Computational Methods in Engineering

---

In this paper, dynamic and stability analysis of a flexible cam-follower system is investigated. Equation of motion is derived considering flexibility of the follower and camshaft. Viscous and Coulomb frictions are considered in the rocker arm pivot. The normalized equation of motion of the system is a 2nd- order differential equation with periodic coefficients. Floquet theory is employed to study parametric stability of the system. Stability diagrams are presented and the effects of varying cam profiles and motion events on the stability of the system are compared. Results show that viscous and Coulomb frictions stabilize the motion of the system

---

In spite of the fact that the effect of earthquake on earth dams has been widely studied during the past decades, the complicated behavior of such earth structures against different seismological characteristics is still unknown. Such ambiguities necessitate more accurate studies using more comprehensive computation tools to achieve new results describing the behavior of such structures subjected to earthquake loading. In the present study, the simple soil model of elastic, perfectly plastic (based on the Mohr-Coulomb criterion), and Rayleigh damping criterion have been adopted for the soil. First, the numerical model employed was verified by dynamic analysis of real cases such as “Long Valley” and “santa Felecia” earth dams. The computational results were then compared with real recorded data or with those reported by other researchers. In addition to evaluating seismic stability of earth dams, their seismic stability was verified using pseudo-static analyses. Therefore, the “Carsington” dam was analyzed to verify the results of pseudo-static analyses and to check the results of FLAC software in calculating the pseudo-static factor of safety. The values of calculated factors of safety in the present study are in good agreement with the published results in the literature. Furthermore, the failure behavior revealed in the analysis shows the ability of FLAC software in defining the failure surface. In the main part of the analyses, a parametric study was conducted for different selected conditions and specially the effect of dam height and the optimum size of crest width were investigated. The results are presented in relevant diagrams.

---

In this research, nonlinear dynamic analysis of concrete shear wall using a new nonlinear model based on damage mechanics approach and considering bond slip effects is presented. Nonlinear behavior of concrete is modeled by a rotational smeared crack model using damage mechanics approach. The proposed model considers major characteristics of the concrete subjected to two and three dimensional loading conditions. These characteristics are pre-softening behavior, softening initiation criteria and fracture energy conservation. The model was used in current research analysis after verification by some available numerical tests. Reinforcements are modeled by a bilinear relationship using two models: Discrete truss steel element and Smeared model. In Discrete model the effects of bond-slide between concrete and rebar is mentioned using the bond-link element model concept. Based on the presented algorithms and methodology, an FEM code is developed in FORTRAN. The validity of the proposed models and numerical algorithms has been checked using the available experimental results. Finally, numerical simulation of CAMUS I and CAMUS III reinforced concrete shear walls is carried out. Comparisons of deduced results confirm the validity of proposed models. The obtained results, both in the expected displacements and crack profiles for the walls, show a good accuracy with respect to the experimental results. Also, using discrete truss element model with respect to the smeared steel model leads to increasing the accuracy of maximum displacement response to 7% in analysis.

---

Since there is no closed-form formula for designing TMD (Tuned Mass Damper) for nonlinear structures, some researchers have proposed numerical optimization procedures such as a genetic algorithm to obtain the optimal values of TMD parameters for nonlinear structures. These methods are based on determining the optimal values of TMD parameters to minimize the maximum response (e.g. inter story drift) of the controlled structure subjected to a specific earthquake record. Therefore, the performance of TMD that has been designed using a specific record strongly depends on the characteristics of the earthquake record. By changing the characteristics of the input earthquake record, the efficiency of TMD is changed and in some cases, it is possible that the response of the controlled structure is increased. To overcome the shortcomings of the previous researches, in this paper, an efficient method for designing optimal TMD on nonlinear structures is proposed, in which the effect of different ground motion records is considered in the design procedure. In the proposed method, the optimal value of the TMD parameters are determined so that the average maximum response (e.g. inter story drift) resulting from different records in the controlled structure is minimized. To illustrate the procedure of the propose method, the method is used to design optimal TMD for a sample structure. The results of numerical simulations show that the average maximum response of controlled structure resulting from different records is reduced significantly. Hence, it can be concluded that the proposed method for designing optimal TMD under different earthquakes is effective.

---

Biot equations that consider fluid and soil interaction at the same time are the most applicable relationships in the soil dynamic analysis. However, in dynamic analysis, due to the sudden increase in the excess pore pressure caused by seismic excitation and the occurrence of high hydraulic gradients, the assumption of the Darcy flow used in these equations is questionable. In the present study, in the u-p form of Biot equations, non-Darcy flow is considered. Also, the nonlinear behavior of soil is modeled by the Pastor-Zienkiewicz -Chan model. For validation, the VELACS No.1 experiment is modeled and the effect of the nonlinear fluid flow assumption on the results is examined. The results indicate that in the low permeability coefficients, the obtained results of the non-Darcy and Darcy flow are in agreement; however, in high permeability coefficients, these two methods differ by time and depth.
 

---

Dynamic stability and liquefaction of tailings dams are great concerns for geotechnical engineers. In this study, the seismic response of the Esphordi mine tailing dam located in Bafgh seismic region of Yazd province is investigated. A finite-difference code (FLAC2D) is used to model the seismic liquefaction applying two constitutive criteria, namely Mohr-Coulomb and Finn-Byrne. For this purpose, a fish function is implemented into the code to simulate the non-linear elasto-plastic Finn-Byrne constitutive model. Horizontal and vertical displacements (subsidence) in the dam body, additional pore pressure, failure zones, and liquefaction due to seismic load were determined using the two selected criteria under the seismic load of the 6.4 magnitude earthquake occurred in 2005. Considering the type of behavioral model, Mohr-Coulomb and Finn-Byrne, the maximum horizontal displacement of 5 and 35 cm in the dam body and downstream, and subsidence of 4 and 23 cm at the dam crest and upstream are observed, respectively. Also, the calculated ratio of excess pore pressure (R_u), for both criteria, was less than the liquefaction limit (0.9), the maximum value of which was 0.7 for the Finn-Byrne criterion and 0.2 for the Mohr-Coulomb criterion. In general, the results show that considering the cumulative effect of the seismic load cycles in the Finn- Byrne model, this criterion provides a better understanding of the liquefaction phenomenon.