V. Kalatjari - Assistant Professor, Shahrood University of Technology
P. Mansoorian - PHD Student, Shahrood University of Technology

Many researches have been carried out in structural topology optimization using the traditional stiffness method in last decades. Similarly, the force method has been employed to optimize the trusses due to its speed and efficiency. In this paper as the first feature, the algebraic force method formulation is developed such that it is possible to utilize it in the topology optimization. The force method has two advantages over the conventional stiffness method. Firstly, number of equations should be solved in the force method is the same as the degree of static indeterminacy in place of the total degrees of freedom used in the stiffness method, resulting in increasing of the optimization speed. Secondly, the matrices corresponding to particular and complementary solutions are formed independently of the members’ mechanical properties. In developed version, these matrices are formed only once for intact truss and used numerous times for trusses of the same topology. Second feature of this study is usage of a cluster of parallel personal computers due to genetic algorithm capability in parallelism. Each processor or island optimizes independently with own parameters and operators, resulting in diversity of solutions and escaping the local optimum.

algebraic force method, topology optimization, distributed parallel genetic algorithm