Correlation between CMBR and Redshift Anisotropies.

On Fri, 18 Jul 2003 12:02:46 GMT, Sam Wormley wrote:

'' wrote:

Henri Wilson (HW@..), in article , wrote:

No they are not. The GPS is not designed to be a verification of GR.

It would hardly be the first time that data had more than one
interpretation, or that an implementation of one branch of science led
to data for another.

Ref: Hartle, "Gravity: An Introduction to Einstein's General
Relativity", Addison Wesley (2003)

The difference between rates at which signals are emitted and received
at two locations with different gravitational potentials is minute in
laboratory circumstances. Yet take these differences into account is
crucial for the operation of the Global Positioning System (GPS) used
every day. If the relativistic effects of time dilation and the
gravitational effects are not properly taken into account. the system
would fail after only a fraction of an hour.

Glasnos works perfectly well without it.


The GPS consists of a constellation of satellites, each in a half
sidereal day orbit about the Earth in a total of six orbital planes.
Each satellite carries accurate atomic clocks that keep proper time on
a satellite to accuracies of a few parts in 10^13 over a few weeks.
Corrections uploaded several times a day from the ground enable
accurate time to be kept over longer periods. The details of operation
of the system are complex, see for example the nearly 800 pages of
detail in Parkinson and Spilker (1996), but the basic idea is easily
explained in an idealization of the real situation.

Imagine an inertial frame in which the center of the Earth is
approximately at rest for the time it takes a signal to propagate from
a satellite to the ground. Periodically each satellite sends out
microwave signals encoded with the time and spacial location of
emission in the coordinates of the inertial frame. An observer that
receives a signal an interval of time later can calculate his or her
distance from the satellite by multiplying that time interval by the
speed of light c. By using the signal from three satellites the
observer's position in space can be narrowed down to the possible
intersection points of three spheres. By using four satellites, the
observer's position in both space and time can be fixed, even without
the observer possessing an accurate clock, giving a complete location
in spacetime. Signals from additional satellites reduce uncertainty
further.

So the only criteria are that all the satellite clocks are in close synch and
that their orbits are precisely known.
Your 'GR correction' is totally unnecessary.


Proper time on the satellite clocks has to be corrected to give the
time of the inertial frame for at least two reasons: time dilation of
special relativity and the effects of the Earth's gravitational field.
to understand this, suppose a GPS satellite emits signals at a constant
rate as measured by its clock. Suppose further that these are monitored
by a distant observer at rest in the inertial frame. A clock of this
observer, at rest and far from any source of gravitational effects,
measures the time of the inertial frame. The signals will be received
at a slower rate than they were emitted.

Absolute bulldust. Are you another believer in the 'tick fairies'?

Every tick emitted by the OC per orbit is counted by the ground receiver.
The GO knows the exact rate of the OC. There is no magical change in rate as
the signals fall to ground.

Time dilation of the moving
satellite clock is one reason. But another is the difference between
the rates of emission an reception because the satellite is lower in
the gravitational potential of the Earth than the distant observer. Two
corrections must therefore be applied to rate of satellite time to get
the time in the inertial frame.

The correction you are referring to is the one which acounts for the fact that
the signals accelerate as they fall. It is numerically the same as the
fictitious 'GR effect'.



These corrections are tiny by everyday standards, but a nanosecond is a
significant time in GPS operation. A signal from a satellite travels 30
cm in a nanosecond. To meet the announced 2-m accuracy for military
applications of the GPS, times and time differences must be known to
accuracies of approximately 6 ns. Keeping time to that accuracy is not
a problem for contemporary atomic clocks, but at these accuracies, both
time dilation and the gravitational redshift become important for GPS
operation.

The GR correction amounts to only 4cms per orbit. The error associated with
that is negligible. Regular corrections made to the clocks, based on empirical
readings, easily compensate for that.


The actual GPS does not employ an inertial frame whose time is defined
by clocks at infinity; rather it uses a frame rotating with the Earth
whose time is defined by clocks on its surface. The rates of the
satellite clocks must be corrected downward to keep the time of that
frame. Further corrections are needed for the relativistic Doppler
effect, the relativity of simultaneity, the Earth's rotation, the
asphericity of the Earth's gravitational potential, the time delays
from the index of refration of the Earth's ionosphere, satellite clock
errors, etc.

The clocks are pre-corrected to account for the fact that they change rates
when in free fall. This makes the system a lot easier to run because it is
obviously better if the OC clock rates are about the same as the ground ones.

Henri Wilson.

See my animations at:
http://www.users.bigpond.com/HeWn/index.htm

On Sat, 19 Jul 2003 10:23:44 +0100, "George Dishman"
wrote:

Henri, I'd like to trim down these posts so I'll press
you for answers on a couple of bits.

I don't see why simultaneity should enter into this.

To compare the rate of two clocks that to see if they
are equal, first synchronise a tick of one clock with
a tick of the other. If the next tick from the clocks
is also in sync, the clocks are ticking at the same
rate. The inequalities for slower or afster rates
follow. In your paradox the clocks are in relative
motion so even if the are co-located for a first
reference tick, they will all be separated for the
next tick, and if you have spatially separated clocks,
you have to take simultaneity into account.
...
You said you didn't see why simultaneity was relevant
so I hope you see the role it plays now.

Have I explained why it is involved?

