A good scientific model should not change?

No, good scientific models should always be open for change. They should be tested and tested repeatedly, so they can be revised, or scrapped in favour of a better hypothesis.