Why a proton beam has many collision energies

The beam energy fixes the energy of each incoming proton. The short-distance collision usually involves one quark or gluon from each proton, each carrying only part of its parent’s momentum. A proton collider therefore samples many partonic collision energies in one run. To predict a rate, we must combine the proton’s momentum distribution with the probability of the hard interaction.

A fraction of each proton enters the hard collision

A quark or gluon treated as the incoming constituent is called a parton. Let $x_1$ and $x_2$ be the longitudinal momentum fractions it carries from the two beams; each lies between zero and one. At high beam energy we neglect the proton mass in this kinematic relation. If $s$ is the squared proton-proton centre-of-mass energy, the squared partonic energy is

This relation already explains why the full beam energy is not available to every hard subprocess. It also explains why the same final-state mass can be produced with different motion along the beam: a larger fraction from one proton and a smaller fraction from the other keep the product fixed.

For a narrow state of mass $M$ produced without recoil at leading order, energy-momentum conservation gives the fractions in terms of its rapidity $Y$. Rapidity is an additive measure of longitudinal motion; the next jet chapter will define it from energy and momentum.

Work out the fractions for a heavy state

Take a hypothetical state of mass 1 TeV in proton collisions at 13 TeV. At zero rapidity both fractions are about 0.0769. At rapidity one they become about 0.209 and 0.0283. Their product is unchanged, so the partonic invariant mass remains 1 TeV. These numbers describe a leading-order, zero-recoil example; energetic additional radiation changes that simple reconstruction.

Now ask whether a rapidity of three is possible in the same example. The first fraction would exceed one. A constituent cannot carry more than the available proton momentum in this collinear setup, so that kinematic point is inaccessible. This is a useful check before calculating any cross section.

Factorization is a weighted sum over incoming possibilities

The parton distribution function $f_a(x,\mu_F)$ describes the distribution of parton species $a$ at a chosen factorization scale $\mu_F$. At leading order it admits a useful density picture; beyond leading order it depends on the factorization convention and should not be treated as an ordinary detector-measured probability curve. The scale specifies how the calculation separates proton-collinear radiation from the hard coefficient.

An integral is the continuous counterpart of a weighted sum. Divide each momentum-fraction axis into small intervals, calculate the hard cross section for each pair, multiply by both densities and interval widths, then add all pairs and species. Shrinking the intervals gives

The two scales organize different parts of the calculation

The hard coefficient $\hat\sigma_{ab}$ has area units. The fractions and the integration measures are dimensionless. The renormalization scale $\mu_R$ sets the scale at which the strong coupling is evaluated. Both scales are calculation choices, not adjustable detector settings. With a complete all-orders prediction their dependence cancels; truncating the perturbative series leaves residual dependence.

What the scale variation can teach us

Suppose two parton-distribution choices change a high-mass prediction and a scale variation changes it again. These variations probe different ingredients: proton structure and omitted perturbative terms. Their uncertainty treatment must say which variations are correlated between bins. Moving every bin independently would discard the smooth changes that a theory calculation actually predicts.

Factorization applies under stated conditions, including an appropriate hard scale and a suitably defined observable. Very restrictive measurements can require additional theoretical structure. The event display illustrates kinematics. Incoming distributions, a parton shower and hadron formation require separately defined calculations.

A task before the next chapter

Use the formulas for the hypothetical 1 TeV state to make a small table at rapidities zero, one and two. Record both fractions and their product. Sketch which proton supplies the larger fraction in each case. The collider Lab supplies a picture of the final-state event; the table is the calculation to compare with that picture.

Next, imagine adding a very low-energy particle to the final state. Should a carefully chosen observable change abruptly? That question determines which final-state quantities can be compared with a perturbative prediction.

Exercises and answers

Exercise 1. Derive the largest allowed magnitude of rapidity for a zero-recoil state of fixed mass.

Answer. Apply the condition that each momentum fraction is at most one. Both inequalities give $|Y|\leq\ln(\sqrt{s}/M)$. For the example the upper magnitude is about 2.56. Near this boundary the available parton density is also strongly suppressed; kinematic permission does not imply a large rate.

Exercise 2. A prediction changes when its factorization scale is varied. Can the change be interpreted as a new detector systematic?

Answer. No. It belongs to the theoretical approximation for the incoming structure and hard process. Detector response is a separate transformation of physical final states into selected records. Both can alter a measured spectrum, which is why their roles must be kept explicit in a fit.

Exercise 3. If the beam energy doubles at fixed final-state mass and zero rapidity, what happens to each required fraction?

Answer. Each fraction halves. The parton densities and hard cross section then determine how many events are produced at these new fractions.