In SR, simultaneity is not absolute. Whether Doppler
is relevant in your theory is not relevant. You cannot
disprove any theory if you start with an assumption that
conflicts with the theory. Your proof assumes absolute
simultaneity hence is invalid for SR.

Of course there is absolute simultaneity.

That is your opinion, I disagree.

It's just that one cannot establish
it if one uses light for communication.

However if two clcoks ar synchedthen moved apart in flat gravity, then they can
be assumed to be in absolute synch

No, we do not make such assumptions. You have to demonstrate
it by experiment or observation and that evidence supports SR.

Do you think rods actually change length when they are moved apart.

I can take a meter rod anywhere in the universe and be confident that any
lengths I measure with it are the same as those back on Earth.
The rod has not changed due to velocity. (It might have changed slightly due to
(measureable) gravitational compression variations).

If 'length' is absolute, then, by the same argument, so are time 'reading' and
time 'rate of change'.
Do you not accept 'length absolutivity'?


and, when at rest wrt each other, will
establish absolute simultaneity at their locations.

Agreed, they would _if_ clocks behaved as you wish.

Why should the clocks have changed their characteristics just because it is
impossible to test for such a change.

Are you claiming that the fact that one cannot prove the clocks DID NOT change
is proof that they did?



Fine, but none of that is relevant to SR so your proof
is still invalid. If you want to prove SR is internally
inconsistent, which is what you attempted above, you
must limit yourself to the rules of SR.

SR is a complete hoax.

Again that is your opinion, but your proof is _still_
invalid because it requires an assumption that is contrary
to what it tries to disprove. Can you confirm you understand
this point so we can move on please.

No. My proof that contractions are not 'real, physical' is perfectly valid.
They are observational effects.


It relies on the misapprehension that falling raindrops take longer too reach
the ground when you view them through your car window because they appear to
move diagonally.

I won't be visiting you then. It must be dangerous living
where the rain falls at the speed of light!

Do you know I could construct an equivalent theory to SR based on the postulate
that falling raindrops always appear to travel at the same speed no matter how
fast your car is moving.


The point remains, the clocks are moving as they pass
that point in orbit so there is a velocity component.

That is a purely Newtonian doppler effect.

Then you admit the GPS clocks are affected by their motion?

No. The clocks are not physically affected by their motion. Their readings have
to be adjusted for transverse doppler when observed from Earth, that is all.

The Doppler shift is not the same as the Newtonian version.
In particular the velocity component is not zero when the
motion is perpendicular to the line of sight.

That is the SRian view.
The Newtonian correction is quite easy to calculate.

Really? Please show me how you get a transverse Doppler
in Newtonian theory.

I was actually referring to the difference in doppler between when the GPS
clocks are on the equator and directly above. I did not actually say that there
was a transverse doppler effect in NM.


You have agreed that the clocks have
not PHYSICALLY changed due to their movement.

I have said they tick at the same rate per unit of
proper time. In my view that means they are not
physically changed but "physically changed" is
not a well defined term and some people would
disagree with me even though they would agree with
my more explicit statement, hence my care with the
wording.

The SRian use of the word 'proper' is just a way around admitting that time
rate of change is absolute.

It is a definition of a measurable quantity. Whether clocks
change rate or not is determined by actual measurements.

Very true. And one can always measure the rate of a clock which is at rest wrt
one.


In my opinion they are not physically affected. The rate
change comes from projecting from the world line of the
clock onto a coordinate axis.

But they clearly emit an increased number of ticks per orbit when they are up
there compared with before launch.

When the ticks are projected onto a ground-based coordinate
axis (i.e. where they are measured), the rate is indeed higher.
When you then use GR to find their rate against the clock's
world lines, we observe that the rate per unit of proper time
is unchanged. Hence I sy the clock is not physically affected
in that sense. In the sense that the ground-measured rate for
a clock in orbit differs from the rate for the same clock prior
to launch, it is physically affected. That is why I am careful
about wording my answer, "physically affected" can be ambiguous.

If the GR correction was right for all orbits, you would have a decent
argument. However there is no reliable information to support this.


That means they have increased their physical ticking rates.
It has nothing to do with 'time changes' or fancy world lines.

snip
[GLONASS] does not use a GR correction.

Really, I haven't found anything on the subject, can
you tell me where you found this information please.

I thought it was pretty common knowledge. I will try to find a reference.

If it were true, it would be headline news. Even the
slightest deviation from GR could be the key to a quantum
theory of gravity. If you can find your source I would
be fascinated (astonished!).

There is nothing to be astonished about. The clock readings are just
continuously software corrected. That wouldn't be all that difficult.



Of course, but it is also vital that they are accurately synchronised
to Earth time in a manner that allows users to use them as a time
reference. Most atomic clocks come with a GPS option for example.

Their signals are simply 'software corrected' for current drift.

The clocks are built with the GR correction in place and
the clock performance is much better than even the smaller
velocity-related component. As long as the mean of those
corrections is less than the uncertainty in the clock rate,
that gives the confirmation of GR.

Whatever the cause, the clocks rates are clearly observed to have physically
changed. I don't see how a maths theory could be responsible for that.


Exactly, and all that could be done with standard amateur
astronomy kit. You see, I really was trying to be helpful,
not just criticise your idea.

You have been helpful and I appreciate that.
Your idea is probably just about as feasible as Romer's light speed
measurements.

Thanks.

Synch two clocks together, move them apart and they should remain in absolute
synch.

