This problem has been giving me all sorts of fits. For one, Taylor states that because the frictional force and normal force are forces of constraint, they produce no work. I'm trying to figure out the right way to think about why friction does no work in this case.
My initial rationalization was that since the friction produces a constraint (i.e. no slipping), its functional form at each moment in time will exactly ensure that the constraint is satisfied, while avoiding any contribution to the kinetic energy.
While trying to couch this in analytical terms, I figured that since dW=F⋅dr, I can also say that dW=F⋅(dr/dt)dt=F⋅vdt. Since the instantaneous velocity at the point of contact is 0, this would give dW=0 in the case of friction. Is this a good argument though?
My second issue was how Taylor equated the length of CB with r$d\theta$. I wrote out my thoughts on why this is possible in the picture, but if someone has an easier way to see this, or if my rationale is flawed, I would love some input.