Can physicists do any physics problem from a university physics textbook?

What if it is an engineering problem rather than a physics one ?

 I find that when you first learn a topic, it's hard, but once you've done a semester of it and you've sat an exam on it, it's much easier a second time round if you need to go back over it. Stuff is always much harder when it's less familiar.

A couple of weeks ago, I was learning magnetism and some concepts left me confuse and some others took more time to sink in. I talked about my doubts with my teacher and I think I kept thinking about the topic at random times during the day. I didn't make a conscious effort to but I just was and things got progressively clearer.

 Quote by CNS92 Out of curiosity, in what year would you study those topics on that paper? How old would the students taking that exam be?
Entrance exam for a selective Indian university. (or Mathematical Institute, as they call themselves) The students taking the exam are generally 17-18. They take it around April, then enroll by August. (or at least, around that time)

 Quote by Phyisab**** Yes, the average person with a PhD in physics would be able to solve that no problem.
I'd have to spend an hour remembering some formulas, but I think I can do it easily if it was open book.

 Recognitions: Gold Member Homework Help If you asked me right after I passed my quals and was still a qual problem solving zombie, I think I would have been able to solve it 5 different ways and give you a lecture on the benefits of one method versus the other! Now, I may need to look up a few formulas, but otherwise, I think I could solve it no problem.
 I know this is almost a year later, but I thought I would state what I think the professor was trying to get the student to understand, or at least what I think is the most important part of the problem. Let C1 be the capacitance just of the region directly above the dielectric, C2 the capacitance just of the dielectric region, and C3 that of the remaining region of the capacitor. Then, the problem is equivalent to C1 and C2 in series, then this pair connected in parallel to C3. That is: C = (1/((1/C1)+(1/C2)))+C3. Each of C1, C2 and C3 is given by its own appropriate version of the parallel-plate capacitance. The derivation of the parallel-plate capacitance is a standard result (c.f. Eqn. 2.54 page 105 of Griffiths).