Again, you assume nature "should" behave the way you want
it to. In fact they always maintain sync in the same way
that Einstein's method produces.

NO THEY DO NOT.

From the Hafele-Keating experiment to the BBC Christmas Lectures
a few years ago, every test I know of has been consistent with
YES THEY DO and outside the error bars of NO THEY DO NOT. Cite
your experimental evidence if you disagree.

Again, can you cite anything on this?

I just explained elsewhere, OWLS differs from TWLS by only 1 in (v/c)^2, where
c is true OWLS wrt a source. v is usually poretty small.


If you consider a rotating, four 45deg mirror configuration, the actual source
velocity doesn't play any part. It is the peripheral velocity of the first
mirror that would contribute to the first c+v effect. The analysis becomes a
bit messy after that because all the angles change slightly during the light
travel time.

They do but the speed of the light leaving each mirror is
c+v and the setup is symmetrical about the centre of each
light path so the light approaches the next mirror at c+v.
Since the mirror is moving at v, it approaches at a relative
speed of c. It leaves at c relative to the mirror hence c+v
again and so on. If the entire path is covered at c+v, there
should be no shift. The change in angles is something to
look at but the symmetry argument tends to cancel them out,
a greater launch angle corresponds to a greater arrival
angle at the next mirror.

But the light doesn't arrive at each mirror at c because the mirrors constantly
change angle so that their velocity is always slightly less that v when the
light reaches them.

In the co-rotating frame, the light arrives at c and the
mirror is not rotating hence there should be no fringe
shift.

I don't think ring gyros would work if we tried to explain them using
co-rotating frames.
I don't believe in rotating frames. Rotation definitely IS absolute.


In the lab frame it is not that simple, the speed would
be c+v at the centre of the beam, less than c+v on one
side and more than c+v on the other because the mirror
is rotating so you have to consider constructing
wavefronts or some equivalent method for predicting the
refelected angle and speed but the result must be the
same as in the co-rotating frame analysis since it is
just another way of predicting the same experiment.

The small angular variation also causes each reflected beam to be be slightly
off 90deg wrt its arrival direction.

Right, in the co-rotating frame the angles are reduced
from 90 degrees because the light paths appear bent. In
the lab frame, the angles are reduced because the path
is no longer square.

However, the change of angles is only needed so that the
beam hits the next mirror. It does not affect the speed
since, in both frames, there is still symmetry about the
normal to the mirror at the point of reflection. The
conclusion remains that there should be no fringe shift
from either of the effects you mention.

The thing is very messy.

It needs care certainly, but the symmetry makes it
tractable and the result from what I have seen is
always a prediction of no fringe shift, contrary to
what we observe.

I understand what you are saying. I will give it some thought. For instance how
does a 'rotating' as distinct from a 'linearly moving' 45 mirror affect the
velocity of light reflected from it?


George



Henri Wilson.

See my animations at:
http://www.users.bigpond.com/HeWn/index.htm

"Henri Wilson" HW@.. wrote in message
...
On Sat, 19 Jul 2003 12:27:13 GMT, "Minor Crank"

GR is not a religion. Tremendous effort is going on in attempts to
discover
where GR fails. It hasn't so far, but not for lack of trying. What do you
know about Gravity Probe B, or STEP? How about tests of the inverse
square
law at small distance?

Minor Crank


Crank you know my attitude. Light simply accelerates down a gravity well
just
like a lump of matter. Equations based on this principle seem to produce
the
same results as GR. So who needs it?

Problem is, your statement is false, as has been explained to you again, and
again, and again... You are totally ignorant of both the experimental data
and the math. That doesn't leave very much that you -do- understand.

Minor Crank

HW@..(Henri Wilson) wrote in message . ..
On Sat, 19 Jul 2003 12:27:13 GMT, "Minor Crank"
wrote:

"George Dishman" wrote in message
...

If it were true, it would be headline news. Even the
slightest deviation from GR could be the key to a quantum
theory of gravity. If you can find your source I would
be fascinated (astonished!).

To Henri:

George's statement is a -very- important point. We -know- that GR must fail
at some point, because GR and QM are incompatible. Also, GR predicts things
(like singularities) that many people are uncomfortable with, and which they
hope might "go away" in a more complete theory.

GR is not a religion. Tremendous effort is going on in attempts to discover
where GR fails. It hasn't so far, but not for lack of trying. What do you
know about Gravity Probe B, or STEP? How about tests of the inverse square
law at small distance?

Crank you know my attitude. Light simply accelerates down a gravity well just
like a lump of matter. Equations based on this principle seem to produce the
same results as GR. So who needs it?

The two are not the same. GPS sends the time to the
user by a radio signal encoded with the time data.
The change of frequency due to the gravitational
effect (as in Pound-Rebka) will alter the received
frequency but does not change the encoded data.

The GPS satellites when on the ground would run
about 38us per day slower than the ground reference
clocks. The time data would therefore drift by that
amount per day if the clocks weren't built with this
in mind and only the carrier frequency change taken
into account.

http://www-astronomy.mps.ohio-state....Unit5/gps.html

In practice, the Doppler effect as the satellites
approach and recede greatly outweighs the gravitational
frequency change of course.

I haven't found a direct reference yet but GLONASS
undoubtedly includes the same corrections for the
reasons given above.

