Finding the curl in diffrent coordinates by transforming variables

[sentinel]

we have a well known and simple equation for curl in cartesian coo. now we want it in let's say cylindrical coordinates.
question is...can we transform every thing to cylinderical and then use the formula for cartesian?I mean writing basis vectors of cartesian in terms of r and theta and z and basis vectors of cylindrical , and then write the x y z components of the vector(which we want its curl) in terms of its r and theta and z (cylindrical) components and then write the partial differentiations of cartesian in terms of cylindrical r and theta and then write the equation in cartesian>>transform everything to cylindrical>>get the desired formula!
I did this but it gave me wrong answer.WHY??!
I can and did derive the curl in different coordinates by using their definition but using the way I said above it should work...why not?

 

[chiro]

Hey sentinel and welcome to the forums.

You should note that the basis vectors for the cartesian system are not going to necessarily be the same (length, orientation, etc) in the new system.

Without speculating, you should post what you did here so you can get some specific feedback and suggestions.