The Naked Scientists

The Naked Scientists Forum

Author Topic: To build a better clock, Ytterbium, Gallium, Germanium Alloy Triple Point Clock  (Read 5280 times)

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
One method of designing a clock would be based on distance, as long as the distance was precisely known.

With the speed of light at 300,000,000 m/s, one could define a clock based on twice the time to traverse 1 meter, or a frequency of 2/(300,000,000)

Alternatively, one could define it as the frequency defined by the length.

So, a 1m wavelength would give one a 2.99x108 hz cycle.
0.25m wavelength would define a 1.199x109 hz cycle.
And etc.

The trick would be to define a "standard" that would be virtually invariant in most frames of reference.

Several notes indicate that Ytterbium, Gallium, Germanium Alloy is one of the most stable substances to thermal expansion.
http://cerncourier.com/cws/article/cern/28977
http://en.wikipedia.org/wiki/YbGaGe
http://www.nature.com/nature/journal/v425/n6959/abs/nature02011.html
http://pubs.acs.org/cen/NCW/8142notw1.html

While YbGaGe apparently exhibits thermal stability over a wide range of temperatures, one still might wish to specify a precise reproducible temperature such as the triple point of water under 1G of acceleration, but realistically one might be able to fudge the temperature standard slightly.

The question is whether it would also be stable under changes in velocity, and changes in gravity.  If it was, then one could potentially create a clock that would defy relativity,  :o and thus would be more accurate than the most accurate clocks today.


 

Offline MikeS

  • Neilep Level Member
  • ******
  • Posts: 1044
  • The Devils Advocate
    • View Profile
As I understand it 'Time' is relative.
The most accurate clock is the speed of light in a vacuum. 

quote CliffordK
"The trick would be to define a "standard" that would be virtually invariant in most frames of reference.
Several notes indicate that Ytterbium, Gallium, Germanium Alloy is one of the most stable substances to thermal expansion."
"The question is whether it would also be stable under changes in velocity, and changes in gravity."

It wouldn't, in comparison to a clock that was not subject to those effects.
« Last Edit: 21/01/2012 13:59:06 by MikeS »
 

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8676
  • Thanked: 42 times
    • View Profile
The problems of thermal stability were addressed some time ago in clock making
http://en.wikipedia.org/wiki/Gridiron_pendulum
however, it's easy to just put the clock in a place where you keep the temperature constant.

Atomic clocks don't really have a temperature dependence like pendulum clocks.
 

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
As I understand it 'Time' is relative.
The most accurate clock is the speed of light in a vacuum. 
Energy is the core principle of relativity.
The energy is manifest in a number of phenomena, perhaps something analogous to heat and mass.

I believe that time should be measured linearly, and that there is in fact a base-frame or state of motion that can be determined.  The inability to differentiate the one-way speed of light from the two-way speed of light doesn't mean that it doesn't exist, but that we just have to try harder to find it.

Our best current generation of clocks are the atomic clocks which are based on the light (or microwave) frequency that cause the energy change in the hyperfine transition state of various atoms including hydrogen and cesium. 

It is not surprising that if energy is the core of relativity, then these clocks would be subject to the same influences as the hyperfine states may vary with the energy changes associated with relativity.

One could also make an angular momentum clock.  However, if an angular momentum would be subject to expansion or an increase of mass, then any clock based on angular momentum would also be predicted to slow down with an increase in velocity.

One might also think of a biological clock.  There is no reason why it should in fact slow down in a higher energy state associated with velocity.  So, humans (or bacteria) traveling at a significant fraction of the speed of light could in fact age faster rather than slower.  Not to mention, of course, other influences of space.

I found the YbGaGe article if you wish.
http://www.cem.msu.edu/~kanatzid/Reprints/nature_YbGaGe.pdf

It isn't perfectly stable either in the X or Y direction, but perhaps it could be cut at an angle which would average out to 0 thermal expansion.  Would it cause a slight twisting effect to do so?  There are other composite materials with nearly zero thermal expansion, created  as a composite from negative and positive thermal expansion materials.  They might also be worth exploring, but the YbGaGe seems to be one of the more stable & predictable ones.  I still need to find some "toughness" data for it.

No, I'm not sure how YbGaGe would change with velocity.  If there is an effect analogous to heat, then one would hope that it would suffer less expansion than other materials.  However, the effect of a mass change may be less predictable.  If velocity creates an expansion analogous to heat, it may be difficult to control for as it may not be measured by temperature with the ordinary methods.  So, while one might control for thermal changes, they may be difficult to detect.

