A parallel level-set based method for topology optimization

Level set method for topology optimization are by now sophisticated and used in a wide variety of research areas. Level set method with mesh requires finite element analysis that is often complex and time consuming due to large sparse coefficient matrix of equations. This paper presents EBE-CG method to parallel compute large sparse linear equations for level set method based on normal conjugate gradient method. Several processes are employed in conjunction with each other. This method is accurate enough and easy to be parallelized. Furthermore, it does not require large memory space. Level set method combined with EBE-CG parallel computing is much more efficient because of evidently abridged computation time.