Banner
    New Luminosity Record At The Tevatron
    By Tommaso Dorigo | April 17th 2010 03:03 AM | 25 comments | Print | E-mail | Track Comments
    About Tommaso

    I am an experimental particle physicist working with the CMS experiment at CERN and the CDF experiment at Fermilab. In my spare time I play chess...

    View Tommaso's Profile
    This is to inform you of the new luminosity record set today by the Tevatron collider at Fermilab. The machine has been working excellently, improving its performance as the machinists found ways to obtain higher stacks of antiprotons, reducing inefficiencies in the transport of the beams from one accelerator to the other in the injection process, or finding better beam tunes. A painstaking work that brought increasing returns, it seems.

    Instantaneous luminosity is a measure of the number of collisions of the colliding beams. It basically depends on the number of particles contained in "packets" circulating in the two opposite directions, the rate at which the packets intersect their orbits in the core of the detectors, and the transverse size of the region where the intersection occurs. One gets higher instantaneous luminosity by managing to get more particles in stable orbits, by increasing the rate of their crossing, and by squeezing the beams in the points where they cross.

    Instantaneous luminosity may be labeled with the letter L. It is measured in units of units per area per time, such that a higher L means that more particles are crossing a given area in a given time. If luminosity is high, so is the rate at which interesting events are collected by the experiments.

    Let me mention what was the record that the Tevatron broke today. The instantaneous luminosity it reached was . At that luminosity, and with the energy of the collisions, about ten W bosons are produced per second. And W bosons are the starting point for many exciting searches of new physics, as well as signatures of possible Higgs boson decays, top quark decays, and a stringent probe of the standard model.


    Above is the up-to-date graph of the instantaneous luminosity performance of the machine. Note that the new point has not been added to the graph yet. And note that they will now need to change the vertical scale to include it -expressed in inverse microbarns (the units of the vertical scale), the new result is 404, an 8% increase from the previous record! The upward jump was quite significant and unexpected -by all but the machinists who made it possible, of course.

    With the Tevatron running so well, one wonders whether the US Department of Energy might reconsider their decision to stop the machine at the end of 2011. I have written elsewhere that after delivering 10 inverse femtobarns of collisions to the CDF and DZERO experiments, it would not make much sense to run for one additional year - the increase would just be of 20% of data. But what if they ran for three more years at this unprecedented luminosity ?









    Comments

    Actually I am stunned at such a jump. We certainly were not trying for a record, so we need to understand the what led to it - and try to repeat it. It certainly caused a lot of excitement!

    dorigo
    Maybe the CLC is the cause :D
    Cheers,
    T.
    Neat data!
    Given the increased luminosity, how much extra slack does the Tevatron have in terms of magnet strength, or power to accelerator magnets? Is there a safety or engineering reserve built in? The thing is, at the end of 2011 run could you try to go to higher beam energies, losing luminosity and increasing the risk of failure? So what if something breaks if it is going to be shut down anyway. You could be at max limits anyway currently, but engineering wise there is always some reserve. Could you get to say 1.5 TeV per beam or 1.2 or something losing 80% luminosity? Would the data collected be worth it or would it be worth it in engineering data as you pushed the thing beyond its limits?

    Just a few thoughts. We always had good learning experiences pushing old well understood medical equipment to and past its limits back in the day.

    lumidek
    Congratulations, increasing the record by 8% for a while is impressive. Now, if you also increase the energy by 250%, you may start to compete with the LHC.
    dorigo
    Hi Markk, Lubos,

    I do not think there is any margin to increase the beam energy of the Tevatron. That is: everything can be done, if you throw enough money at it. The issue is whether it is worthwhile, and I think it is not.

    Cheers,
    T.
    lumidek
    Of course there's no big sense. I was just hinting that the Fermilab's increases in luminosity won't help it much to compensate for the energy gap that makes it by far the weaker among the 2 large colliders. The right question is whether it really makes sense to continue to run the Tevatron now.
    dorigo
    Yes, I agree. And in fact I think the consensus on that one question is: until the LHC has clearly surpassed the Tevatron reach on all the interesting physics searches, the Tevatron collider experiments should not be decommissioned.

