• Member Since 28th Oct, 2012
  • offline last seen 2 hours ago

Pineta


Particle Physics and Pony Fiction Experimentalist

More Blog Posts440

  • 6 weeks
    End of the Universe

    I am working to finish Infinite Imponability Drive as soon as I can. Unfortunately the last two weeks have been so crazy that it’s been hard to set aside more than a few hours to do any writing…

    Read More

    6 comments · 143 views
  • 9 weeks
    Imponable Update

    Work on Infinite Imponability Drive continues. I aim to get another chapter up by next weekend. Thank you to everyone who left comments. Sorry I have not been very responsive. I got sidetracked for the last two weeks preparing a talk for the ATOM society on Particle Detectors for the LHC and Beyond, which took rather more of my time than I

    Read More

    1 comments · 130 views
  • 10 weeks
    Imponable Interlude

    Everything is beautiful now that we have our first rainbow of the season.

    What is life? Is it nothing more than the endless search for a cutie mark? And what is a cutie mark but a constant reminder that we're all only one bugbear attack away from oblivion?

    Read More

    3 comments · 188 views
  • 12 weeks
    Quantum Decoherence

    Happy end-of-2023 everyone.

    I just posted a new story.

    EInfinite Imponability Drive
    In an infinitely improbable set of events, Twilight Sparkle, Sunny Starscout, and other ponies of all generations meet at the Restaurant at the end of the Universe.
    Pineta · 12k words  ·  49  0 · 827 views

    This is one of the craziest things that I have ever tried to write and is a consequence of me having rather more unstructured free time than usual for the last week.

    Read More

    2 comments · 137 views
  • 15 weeks
    Secrets of Auroras

    Read More

    3 comments · 148 views
May
20th
2021

g minus two · 3:06pm May 20th, 2021

This post is a long read. A rambling personal story with a big dose of particle physics.

Did you catch the big particle physics news story that came out last month? The Muon g−2 collaboration at the Fermilab laboratory have measured the muon g-factor to a higher precision and confirmed that it doesn’t match the prediction of the Standard Model. Isn’t this exciting? Are you excited? I’m excited, but I have a personal connection with this one. I’m one of some 230 authors on the paper now published in Physical Review Letters.

What is g−2? It got a brief explanation by Pinkie Pie in my old post: Twilight and Pinkie visit Fermilab (Batavia, Illinois). Take the muon g-factor, which is about 2.00233184, and subtract 2. Or, as it is more commonly expressed a_\mu=\frac{g-2}{2}. The latest value has been measured giving a new world average of a_\mu = 116592061 \times 10^{−11}. The real exciting bit is when we compare this to the theoretical calculated value, which is just a_\mu = 116591810 \times 10^{−11}. The difference in the last few digits of these two numbers could be a sign of some exciting new scientific phenomena. A hint of some unknown particles waiting to be discovered.

The g-factor is a measure of a measure of how magnetic a particle is. The simple version of quantum mechanics predicts g = 2 for the muon, but in the full calculation done with Quantum Field Theory, the value is shifted slightly away from 2 by the cloud of virtual particle and antiparticles surrounding it, which are forever being created momentarily and then disappearing, but can change its properties. Thus, the g-factor is linked to every particle and force, known and unknown, and such precision measurement can give clues to new scientific phenomena. The muon is a good particle to use to look for such clues, as it is relatively easy to make a beam of them, they live long enough to see what they do, but they are exotic and heavy enough to be sensitive to New Science.

My involvement with this goes back to 2012 or so. I was the leader of a Muon g−2 Research Group at the University of Oxford. That’s what I put on my CV, which makes it sound like I was a high-flying professor, directing a team of minions in my shiny high-tech laboratory. Actually, it was me and a couple of friends desperately trying to find a way to be involved with an exciting project, to keep our jobs, and do something interesting, while also keeping up with all the other stuff we had to do. We were a group of misfits. We had enthusiasm. We had a unique set of skills in hacking electronics, building scientific instrumentation, and writing fan fiction. But we had a few problems.

