Debunked by Proof: The Hafele&Keating RT Time Dilation Experiment of 1971

The famous RT Time Dilation Experiment by Hafele&Keating
was performed 1971. They flew in planes around the world and
measured the time dilation at an altitude about 10 km. They made two
roundtrips around the globe; one Eastwards and another westwards.

Hafele&Keating in their paper said they have measured the following values:
Eastflight: -59 nsec in 65.42 hours flight time
Westflight: +273 nsec in 80.33 hours flight time
(see http://tycho.usno.navy.mil/ptti/1971/Vol%2003_17.pdf )

But at that time the uncertainty rate of the atomic clocks
was 27.38 nanoseconds per day (see below and
http://tf.nist.gov/general/pdf/2056.pdf )

Because of that fact alone can their result not be accurate.
Let's do the math that proves them wrong:

For the EastFlight the uncertainty amounts to 65.42 / 24 * 27.38 = 74.63 nanoseconds :-)
-- this is even 1.26 factors more than what they claim to have measured!!!

For the Westflight the uncertainty amounts to 80.33 / 24 * 27.38 = 91.64 nanoseconds
-- this is even 1/3 of the allegedly measured value!!!

Ergo: the H&K results are useless trash as no statistician and
no statistical significance test can accept this inaccuracy!
For 36 years nobody saw this error in the data!

Even at wikipedia it is classified as "Problematic Physics Experiments":

"Hafele-Keating (1971) - time dilation in clocks flown around the world
The effect itself does not seem to be generally in dispute, but
questions have been raised about their approach and statistical
analysis, given the large degree of variation between clocks.
It is claimed (Kelly) that under the revised USNO guidelines issued
the following year, the H-K results would have had to have been
REJECTED AS UNRELIABLE. This does not seem to be disputed."
http://en.wikipedia.org/wiki/Problem...cs_experiments

"That result was contested by Dr. A. G. Kelly who examined the raw data:
according to him, the final published outcome had to be averaged in a
biased way in order to claim such a high precision. Also, Louis Essen,
the inventor of the atomic clock, published an article in which he discussed
the (in his opinion) inadequate accuracy of the experiment."

---------------------------------------------------------------
History Of Atomic Clocks:
Accuracy of NIST NBS-4 and NBS-5 (year 1968 to 197x or so)

From a text from the mid-1970'ies entitled
"Twenty-Five Years Later... tick... tick... tick... - Atomic Timekeeping"
http://tf.nist.gov/general/pdf/2056.pdf
"[...]
Since 1960, several generations of NBS cesium beam devices
have provided our nation and NBS with a primary frequency standard.

NBS-II, successor of the experimental device NBS-I, furnished
this standard from 1960 until 1963, being replaced by NBS-III,
which was used until 1969.

NBS-5 became the latest generation of NBS cesium atom primary
frequency standards in 1972.

NBS-4, although initially completed in 1970, has been revised and
is presently used as an independent primary frequency standard.

Forming a system which is mutually supportive, NBS-4 and NBS-5
provide accuracies approximately 100,000 times better than the
second of time as measured by our revolving earth.

If NBS-5 was allowed to run constantly for one million years
without adjustment, it would still be accurate to better than
ten seconds!

Future primary frequency sources may someday replace these
cesium beam standards, as cesium devices replaced the
ammonia molecular clock, but the concept of atomic clocks
will continue.
[...]"

So, in 1971/1972 the NBS-5 atomic clocks had an uncertainty of
10 / (1E6 * 365.25) = 27.38 nanoseconds per day.

---------------------------------------------------------------

"qbit" wrote in message
...
Debunked by Proof: The Hafele&Keating RT Time Dilation
Experiment of 1971

The famous RT Time Dilation Experiment by Hafele&
Keating was performed 1971. They flew in planes
around the world and measured the time dilation at an
altitude about 10 km. They made two roundtrips around
the globe; one Eastwards and another westwards.

