Thanks a lot for your reply.
I did not present further examples because it is needless as the problem occurs in the first simple example as well.
You can see that when we wrote with(Physics) at the top of our codes, the answer changes.
As you stated, It seems that I must change the frozen part, but "Physic:-diff" does not work.
I really wonder about "freeze" in general; I think that freeze is one of the maple's Bugs, I mean that maple should identify these automatically because the issue does not just occur about that; For example, in Taylor's expansion of
taylor(cos(theta(t)),theta(t)=0 , 7);
maple returns error while when I remove the t-dependence as
taylor(cos(theta),theta=0 , 7)
I can get the result. Indeed, maple is incapable of recognizing such things !!
I feel that for this problem, I have to use freeze as well to reach a conclusion. However, I failed for this problem as well.
Anyway, if I find a solution for both, I shall write here.
Thanks a lot for your evergreen helps.