    Cheers,
    T.
    No, we could not raise the Tevatron energy very much at all at this point. Higher energy means higher electrical current in the magnets which requires more liquid helium to cool them and prevent them from quenching (becoming non-superconducting); we already push the cryogenic system close to its limit. Running at 1.0 TeV rather than 0.98 TeV per beam may be possible for a short time, but likely not for extended periods. Bigger jumps to 1.2 or 1.5 TeV per beam as suggested just isn't possible with the Tevatron. (Note that the Tevatron energy was raised from 0.90 to 0.98 TeV between Run 1 and Run 2 in part by shuffling magnets around to put the weaker magnets in the "colder" areas of the machine where the cryogenic system is more efficient.)

    dorigo
    Hi Ron,
    thanks for this information. I actually did not know some of these details.

    Cheers,
    T.
    Hfarmer
    That makes logical sense.  It was engineer'd for a certain energy by physicist.  What one commenter was suggesting is that a common engineering practice may have been done when it was designed.  That being to design a system that has a 15-20% margin for safety built in.  To over engineer the system.  
    It's still a lovely thought.  Sending such a wonderful machine out with a bang of sorts. 
    Science advances as much by mistakes as by plans.
    Being a native of Chicago, naturally I have personal pride in the accomplishments of Fermilab. I have actually had the pleasure of visiting the facility. It is quite fascinating the work that they do there.
    dorigo
    Chicago is a wonderful city. I spent about four years of my life there, integrating over the 18 years of my participation in the CDF experiment.

    Cheers,
    T.
    Tommaso, would you consider doing a future post, if only a quick one, explaining how the Tevatron circulates counter-rotating beams of p / p-bars within the SAME BEAM TUBE? As a non physicist, I understand that that this is possible because of the way their opposite electric charges "see" the magnetic field of the superconducting dipoles, but I do not understand how exactly you keep them from colliding at all points around around the ring. I hear things about "pretzel orbits" and such, but all the literature seems to be far beyond the layman (ie. meee) and I think it would be interesting to understand how this clever hack is done (was it an afterthought of the original design?) and how stuff like beam/beam interaction instabilities are suppressed when such high energy particle beams are passing so close to one another in the same tube. Just a thought.

    dorigo
    Hi Blake,

    it would require me some work to put together a post like the one you have in mind. I don't see it happening in the near future.
    Instead, let me see if I can answer here sufficiently well in short. The beam pipe of the Tevatron has, in the core of the detectors, a width of two inches (elsewhere it may be different, but that is the order of magnitude anyway). The beams are "squeezed" when they cross in the detectors by dedicated magnets, called "low-beta quadrupoles". These ensure that the beam presents, at interaction, a wider spread in momenta and a smaller spread in transverse coordinates. The resulting beams have a width of a few tens of microns. They are brought to intersect at a very small angle, which guarantees a "luminous region" where interactions happen between the packets of protons and antiprotons along a longitudinal area spanning about 30 centimeters. This is an optimum to maximize the number of collisions without getting them too far out from the center of the detector, where event reconstruction is most efficient.

    Away from the interaction points, the beams travel separated in the same vacuum tube. They are like two electrical currents traveling in opposite directions in paired wires.  Of course there are beam-beam interactions, but there are as well beam interactions with the walls of the accelerator too. The whole system is strongly coupled and finding the proper orbit is a very compex task. But it can be done.

    Cheers,
    T.
    Let me jump back in here - the beams collide head-on in the Tevatron (no crossing angle). The beams are kept separated by using "electrostatic separators" spread around the ring. The separators are essentially parallel-plate capacitors with a large voltage across the plates. When the beams pass through that electric field, the protons get kicked one way and the antiprotons get kicked in the opposite direction (because of their opposite electric charges). These kicks cause the beams to spiral around each other in a double-helix pattern as they go around the ring (we call it "the helix"). We can turn on/off collisions at CDF and D0 by adjusting some of the separators.

    The beams only collide at CDF and D0, but there are numerous "parasitic crossings" around the machine where long-range beam-beam effects do disrupt the beams a little and cause the lifetime to diminish. It's a delicate balance between the physical aperture of the beam pipe and how hard we can operate the separators before they spark between plates and cause problems.

    I have been thinking about adding a discussion on separators to the Tevatron's Facebook page. Stay tuned.
    In the meantime, I will send Tommaso a schematic image of how a separator works - maybe he can post it here since I can not.

    dorigo
    Very good, I will be glad to do it. You should be able to post it yourself though.

