Naked Science Forum
Non Life Sciences => Technology => Topic started by: matthewh on 11/06/2019 11:03:47
-
David has been weighing this question carefully:
How reliable is the Kibble Balance? Will there be only one Balance to make the definitive measurement or others throughout the world; if the latter, how certain will it be that they will all measure identically?
Let's see what can be measured up for responses!
-
Just as a heads up, the use and precision of the Kibble Balance was touched upon in the news article The Kilogram is Not Dead Weight Anymore, found right here on The Naked Scientists' website. There is also more detailed information referenced at the end of the article if needed!
-
How reliable is the Kibble Balance?
Well, it is a very delicate piece of equipment:
- For accuracy, it needs to operate in a vacuum
- Most lubricants fail in this environment
- At carefully controlled temperatures, so that physical components don't change shape, and thermoelectric effects don't change the results
- Some parts need to be held at cryogenic temperatures
- Isolated from vibration, which will move the mechanical components
- etc
- Not something that everyone could build, and achieve world-class results.
Fortunately, most of the variables can be related back to time or frequency, which are the units we can measure with the greatest precision.
See: https://en.wikipedia.org/wiki/Kibble_balance
Will there be only one Balance to make the definitive measurement or others throughout the world
One of the fundamental goals of metrology is to define standards that can be reproduced and used in many places around the world, to facilitate world trade (among other things).
That means that there is an ongoing programme of comparison of standards between different countries.
So in practise, every country will have its own Kibble Balance. The US and UK have built their own, and these have been compared.
Now that the Kibble Balance is written into standards, commercial manufacturers will start to produce them in (small) volumes, so that smaller countries will have a choice: build one themselves, or buy one on Amazon.
Large companies needing accurate mass measurements may decide to buy their own Kibble Balance.
Smaller companies will calibrate their test equipment at a national laboratory against the national Kibble Balance (for a fee).
See: https://en.wikipedia.org/wiki/International_Bureau_of_Weights_and_Measures
-
The general requirement for primary standards laboratories is to work to a total error (sum of random and systematic uncertainties) about two orders of magnitude better than current customer demand. Secondary standards are usually acceptable one order of magnitude better than required by the field instruments they calibrate.
Replacing the prototype kilogram by a universal definition of mass has removed the uncertainties of the unit itself. It isn't difficult to make a Kibble balance that meets all present demands, though the definition of "difficult" is somewhat different between your local blacksmith and the clean workshops in a primary standards laboratory. Since mass is now defined and measured entirely in terms of fundamental constants, there should be no need for international comparisons to eliminate systematic uncertainties, and all that remains are, in principle, random errors that can be assessed from local reproducibility.
-
It's a great story. I have discovered a website that can provide useful information for me.
-
For a peek inside the US National Kibble Balance, and to get instructions on building your own, have a look at this 10-minute video:
https://www.nist.gov/si-redefinition/nist-do-it-yourself-kibble-balance-made-lego-bricks
If you find the hype a bit overwhelming, move to 2:45 for how it works, then to calibration starting at 7:45.
PS: I couldn't find any commercial suppliers at the moment, so small countries will need to rely on test masses calibrated at their nearest nation with a Kibble Balance.
-
I vaguely knew Bryan Kibble at the UK National Physical Lab, and Ken Hursey, the engineer who built the first one, played piano in my band - taking over from Brian Jones, the engineer who built the first CT scanner and played the Hammond organ!
I doubt there will be many commercial suppliers. The big national and international standard labs (USA, UK, Germany, Russia, China, Australia, BIPM (Paris)) tend to build their own primary standards with local "tweaks", and since the "product" is generally 100 times more accurate than the most demanding application, the old national lumps of platinum will almost certainly remain in use for a long time.
Whilst the entire concept is a work of genius and beauty (Ken made clocks in his spare time) I still think it odd that it depends on a local measurement of g, made by timing a falling weight in a vacuum or something equally crude, to umpteen decimal places, with no reason why it should be constant or even slow-moving!
-
nice information.
-
I still think it odd that it depends on a local measurement of g, made by timing a falling weight in a vacuum or something equally crude, to umpteen decimal places, with no reason why it should be constant or even slow-moving!