We didn’t have any money. The project was very well funded by the US Department of Energy, and other agencies, to the order of $20 million, depending on how you count it. The Fermilab budget is over half a billion. But building high-intensity muon beams, precision high-field magnets, and shipping a 14-metre superconducting ring from New York to Illinois, doesn’t come cheap. The collaboration had a list of extras bits that they could really use, and this provided opportunities for outsiders like us to get involved, if we could supply them.

I used my creative writing and world building skills to craft an exciting research proposal describing how by building a helium-3 calibration magnetometer we would provide a full cross-check of a key uncertainty in the measurement, and provide a blue print for a future magnetometry standard. This was submitted as part of a wider proposal for a UK g-2 research programme, which was funded, but not my bit of it. There was only so much funding available, there were other UK scientists trying to get involved, and we weren’t high enough up the academic pecking order to get a cut.

There then followed a process of negotiation where I cut a deal. If we moved some resources from another project, combined this with a small grant to develop magnetometers for geophysics, bought a rubidium frequency reference cheap on ebay, and salvaged an old test magnet from the scrap yard, then we would just need a few thousand to buy a Class IV fibre laser and a decent data acquisition unit and we could get to work. I persuaded the Dark Matter group to let us convert a tiny room they weren’t using into a laser lab, and begged and borrowed components from other groups. We designed a prototype probe. Found a way to make it cheaply by 3D-printing components. Did some first tests in someone else’s test magnet. It sort-of worked, but we still had a long way to go.

If this had been a fic, we would have succeeded after a sequences of plot twists involving villains, super heroes, mysterious creatures, and magical powers. But as this was real life, we just had the villains. In the end it wasn’t technical problems but management that finished it. Some senior assholes in the university were out to get us from the start. They were probably hoping we would quit, and when we didn’t, things got nasty. I was summoned to meetings, at short notice, to defend my project, on dates when I had booked flights to the US. Reviewers who had previously told us to diversify, now attacked us for trying to do too many things, or attacked details of our design that we had never even proposed. Meanwhile people in the US, who had initially supported us, became less helpful when they realised we didn’t have secure support. Simple requests for access to laboratories were delayed for months. I’m sure we also had friends in the management who were trying to help us. There were things happening behind closed doors that I will probably never know. But in the end, things got so impossible that we had to stop. I sometimes think I should have fought harder, but it took a toll on my mental health, and things reached a point where it was so unpleasant, I didn’t have the motivation to push on. Academia can be like this.

We wrote up the test results from our prototype as an internal report. Shipped our laser to the University of Michigan, where they took our design and finished the project, and took the credit. We got a brief mention in a footnote at the end of the paper.

Things moved on. On the bright side, despite everything, we all kept our jobs, but I’ve not worked on anything so interesting since then. My geophysics work didn’t lead anywhere. I now managing code for a component database for an experiment at CERN. And run all the wacky public engagement projects I can.

Once we were out of the collaboration, we were cut off the mailing list. As the project moved into data-taking, an atmosphere of secrecy developed, and were heard very little about what was going on. We just knew that the schedule had slipped by a year, then another. Then, just when we had almost forgotten about it all, we received an invitation to attend the Unblinding.

What is an Unblinding? Well, most experiments of this size, now do some sort of data blinding. When analysing the data, it is very easy to subconsciously fudge things to give the answer you are expecting, or hoping for. To avoid this, and give more integrity to the results, the data are ‘blinded’, like in medical trials, so you can’t see the result until the very final step, by which point you are committed to publishing it whatever you get.

The g−2 measurement is done by measuring the muon spin precession frequency (which Izzy explained in my last post) – the frequency with which they dance around their direction of travel as they whiz around inside a ring magnet. This is measured with an array of detectors that count the positrons produced when the muons decay.

You count the number of times the positron rate oscillates up and down per second. To do this, you need an accurate clock (linked by GPS to the worldwide atomic time standard.) To blind the data, you get someone from outside the collaboration to program a slight shift on the clock frequency, write down the number, put it in a sealed envelope, and lock it in a safe.

The reason for all this drama is that there are stories of experiments in the past that tried to blind their data, only to see it accidentally unblinded when someone opened the wrong file.

The Unblinding ceremony was held on Zoom. We all watched as the project leaders made a show of opening the envelop and entering the number into a spreadsheet, where a pre-written code then revealed the new measured value for g−2. It was in line with the last experiment to measure it, and out of line with the theory. We had done it.

