Some designs fit some data better !! .regards

Many researchers use CCDs as screening designs for response surface methods to optimize say chemical reaction yields. You might wish to modify your research questions to study this method in the system you are following. There's a big literature in this area. Best wishes David Booth

David Eugene Booth : Any suggestion for the literature?

Try the two books that I indicated in the attachments. They are classic references

They are both available for download from the z-library Best wishes David Booth

There are few question necessery to ask.

Are you sure you have used Central Composite Design? CCD requires 5 levels for each factor: -axial, -1, 0, 1, axial. Perhaps you used Box-Behnken Design which requires -1, 0, 1?

Nextly, what do you mean you used different data, which gave you a good model? You already had responses for this exact design? If not, maybe previous data does not represent your current experiment?

Eventually, are you sure your factors significantly affect the response? They might not, therefore it can not find significant model.

Jakub has already mentioned a few important factors that can lead to unsatisfactory results. It is usually not due to the methods used, but to the users. When designing 31 experiments, I should also be aware that the variables chosen are significant to the process, are independent of each other, and all other variables are constant. If this is not adhered to, especially the constancy of the parameters not taken into account and, in the case of comparisons, the agreement of this constancy, then a good, ideal mathematical analysis model is of no use to me.