Over the last decades, heuristic optimization methods based on imitating natural, biological, social or cultural processes on a computational level have aroused the interest of the scientific community due to their ability to effectively explore multimodal and multidimensional solution spaces. Despite all the papers published in the international literature, most heuristic algorithms still have low precision and accuracy. In this context, a modified Simulated annealing algorithm (MSAA) is proposed and validated for solving optimization problems. Performance evaluation was performed on test functions (benchmark functions) with and without restrictions reported in the international literature and practical design problems in civil engineering. In all cases analyzed MSAA obtained equal or better results than those reported by other authors, illustrating the applicability of the proposed algorithm.