There is ample evidence that g doesn't change much, or quickly.
https://en.wikipedia.org/wiki/Shortt%E2%80%93Synchronome_clock
-
Except for the last sentence of that reference. Comparing a synchronome pendulum with an atomic clock
His data revealed the clock was so sensitive it was detecting the slight changes in gravity due to tidal distortions in the solid Earth caused by the gravity of the Sun and Moon.
And if my tide tables are to be believed, that's a twice-a-day sinusoid with monthly and quarterly variations of amplitude. So to use the Kibble balance you have to measure g each side of your mass determination, and before you know it, you have a lab full of vacuum pumps, rubidium clocks, lasers and GKW, plus your ultrastable electrical supply, cryogen handling kit...and then somebody stores surplus lead blocks* in the basement (see discussion of Cavendish, elsewhere on this site).
*no joke! "Pre-nuclear" lead and steel are priceless if you are in the business of measuring fallout. Within my undistinguished career, one overzealous administrator sold 150 tons of prenuclear lead for scrap without asking why we kept "a roomful of junk" and a manufacturer petitioned Parliament for permission to salvage a 1918 18-inch naval gun barrel from a war grave.
-
And if my tide tables are to be believed...
...Then they can be used to supply the correction needed.
So to use the Kibble balance you have to measure g each side of your mass determination
Which is fairly easy.
What would you use instead (to define the Kg)?
"Pre-nuclear" lead and steel are priceless if you are in the business of measuring fallout.
I have some old lead for screening stuff.
It also helps to get it from a mine where the background uranium levels are low.
-
And if my tide tables are to be believed...
...Then they can be used to supply the correction needed.
Alas no, they are crude predictions, good enough for entering a harbor or transiting a narrow passage, but don't claim to tabulate g to parts per billion on land.
So to use the Kibble balance you have to measure g each side of your mass determination
Which is fairly easy.
What would you use instead (to define the Kg)?
A Kibble balance.
"Pre-nuclear" lead and steel are priceless if you are in the business of measuring fallout.
I have some old lead for screening stuff.
It also helps to get it from a mine where the background uranium levels are low.
Sadly, manmade contaminants like plutonium and iodine get into the refining process and screw up the best forensic gamma spectrometry. I'm currently watching "Chernobyl" - somewhat oversimplified but it shows the importance of precise analysis. We used pre-nuclear shielding to measure the acquired activity of UK civilians immediately after the event, with some very interestng results.
-
I wonder if anyone has looked into this as a possible source of "clean" steel.
https://en.wikipedia.org/wiki/Direct_reduced_iron
You can dig ore from so deep underground that it wouldn't be affected by any nuclear history.
Then reduce it with nice clean syn gas.
(and you can do similar things with lead but you have to worry about natural radioisotopes of lead)
-
And if my tide tables are to be believed...
Unfortunately, tide tables measure a lot of things apart from the influence of the Sun and the Moon...
- They measure the sloshing of oceans in their basin
- Which depends on the shape and depth of the basin
- And local resonance effects (especially if the resonance is near 12-13 hours, the driving period of the Moon & Sun)
- And local factors like the age of the tide: "the tide keeps getting stronger for a few days after the Sun and Moon are aligned", which is the point of maximum driving force
See: https://en.wikipedia.org/wiki/Tide#Timing
In fact, the motions of the Earth, Sun and Moon are dominated by some fairly simple ellipses.
Unlike oceans, gravity travels at the speed of light, doesn't have inertia, and doesn't bounce off shorelines, so a fairly simple calculation devised by Kepler will give you the relative positions of the center of gravity of Sun, Moon and Earth
- As observed by Newton, gravity has an inverse-square law (and tide has an inverse-cube law)
- And a fairly simple clock will tell you where your Kibble balance is pointing, relative to the Sun.
- This will allow you to calculate the correction to local g continuously while doing mass measurements.
- With current measurement techniques, gravity corrections for Jupiter or the other planets are way down in the noise...
You still need to drop your test mass in a vacuum periodically, to correct for the concrete absorbing CO2, the gardener digging up the roses and continental drift.