aspodkfpo said:
? How am I thinking about this wrong? E = V/d, d decreases and E becomes larger.
There is no more "d" after the central plate is inserted. There are two smaller "d"s, each smaller than d/2 since the central plate has finite thickness ##\delta##.
The fact that the E field does not change can be ascertained by showing that the charge densities on each inside face = Q/A or -Q/A, A = plate area, same as before, so that ## Q/(\epsilon_0 A) = ## E, the same before & after insertion.
Remember that we are ignoring fringing effects so if you know ##\sigma## on one inside surface you know that the E field is ##\sigma/\epsilon_0## all the way from one inside surface to the opposing inside surface. Plate separation enters the problem only if you want to compute potentials, which we don't need to (but just FYI the new V changes from Ed to E(d-##\delta##). )
So anyway you end up with the same E field but with smaller volume, changing from dA to ##(d-\delta)## A.
EDIT: Or you can go as follows:
Let unprimed = before 3rd plate insertion
primed = after insertion
Then,
## Q'=Q \Rightarrow C'V' = CV. ##
## C = \epsilon_0 A/d ##
## V = Ed ##
## C' = \epsilon_0 A/(d-\delta) ## (2 capacitors in series)
## V' = 2E'(d-\delta)/2 = E'(d-\delta) ##
So ## C'V' = CV \Rightarrow E'=E. ##