Journal of Engineering Geology

---

All rockfill materials subjected to stresses above the normal geotechnical ranges exhibit considerable particle breakage. Particle breakage and crushing of the large particles to smaller ones result in a lower strength and higher deformability. The breakage of particle that is observed in the large scale triaxial tests, is usually expressed quantitatively by the Marsal breakage index, . This paper presents a method for calculating at any axial strain level in the large triaxial tests. The model used Rowe’s minimum energy principle ratio. The key parameter in modeling , is the friction angle which excludes dilation and breakage effects, . The results indicate that the internal friction angles at confining pressure equal and less than 200 kPa at the constant volume state is a unique value. Moreover, there is a linear relationship between the variation of energy spent on particle breakage to Marsal Breakage index with confining pressure, at failure axial strain.

---

This paper presented the feasibility of developing and using artificial neural networks (ANNs) for modeling the monotonic large scale triaxial tests over angular, rounded rockfill and materials contained various percentages of fines as a construction material in some dams in Iran. The deviator stress/excess pore water pressure versus axial strain behaviors were firstly simulated by employing the ANNs. Reasonable agreements between the simulation results and the tests results were observed, indicating that the ANN is capable of capturing the behavior of gravely materials. The database used for development of the models comprises a series of 52 rows of pattern of strain-controlled triaxial tests for different conditions. A feed forward model using multi-layer perceptron (MLP), for predicting undrained behavior of gravely soils was developed in MATLAB environment and the optimal ANN architecture (hidden nodes, transfer functions and training) is obtained by a trial-and-error approach in accordance to error indexes and real data. The results indicate that the ANNs models are able to accurately predict the behavior of gravely soil in CU monotonic condition. Then, the ability of ANNs to prediction of the maximum internal friction angle, maximum and residual deviator stresses and the excess pore water pressures at the corresponding strain level were investigated. Meanwhile, the artificial neural network generalization capability was also used to check the effects of items not tested, such as density and percentage smaller of 0.2 mm.

---

Consolidated-drained triaxial compression tests were conducted to compare the stress-strain and volume change response of sands and clayey sands reinforced with discrete randomly distributed poly-propylene fibers. The influence of various test parameters such as fiber content (0.0%, 0.5% and 1.0% by weight), clay content (0%, 10% and 20% by weight), relative density (50% and 90%) and confining pressure (100 kPa, 200 kPa and 300 kPa) were investigated. It has been observed that addition of clay particles to the sands decreased the shear strength of samples. Also, increase in clay content reduced dilation and increased compressibility of the mixed soil. Addition of the fiber to both sands and clayey sands samples improved the shear strength and increased ductility and axial strain at failure point. 

---

of initial texture and plastic fine content has been investigated. In order to model the different deposition condition four different specimen preparation methods are used. The tests are conducted on sands and its combination with to 25 fine percent using static and cyclic method. Four different specimen preparation methods are used consisting of Dry Funnel Deposition (DFD), Water Sedimentation (WS), Moist Tamping (MT) and Air Pluviation (AP). Each method may be useful to model the behavior of some special deposited soils. For example, DFD is the best method to model the natural deposition of silty sands, while the WS method is suitable for simulating the sands natural deposition in the rivers. MT method is the best method to simulate the behavior of compacted embankments while SD method is suitable to prevent the segregation of particles with respect to other methods like Water Pluviation. Effect of sample preparation methods on the sands and silty sands behavior is investigated to some extent, while the clayey sands are rarely considered. The results show that water sedimentation method, Air pluviation method and Dry funnel deposition method tends to create dilative textures with continuous dilative behavior even in high clay fine content. In contrast, the wet tamping method shows the dilative behavior only for clean sands specimens and increasing fines will results in instability and complete strain softening behavior. In cyclic loading the difference between different methods are less with respect to static loading. It seems that the results of different methods tends to be identical with increasing in fine content. In general, increasing fine content up to 20 percent results in more instability. After that a threshold value can be estimated bout 2 percent, after that the instability becomes less with continuous increase in fines content. The valuable result is that the threshold value is not dependent on the specimen preparation method or initial texture of specimens.