What makes colliding neutrons bounce off one another?

[andrewkirk]

When two free neutrons collide at high speed do they ricochet off one another like billiard balls?
If so, what is the force that they exert on one another in order to change their velocities?
It clearly isn't gravitational or electromagnetic.
The only other fundamental forces are the strong and weak nuclear forces. Is it one of those?
If not, is it some other principle (eg Pauli exclusion principle) that somehow circumvents the need for a force to explain the ricochet?

 

[bcrowell]

You need both the strong nuclear force and the exclusion principle to give a good account of neutron-neutron scattering. At moderate impact parameters, the force is going to be attractive, so it's not just like billiard balls.

 

[Bill_K]

And electromagnetic forces also. Neutrons have a magnetic dipole moment.

 

[andrewkirk]

Thank you bcrowell and Bill_K for the answers. May I ask supplementary questions?

1. I can imagine how a neutron would have a dipole moment from being made of three quarks, two with charge -1/3 and one with charge +2/3. Does it make sense to think of these quarks being arranged at the vertices of an equilateral triangle, so that the -ve pole is between the two negative quarks and the +ve pole is at the +ve quark, or are quarks not arranged in a geometric configuration within a neutron?

2. Given that the exclusion principle can accelerate the neutrons and they have mass, it would appear that the neutrons impart equal and opposite forces on one another which, when you subtract the effect of the four fundamental forces, leaves you with a pair of net residual forces that are explicable only in terms of the exclusion principle. Why is this principle not included as a fundamental force along with the other four? Is it just a choice of terminology arising from the historical development of our understanding of these issues or is there a way of interpreting it to not be a force?

 

[Bill_K]

andrewkirk, Protons and neutrons are rather complicated things. Although it is well known that they contain three quarks, this is only a rough description. These three quarks are called "valence quarks". But in addition inside the nucleon you will find virtual particles: gluons, and quark-antiquark pairs. No one has yet succeeded in calculating the nucleon's properties from first principles and the properties of its constituents.

From your question it sounds like you're thinking of an electric dipole moment rather than a magnetic dipole moment. Although the neutron may indeed have an electric dipole moment, it is expected to be very small and has not yet been seen. One reason the neutron has a magnetic dipole moment is that the quarks themselves carry intrinsic magnetic dipole moments.

Your picture of the quarks at the vertices of an equilateral triangle is not the way quantum systems work. They can't be said to occupy specific positions. In crude terms the quarks form a cloud, like the electrons do in an atom.

Finally, about the exclusion principle. Although it requires identical particles to remain apart, and may effectively cause an increase in their total energy, it can't really be interpreted as a force. If both particles can't occupy the same state, that just means that one of them must occupy the next state. And the system is what determines the energy of the next state, not the exclusion principle.

 

[andrewkirk]

Thanks for your explanation Bill. Every increase in my understanding of these issues is a joy.

 

[daschaich]

No one has yet succeeded in calculating the nucleon's properties from first principles and the properties of its constituents.

Not analytically, of course -- but such calculations are the point of lattice QCD.

 

[Bill_K]

Not analytically, of course -- but such calculations are the point of lattice QCD.

And they have not yet succeeded.

 

[daschaich]

And they have not yet succeeded.

This is dead wrong.

My initial response to this error was more to the point and seems to have been removed without notice by the powers that be for that reason. So allow me to elaborate without dragging this thread too far off topic.

Lattice QCD is a solved problem. The systematic effects of working on the lattice are well understood, well under control for many calculations, and straightforwardly addressable for most others. Any question you care to ask about nucleons, other hadrons, their structure or their interactions can be answered through lattice QCD, given sufficient computing power. Because "sufficient" computing power is extreme computing power, most work in lattice QCD focuses on improving the calculations possible with currently-available resources.

Unless you are considering exact closed-form expressions as your metric of success, the lattice approach has already succeeded, and continues to improve.

 

[Bill_K]

Efforts to calculate the way that nucleon spin is distributed between quarks, gluons and orbital angular momentum have not succeeded. This is a fundamental property of the nucleon and remains a major outstanding problem. You may consider lattice QCD "solved" and only lacking in computing power, and perhaps that is so. Nevertheless it has not yet yielded the answer to this question.

 

[daschaich]

Fair enough, that's not my field of expertise (though I vaguely recall some debate regarding whether or not such a decomposition is really well-defined, i.e., unique and gauge-invariant).

Quick Google searches give me plenty of lattice results on the spin structures of hadrons. For instance
http://arxiv.org/abs/0903.4080
very briefly reviews the quark spin and total quark contribution to the nucleon's spin, as well as moments of transverse spin densities.
http://arxiv.org/abs/0912.5483
provides considerably more information, none of it suggesting that there is any fundamental obstacle to these calculations.