Yes, the two-way speed of light should be invariable in all frames.  But, to use it as a "clock", one needs to find a standard to measure it with.  I.E.  light travels 300,000,000 m/s.  Or, traverses a meter and returns in 2/300,000,000 s.  However, one still needs to know precisely how long a meter is (which is now defined by the distance traveled by light in a specific time). 

If one can define a new meter standard that is invariable with velocity (no expansion due to relativity), then one has a better clock.
 

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
The problems of thermal stability were addressed some time ago in clock making
http://en.wikipedia.org/wiki/Gridiron_pendulum
however, it's easy to just put the clock in a place where you keep the temperature constant.

Atomic clocks don't really have a temperature dependence like pendulum clocks.
I like the idea of the Gridiron pendulum.  Not necessarily for making a pendulum, but for making a length standard that would be less affected by possible velocity related expansion.

It does appear as if it is flawed, with always having an odd number of elements (2 downward and 1 upward element, etc), which the engineers have compensated for by using different metals.

The YbGaGe shows different expansion in different directions.  Perhaps it could be used to correct itself in a similar device.

Unfortunately, I don't think we have enough data to know if there is in fact a velocity related expansion, and how that relates to thermal expansion.

While the Atomic clocks may not be affected by distance, they have an energy dependence on the hyperfine transition states of various atoms that could be fouled up by the energy changes related to velocity and gravity.
« Last Edit: 21/01/2012 18:26:39 by CliffordK »
 

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8676
  • Thanked: 42 times
    • View Profile
"It does appear as if it is flawed, with always having an odd number of elements (2 downward and 1 upward element, etc), which the engineers have compensated for by using different metals."
That's not a flaw, it's how it works.
If you didn't use different metals then the arrangement of the rods would not matter- it would just expand when you heated it.
 

Offline Geezer

  • Neilep Level Member
  • ******
  • Posts: 8328
  • "Vive la résistance!"
    • View Profile
The most accurate clock is the speed of light in a vacuum. 

Presumably you mean the distance light travels (speed requires knowledge of time), but then you're back to the problem of distance, so it's really a pretty lousy clock.
 

Offline Geezer

  • Neilep Level Member
  • ******
  • Posts: 8328
  • "Vive la résistance!"
    • View Profile
they have an energy dependence on the hyperfine transition states of various atoms that could be fouled up by the energy changes related to velocity and gravity.

Precisely! If they didn't, they would not be measuring time.
 

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
The most accurate clock is the speed of light in a vacuum. 

Presumably you mean the distance light travels (speed requires knowledge of time), but then you're back to the problem of distance, so it's really a pretty lousy clock.
Velocity is a calculation of distance/time.

Knowing the velocity, as well as precisely knowing either distance, or time, the remaining factor can be calculated.

So, if one calculates:

c= d/t,  or better, c=Δd/Δt

Relativity would indicate that if the unit for time gets larger (slower time), then the unit for distance would also have to increase (longer distance). 

And, it would indicate that any measurements of the time-expanded distance would be the same for all materials in all directions.

One would still need a mechanism.  We've noted that thermal expansion is different for all materials.  Has this speed dependent expansion been directly observed in space, and been observed to be constant for all materials?

Noting, of course, that mass (also predicted to increase) would tend to increase with volume, and thus is essentially a 3-D phenomenon.  Distance is a 1-D phenomenon, so one wouldn't necessarily expect them to expand at the same rate.
« Last Edit: 21/01/2012 20:11:46 by CliffordK »
 

Offline Geezer

  • Neilep Level Member
  • ******
  • Posts: 8328
  • "Vive la résistance!"
    • View Profile

as well as precisely knowing either distance, or time, the remaining factor can be calculated.


But you don't know the distance precisely which is why light does not make a very good clock. Atomic clocks avoid this problem by using some sort of atomic activity instead.

 

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
Let me introduce a new measurement.

Call it the relativistic linear expansion coefficient (RLEC).

One could then set an arbitrary material (compound, element, etc) to being 1 RLEC. 
Then just compare materials and compounds.

As BC pointed out, if there is any deviation from the standard, 1, then one can build a Gridiron device that would be exactly one meter under the specified conditions.

c=d/t

If one can define d as being invariant (under a set of conditions), then t must also be invariant under those same conditions. 

Whether or not we can build a linear distance clock to the same accurate to the same number of decimal points as the hyperfine transition clocks, we would be able to build a clock that was independent from the relativity effect.

And thus, the variability of time with respect to space would be disproven.

Do we have the ability to measure lengths to an adequate degree of accuracy, and compare the lengths at high velocity, low gravity in LEO and what is found on the surface of Earth?  Or, does it require building a length-based clock to compare the reaction of different materials?
« Last Edit: 22/01/2012 00:02:30 by CliffordK »
 

Offline Geezer

  • Neilep Level Member
  • ******
  • Posts: 8328
  • "Vive la résistance!"
    • View Profile

Whether or not we can build a linear distance clock to the same accurate to the same number of decimal points as the hyperfine transition clocks, we would be able to build a clock that was independent from the relativity effect.