Hafele&Keating in their paper said they have measured
the following values:
Eastflight: -59 nsec in 65.42 hours flight time
Westflight: +273 nsec in 80.33 hours flight time
(see http://tycho.usno.navy.mil/ptti/1971/Vol%2003_17.pdf )

But at that time the uncertainty rate of the atomic clocks
was 27.38 nanoseconds per day (see below and
http://tf.nist.gov/general/pdf/2056.pdf )

Because of that fact alone can their result not be accurate.
Let's do the math that proves them wrong:

Why? It has been repeated, even by a fellow on vacation, with
the family car and a mountaintop, with sufficient accuracy. GR
is correct in this respect, and you are still trolling...

David A. Smith

On Aug 8, 7:09 pm, "qbit" wrote:
[...]


"That result was contested by Dr. A. G. Kelly [...]

Just another retired engineer ****ting on relativity - nothing to see
here.

qbit wrote:
Debunked by Proof: The Hafele&Keating RT Time Dilation Experiment of 1971

The famous RT Time Dilation Experiment by Hafele&Keating
was performed 1971. They flew in planes around the world and
measured the time dilation at an altitude about 10 km. They made two
roundtrips around the globe; one Eastwards and another westwards.

I'm afraid that there is a huge mistake in your analysis:

Hafele&Keating in their paper said they have measured the following values:
Eastflight: -59 nsec in 65.42 hours flight time
Westflight: +273 nsec in 80.33 hours flight time
(see http://tycho.usno.navy.mil/ptti/1971/Vol%2003_17.pdf )

But at that time the uncertainty rate of the atomic clocks
was 27.38 nanoseconds per day (see below and
http://tf.nist.gov/general/pdf/2056.pdf )

Because of that fact alone can their result not be accurate.
Let's do the math that proves them wrong:

For the EastFlight the uncertainty amounts to 65.42 / 24 * 27.38 = 74.63 nanoseconds :-)
-- this is even 1.26 factors more than what they claim to have measured!!!

For the Westflight the uncertainty amounts to 80.33 / 24 * 27.38 = 91.64 nanoseconds
-- this is even 1/3 of the allegedly measured value!!!

Ergo: the H&K results are useless trash as no statistician and
no statistical significance test can accept this inaccuracy!
For 36 years nobody saw this error in the data!

You state that for 36 years no-one saw this error. But yet the article
you cite below clearly references Kelly's paper as published in 1996.
But simple arithmetic shows: 1996-1971=25.

Now let us analyze this data.

True time before error discovered: 25 years
Time you claim before error was discovered: 36 years
Percentage error in figure you claim: (36-25)/25*100% = 44%

That is, you are 1.44 times too large!!!!

Ergo: your results are useless trash and no statistician can accept this
inaccuracy!


Even at wikipedia it is classified as "Problematic Physics Experiments":

"Hafele-Keating (1971) - time dilation in clocks flown around the world
The effect itself does not seem to be generally in dispute, but
questions have been raised about their approach and statistical
analysis, given the large degree of variation between clocks.
It is claimed (Kelly) that under the revised USNO guidelines issued
the following year, the H-K results would have had to have been
REJECTED AS UNRELIABLE. This does not seem to be disputed."
http://en.wikipedia.org/wiki/Problem...cs_experiments

"That result was contested by Dr. A. G. Kelly who examined the raw data:
according to him, the final published outcome had to be averaged in a
biased way in order to claim such a high precision. Also, Louis Essen,
the inventor of the atomic clock, published an article in which he discussed
the (in his opinion) inadequate accuracy of the experiment."

---------------------------------------------------------------
History Of Atomic Clocks:
Accuracy of NIST NBS-4 and NBS-5 (year 1968 to 197x or so)

From a text from the mid-1970'ies entitled
"Twenty-Five Years Later... tick... tick... tick... - Atomic Timekeeping"
http://tf.nist.gov/general/pdf/2056.pdf
"[...]
Since 1960, several generations of NBS cesium beam devices
have provided our nation and NBS with a primary frequency standard.

