PDA

View Full Version : Proper Time in a Rotating Relativistic Frame


Murat Ozer
Sep12-04, 02:23 AM
<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>Hello!\nI would like to bring to your attention an issue that has been\nbothering me for some time. It concerns the definition of proper time\nin a rotating system, like a turntable. In Part 1 of the following, I\nwill go over how proper time is defined in Special Relativity. In Part\n2, I will treat the relativistic rotating frame and point out that\nproper time for this case is not quite defined as in Part 1.\nPart 1: Special Relativity\nAs usual, consider two reference frames S and S\', S\' moving linearly\nin the x direction away from S with a velocity v. The infinitesimal\nLorentz transformations between S\' and S are:\n\ndt\'= (gamma)(dt - vdx), (1)\n\ndx\' = (gamma)(dx - vdt), (2)\n\ndy\' = dy, dz\' = dz, (3)\n\nwhere (gamma) = 1/sqrt(1-v^2) and c has been set to 1.\n\nThe invariant interval in S and S\' are\n\nds^2 = dt^2 - dx^2-dy^2 - dz^2, (4)\n\nds\'^2 = dt\'^2 - dx\'^2 -dy\'^2 - dz\'^2. (5)\n\nFactoring out dt^2 , equation (4) can be cast into\n\nds =sqrt(1 - v^2) dt (6)\n\nBy definition, the proper time dT of a clock at rest in S\' is found by\nsetting\ndx\' = dy\' = dz\' =0, and dt‘ = dT which results in\n\nds\' = dT (T = Tau). (7)\n\nComparing equations (6) and (7) gives, due to the invariance of the\ninterval,\n\ndT = sqrt(1 - v^2) dt. (8)\n\nIncidentially, the time dilation equation (8) can be obtained directly\nfrom the Lorentz transformations (1,(2), and (3). From (2),\n\n(gamma)dx = dx\' + (gamma)vdt.\n\nInserting this in (1) yields\n\ndt\'=sqrt(1- v^2)dt - vdx\'.\n\nNow setting dx\' =0 (along with dy\' = dz\' =0) and dt\' = dT gives (8).No\nsurprise…\n(End of part 1)\nPart2: Motion in a Rotating Relativistic Frame\nLet us now consider a frame at rest, S, and a second frame S\' rotating\nabout the z axis - a turntable - with a constant angular velocity w.\nIn cylindrical coordinates the interval in S is\n\nds^2 = dt^2 - dr^2 - r^2d(phi)^2 - dz^2. (9)\n\nThe coordinates of S\' are related to those of S by the\ntransformations\n\ndt\' = dt, (10)\n\nd(phi)\' = d(phi) - wdt, (11)\n\ndr\' = dr, dz\' = dz, (12)\n\nin terms of which the interval in S\' becomes\n\nds\'^2 = (1 - w^2r‘^2)dt‘^2 - dr‘^2 - r\'^2d(phi)\'^2\n- 2wr\'^2d(phi)‘dt\' - dz\'^2, (13)\n\nwhere S and S\' have the same common origin so that r = r\' and z = z\'.\n\nEquation (9) can be cast into\n\nds^2 = (1 - dr^2/dt^2 - r^2d(phi)^2/dt^2 - dz^2/dt^2) dt^2,\n= (1 - 0 - r^2w^2 - 0) dt^2 = (1 - w^2r^2) dt^2.\nOr,\n\nds = sqrt(1- w^2r^2)dt, (14)\n\nwhich corresponds to equation (6) above.\n\nNow, for a clock at rest on the turntable dr\' = d(phi)\' = dz\' =0 so\nthat\n\nds\'^2 = (1 - w^2r^2) dt\'^2, (15)\n\nand\n\nds\' = sqrt(1 - w^2r^2) dt\' (16)\n\nIn the literature the proper time of the clock at rest on the\nturntable is defined as\n\nds = ds\' = dT (17)\n\nso that (14) becomes\n\ndT = sqrt(1 - w^2r^2) dt (18)\n\nand (16) becomes\n\ndT =sqrt(1 - w^2r^2) dt\' (19)\n\nEquation (18) is the one quoted in the literature. Defining the proper\ntime this way, dt\' looses its meaning that for a clock at rest on the\nturntable, dt\' is its proper time! Indeed, equation (19) reveals that\nfor a clock at rest on the turntable dt\' is not the proper time. Both\ndT and dt‘ refer to the same clock at rest, and they have different\nmeanings. IS NOT THIS IN CONTRADICTION WITH THE DEFINITION OF PROPER\nTIME?