0

Can someone prove that the minimum distance between two non-intersecting parabolas is along their common normal (without calculus)? (If I understand how to prove it, I'll prove it myself for other curves too.)

Archer
  • 6,051

1 Answers1

3

Assuming the existence of a shortest distance, let the shortest distance join points $A$ and $B$ and construct a circle with diameter $AB$. If the circle cuts either parabola at a point other than $A$ or $B$ there is a line joining the two curves which is shorter than the diameter. Else the circle must be tangent at points $A$ and $B$ and the diameter $AB$ is normal to both curves (because the diameter is perpendicular to the tangent).

There are some implicit assumptions here, but would this do? Note that if $A=B$ then the diameter doesn't have a fixed direction and the situation is different.

Mark Bennet
  • 100,194
  • You don't explain how to find points $A,;B$ – Raffaele Oct 23 '17 at 13:11
  • 1
    @Raffaele No - I assume their existence (see very first comment in answer) and show that if they exist they are joined by a common normal. (Implicit assumptions include that the parabolas do not cross). If there is a shortest distance, there are points which are that distance apart. – Mark Bennet Oct 23 '17 at 14:01