliff made of hard rocks superimposed on soft rocks is prone to rock slump, characterized by shearing through the underlying weak strata along a curved surface and backward tilting. When a slope contains weak interlayers rather than a soft basal, a rock collapse could occur from the compression fracture and tensile split of the rock mass near the interfaces. A rockslide might shear through a hard rock mass if no discontinuities are exposed in the cliff slope, and sliding may occur along a moderately inclined rupture plane. The "toe breakout" mechanism mainly depends on the strength characteristics of the rock mass.

Download Full-text