For a few weeks, the news was embargoed, and we had to resist the urge to tell everyone about it. Then the result was announced on April 7. Immediately after, theorists released papers giving their explanations for the anomaly. While a discrepancy is a sure sign that there must be some new scientific phenomena, it gives only the vaguest clue of exactly what that could be. Theories of possible New Physics beyond the Standard Model have been shaken up in recent years as the Large Hadron Collider has failed to find any particles predicted by popular models. New ideas are gaining ground, and the muon g−2 anomaly is fuelling the imagination of theoretical physicists.

While the experimental result is now looking reliable (as the new experiment matched previous results), another less-exciting possibility, is that there is an error in the calculated value. This is a very complicated calculation, undertaken by an international team of computational physicists. A new paper came out the same week, which calculates an especially tricky bit – the hadron vacuum polarization term – using a new approach. This new calculation is in line with the experimental measurement. This explanation would appear to dampen all our enthusiastic talk of new physics, but explaining how this new calculation is compatible with the well-established one could be an even bigger mystery, also requiring some new theories. At the moment the particle physics world is taking a wait-and-see attitude.

And there is still more data to analyse.

Comments ( 17 )

Ah, a nice dose of particle physics to start my morning. Sorry to hear about the douchebags higher up.

Georg #2 · May 20th, 2021 · · 1 ·

bought a rubidium frequency reference cheap on ebay

Mind boggled. I knew they sold about everything there, but...

Yes, you now go into the gigantic furball where Wise Bearded Scientists Who Came Up With Theories Before You Were Born And Are Therefore Far Wiser Than You tutt and humm over your data, bringing up points that they have done for the last several decades about how their Great Theories are still perfect despite these new observations that only seem to contradict previous observations but in Their Opinion can be simply discarded because they can't possibly be right and get off my lawn.

Office politics, bleh. But very interesting explanation regardless, thank you.

5522098
Rubidium frequency standards, and a lot of precision electronics kit, are very expensive if you buy a new laboratory system - but you can get cheap units salvaged from old telecoms equipment. Before ebay, we had radio and computer rallies.

Office politicking is as fundamental a force as any of them. Don't let the bastards grind you down.

Hooflon's razor? "Never attribute to undiscovered phenomena what can be sufficiently explained by calculation error"?

How accurate, or inaccurate, are the old colour clock frequency crystals used in all CRT TVs and video systems? PAL 4.43..... Mhz, supposedly cut to an accuracy of 0.01 Hz to reduce dot crawl, or 4 parts per billion? I managed to find a quad frequency, 17.76 Mhz to use in a 20 Mhz system for generating images, but being able to pick up shifts in vertical lines has a similar phase effect? Like the method used to photograph lightning for range?

There was some mention recently about the tighter you make your clock jitter, the higher the local entropy? Isnt that Heisenburg operating, so you squeeze the oranges and it gives you lemons?

Thats a lot of effort goes on there, and typical the academic crab bucket is cooking nicely, slowly roasting everyone at the bottom, but still incapable of climbing out of the pot? :pinkiesad2:

In this measurement, I thought I remembered seeing many years ago about a CD player FFT on a 14 Mhz system that had jitter in the sub nanosecond range, maybe 100 pico seocnds. Would this variation show up reliably on that scale, not verifyably, or would it just be buried in the noise?

Have you had any experience or intrest in the recent release of the Raspberry Pi Pico, for its Digital HDMI accuracy bit banging , for sampling and processing?

Oh, this is very cool indeed. The discrepancy, not upper management out to get you. I am sorry about that. Here's hoping there's as little of that as possible in the future.

5522147
But then there’s Holmes’ Precept: “How often have I said to you (Dr. Watson) that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?”

5522172
I don’t think that’s good enough. A typical modern rubidium frequency reference is accurate to <±5 × 10-11, or roughly two orders of magnitude better than the clock crystals you’re referencing. They’ve become considerably less expensive with the explosive growth of the development of devices incorporating GPS receivers in some fashion, but that’s not to say that they’re cheap. Heck, they even make full-blown atomic clocks with that level of accuracy as (admittedly large) PCB components now if you need something compact and can live without recalibrating it every few months as it ages.

:O I have a colleague on FimFic.

My life is complete. \o/

Glad you remained on the author list despite all the backroom drama. The editing of those papers was an insane process.

5522374

A persistant measurement error that needs correcting at those accuracies? Is there a isochronal map available for the world, with published correction values for daily, monthly, yearly deformations and maybe even for extreme weather influences etc?

I mean, Ive seen reports of Aluminium Ion clocks that are so accurate, they vary due to the gravity of the researchers and bench height?:twilightoops:

5522539

A persistant measurement error that needs correcting at those accuracies? Is there a isochronal map available for the world, with published correction values for daily, monthly, yearly deformations and maybe even for extreme weather influences etc?

Those, I don’t know. But there are certainly less expensive frequency references albeit of less accuracy and precision, so Pineta will have to answer why a rubidium one was required.

I mean, Ive seen reports of Aluminium Ion clocks that are so accurate, they vary due to the gravity of the researchers and bench height?:twilightoops:

If Wikipedia and the linked articles are to be believed, the aluminum ion quantum clock accuracy is down to <±2.1 × 10-18, which is theoretically sufficient to observe the gravitational time dilation delta for an elevation change of 2 cm, so — yes. And this has been empirically demonstrated with a pair of slightly less accurate clocks with an elevation difference of 30.48 cm (one foot).

Of course, nothing is ever good enough, so work continues on nuclear clocks to get the accuracy down to <±1 × 10-19.

5522548

Couldnt you place two clocks, at 1 meter and 2 meter from a 1kg mass, and claim that the mass of 1 kg is the mass that causes a change in relativistic time due to gravity of a given value for that perfect isolated set up?

Therefore making a direct relation between mass, space, time, without needing electric or magnetic fields etc?

I was just wondering because the 50 year old TV crystals apparently can reach that 1st or even 2nd place deviation, even before oven temperature control and signal processing etc, and the longer the experiement and more experiments the smaller the errors? And theyre definitly, back then, a lot cheaper and more compact that atomic ion or fountain clocks? If they were used in networks, how much improvement in accuracy would there be averaging over several billion, possibly, and theoretically?

5522551

Couldnt you place two clocks, at 1 meter and 2 meter from a 1kg mass, and claim that the mass of 1 kg is the mass that causes a change in relativistic time due to gravity of a given value for that perfect isolated set up?

I think the biggest problem would be dealing with environmental noise. I mean, LIGO had it bad enough with trucks on the road and animals on the 4km-long vacuum tubes, but they found ways to actively and passively damp out or otherwise measure and compensate for vibrational transients. But how would one shield the experiment from other nearby gravity sources that would interfere with the time dilation measurements? The smaller the experimental mass, the greater the gravitational time dilation noise issue will become; at some point, you’ll need to compensate for magma displacements, the position and orientation of the Moon/Earth in its orbit around the Earth/Sun along with the Earth’s rotation and orientation, the crust displacement of earthquakes, rainfall (never even over an area), etc. Heck, you might need aLIGO measurements to compensate for the gravity waves from the occasional black hole collision on the other side of the universe rattling the spacetime between the experimental mass and the clocks.

I’m not saying the problems with this approach are insoluble, merely — extremely challenging. :eeyup: Perhaps if an enormously more massive, precisely fabricated, homogenous test mass were used, say, a 106kg lead alloy sphere? Tungsten would be ideal due to its strength and density, but the quantity is a problem, along with fabricating something of such size out of it.

5522551
5522567

Here are some precision timekeeping figures:

30ppm Harrison H4 Marine Chronometer (1761)
0.1ppm Quartz resonators (1927)
1ppb Rubidium clock
1:1016 NIST-F2 Caesium clock (2014)

We were aiming for a precision of better than 50ppb for our magnetometer, so quartz wasn't good enough. Rubidium clocks are the industry standard. They are linked by GPS to the worldwide network of atomic clocks to give long term stability. We weren't at the level where we had to worry about time dilation effects on the clock time.

Takes me back to when I wrote all the footnotes in Time on Their Hooves

Defining the kilogram by the time dilation caused by a mass is theoretically possible, but not a very accurate or practical way to do it.

5522098
Sounds like an interesting take on Starswirl the Bearded

Login or register to comment