NBS-II, successor of the experimental device NBS-I, furnished
this standard from 1960 until 1963, being replaced by NBS-III,
which was used until 1969.

NBS-5 became the latest generation of NBS cesium atom primary
frequency standards in 1972.

NBS-4, although initially completed in 1970, has been revised and
is presently used as an independent primary frequency standard.

Forming a system which is mutually supportive, NBS-4 and NBS-5
provide accuracies approximately 100,000 times better than the
second of time as measured by our revolving earth.

If NBS-5 was allowed to run constantly for one million years
without adjustment, it would still be accurate to better than
ten seconds!

Future primary frequency sources may someday replace these
cesium beam standards, as cesium devices replaced the
ammonia molecular clock, but the concept of atomic clocks
will continue.
[...]"

So, in 1971/1972 the NBS-5 atomic clocks had an uncertainty of
10 / (1E6 * 365.25) = 27.38 nanoseconds per day.

---------------------------------------------------------------

"Stephen Montgomery-Smith" wrote
qbit wrote:

Debunked by Proof: The Hafele&Keating RT Time Dilation Experiment of 1971

The famous RT Time Dilation Experiment by Hafele&Keating
was performed 1971. They flew in planes around the world and
measured the time dilation at an altitude about 10 km. They made two
roundtrips around the globe; one Eastwards and another westwards.

I'm afraid that there is a huge mistake in your analysis:

Hafele&Keating in their paper said they have measured the following values:
Eastflight: -59 nsec in 65.42 hours flight time
Westflight: +273 nsec in 80.33 hours flight time
(see http://tycho.usno.navy.mil/ptti/1971/Vol%2003_17.pdf )

But at that time the uncertainty rate of the atomic clocks
was 27.38 nanoseconds per day (see below and
http://tf.nist.gov/general/pdf/2056.pdf )

Because of that fact alone can their result not be accurate.
Let's do the math that proves them wrong:

For the EastFlight the uncertainty amounts to 65.42 / 24 * 27.38 = 74.63 nanoseconds :-)
-- this is even 1.26 factors more than what they claim to have measured!!!

For the Westflight the uncertainty amounts to 80.33 / 24 * 27.38 = 91.64 nanoseconds
-- this is even 1/3 of the allegedly measured value!!!

Ergo: the H&K results are useless trash as no statistician and
no statistical significance test can accept this inaccuracy!
For 36 years nobody saw this error in the data!

You state that for 36 years no-one saw this error. But yet the article
you cite below clearly references Kelly's paper as published in 1996.
But simple arithmetic shows: 1996-1971=25.

Now let us analyze this data.

True time before error discovered: 25 years
Time you claim before error was discovered: 36 years
Percentage error in figure you claim: (36-25)/25*100% = 44%

That is, you are 1.44 times too large!!!!

Ergo: your results are useless trash and no statistician can accept this
inaccuracy!

You are sort of off-topic and your math is based on wrong facts:
Kelly's criticizm is about how the data was compiled (ie. the
"averaging" of the data sets).
I on the other hand show that the accuracy of the atomic clocks in 1971
was not good enough to have really measured the said values.
It is impossible to measure such small values with such clocks
which have a such high uncertainty rate. Just do the math yourself! See above.

Even at wikipedia it is classified as "Problematic Physics Experiments":

"Hafele-Keating (1971) - time dilation in clocks flown around the world
The effect itself does not seem to be generally in dispute, but
questions have been raised about their approach and statistical
analysis, given the large degree of variation between clocks.
It is claimed (Kelly) that under the revised USNO guidelines issued
the following year, the H-K results would have had to have been
REJECTED AS UNRELIABLE. This does not seem to be disputed."
http://en.wikipedia.org/wiki/Problem...cs_experiments

"That result was contested by Dr. A. G. Kelly who examined the raw data:
according to him, the final published outcome had to be averaged in a
biased way in order to claim such a high precision. Also, Louis Essen,
the inventor of the atomic clock, published an article in which he discussed
the (in his opinion) inadequate accuracy of the experiment."

