Two Jobs, Done a Million Times a Second
By now you know why we want a ring: a synchrotron reuses the same stretch of accelerating hardware lap after lap, so a modest push given thousands of times can add up to colossal energy. But that picture hides a quiet demand. A particle moving in a straight line does not magically curve back; left alone it flies off tangent and smashes into the wall. So a ring really only asks two things of its hardware, and it asks them relentlessly. Bend the beam so it returns to where it started, and push the beam so each lap leaves it with a little more energy than before.
These two jobs split cleanly by physics, and that split is the master key to the whole machine. A magnetic force always pushes *sideways* to the motion — it can turn a particle but never speed it up, because it never pushes along the direction of travel. An electric force, by contrast, pushes *along* whatever direction you point it, so it can add energy. So magnets do all the steering and focusing, and electric fields do all the accelerating. Recall from the earlier guide that this is exactly why we use radio-frequency acceleration rather than a single static voltage: the pushing has to be split into many small kicks spread around the ring.
Dipoles Bend, Quadrupoles Focus
The simplest magnet is a dipole — a north pole above the beam, a south pole below, so the field points straight down across the pipe. A charged particle flying through such a uniform field feels a constant sideways shove and curves along a smooth arc, exactly like a stone whirled on a string. String enough dipoles in a circle and the beam threads from one arc to the next, tracing out the ring. The LHC needs over a thousand of these bending magnets, each fifteen metres long, just to close the 27-kilometre loop. This is the workhorse role named by dipole and quadrupole magnets: the dipole is the steering wheel.
But bending alone is not enough. A real beam is a swarm of millions of particles, each with a slightly different angle and energy, so the swarm naturally spreads out — within metres it would smear across the pipe and be lost. We need to *focus* it, like a lens squeezing a flashlight beam to a spot. That is the job of the quadrupole, a magnet with four poles arranged in a square. Its field is zero on the exact centre line and grows stronger the farther out a particle strays, so it pulls wandering particles back toward the axis — a restoring force that keeps the swarm tight.
There is a lovely catch, though: a single quadrupole that focuses *horizontally* always *defocuses* vertically, and vice versa — it squeezes in one direction while letting the beam fan out in the other. The trick, borrowed straight from optics, is to alternate them: a horizontally focusing quadrupole, then a vertically focusing one, then another, and so on around the ring. The net effect of an alternating series is focusing in *both* planes, a scheme so central it has its own name ("strong focusing"). So the standard layout of a ring is a repeating pattern — bend, focus, bend, focus — dipoles doing the steering, quadrupoles keeping the swarm from blowing apart.
Why the Magnets Have to Be Superconducting
How hard you can bend a beam comes down to one number: the strength of the magnetic field. The faster and more energetic the particle, the stiffer it is to turn, so for a ring of fixed size, higher energy demands a stronger field. There is a clean rule of thumb worth carrying: a proton's momentum, the ring's radius, and the field all lock together. Squeeze a 7-TeV proton into the LHC's tunnel and you need a field of about 8 teslas — well over a hundred thousand times Earth's magnetic field, and far beyond what an ordinary iron electromagnet can deliver.
A field that strong takes an enormous electric current running through the magnet's coils. In ordinary copper wire that current would meet electrical resistance, heat the wire red-hot, and waste power on a scale no laboratory could pay for. The way out is a [[superconducting-magnet|superconducting magnet]]: certain metal alloys, cooled to a few degrees above absolute zero, lose *all* electrical resistance and carry vast currents with no heating at all. The LHC's dipoles are wound from a niobium-titanium alloy and bathed in superfluid helium at 1.9 kelvin — colder, famously, than the deep space between the stars. Only then can a current of over 11,000 amps loop through the coils year-round to hold up that 8-tesla field.
This is why field strength, not tunnel size, is the real frontier. If you cannot dig a bigger ring, the only way to higher energy is stronger magnets — and that pushes superconductor technology to its limits. Proposed future machines aim for 16 teslas using niobium-tin, a more brittle and demanding material. It also explains a sober operational fact: if any tiny patch of coil warms up and stops superconducting, the stored magnetic energy dumps into that spot in an instant. Such a "quench" must be detected in milliseconds and the energy safely drained, or the magnet can destroy itself — a real 2008 incident delayed the LHC's start by more than a year.