George

HW@..(Henri Wilson) wrote in message . ..
Do you think rods actually change length when they are moved apart.

No. Do you think that relativity predicts that a rod changes
length when translated in position?

Perhaps you've forgotten key words like "reference frame"
and "relative velocity".

I can take a meter rod anywhere in the universe and be confident that any
lengths I measure with it are the same as those back on Earth.

Sure, so long as you keep it in your rest frame.

The rod has not changed due to velocity.

As you drag the rod around the universe, you aren't giving
it a relative velocity. You aren't doing the experiment.

If 'length' is absolute, then, by the same argument, so are time 'reading' and
time 'rate of change'.
Do you not accept 'length absolutivity'?

Nope. As Lorentz noted, the astronomical evidence is that
distances change when the rulers are in relative motion
to the observer, in an amount predicted by the Lorentz
transformation.

- Randy

"Henri Wilson" HW@.. wrote in message ...
On Sat, 19 Jul 2003 10:23:44 +0100, "George Dishman"
wrote:

Henri, I'd like to trim down these posts so I'll press
you for answers on a couple of bits.

I don't see why simultaneity should enter into this.

To compare the rate of two clocks that to see if they
are equal, first synchronise a tick of one clock with
a tick of the other. If the next tick from the clocks
is also in sync, the clocks are ticking at the same
rate. The inequalities for slower or afster rates
follow. In your paradox the clocks are in relative
motion so even if the are co-located for a first
reference tick, they will all be separated for the
next tick, and if you have spatially separated clocks,
you have to take simultaneity into account.
...
You said you didn't see why simultaneity was relevant
so I hope you see the role it plays now.

Have I explained why it is involved?

I'm sure you follow the above but can you please
confirm and then we can trim it out.

snip
Do you think rods actually change length when they are moved apart.

I take it you mean a change of speed rather than a change
of position. If so the answer is similar to that I gave
for clocks; the length of the rod measured in its own rest
frame will be unchanged. That is, speed does not physically
alter the rod. You can argue this easily, an observer
standing beside the rod and another moving past it will
measure different lengths for the same rod.

I can take a meter rod anywhere in the universe and be confident that any
lengths I measure with it are the same as those back on Earth.
The rod has not changed due to velocity. (It might have changed slightly due to
(measureable) gravitational compression variations).

... and temperature and air pressure and transient effects
of acceleration etc., etc. of course.

If 'length' is absolute, then, by the same argument, so are time 'reading' and
time 'rate of change'.
Do you not accept 'length absolutivity'?

I do not. Hold a ruler in front of a wall and measure the
length of the shadow. Now incline the ruler and measure
again. The ruler hasn't changed but the length of the
shadow has. The change is quite real ('physical' if you
like) but is not caused by a change in the length of the
ruler.

and, when at rest wrt each other, will
establish absolute simultaneity at their locations.

Agreed, they would _if_ clocks behaved as you wish.

Why should the clocks have changed their characteristics just because it is
impossible to test for such a change.

Are you claiming that the fact that one cannot prove the clocks DID NOT change
is proof that they did?

No, I am saying we _can_ measure them and we observe that they
produce the same number of ticks per unit of proper time (to
within the accuracy of the clock) regardless of their motion.
For example we _can_ and _do_ measure the GPS clocks and
correct them if required, and we find that the corrections have
no mean error compared to the GR prediction.

Fine, but none of that is relevant to SR so your proof
is still invalid. If you want to prove SR is internally
inconsistent, which is what you attempted above, you
must limit yourself to the rules of SR.

SR is a complete hoax.

Again that is your opinion, but your proof is _still_
invalid because it requires an assumption that is contrary
to what it tries to disprove. Can you confirm you understand
this point so we can move on please.

No.

OK, suppose there is a theory that all fairies are blue and
you want to prove it false:

Let us assume that some fairies are pink.

If some fairies are pink then not all fairies are blue.

Hence the theory "all fairies are blue" is false. QED.

This proof is invalid because it attempts to show internal
inconsistency but uses an assumption that contradicts the
original theory. Do you see what I am getting at?

The same applies to your proof. It attempts to show an
internal inconsistency in SR but to do so assumes
absolute simultaneity which is contrary to SR. Your
proof is therefore invalid.

My proof that contractions are not 'real, physical' is perfectly valid.
They are observational effects.

The length of a rod can be defined as the distance between
its ends. If you are not measuring in its rest frame, you
have to be clearer and say it is the distance between the
location of its ends noted *at the same time*. Again
simultaneity comes into play any attempt at falsification
that assumes absolute simultaneity would similarly be
invalid. I would not call the effect observational since
that implies a flaw in the measurement. In the case of
the ruler and shadow, the change of the length of the
shadow is real, not an error in the way we measure it.

It relies on the misapprehension that falling raindrops take longer too reach
the ground when you view them through your car window because they appear to
move diagonally.

I won't be visiting you then. It must be dangerous living
where the rain falls at the speed of light!

Do you know I could construct an equivalent theory to SR based on the postulate
that falling raindrops always appear to travel at the same speed no matter how
fast your car is moving.

You can construct a model on pretty much any set of
postulates but it only becomes a theory if you can show
it is self-consistent and matches reality within some
defined range of validity.

No. The clocks are not physically affected by their motion. Their readings have
to be adjusted for transverse doppler when observed from Earth, that is all.