    About the crossing angle: Are you sure it is exactly zero ? I seemed to remember a small but nonzero value. Otherwise I got confused with LHC.

    Cheers,
    T.
    Hank
    Very good, I will be glad to do it. You should be able to post it yourself though.
    He just has to create an account and he can post anything he wants.  

    Idaho National Lab, for example, has an account here and when one of their feature writers creates something they want to put out to a broader audience, they post those here also.

    Facebook is the format du jour but in terms of actual attention, being here is 4X better than Facebook and 25X better than Twitter.
    Yes, it's zero at the Tevatron for 36x36. If we had implemented the 132 ns bunch spacing we would have needed a small crossing angle. For normal 36x36 operation, the nearest parasitic crossings are ~57m away from the interaction points outside of the straight sections.

    dorigo
    Aha. Thanks for giving me a honorable way out ;-)

    Anyway, for the sake of everybody else reading this thread: here is a schematic from Ron Moore showing the electrostatic separators, that divert the beams using 100 kV electrostatic plates:


    I think the drawing is quite clear... If you have questions, please ask.

    Cheers,
    T.
    Cool! Thanks you guys! :)) Kindof funny that using 8 foot long plates charged to over a hundred kilovolts you only get like a thousandth of a degree of deviation in the beams. I guess protons at 1 TeV are going pretty fast after all!

    Given the continued problems at the LHC I would think it worthwhile to run the Tevatron at least one additional year. T what is the approximate cost? And how long would a fully running LHC (after the next repair period) require to accumulate similar statistics?

    dorigo
    Snowboarder, which problems ? Now the LHC is running well, it is ramping up in luminosity according to the predictions.

    As I wrote somewhere, one more year of Tevatron data only improves the precision of measurements by 10%. It is not going to modify the reach. Three years may be worthwhile but, as somebody at Not Even Wrong mentioned, if the Tevatron decides to continue running, the LHC may decide to continue too. And once LHC is at reasonable power, one year of LHC data makes all the difference, while one year of Tevatron data does not.

    Cheers,
    T.
    Thanks for the news Tommaso,
    but I'd have loved to go more straight to the heart of the achivements. How was the miracle of the 8% lumi increase possible? Which is the parameter that was improved, and why this results is fairly better than expected? Clearly you can also say that this the Tevatron best secret, that people from the machine do not reveal fearing it could be hacked by LHC friends...

    dorigo
    Hi Leonardo,

    in fact I did not know at the time -had no data to study. However, if I look at the details of the best stores here (scroll to the bottom of the page), I see that there was a combination of factors walking in the right direction: a low emittance, a high transfer efficiency, high number of protons and pbars. I believe it was not a single factor which fluctuated weirdly, but a display of what the Tevatron can achieve: in a sense it says there is still some further room for improvement.

    Cheers,
    T.
    There is nothing mysterious going on here. The recent high initial luminosities resulted mainly from excellent proton beam intensities and emittances (transverse beam size, smaller is better ) from the lower energy injectors. In addition, faster (~3 minutes) "scraping" of the beam (to reduce backgrounds in experiments) has decreased the time with colliding beams before letting the experiments turn on, so the luminosity has not decreased so much. The antiproton beam is a little bit smaller since we have turned off a noise source in the Tevatron that caused them to grow. We had been deliberately blowing them up a bit so we could reduce proton beam loss that had been causing occasional quenches; we have been able to control/handle that beam loss better.

    Every little improvement adds up!

    Add a comment

    The content of this field is kept private and will not be shown publicly.
    • Allowed HTML tags: <sup> <sub> <a> <em> <strong> <center> <cite><TH><ul> <ol> <li> <dl> <dt> <dd> <img> <br> <p> <blockquote> <strike> <object> <param> <embed> <del> <pre> <b> <i> <table> <tbody> <div> <tr> <td> <h1> <h2> <h3> <h4> <h5> <h6> <hr> <iframe><u><font>
    • Web page addresses and e-mail addresses turn into links automatically.
    CAPTCHA
    If you register, you will never be bothered to prove you are human again. And you get a real editor toolbar to use instead of this HTML thing that wards off spam bots.