I'm guessing that when the LHC ramps up to 4000 GeV this means they are increasing the current in the superconducting magnets as RF fields accelerate the beams. Where does this current go when they ramp down? Is it dissipated as heat? Is it fed back into the grid?

P.S.- by 'ramp' I mean the normal increasing/decreasing of the current in the magnet at the beginning and ending of a fill. It builds gradually (ramps) and decreases gradually. The whole time the magnet is still superconducting.

  • $\begingroup$ Typically you don't ramp down ("quench") electromagnets that powerful, since it takes a decent amount of time to bring them up in the first place. $\endgroup$ – Ignacio Vazquez-Abrams May 15 '12 at 15:49
  • 2
    $\begingroup$ @IgnacioVazquez-Abrams Quench generally implies a unplanned, fast failure of the superconductivity of the magnet. Quenches are bad. Ramping magnets down is a fairly common occurrence and is performed under control. Though I don't actually know the answer to the question. My experience is that you go to the magnet control GUI, set the desired current and push the "RAMP" button... $\endgroup$ – dmckee --- ex-moderator kitten May 15 '12 at 16:42
  • $\begingroup$ I think that is what ramp means. At the end of the fill the beam is dumped and then the magnets are ramped down. $\endgroup$ – anna v May 15 '12 at 19:31

This link might enlighten you, also this.

The cryogenic technology chosen for the LHC uses superfluid helium, which has unusually efficient heat transfer properties, allowing kilowatts of refrigeration to be transported over more than a kilometre with a temperature drop of less than 0.1 K. LHC superconducting magnets will sit in a 1.9 K bath of superfluid helium at atmospheric pressure. This bath will be cooled by low pressure liquid helium flowing in heat exchanger tubes threaded along the string of magnets. The LHC cryogenic system is very large as well as very cold. Refrigeration power equivalent to over 140 kW at 4.5 K is distributed around the 27 km ring. To save costs, the four existing LEPII 12 kW, 4.5 K cryoplants will be reused. Their cooling power will be increased by 50% and 1.9 K stages will be added. In all, LHC cryogenics will need 40,000 leak-tight pipe junctions. 12 million litres of liquid nitrogen will be vaporised during the initial cooldown of 31,000 tons of material. The total inventory of liquid helium will be 700,000 litres.

Unfortunately the FAQ page at the CERN page is out of order at the moment and cannot be linked. Here is the LHC machine outreach page.

During normal beam dumps there is controlled quenching, using the systems set up for emergency quenching when there are problems with a magnet. Here is a book on LHC like magnet design.

  • $\begingroup$ I read the article and didn't find any reference to a normal shutdown of the magnets. It says that the beam dump system is the same in a normal dump as in a quench. Or did I misread it? Certainly other superconducting magnets systems I've worked with had the ability to ramp down in a controlled way. $\endgroup$ – dmckee --- ex-moderator kitten May 15 '12 at 17:53
  • $\begingroup$ @dmckee I think because of the danger of uncontrolled quenches the system is over-designed as far as the normal ramp down, the heat absorbing mechanisms are there already. A TeV beam is much more penetrating that a GeV one and the danger of a quench if some of it hits a magnet is exponential I would think. $\endgroup$ – anna v May 15 '12 at 18:01
  • $\begingroup$ Paging through the big PDF you linked seems to confirm that. Thanks. $\endgroup$ – dmckee --- ex-moderator kitten May 15 '12 at 18:03
  • 1
    $\begingroup$ The dimensions of CERN are still unbelievable to me. It takes here a day to cool down less than 50 kg but 31000 tons that is incredible. Thank you for sharing that bit. $\endgroup$ – Alexander May 15 '12 at 23:36
  • $\begingroup$ @Alexander: I know what you mean. At JLAB we occasionally had to argue with the other halls over a few grams per second of excess cold helium to keep our magnets and cryotargets cool. Admittedly the accelerator used much larger quantities for the klystrons, but we got used to thinking in those quantities. $\endgroup$ – dmckee --- ex-moderator kitten May 16 '12 at 1:22

"Each dipole magnet is connected to 153 neighbors, and their energy also has to be immediately removed. A switch sends the energy into large resistors, where it heats eight tons of steel to a temperature of 300°C (570°F) in less than two minutes."

There are actually deliberate heaters in the magnet windings. So if a point quench is detected it rapidly heats the whole magnet out of superconducting and the energy is spread evenly.


  • $\begingroup$ Like I noted in a comment on the question: a normal operational ramp down is different from a quench. $\endgroup$ – dmckee --- ex-moderator kitten May 15 '12 at 17:21
  • $\begingroup$ @dmckee - I assumed that you still connected the magnets to dump resistors - where else would the energy go? $\endgroup$ – Martin Beckett May 15 '12 at 17:52
  • 1
    $\begingroup$ I don't know. Perhaps with the LHC dipoles it makes sense to dump it all every time, but many superconducting magnets want to be able to adjust the field up and down. In any case I would be very surprised to learn that a normal ramp down (even a complete one) involved heating the magnets. You generally like to keep the superconductors cold rather than cycling them. $\endgroup$ – dmckee --- ex-moderator kitten May 15 '12 at 17:57

I am not familiar with the exact design of the LHC magnets. For 'normal' superconducting magnets the procedure to ramp up or ramp down the field is quite similar:

  1. You ramp up/down the power supply to match exactly the current that is at the moment in the superconducting magnet.
  2. The switch heater is turned on and will create a small portion of the coils that is above the critical temperature. Now you have a large superconducting coil which is shunted by a normal resistor and the power is supplied externally.
  3. The external supply is ramped up/down to the target field. This can take some time, as the magnet will have a very large inductance and you don't want to heat it up by vortex movement.
  4. At the target field the switch heater is deactivated, the whole magnet becomes superconducting again.
  5. The external power supply is ramped down and shut off.

This is the normal mode of operation, if there is an emergency, i.e. a quench the energy is dissipated rapidly and Martin Beckett's post explains where that energy flows into (in the ideal case, otherwise the magnet can blow up, which happened at CERN earlier).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.