The Doppler shift is not the same as the Newtonian version.
In particular the velocity component is not zero when the
motion is perpendicular to the line of sight.

That is the SRian view.
The Newtonian correction is quite easy to calculate.

Really? Please show me how you get a transverse Doppler
in Newtonian theory.

I was actually referring to the difference in doppler between when the GPS
clocks are on the equator and directly above. I did not actually say that there
was a transverse doppler effect in NM.

Oh well, I don't quite follow your argument over those
paragraphs in that case then but let's not dwell on it.
All I was saying is that since there is a Doppler effect,
you cannot treat the craft as stationary. They do, as you
said, return to the same point above the observer but
they are always moving at that point.

snip
In my opinion they are not physically affected. The rate
change comes from projecting from the world line of the
clock onto a coordinate axis.

But they clearly emit an increased number of ticks per orbit when they are up
there compared with before launch.

When the ticks are projected onto a ground-based coordinate
axis (i.e. where they are measured), the rate is indeed higher.
When you then use GR to find their rate against the clock's
world lines, we observe that the rate per unit of proper time
is unchanged. Hence I sy the clock is not physically affected
in that sense. In the sense that the ground-measured rate for
a clock in orbit differs from the rate for the same clock prior
to launch, it is physically affected. That is why I am careful
about wording my answer, "physically affected" can be ambiguous.

If the GR correction was right for all orbits, you would have a decent
argument.

My reason for being careful is simply to avoid misunderstandings.
I have seen protracted aruments in the past over this between
people who held exactly the same views on what happens but
disagreed about the meaning of "physically affected".

However there is no reliable information to support this.

There is reliable information for GPS which supports it. I don't
know what is available for GLONASS, it may not be public. Certainly
there is no evidence against it.


[GLONASS] does not use a GR correction.

Really, I haven't found anything on the subject, can
you tell me where you found this information please.

I thought it was pretty common knowledge. I will try to find a reference.

If it were true, it would be headline news. Even the
slightest deviation from GR could be the key to a quantum
theory of gravity. If you can find your source I would
be fascinated (astonished!).

There is nothing to be astonished about. The clock readings are just
continuously software corrected. That wouldn't be all that difficult.

No, I mean I would be astonished if someone had such proof
that relativity was wrong and hadn't cashed in on what would
be the discovery of the century. Do you have a source for
your claim that GLONASS does not use relativistic correction?

The clocks are built with the GR correction in place and
the clock performance is much better than even the smaller
velocity-related component. As long as the mean of those
corrections is less than the uncertainty in the clock rate,
that gives the confirmation of GR.

Whatever the cause, the clocks rates are clearly observed to have physically
changed. I don't see how a maths theory could be responsible for that.

True, the rates when observed on Earth are not the same as
when the clock was on the Earth. It almost is the same as the
shadow of the ruler I described earlier, the length of the
shadow really, really changes even though the ruler doesn't.
Well the rate measured on Earth really, really changes even
though the rate in the clock's rest frame doesn't. (I can
show it for SR but my skills are too limited for GR).

Again, you assume nature "should" behave the way you want
it to. In fact they always maintain sync in the same way
that Einstein's method produces.

NO THEY DO NOT.

From the Hafele-Keating experiment to the BBC Christmas Lectures
a few years ago, every test I know of has been consistent with
YES THEY DO and outside the error bars of NO THEY DO NOT. Cite
your experimental evidence if you disagree.

Again, can you cite anything on this?

I just explained elsewhere, OWLS differs from TWLS by only 1 in (v/c)^2, where
c is true OWLS wrt a source. v is usually poretty small.

Before you can show that, you need to define how you define
"OWLS". As I said at the beginning it depends on a definition
of simultaneity.

That said, to paraphrase your comment above "Are you claiming
that the fact that one cannot prove that OWLS differs from TWLS
is proof that they does?" Isn't that effectively what you were
claiming when you said so emphatically "NO THEY DO NOT."

snip: more later

George

"Henri Wilson" HW@.. wrote in message ...
On Sat, 19 Jul 2003 10:23:44 +0100, "George Dishman"
wrote:

In the co-rotating frame, the light arrives at c and the
mirror is not rotating hence there should be no fringe
shift.

I don't think ring gyros would work if we tried to explain them using
co-rotating frames.

I think they would still work if we tried to explain them
by throwing fish at a bicycle.

I don't believe in rotating frames. Rotation definitely IS absolute.

Agreed, but you can describe the situation and do the
maths in any frame you like provided you translate
between the frames correctly. You, I presumed, would use
Galilean relativity to do that. The maths is simply
easier in the co-rotating frame in this particular case.

Incidentally, that is the basis of the principle of
equivalence, for a self-consistent theory the conclusion
must be the same in any frame since there can be only one
outcome in reality.

In the lab frame it is not that simple, the speed would
be c+v at the centre of the beam, less than c+v on one
side and more than c+v on the other because the mirror
is rotating so you have to consider constructing
wavefronts or some equivalent method for predicting the
refelected angle and speed but the result must be the
same as in the co-rotating frame analysis since it is
just another way of predicting the same experiment.

The small angular variation also causes each reflected beam to be be slightly
off 90deg wrt its arrival direction.

Right, in the co-rotating frame the angles are reduced
from 90 degrees because the light paths appear bent. In
the lab frame, the angles are reduced because the path
is no longer square.