Perhaps not. I don't think distance is absolute. If gravity can "bend" space, everything in that space is bent too.
 

Offline MikeS

  • Neilep Level Member
  • ******
  • Posts: 1044
  • The Devils Advocate
    • View Profile
The most accurate clock is the speed of light in a vacuum. 

Presumably you mean the distance light travels (speed requires knowledge of time), but then you're back to the problem of distance, so it's really a pretty lousy clock.

Lousy, (read compromise) it may be but I still think it is the best we will ever have. 
I believe, in a sense, the speed of light is the universes master clock.  All other clocks, including atomic, are subject to local variations due to gravity and possibly velocity.


as well as precisely knowing either distance, or time, the remaining factor can be calculated.


But you don't know the distance precisely which is why light does not make a very good clock. Atomic clocks avoid this problem by using some sort of atomic activity instead.



Atomic clocks are subject to local effects like gravity and velocity which CliffordK was trying to avoid.

Geezer you summed it up in your last post
"I don't think distance is absolute. If gravity can "bend" space, everything in that space is bent too."

As space and time are intimately intertwined, it is not just space that is bent but time as well. 
It is impossible to divorce 'time' from the effects of gravity and velocity.
« Last Edit: 22/01/2012 09:35:11 by MikeS »
 

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8676
  • Thanked: 42 times
    • View Profile
We already have clocks that (just about) run at measurably different rates on the top floor of the building, compared to the ground floor.
Any proper clock will do the same. What's the point of using some exotic alloy?
Surely it's easier to use something that isn't so critically dependent on the physical size of something?
For example, the wavelength of microwaves emitted or absorbed by some bunch of atoms is practically independent of the size of the container they are in.
The effect of gravity can be accounted for by using atoms that are in free fall at the time of the measurement.
http://en.wikipedia.org/wiki/Atomic_fountain
 

Offline MikeS

  • Neilep Level Member
  • ******
  • Posts: 1044
  • The Devils Advocate
    • View Profile
We already have clocks that (just about) run at measurably different rates on the top floor of the building, compared to the ground floor.
Any proper clock will do the same. What's the point of using some exotic alloy?
Surely it's easier to use something that isn't so critically dependent on the physical size of something?
For example, the wavelength of microwaves emitted or absorbed by some bunch of atoms is practically independent of the size of the container they are in.
The effect of gravity can be accounted for by using atoms that are in free fall at the time of the measurement.
http://en.wikipedia.org/wiki/Atomic_fountain


I can't remember where I found it (think the link is in one of my posts somewhere in these forums) but apparently the very latest atomic clocks (optical, I think) can measure the difference in height to about a centimetre if memory serves me correctly.  This is not the link I refered to but this link is about atomic cocks with an accuracy of less than a meter height measurement.
http://www.sciencemag.org/content/329/5999/1630.abstract
And this one 12in.
http://www.dailymail.co.uk/sciencetech/article-1314656/Scientists-prove-time-really-does-pass-quicker-higher-altitude.html
and this
http://news.sciencemag.org/sciencenow/2010/09/superaccurate-clocks-confirm-you.html

This takes into account the local gravitational field but does not (I believe) take into account the residual global (universal) gravitational field.

This is an interesting article on the latest optical lattice clocks
http://www.nature.com/nphoton/journal/v5/n4/full/nphoton.2011.45.html
 

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
All the "atomic" clocks measure the energy of the hyperfine transition state of various atoms, Hydrogen, Cesium, or Rubidium. 

I believe that it is not Time that is subject to the effects of relativity, but rather Energy that is affected by relativity.  Clocks based on the hyperfine transition state are extremely sensitive to the subtle energy changes with relativity, and thus have been offered as proof that time changes with relativity. 

The hyperfine transition state clocks are considered accurate to 15 decimal points, but all fail to remain synchronized if moved more than a few feet apart, and are fundamentally flawed for running tests like the speed of Neutrinos in the Opera-Gran Sasso experiment.

If one could find a material that did not expand with relativity, or could construct a Gridiron device that did not lengthen with relativity, then a clock based on length (reflecting light through a series of mirrors, then either measuring velocity or wavelength), would not be subject to relativity.

