Sharif Digital Repository

Demand to structural and industrial materials with unique properties lead to produce unorthodox combination of metals. One of the most useful combinations is composites of Aluminum-Steel and this composites is mostly used as a sheet. Due to differences in melting point of two metals, preferred method for welding of this two metals is Cold Roll Bonding. For enhance of mechanical properties of this two metals, preferred methods is Severe Plastic Deformation. By the way, in SPD methods, two method was established for sheets. Accumulative Roll Bonding and Constrained Groove Pressing but ARB method produce multi-layer composites so preferred method to produce bilayer sheets is CGP. In this study,... 

In this thesis, we study the mean graph isoperimetry problem using an optimization approach. The k-th isoperimetric constant of a graph is defined as the minimum of an objective function (p-norm of the vector consisting of normalized flow) over k-subpartitions of vertices. We note that the normalized cut problem can be formulated as a semidefinite programming problem and utilizing the relaxation methods for semidefinite programs, the problem can be solved in approximately polynomial time. Finally, we model the isoperimetry problem as an optimization problem with orthogonality constraints and utilizing Wen and Yin’s efficient method for finding local minima of the problem, we extract a... 

Abstract: In this thesis, we tackled the problem of nonlinear activation function in optical artificial neural networks, and in particular in convolutional and recurrent neural networks. In the end, we propose an all-optical recurrent neural network in free-space optics for the first time. Artificial neural networks are a branch of artificial intelligence, which can be adopted to solve a wide variety of problems. While very powerful, these networks can be very power hungry and slow when it comes to solving very complicated problems. Optical versions of these networks bring the promise of solving both of these issues and provide a fast and power efficient platform for these networks. However,... 

Orthogonal Nonnegative Matrix Factorization (ONMF) with orthogonality constraints on a matrix has been found to provide better clustering results over existing clustering problems . Because of the orthogonality constraint , this optimization problem is difficult to solve . Many of the existing constraint-preserving methods deal directly with the constraints using different techniques such as matrix decomposition or computing exponential matrices . Here , we propose an alternative formulation of the ONMF problem which converts the orthogonality constraints into non-convex constraints . To handle the non-convex constraints , a penalty function is applied . The penalized problem is a... 

Lateral spreading is defined as finite lateral displacement of mildly sloping grounds or those ending in free faces induced by liquefaction. The phenomenon of lateral spreading caused by liquefaction in coastal areas and mildly sloping grounds has caused significant damage to deep foundations of engineering structures such as bridge and buildings in severe earthquakes. Since earthquake is unavoidable, therefore, it is necessary to provide appropriate solution to reduce the effects of liquefaction induced lateral spreading. Despite conducting various laboratory and field studies by previous researchers, there is still no comprehensive approach to evaluate the effects of lateral spreading on... 

Due to inevitable uncertain sources in modeling, operational and environmental conditions, finite element model and structure’s response to a same load pattern differs drastically. To reduce this difference and make the model’s response close to the real structure, finite element model updating procedure is essential. Updated model can be used for structural assessment, damage identification, remaining service life estimation, and structural control. Model updating methods are categorized into two groups in term of information domain used for model updating; time-domain methods and frequency-domain approaches. Time-domain methods have a preference because of the main drawbacks of frequency...