However, the change of angles is only needed so that the
beam hits the next mirror. It does not affect the speed
since, in both frames, there is still symmetry about the
normal to the mirror at the point of reflection. The
conclusion remains that there should be no fringe shift
from either of the effects you mention.

The thing is very messy.

It needs care certainly, but the symmetry makes it
tractable and the result from what I have seen is
always a prediction of no fringe shift, contrary to
what we observe.

I understand what you are saying. I will give it some thought. For instance how
does a 'rotating' as distinct from a 'linearly moving' 45 mirror affect the
velocity of light reflected from it?

I think it affects the angle of reflection more than the
speed but it all depends on your model. If you assume the
light is re-emitted from the mirror at c wrt the mirror
regardless of incident speed then you might get one
answer while if the outgoing speed matched the incoming
speed then you might get another.

In Sagnac, there is a symmetry that solves a lot of this
as I said. Since the light approaches each mirror at c
wrt the mirror, both approaches give a reflected speed
of c wrt the mirror.

To determine the angle you could draw wavefronts arriving,
step them forward a fraction of a wavelength at a time
and adjust the position of the mirror to allow for the
movement. The angle might change slightly, you would need
to work through that, but that will almost certainly give
a second order effect since the change of angle will depend
on the direction of rotation (I'm guessing but fairly
confident).

The same reasoning applies to the apparent increase in
path length because of the curvature of the paths in
the co-rotating frame. The path length is a minimum when
the paths are straight (v=0) hence must be of the form:

L = L_0 + k * v^2

so any effects must be second order at least.

Of course if you have a ballistic particle-like model for
the light then the speed and angle would be determined
solely by the angle at the instant of impact and motion
is irrelevant.

As far as I can see, all of these predict no fringe shift.

George

On 22 Jul 2003 05:44:46 -0700, (George G. Dishman)
wrote:

HW@..(Henri Wilson) wrote in message . ..
On Sat, 19 Jul 2003 12:27:13 GMT, "Minor Crank"
wrote:

"George Dishman" wrote in message
...

If it were true, it would be headline news. Even the
slightest deviation from GR could be the key to a quantum
theory of gravity. If you can find your source I would
be fascinated (astonished!).

To Henri:

George's statement is a -very- important point. We -know- that GR must fail
at some point, because GR and QM are incompatible. Also, GR predicts things
(like singularities) that many people are uncomfortable with, and which they
hope might "go away" in a more complete theory.

GR is not a religion. Tremendous effort is going on in attempts to discover
where GR fails. It hasn't so far, but not for lack of trying. What do you
know about Gravity Probe B, or STEP? How about tests of the inverse square
law at small distance?

Crank you know my attitude. Light simply accelerates down a gravity well just
like a lump of matter. Equations based on this principle seem to produce the
same results as GR. So who needs it?

The two are not the same. GPS sends the time to the
user by a radio signal encoded with the time data.
The change of frequency due to the gravitational
effect (as in Pound-Rebka) will alter the received
frequency but does not change the encoded data.

Oh Gord! We have another 'tick fairy' addict.

The GO counts every tick emitted by the OC per orbit, moron.
How can the frequency change on the way down?
Yes I know. The tick fairies.


The GPS satellites when on the ground would run
about 38us per day slower than the ground reference
clocks. The time data would therefore drift by that
amount per day if the clocks weren't built with this
in mind and only the carrier frequency change taken
into account.

The clocks speed up fractionally when in orbit, presumeably through being
relieved of internal gravitational stresses.


http://www-astronomy.mps.ohio-state....Unit5/gps.html

In practice, the Doppler effect as the satellites
approach and recede greatly outweighs the gravitational
frequency change of course.

I haven't found a direct reference yet but GLONASS
undoubtedly includes the same corrections for the
reasons given above.

It does not.
The 'GR correction' is totally unnecessary.


George


Henri Wilson.

See my animations at:
http://www.users.bigpond.com/HeWn/index.htm

On 22 Jul 2003 06:58:53 -0700, (Randy Poe) wrote:

HW@..(Henri Wilson) wrote in message . ..
Do you think rods actually change length when they are moved apart.

No. Do you think that relativity predicts that a rod changes
length when translated in position?

You people keep repeating that but you don't practice what you preach.
You say rods don't physically change but clocks apparently DO.
You are very inconsistent Randy.


Perhaps you've forgotten key words like "reference frame"
and "relative velocity".

I can take a meter rod anywhere in the universe and be confident that any
lengths I measure with it are the same as those back on Earth.

Sure, so long as you keep it in your rest frame.

Frames have nothing to do with the fact that the rod DOES NOT physically
change. (It might if there is an aether, of course)


The rod has not changed due to velocity.

As you drag the rod around the universe, you aren't giving
it a relative velocity. You aren't doing the experiment.

THE ROD DOES NOT CHANGE. NOR DO CLOCK RATES (in flat gravity).


If 'length' is absolute, then, by the same argument, so are time 'reading' and
time 'rate of change'.
Do you not accept 'length absolutivity'?

Nope. As Lorentz noted, the astronomical evidence is that
distances change when the rulers are in relative motion
to the observer, in an amount predicted by the Lorentz
transformation.

Bull. That is purely an observational effect.
The LT's are also Bull! The linear doppler term gives the correct answer.

- Randy


Henri Wilson.