The Push: Radio-Frequency Cavities
Now for the pushing. The energy comes from an [[rf-cavity|radio-frequency cavity]] — a hollow metal chamber, shaped like a series of bells, in which a powerful oscillating electric field is set up by feeding in radio waves. As a particle crosses the gap at just the right instant, the field points forward and gives it a kick. By the time the field has flipped to point backward, the particle is gone and the next one has not yet arrived. The cavity's oscillation is precisely tuned so that every passing particle always meets a forward push — this exact timing is the heart of RF acceleration.
This timing has a profound side effect: it explains why the beam is not a smooth continuous stream but a train of discrete clumps. Only particles that arrive in the narrow window of forward push survive; one arriving slightly early or late gets a smaller kick — or even a nudge that pulls it back toward the right moment. The cavity therefore *herds* particles into tidy packets called bunches, each riding the crest of the RF wave like a surfer. This is exactly the bunch structure described under beam bunches and the beam pipe: the LHC's beam is some 2,800 bunches, each a few centimetres long, spaced out around the ring.
Like the bending magnets, the most demanding RF cavities are also superconducting — chilled niobium chambers that store the radio-wave energy with almost no loss, so a modest input power can sustain a strong field. And the numbers are surprisingly gentle. The LHC's cavities raise each proton's energy by only about 0.5 MeV per turn, yet a proton loops the ring some 11,000 times a second. A few thousand turns of patient nudging carries it from injection energy all the way to 7 TeV — the ring's whole reason for being, paying off one small kick at a time.
Synchrotron Radiation: The Tax and the Gift
There is a price for all that bending, and it is dictated by physics, not engineering. Any electric charge that is *accelerated* radiates electromagnetic waves — and turning a corner counts as acceleration, because the velocity is changing direction even if its speed is steady. So every time the dipoles bend the beam, the particles bleed off energy as light. This glow is [[pp-synchrotron-radiation|synchrotron radiation]], and it is the standing tax a circular machine must keep paying.
The cruel part is how steeply the tax rises. The energy lost per turn grows as the *fourth power* of the particle's energy, and is far worse for light particles than heavy ones — it scales as one over the mass to the fourth power. An electron is about 1,800 times lighter than a proton, so at the same energy it radiates roughly 10 trillion times more fiercely. This single fact is why the LHC collides protons, not electrons: an earlier electron ring in the very same tunnel topped out near 100 GeV, its RF cavities barely able to refill the energy the beam was hosing away as light. Protons, being heavy, hardly radiate at all and can climb seventy times higher in energy.
And yet the very same loss is one of the most useful tools in modern science. The light a bent beam throws off is brilliant, finely tunable, and pours out in an exquisitely narrow forward cone — a consequence of the relativistic near-light-speed motion we met early in this ladder, which sweeps the radiation forward like a searchlight. Whole laboratories, called synchrotron light sources, are built *on purpose* to bend electron beams and harvest this radiation, then use it to film proteins, map materials atom by atom, and probe chemistry in real time. What is a nuisance to the particle physicist is a precision X-ray flashlight for everyone else.
Keeping It All in Step
We can now close the loop on the word "synchrotron" itself. Recall the bending rule: for a fixed ring radius, the magnetic field needed to hold a particle on its circle grows in lockstep with the particle's momentum. But the RF cavities are *raising* that momentum on every single turn. So if the dipole field stayed fixed, a beam that gained energy would immediately need a tighter bend than the magnets could provide, and it would spiral outward into the wall.
p [GeV/c] ~ 0.3 * B [tesla] * r [metre] (bending rule) B ramps UP in step with p as the RF cavities push => the beam stays on the SAME radius the whole time
The answer is in the name. As the cavities accelerate the beam, the operators ramp the dipole field *up in synchrony* — turning up the magnet current at exactly the rate that keeps the bend matched to the growing momentum, so the beam holds the same fixed radius the whole way up. This delicate duet between rising field and rising energy is what "synchro-" means. The injection process from the previous guide hands the ring a beam already at a sensible starting energy; from there the magnets and cavities ramp together, in a choreography held to a fraction of a percent, until the beam reaches full energy and the cavities switch to merely topping up what synchrotron radiation and other small losses nibble away.