Why are gluons & the strong force necessary to keep proton together? This is something that's puzzled me for a while. Say you have a proton with quarks uud, with the u quarks having +2/3 charge and the d quark -1/3. Doesn't this make quite a neat fit for keeping the quarks confined within the proton, in terms of the electrostatic forces between the three quarks. At what point, and for what reason did it become necessary to invoke gluons to provide the extra 'colour charge' for ensuring quark confinement. Why is the ordinary electrostatic force not enough?
 A: The uncertainty principle tells us that the more tightly you confine an object the greater the uncertainty in its energy. For example you can use this in a hand waving sort of way to calculate the size of a hydrogen atom from its ground state energy or vice versa.
And when we do this we find the electrostatic force is simply not strong enough to confine the quarks inside a nucleus. If we had three quarks interacting just by the electrostatic force we'd get an object roughly the same size as an atom. The only way to get a bound state as small as a nucleus is to have a vastly stronger force binding the quarks together. That vastly stronger force is of course the strong force.
A: There is another issue related to the need for color charge and the associated gauge fields.
I'll quote from section 9.1 The colour degrees of freedom of "Gauge Theories in Particle Physics":

For a baryon made of three spin-1/2 quarks, the original
  non-relativistic quark model wavefunction took the form
$$\psi_{3q} = \psi_{space}\psi_{spin}\psi_{flavour}\qquad\qquad (9.1)$$
It was soon realised (e.g. Dalitz 1965) that the product of these
  space, spin and flavour wavefunctions for the ground state baryons was
  symmetric under interchange of any two quarks.
...
But we saw in $\S 4.5$ that quantum field theory requires fermions to
  obey the exclusion principle - i.e. the wavefunction $\psi_{3q}$
  should be antisymmetric with respect to quark interchange.  A simple
  way of implementing this requirement is to suppose that the quarks
  carry a further degree of freedom, called colour, with respect to
  which the 3q wavefunction can be antisymmetrised, as follows.
...
With the addition of this degree of freedom, we can certainly form a
  three-quark wavefunction which is antisymmetric in colour by using the
  antisymmetric symbol $\varepsilon_{\alpha\beta\gamma}$, namely
$$\psi_{3q,\mathrm{colour}} =
 \varepsilon_{\alpha\beta\gamma}\psi^\alpha\psi^\beta\psi^\gamma$$
and this must be then be multiplied into (9.1) to give the full 3q
  wavefunction.