See my animations at:
http://www.users.bigpond.com/HeWn/index.htm

On Tue, 22 Jul 2003 19:55:16 +0100, "George Dishman"
wrote:


"Henri Wilson" HW@.. wrote in message ...
On Sat, 19 Jul 2003 10:23:44 +0100, "George Dishman"
wrote:

Henri, I'd like to trim down these posts so I'll press
you for answers on a couple of bits.

I don't see why simultaneity should enter into this.

To compare the rate of two clocks that to see if they
are equal, first synchronise a tick of one clock with
a tick of the other. If the next tick from the clocks
is also in sync, the clocks are ticking at the same
rate. The inequalities for slower or afster rates
follow. In your paradox the clocks are in relative
motion so even if the are co-located for a first
reference tick, they will all be separated for the
next tick, and if you have spatially separated clocks,
you have to take simultaneity into account.
...
You said you didn't see why simultaneity was relevant
so I hope you see the role it plays now.

Have I explained why it is involved?

I'm sure you follow the above but can you please
confirm and then we can trim it out.

The RoS is purely a result of the fact that light takes time to travel. With
instantaneous communication, it disappears.

I.C. can be simulated by using a 'Wilsonian Rest Frame'.


snip
Do you think rods actually change length when they are moved apart.

I take it you mean a change of speed rather than a change
of position. If so the answer is similar to that I gave
for clocks; the length of the rod measured in its own rest
frame will be unchanged. That is, speed does not physically
alter the rod. You can argue this easily, an observer
standing beside the rod and another moving past it will
measure different lengths for the same rod.

Observers don't enter the picture. The rod length does not physically change no
matter how or where it goes(assuming no absolute aether). I can use the same
rod to measure length anywhere in the universe and it will give me a perfect
comparison with lengths on Earth..


I can take a meter rod anywhere in the universe and be confident that any
lengths I measure with it are the same as those back on Earth.
The rod has not changed due to velocity. (It might have changed slightly due to
(measureable) gravitational compression variations).

.. and temperature and air pressure and transient effects
of acceleration etc., etc. of course.

Assume it is a 'perfect' rod.


If 'length' is absolute, then, by the same argument, so are time 'reading' and
time 'rate of change'.
Do you not accept 'length absolutivity'?

I do not. Hold a ruler in front of a wall and measure the
length of the shadow. Now incline the ruler and measure
again. The ruler hasn't changed but the length of the
shadow has. The change is quite real ('physical' if you
like) but is not caused by a change in the length of the
ruler.

That's silly.

One uses the calibrations along the rod to compare lengths. Theoretically that
can be done anywhere in space.

and, when at rest wrt each other, will
establish absolute simultaneity at their locations.

Agreed, they would _if_ clocks behaved as you wish.

Why should the clocks have changed their characteristics just because it is
impossible to test for such a change.

Are you claiming that the fact that one cannot prove the clocks DID NOT change
is proof that they did?

No, I am saying we _can_ measure them and we observe that they
produce the same number of ticks per unit of proper time (to
within the accuracy of the clock) regardless of their motion.
For example we _can_ and _do_ measure the GPS clocks and
correct them if required, and we find that the corrections have
no mean error compared to the GR prediction.

The clocks obviously PHYSICALLY change when in free fall. That is what is
observed. I don't care what the reason is - but it certainly ain't a bit of
maths.


Fine, but none of that is relevant to SR so your proof
is still invalid. If you want to prove SR is internally
inconsistent, which is what you attempted above, you
must limit yourself to the rules of SR.

SR is a complete hoax.

Again that is your opinion, but your proof is _still_
invalid because it requires an assumption that is contrary
to what it tries to disprove. Can you confirm you understand
this point so we can move on please.

No.

OK, suppose there is a theory that all fairies are blue and
you want to prove it false:

Let us assume that some fairies are pink.

If some fairies are pink then not all fairies are blue.

Hence the theory "all fairies are blue" is false. QED.

This proof is invalid because it attempts to show internal
inconsistency but uses an assumption that contradicts the
original theory. Do you see what I am getting at?

The same applies to your proof. It attempts to show an
internal inconsistency in SR but to do so assumes
absolute simultaneity which is contrary to SR. Your
proof is therefore invalid.

Absolute simultaneity can be theoretically achieved using one of my rest
frames.
One can make instantaneous readings without EM then analyse the data
afterwards.


My proof that contractions are not 'real, physical' is perfectly valid.
They are observational effects.

The length of a rod can be defined as the distance between
its ends. If you are not measuring in its rest frame, you
have to be clearer and say it is the distance between the
location of its ends noted *at the same time*. Again
simultaneity comes into play any attempt at falsification
that assumes absolute simultaneity would similarly be
invalid. I would not call the effect observational since
that implies a flaw in the measurement. In the case of
the ruler and shadow, the change of the length of the
shadow is real, not an error in the way we measure it.

I don't wish to measure the rod at all. I merely want to use it to measure
other lengths. I can be confident that it remains physically the same no matter
how it moves. By aligning it with an unknown length, I can give a magnitude to
that unknown by using the rod's calibrations. That will give me an exact
comparison with lengths back on Earth. I can give the rod to any other observer
and he/she/it can do the same ANYWHERE, ANYTIME AND AT ANY SPEED.


It relies on the misapprehension that falling raindrops take longer too reach
the ground when you view them through your car window because they appear to
move diagonally.

I won't be visiting you then. It must be dangerous living
where the rain falls at the speed of light!