---------------------------------------------------------------
History Of Atomic Clocks:
Accuracy of NIST NBS-4 and NBS-5 (year 1968 to 197x or so)

From a text from the mid-1970'ies entitled
"Twenty-Five Years Later... tick... tick... tick... - Atomic Timekeeping"
http://tf.nist.gov/general/pdf/2056.pdf
"[...]
Since 1960, several generations of NBS cesium beam devices
have provided our nation and NBS with a primary frequency standard.

NBS-II, successor of the experimental device NBS-I, furnished
this standard from 1960 until 1963, being replaced by NBS-III,
which was used until 1969.

NBS-5 became the latest generation of NBS cesium atom primary
frequency standards in 1972.

NBS-4, although initially completed in 1970, has been revised and
is presently used as an independent primary frequency standard.

Forming a system which is mutually supportive, NBS-4 and NBS-5
provide accuracies approximately 100,000 times better than the
second of time as measured by our revolving earth.

If NBS-5 was allowed to run constantly for one million years
without adjustment, it would still be accurate to better than
ten seconds!

Future primary frequency sources may someday replace these
cesium beam standards, as cesium devices replaced the
ammonia molecular clock, but the concept of atomic clocks
will continue.
[...]"

So, in 1971/1972 the NBS-5 atomic clocks had an uncertainty of
10 / (1E6 * 365.25) = 27.38 nanoseconds per day.

---------------------------------------------------------------

"qbit" wrote in message
...
I on the other hand show that the accuracy of the atomic clocks in 1971
was not good enough to have really measured the said values.

What about since then .. has accuracy increased?

"Jeckyl" wrote
"qbit" wrote in message

I on the other hand show that the accuracy of the atomic clocks in 1971
was not good enough to have really measured the said values.

What about since then .. has accuracy increased?

Sure, see here http://tf.nist.gov/timefreq/cesium/atomichistory.htm

But the topic is about the accuracy of the H&K experiment only,
because that H&K experiment gets unethically marketed and
referenced by the RT lobbyists.
I mean RT people should be ashamed when they still give H&K as a reference...

"qbit" wrote in message
...
"Jeckyl" wrote
"qbit" wrote in message

I on the other hand show that the accuracy of the atomic clocks in 1971
was not good enough to have really measured the said values.

What about since then .. has accuracy increased?

Sure, see here http://tf.nist.gov/timefreq/cesium/atomichistory.htm

But the topic is about the accuracy of the H&K experiment only,

That experiment has been repeated with more accurate clocks and the results
confirmed.

because that H&K experiment gets unethically marketed and
referenced by the RT lobbyists.
I mean RT people should be ashamed when they still give H&K as a
reference...

From a cursory inspection of the paper, it seems that the definitely took
the accuracy issues etc into account. Of course, today we can get more
accurate readings and, as I understand, they confirm the relativity
predictions. Of course, with your preferred classical model,there would be
no difference, so your cases is looking even worse now than it was at the
time of H&K.

qbit wrote:
"Stephen Montgomery-Smith" wrote
qbit wrote:

Debunked by Proof: The Hafele&Keating RT Time Dilation Experiment of 1971

The famous RT Time Dilation Experiment by Hafele&Keating
was performed 1971. They flew in planes around the world and
measured the time dilation at an altitude about 10 km. They made two
roundtrips around the globe; one Eastwards and another westwards.
I'm afraid that there is a huge mistake in your analysis:

Hafele&Keating in their paper said they have measured the following values:
Eastflight: -59 nsec in 65.42 hours flight time
Westflight: +273 nsec in 80.33 hours flight time
(see http://tycho.usno.navy.mil/ptti/1971/Vol%2003_17.pdf )

But at that time the uncertainty rate of the atomic clocks
was 27.38 nanoseconds per day (see below and
http://tf.nist.gov/general/pdf/2056.pdf )

Because of that fact alone can their result not be accurate.
Let's do the math that proves them wrong:

For the EastFlight the uncertainty amounts to 65.42 / 24 * 27.38 = 74.63 nanoseconds :-)
-- this is even 1.26 factors more than what they claim to have measured!!!

