Multi-Objective Optimization of Manufacturing Processes Using Evolutionary Algorithms with a crossover probability of 0.8 (Figure 7f). The crossover probability has usually a higher value, as generation of new chromosomes is to be encouraged. Then bitwise mutation is performed subsequently with a mutation probability of 0.1, which is commonly taken to be a smaller value (Table 3, Col.(s) and Figure 7g). The chromosomes obtained after mutation in the first generations are used as input for next generations and so on (Table 3, Col.(t)). The above procedure is continued until the termination criterion (200 generation) is over. The values shown in Table 3 are obtained after the 200 th generation. The finally generated chromosome is expected to have better solutions than that of the previous generations (Table 3, Col.(t)). The chromosomes obtained in the final generation with rank 1 are considered as Pareto-optimal solutions. Typical results of the Pareto-optimal solutions (1 st rank) obtained at the 200 th iteration is shown GA. In this algorithm, selection is carried out with the help of crowded-comparison operator based on the ranking and crowding distance. The offspring population is first created by using the parent population. Instead of finding the non-dominated front, the two populations are combined to form of size 2N. Then, a non- dominated sorting is used to classify the entire population. The main advantage of maintaining non-dominated solutions in the population is straightforward implementation. In this strat- egy, the population size is an important GA parameter, since no external archive is used. Although this requires more effort, however it allows a global non-dominance check among the parent and offspring solutions. Once the non-dominated sorting is over, the new popu- lation is filled by solutions of different non- dominated front, one at a time, starting with the best non-dominated front, followed by the