Do you know I could construct an equivalent theory to SR based on the postulate
that falling raindrops always appear to travel at the same speed no matter how
fast your car is moving.

You can construct a model on pretty much any set of
postulates but it only becomes a theory if you can show
it is self-consistent and matches reality within some
defined range of validity.

SR has to be self consistent because it is entirely circular. That doesn't mean
it matches reality.

The velocity addition equation is Einstein's masterstroke.
See my demo, www.users.bigpond.com/hewn/photons.exe


No. The clocks are not physically affected by their motion. Their readings have
to be adjusted for transverse doppler when observed from Earth, that is all.

The Doppler shift is not the same as the Newtonian version.
In particular the velocity component is not zero when the
motion is perpendicular to the line of sight.

That is the SRian view.
The Newtonian correction is quite easy to calculate.

Really? Please show me how you get a transverse Doppler
in Newtonian theory.

I was actually referring to the difference in doppler between when the GPS
clocks are on the equator and directly above. I did not actually say that there
was a transverse doppler effect in NM.

Oh well, I don't quite follow your argument over those
paragraphs in that case then but let's not dwell on it.
All I was saying is that since there is a Doppler effect,
you cannot treat the craft as stationary. They do, as you
said, return to the same point above the observer but
they are always moving at that point.

Yes. The doppler shift changes as the clocks move from the horizon to the
vertical. Straight Newton.


snip
In my opinion they are not physically affected. The rate
change comes from projecting from the world line of the
clock onto a coordinate axis.

But they clearly emit an increased number of ticks per orbit when they are up
there compared with before launch.

When the ticks are projected onto a ground-based coordinate
axis (i.e. where they are measured), the rate is indeed higher.
When you then use GR to find their rate against the clock's
world lines, we observe that the rate per unit of proper time
is unchanged. Hence I sy the clock is not physically affected
in that sense. In the sense that the ground-measured rate for
a clock in orbit differs from the rate for the same clock prior
to launch, it is physically affected. That is why I am careful
about wording my answer, "physically affected" can be ambiguous.

If the GR correction was right for all orbits, you would have a decent
argument.

My reason for being careful is simply to avoid misunderstandings.
I have seen protracted aruments in the past over this between
people who held exactly the same views on what happens but
disagreed about the meaning of "physically affected".

However there is no reliable information to support this.

The orbiting clocks are observed to have changed rates. This is a REAL change,
not something that happens as the 'ticks' fall to ground.




If it were true, it would be headline news. Even the
slightest deviation from GR could be the key to a quantum
theory of gravity. If you can find your source I would
be fascinated (astonished!).

There is nothing to be astonished about. The clock readings are just
continuously software corrected. That wouldn't be all that difficult.

No, I mean I would be astonished if someone had such proof
that relativity was wrong and hadn't cashed in on what would
be the discovery of the century. Do you have a source for
your claim that GLONASS does not use relativistic correction?

No. It has been mentioned here quite regularly though.


The clocks are built with the GR correction in place and
the clock performance is much better than even the smaller
velocity-related component. As long as the mean of those
corrections is less than the uncertainty in the clock rate,
that gives the confirmation of GR.

Whatever the cause, the clocks rates are clearly observed to have physically
changed. I don't see how a maths theory could be responsible for that.

True, the rates when observed on Earth are not the same as
when the clock was on the Earth. It almost is the same as the
shadow of the ruler I described earlier, the length of the
shadow really, really changes even though the ruler doesn't.
Well the rate measured on Earth really, really changes even
though the rate in the clock's rest frame doesn't. (I can
show it for SR but my skills are too limited for GR).

I have pointed out before, there exists a common clock: ONE COMPLETE ORBIT.
Both the GO and the OO can count the exact number of ticks emitted by the OC
per exact orbit.
They MUST get the same answer (assuming no tick fairies) and it will be
different from the ticks emitted by the GC per exact orbit.


Again, you assume nature "should" behave the way you want
it to. In fact they always maintain sync in the same way
that Einstein's method produces.

NO THEY DO NOT.

From the Hafele-Keating experiment to the BBC Christmas Lectures
a few years ago, every test I know of has been consistent with
YES THEY DO and outside the error bars of NO THEY DO NOT. Cite
your experimental evidence if you disagree.

Again, can you cite anything on this?

I just explained elsewhere, OWLS differs from TWLS by only 1 in (v/c)^2, where
c is true OWLS wrt a source. v is usually poretty small.

Before you can show that, you need to define how you define
"OWLS". As I said at the beginning it depends on a definition
of simultaneity.

You can forget simultaneity if you use one of my rest frames.
If you are not familiar with 'Wilsonian rest frames' they essentially consist
of an infinite grid of presynched and perfect clocks that are moved into their
positions with the same acceleration regimes. The spacing of the grid is
infinitesimal and is set using identical rods.
The clocks peform all my measurements and relay their times back to my central
computer which then constructs an 'instantaneous picture' of whatever happened.


That said, to paraphrase your comment above "Are you claiming
that the fact that one cannot prove that OWLS differs from TWLS
is proof that they does?" Isn't that effectively what you were
claiming when you said so emphatically "NO THEY DO NOT."

Einstein adjusts his clocks so that Tab always =Tba.

That ensures that OWLS =TWLS, but by definition only.


snip: more later

George



Henri Wilson.

See my animations at:
http://www.users.bigpond.com/HeWn/index.htm