For the Westflight the uncertainty amounts to 80.33 / 24 * 27.38 = 91.64 nanoseconds
-- this is even 1/3 of the allegedly measured value!!!

Ergo: the H&K results are useless trash as no statistician and
no statistical significance test can accept this inaccuracy!
For 36 years nobody saw this error in the data!
You state that for 36 years no-one saw this error. But yet the article
you cite below clearly references Kelly's paper as published in 1996.
But simple arithmetic shows: 1996-1971=25.

Now let us analyze this data.

True time before error discovered: 25 years
Time you claim before error was discovered: 36 years
Percentage error in figure you claim: (36-25)/25*100% = 44%

That is, you are 1.44 times too large!!!!

Ergo: your results are useless trash and no statistician can accept this
inaccuracy!

You are sort of off-topic and your math is based on wrong facts:
Kelly's criticizm is about how the data was compiled (ie. the
"averaging" of the data sets).
I on the other hand show that the accuracy of the atomic clocks in 1971
was not good enough to have really measured the said values.
It is impossible to measure such small values with such clocks
which have a such high uncertainty rate. Just do the math yourself! See above.

OK, you have succeeded in sucking me into a scientific argument (which I
shall shortly give). But let me first commend you on presenting a
rather better case than your previous effort (you remember, where you
express mock surprise because two incompatible theories give two
different answers!).

Now to the scientific argument - I must readily admit that I don't have
enough expertise to comment on the veracity of your statements
concerning 1971 atomic clocks. But the fact that Kelly spent his
efforts attacking the averaging process used in analyzing the data,
rather than attacking the fundamental limits of the accuracy of the
atomic clocks, is telling. Contrary to your earlier statement,
obtaining time differences of about 250ns with equipment with total
possible errors of up to 150ns is NOT statistical trash, but merely data
that requires statistical care.

If you look at the data presented in
http://en.wikipedia.org/wiki/Hafele-Keating_experiment, it is clear from
the fact that the scientists gave estimates of the errors, that they
were only too aware of the potential systematic errors introduced by
inaccuracies in the atomic clocks.

Qbit - I suspect that you haven't read the original paper of Hafele and
Keating. If you do read it, and it becomes apparent that they simply
didn't consider that the clocks were not totally accurate, then I'll be
more inclined to take you seriously. But really, the notion that you
are the first person to consider this possibility, and then after 36
whole years, is incredibly implausible.

And to go back off topic - I concede that you did win the verbal banter
this time. You did a good job keeping your cool. But next time - just
you watch out!!!

Best regards, Stephen

"Jeckyl" wrote
"qbit" wrote in message
"Jeckyl" wrote
"qbit" wrote in message

I on the other hand show that the accuracy of the atomic clocks in 1971
was not good enough to have really measured the said values.

What about since then .. has accuracy increased?

Sure, see here http://tf.nist.gov/timefreq/cesium/atomichistory.htm

But the topic is about the accuracy of the H&K experiment only,

That experiment has been repeated with more accurate clocks and the results
confirmed.

Nobody doubts the time dilation effect, at least not myself.
But I'm concerned that in RT it is not calculated correctly.
That's exactly another problem of RT I want to point to, see below.

because that H&K experiment gets unethically marketed and
referenced by the RT lobbyists.
I mean RT people should be ashamed when they still give H&K as a
reference...

From a cursory inspection of the paper, it seems that the definitely took
the accuracy issues etc into account.

You will need much more time to study their data.

Of course, today we can get more accurate readings and,
as I understand, they confirm the relativity predictions.

Time dilation is real, but the RT calculation method is wrong too, see below.

Of course, with your preferred classical model,there would be
no difference, so your cases is looking even worse now than it was at the
time of H&K.

You don't know the truth :-)
Soon I will debunk also the RT time dilation calculation method
that was and still gets used in all such tests, and even in the GPS system.
Stay tuned! :-)