Which design is better depending on the specific objectives of your experiment and the nature of the factors involved. If you have three factors and want to investigate their main effects and two-factor interactions while keeping the number of experimental runs low, Box-Behnken design can be a good choice. If you have more than three factors, want to capture higher-order interactions or quadratic effects, and have the resources for conducting additional experimental runs, Central Composite design is more suitable.

Thank you for your question.