What could be a good reason for substituting a scientific paradigm (as Kuhn defined it) with a new one? For example why copernican astronomy was substituted by newtonian's?
Good reasons include if the new theory is simpler, has a wider explanatory scope, is more accurate, and/or produces a new set of testable predictions that are subsequently confirmed.
The way you've framed the question, there are no such reasons!
In Chapters 6–9 of Structure of Scientific Revolutions, Kuhn lays out the process by which one paradigm replaces another as a sociological process. The process is typically promoted by increased attention to anomalies in the existing paradigm, which is itself promoted by the sociological phenomenon of younger generations looking for new research projects, and by the development of technology aimed at investigating anomalies, which in turn creates new anomalies and amplifies the visibility of existing ones. Other factors contribute.
Within a particular paradigm, the emergence of anomalies will seem like a good reason to abandon the current paradigm, but these can never be real reasons, in the sense of ideas that one ought to accept. That is because there are, for Kuhn, no paradigm-independent truths, and because shifting from one paradigm to another is a matter of Gestalt-shift, or change in one's way of looking at things. It is not a rational, or reason-driven process. It is produced by crisis in an existing paradigm that makes scientists want to flee it for something else, rather than rational or reason-driven evaluation of two paradigms.