Journal of Engineering Geology

---

Evaluation of the excavation-induced ground movements is an important design aspect of supporting system in urban areas. This evaluation process is more critical to the old buildings or sensitive structures which exist in the excavation-affected zone. Frame distortion and crack generation are predictor, of building damage resulted from excavation-induced ground movements, which pose challenges to projects involving deep excavations. Geological and geotechnical conditions of excavation area have significant effects on excavation-induced ground movements and the related damages. In some cases, excavation area may be located in the jointed or weathered rocks. Under such conditions, the geological properties of supported ground become more noticeable due to the discontinuities and anisotropic effects. This paper is aimed to study the performance of excavation walls supported by nails in jointed rocks medium. The performance of nailed wall is investigated based on evaluating the excavation-induced ground movements and damage levels of structures in the excavation-affected zone. For this purpose, a set of calibrated 2D finite element models are developed by taking into account the nail-rock-structure interactions, the anisotropic properties of jointed rock, and the staged construction process using ABAQUS software. The results highlight the effects of different parameters such as joint inclinations, anisotropy of rocks and nail inclinations on deformation parameters of excavation wall supported by nails, and induced damage in the structures adjacent to the excavation area. The results also show the relationship between excavation-induced deformation and the level of damage in the adjacent structure.

---

./files/site1/files/7Extended_Abstract.pdfExtended Abstract
 (Paper pages 135-156)
Introduction
Many civil structures (e.g. tunnel walls, bridge pillars, dam abutments and road foundations) are subjected to both static and dynamic loads. Cyclic loading leads to occurring fatigue phenomenon. Fatigue is the tendency of materials to break, or the process of damage accumulation, under cyclic loading. It was found that the dynamic fatigue strength can be reduced by 30-70 percent on average compared to uniaxial compression strength. Different materials show different response when they are subjected to cyclic loading. Some materials become stronger and more ductile, while others become weaker and more brittle. Although it is clear that the mechanical properties of rock under dynamic loads varied dramatically from those under static loads, the nature of dynamic failure in rock remains unclear, especially in cyclic loading condition. Fatigue behavior of rocks was rarely studied in respect to other materials such as steel and soil. The performed researches on fatigue behavior of rocks indicated that fatigue life will be decreased by increasing load amplitude in logarithmic and exponentially pattern. Also, strain softening is the dominated behavior of rocks against cyclic loading. Furthermore, some parameters such as maximum load level, confining pressures, amplitude, and loading frequency have considerable effects on fatigue behavior of rocks. However, available data on fatigue behavior remain insufficient for solving the practical tasks of predicting rock bursts and earthquakes. Obtained results are inconclusive and sometimes discordant. The aim of the current work was to assess tonalite rock fatigue behaviour under different loading conditions to describe the fatigue damage process of the granitic rock.
Material and methods
Several core samples were prepared to perform this research. The core samples were prepared with a L/D ratio of 2.5 with an average diameter of 54 mm. Before the fatigue tests, the physical and mechanical properties of the rocks were measured. Uniaxial compressive strength test (UCS) has been done on 5 core samples. The tests were performed in the load-control mode with a 1.6 kN/s loading rate. The tests were conducted to obtain the physico-mechanical parameters of the rocks in static loading condition, and provided a reference for subsequent dynamic tests. The cyclic tests were performed in both load and displacement control modes. To record axial and lateral strains during the fatigue tests, four strain gauges have been employed with arrangement of two axial and two laterals. Also, three acoustic emission sensors were installed on top, mean and bottom of the core samples to record cracking sound. In order to doing the tests a servocontrol Instron machine with 500 kN capacity was employed. The fatigue tests were conducted with three different maximum loads, 1 Hz frequency, and constant amplitude (0.82 of uniaxial compressive strength). The maximum stress level (the ratio of maximum cyclic stress to static strength) was varied 0.80, 0.85, and 0.90. The amplitude level (the ratio of amplitude stress to static strength) ranged from 0.50 to 0.70 and 0.90. Finally, Multi stages loading with increasing amplitude were applied for the displacement control tests. The results of fatigue tests have been evaluated by fatigue damage parameters including maximum and minimum axial strain, maximum and minimum lateral strain, tangent and secant modulus, toughness and hysteresis energy.
Results and discussion
The obtained results indicated that during fatigue process failure occurs below the maximum strength loading condition as a result of accumulative damage. Analysis of the fatigue test results showed that the fatigue failure consisted of three stages: fatigue crack formation (initiation phase I), stable crack propagation (uniform velocity phase II), and unstable crack propagation resulting in a sudden breakdown (accelerated phase III). By comparing the axial and lateral deformation, it was found that lateral deformation is more sensitive to fatigue. At higher stress levels, considerable part of fatigue life is response to crake development, whereas at lower stress levels, crack acceleration phase of fatigue life is distinguishable. Descending trend of loading and unloading tangent modulus shows a scatter pattern. This behavior may be related to the calculation method and loading condition, as well as microstructure and behavior of the rock mass. In spite of tangent modulus results, the three-stages of damage process (especially phase I and II) for secant modulus in both loading and unloading conditions are clear. The result is due to the method of calculation and increase in axial strain with increasing number of cycles. Brittle behavior of this type of rock leads acceleration phase to be hidden and unclear in most of fatigue damage parameters. A dramatic decrease of toughness and hysteresis energy in the first few cycles is due to the closing of pre-existing micro fractures. In fact, during the initial cycle, the rock behaves in a more ductile fashion than in the next few cycles. Thereafter, toughness begins to increase slowly, then steadily, and finally rapidly. A similar behavior was found for hysteresis energy as well. This fact indicated that cracks generated in parallel to loading direction. Fatigue displacement control tests show a strain softening behavior for the granitic rocks. This behavior is highlighted in variation of maximum stress during the tests. This parameter, especially in final step of loading, shows distinguishable decreasing trend.
Conclusion
The tonalite rocks were subjected to uniaxial cyclic loading in both load and displacement control mode. The following conclusions were drawn from this research.
-Accumulated fatigue damage occurs in an obvious three-stage process. This is the result of the micro-fracturing mechanism in the fatigue process.
-By comparing axial and lateral strain damages, it was found that crack propagation occurred in the loading direction and crack opening occurred in the lateral direction. So, among fatigue damage parameters, lateral strain shows the best three-stage fatigue damage behavior.
- Strain softening was found as rock response to cyclic displacement control loading.

---

In recent decades many researchers have studied on the damage assessment of structures after a seismic event. To assess the damage of structures under an earthquake, it is so important to study the correlations between earthquake parameters and damages of the structures. A lot of seismic parameters have been defined by researchers to characterize an earthquake. Spectral parameters of an earthquake convey a variety of information about ground motion, so they can properly characterize an earthquake. Also a lot of damage indices were proposed by researchers to quantify the damage of the structures or to rank their vulnerability relative to each other. Park-Ang index is one of the best indices to describe the damage of a structure. In this paper, the correlations between spectral parameters of earthquakes and Park-Ang indices are studied. Three RC frames with different height are analyzed under far-fault earthquake records by nonlinear dynamic analyses. The correlations between spectral parameters and Park-Ang indices of the frames are calculated. The results show that in all the frames most of spectral parameters have strong correlations with damage intensity. In order to estimate the damage potential of an earthquake, some spectral parameters which have high correlations with damage intensity can be proper indices. Housner intensity, acceleration spectrum intensity and velocity spectrum intensity are shown to have strong correlations with damage intensity. In this paper, a new spectral parameter which has high correlation with damage intensity is achieved.