A device 1 meter long could reflect a laser through a series of mirrors, perhaps 2 mirrors 100 times, giving it an effective length of 100 meters.  A 300 GHZ oscillator would create an IR wavelength of about 1mm.  Reflecting it through 100 meters, it would effectively have 200,000 repeats.  If the wave synchronization could then be measured to a resolution of 50, then it would be accurate to 10,000,000, or 107.  Not too good compared Atomic Clocks, but perhaps one could pick up a couple extra zeros. 

The GPS clocks are adjusted about 1ns/day (10-9 seconds), so if one could make a length based clock accurate to 1010, it would be more accurate than the unadjusted GPS clocks, and likely could be used for light synchronization experiments which the hyperfine clocks don't do well with.
 

Offline MikeS

  • Neilep Level Member
  • ******
  • Posts: 1044
  • The Devils Advocate
    • View Profile
Personally, I believe that energy is one of the ingredients of time (along with gravity), so altering the energy level is likely to alter the time dilation factor.  As I see it, the going rate of time is still relativistic and I can't imagine any scenario that can divorce it from that.

I think the very best atomic clocks are accurate to 18 (?) decimal places.

 

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8676
  • Thanked: 42 times
    • View Profile
"The GPS clocks are adjusted about 1ns/day (10-9 seconds), so if one could make a length based clock accurate to 10^10, it would be more accurate than the unadjusted GPS clocks, "
Since the GPS clocks are adjusted, the " unadjusted GPS clocks" don't exist.
There are more than 10 seconds in a day.

1 nanosecond in a day is 1 in 86,400,000,000,000
About 1 in 10^15.
The best clocks are a good deal better, MikeS's figure of 1 in 10^18 sounds right.
The point is that if you made a "mechanical" clock that was accurate enough, you would find that it too varied with gravity and velocity in the same way as an atomic clock.

"If one could find a material that did not expand with relativity"
What are you going to measure it with?
You need to compare it to the definition of the metre.
That definition is expressed in terms of the speed of light.
Light is affected by relativity.

"The hyperfine transition state clocks are considered accurate to 15 decimal points, but all fail to remain synchronized if moved more than a few feet apart, "
Not if you move them slowly enough and keep them at the same height.

 

Offline CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6321
  • Thanked: 3 times
  • Site Moderator
    • View Profile
"If one could find a material that did not expand with relativity"
What are you going to measure it with?
You need to compare it to the definition of the metre.
That definition is expressed in terms of the speed of light.
As mentioned, if one knew proportionally how much different materials expanded, then one should be able to construct a Gridiron device to counter the linear expansion.  Or, one could target using a material that would be predicted to have low linear expansion due to relativity.  Then, if the clock was dependent on the linear expansion, it would be obvious how it was affected if placed into low earth orbit.
One would not have to re-measure the length standard at different orbit as the clock itself would make that measurement.
Knowing the correction factors for the clocks in LEO, one could, of course, use the corrected clock to measure length based on the current SI standards, then average it over a 24 hr period.
That definition is expressed in terms of the speed of light.
Light is affected by relativity.
According to relativity, the (two-way) speed of light is invariant.  The complex time stuff comes into play with the one-way speed of light.
"The hyperfine transition state clocks are considered accurate to 15 decimal points, but all fail to remain synchronized if moved more than a few feet apart, "
Not if you move them slowly enough and keep them at the same height.
You've already convinced me that I can't measure the one-way speed of light between two hilltops because the clocks fail to remain synchronized when moved, no matter how carefully it is done.  And, I can't use the clocks to reverse directions on a 12 hr half-day because, while they may be accurate to 10-18 seconds on a 24 hr day, they aren't on a 12 hr half-day.

Presumably a hyperfine transition based clock placed near the 45th parallel in the USA, one placed at the equator, and one placed at the South Pole would all drift apart, and would require adjustments based on their latitude. 

The south pole might not be a bad place to put one's standard because at least one could subtract out one additional velocity factor, but one would still have to communicate that data to other locations.
« Last Edit: 22/01/2012 19:24:43 by CliffordK »
 

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8676
  • Thanked: 42 times
    • View Profile
"The south pole might not be a bad place to put one's standard because at least one could subtract out one additional velocity factor, but one would still have to communicate that data to other locations."
What would be the point?
At that level of precision, time is a local thing. The time somewhere else passes at a different rate.
 

Offline yor_on

  • Naked Science Forum GOD!
  • *******
  • Posts: 12001
  • Thanked: 4 times
  • (Ah, yes:) *a table is always good to hide under*
    • View Profile
The speed of light in a vacuum will always be of the same measure, no matter from what frame of reference you measure it locally. And locally is the only way you can measure it directly.. So you can in fact split that in even chunks and find those chunks, aka 'clock beats', to have a same balance relative your life expectancy, no matter where you go, or how fast.

So it works.
 

The Naked Scientists Forum


 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums