> We find that any gravitational-strength Yukawa interaction must have λ < 38.6 μm. This implies that the dilaton or heavy graviton mass and the radion unification mass must be greater than 5.1 meV and 7.1 TeV, respectively, and that the largest extra dimension must have a toroidal radius less than 30 μm. These are the tightest existing lab constraints on “string inspired” new gravitational phenomena.
But the key words in it mean ~nothing to me :(
So do they do anything more than constrain parameters?
The first sentence above is, for a physicist, the key result. It says that, for the most common way one could modify gravity to incorporate a new force-mediator, the new interaction has to turn on at distances shorter than 39 microns.
At distances longer than that, these measurements show that gravity acts, and in a way that roughly follows the inverse-square law. (at distances of ~100 microns it very-much follows the ISL, and at distances longer than that, it really follows the ISL).
The easy way to think about this result: we are on the cusp of showing that gravity acts over distances shorter than the width of a human hair. Only the finest human hairs are smaller in diameter.
For many years, my goal in life (see HN handle) was to push this limit below 10-microns. Turns out that is really hard to do :).
You are probably thinking of galaxy rotation curves which do not agree with theoretical predictions that assume only visible matter exists.
Gravity not following the ISL at large scales would indeed be a possible explanation for this. However, most alternative theories of gravity clash with other observations, such as the motion of binary pulsars or gravitational waves, whereas (unmodified) General Relativity is in excellent agreement. This is one of the reasons why the existence of dark matter is considered to be a much better explanation for the deviation of the rotation curves.
It's pretty simple. You can use the known laws of gravitation and the visible distribution of matter in a galaxy to predict how fast stars at the edge of the galaxy should move. But when we do this we find that the stars are moving much faster than expected. So EITHER the formula for gravity is wrong OR there is extra matter there which we can't see, i.e. dark matter. Dark matter is currently the preferred explanation because it also explains a bunch of other, unrelated, phonenomena, but it has as yet never been directly detected or measured.
Cool to see you guys here - as a UW undergrad I visited the lab once and was amazed by all of the effects you had to take into account to reduce error. Wasn't it something like the experiment can pick up the sun shining on the bricks of the building and shifting it, and the changes in the level of groundwater after it rains? I never understood why you didn't have the apparatus in the middle of the desert.
What sort of error reductions and changes allowed the new distance measurements?
We contemplate moving elsewhere from time-to-time, but the proximity of our colleagues, machine shop, and an awesome lab keep us here.
For some of our experiments, we need the hill we're built into, and we can handle the tilts.
This short-range measurement was greatly benefited by a fortuitous snowstorm that brought Seattle's buses to a halt right as John was taking data with the fully-commissioned instrument. If we had foreseen with 6-8 weeks of foresight how much of a societal disruption Covid-19 would cause, we would have crushed it to get this short-range instrument back in operation in time.
We're working on new seismic-isolation systems for our next upgrade. If that doesn't work out, we'll be looking for a remote hole in the ground to put a short-range instrument in.
The big steps forward on this cycle were improved alignment systems, improved flatness of multiple critical parts, improved cleanliness, a new gluing arrangement that reduced Newtonian-gravity systematic uncertainty, even greater care to ensure non-magnetic fabrication, and another ~5 years of experience with the instrument.
The next big jump for us is active seismic isolation. A small improvement there, with other tweaks, may buy us another ~5-10 microns.
> We're working on new seismic-isolation systems for our next upgrade. If that doesn't work out, we'll be looking for a remote hole in the ground to put a short-range instrument in.
Just put it on top of one of the LIGO mirrors? :P
If you magically got an couple of order of magnitudes improvement (~0.5um), would you have to start to worry about van der Waals forces, or is that still weak enough? Or does it not matter at all for this kind of apparatus?
We actually work on LIGO's seismic isolation, too. We are adapting some of what we know from that work to improve our own isolation systems. Torsion balances have slightly different requirements, so it's not just drop-in use of that technology, but it is pretty close.
Regarding van der Waals -- yes. At the moment, we block all electrostatic/Casimir interactions with a metal foil. When the distances involved are "sufficiently short", we won't get complete screening any more. That day will be an interesting day.
On the bright side, if we start to see Casimir leakage, that will be a boon to the people who try to understand the Casimir force at finite temperature (another open problem), so we'll learn something useful no matter what...
In the short term, though, I'd love to have that problem. It will take a lot of work for us to get there.
Thanks! Been viewing presentations over at pirsa.org about LIGO and other experiments that requires high accuracy, and it's just jaw dropping the extremes one has to go through to eliminate the various noise sources. Highly impressive work, that's for sure!
Speaking of, do you have a link to share of a recorded presentation about this or one of the previous experiments?
We're looking for any deviation from that plan, in the form of new forces or a change in the behavior of gravity itself. Either one would be really cool to find, and the fact that we keep not finding anything places increasingly tight constraints on theories that attempt to extend the Standard Model or build a bridge from the SM to gravity.
Gravity and the standard model are, as written, very different theories. Our short-range gravity experiments are simultaneously at the forefront of some classes of high-energy physics one might explore at the LHC and interesting astrophysical/cosmological questions.
With our entire program of measurement, we are pushing as hard as we can on the theory of gravity, looking for a gap in the armor to see what's next. The draw of these short-range experiments is that it is totally-unexplored territory, and with every step forward, we continue to check whether or not our oldest-known force behaves as predicted.
It's late, so I'm beginning to ramble: There are viable theories that predict that gravity should get stronger at these scales, and viable theories that gravity should get weaker at these scales. With this result, those theories just got squeezed into a smaller box.
While this isn't my result (I helped with this one, but am not a prime-mover), for me, the big draw is to study gravity until we push it, like all the other interactions, into the realm of the microscopic.
Ah, I realized my comment sounded a little like I was trying to downplay the work! Wasn't my goal -- I truly find this work extremely interesting and at one point wanted to study physics so that I could do particle physics + cosmology. Parameters finding is an absolute must for us as a species if we want to push forward.
On quora there is prolific nutjob who spreads misinformation about gravity claiming gravity as mass/distance² was never experimentally proved and instead claims gravity is volume/distance², his responses are often quite highly ranked, e.g. 2nd answer here: https://www.quora.com/Why-is-a-lunar-orbit-unstable
We would be incredibly stoked if we could drop that number to ten microns; nobody knows yet if that can be done.
I'd be surprised if either we or the group at Huazhong University don't reach at least 35 microns by 2030. We all aspire to ~20 microns.
These all sound like small improvements, but Figure 5 tells the real story: reaching 10-microns requires a 3,000x improvement in sensitivity, in an environment where almost every experimental difficulty gets exponentially or power-law worse at short distances. That's why it is hard.
The field is ripe for a totally-new technology, but whatever it is will be deeply clever and revolutionary. I've spent the last 16 years of my life trying to figure out how to do it with limited success.
You say seismic vibrations increase the noise a lot but not, that I saw, anything about what measures you take against them?
Presumably you do some sort of active noise cancelling for vibrations (like noise cancelling headphones)?
Seismic noise must be a common problem in physics experiments, which other experiments have the best environmental vibration cancelling systems ... do people pay for use of them like they would for use of a space telescope?
We are moving from passive to active isolation now. The isolation that we need right now isn't too challenging. The hard part is ensuring that it doesn't disturb other properties of our instruments. We are sensitive to nanoradian rotations at millihertz frequencies, so we have to ensure that our isolation systems don't add any rotation noise.
Regarding isolation-as-a-service: Most of these kinds of experiments run for many years in a university setting. The environmental requirements for each instrument vary widely, too. "Isolated-environment laboratories as a user facility" is an interesting idea to consider. I'm not sure if it would soar in the academic-research marketplace, but there might be a niche somewhere.
First of all - very interesting work. I am gonna use the opportunity to ask question to authors directly. Keep in mind that I don’t know a lot about gravity ;)
As far as I understand speed of gravity interaction is limited by speed of light. So my understanding is that the 1/r^2 law is only approximately true since real interactions are retarded. Is that correct? If yes then at what level (what order of magnitude) those corrections affect your result (I understand that at such short distances it’s very minor but I suppose it can be an somehow is estimated)?
I wonder what scales string theories would expect? As in — is this getting uncomfortably close for string theory to manage or still orders of a magnitude away?
“New” a hundred years ago, yes. No one has been happy with our model of gravity since Newton’s time, and people were only happy with it then because they didn’t know any better.
I have a hard time believing how gravity could be measured at such a small scale. Is it possible to video record the experiment? Maybe with some layman explanation.
I would like to make my own gravity experimentation. Is this something I could do at home?
That will give you a flavour of the difficulty involved.
Below those sizes, electrical forces start making your experiment very difficult (electrostatic, van de Waals), and any vibration will pollute your data. This is where the "can't do it at home" starts to read its head (something to compensate the vibration of buses driving nearby, a lab with no static electricity or magnetic fields anywhere, etc, etc)
Beyond ... the Casimir effect places a limit to the lengths you can reach.
Hopefully the lead author (an HN-lurker, as far as I know) will jump in to bask in the limelight and answer your questions :).