\n\nOn the other hand, if we stick to the definition of proper time and\ndefine it for a clock at rest on the turntable as dt\' = dT, equation\n(16) becomes\n\nds\'= sqrt(1 - w^2r^2)dT , (20)\n\nwhich when compared with (14) gives\n\ndT = dt (21)\n\nIn other words, there is no time dilation! This is consistent with the\ntransformation equation (10) above, which would directly give dT = dt.\n\nJohn B. Kogut, in his book "Introduction to Relativity" treats motion\nin a rotating, relativistic frame in Section 7.2, p.100. Having\neliminated the d(phi)\'dt\' cross term in equation (13) above he writes\n\nds^2 = (c^2 - w^2r^2)dt\'^2-(dr^2 + c^2r^2d(phi)\'^2/(c^2-w^2r^2)+\ndz^2),\n\nhis equation (7.2.6). He then says\n\n" In order to interpret Eq.(7.2.6), we can compare it to the invariant\ninterval of an inertial frame of reference chosen to approximate the\ntransverse velocity wr locally. For example, taking a clock at rest in\nthe rotating frame,\n\nds^2 = (c^2 - w^2r^2)dt\'^2,\n\nand comparing it to a clock at rest in a locally inertial frame,\n\nds^2 = c^2dT^2,\nwe have time dilation again,\n\ndT=sqrt(1-w^2r^2/c^2)dt\'. "\n\nIs this a bad phrasing or a bad mistake?\nNote that even though the last expression in Kogut\'s statement has the\nsame form as my equation (19) above, the meaning of dT in the two\nequations are quite different.\nComments on Kogut\'s statement and my thinking above will be\nappreciated very much.\n\nRegards,\n\nMurat Ozer\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">&nbsp;&nbsp;View this Usenet post in original ASCII form </a></div><P></jabberwocky>Hello!
I would like to bring to your attention an issue that has been
bothering me for some time. It concerns the definition of proper time
in a rotating system, like a turntable. In Part 1 of the following, I
will go over how proper time is defined in Special Relativity. In Part
2, I will treat the relativistic rotating frame and point out that
proper time for this case is not quite defined as in Part 1.
Part 1: Special Relativity
As usual, consider two reference frames S and S', S' moving linearly
in the x direction away from S with a velocity v. The infinitesimal
Lorentz transformations between S' and S are:

dt'= (\gamma)(dt -[/itex] vdx), (1)

dx' = (\gamma)(dx - vdt), (2)

dy' = dy, dz' = dz, (3)

where (\gamma) = 1/\sqrt(1-v^2) and c has been set to 1.

The invariant interval in S and S' are

ds^2 = dt^2 - dx^2-dy^2 - dz^2, (4)

ds'^2 = dt'^2 - dx'^2 -dy'^2 - dz'^2. (5)

Factoring out dt^2 , equation (4) can be cast into

ds =\sqrt(1 - v^2) dt (6)

By definition, the proper time dT of a clock at rest in S' is found by
setting
dx' = dy' = dz' =0, and dt‘ = dT which results in

ds' = dT (T = \Tau). (7)

Comparing equations (6) and (7) gives, due to the invariance of the
interval,

dT = \sqrt(1 - v^2) dt. (8)

Incidentially, the time dilation equation (8) can be obtained directly
from the Lorentz transformations (1,(2), and (3). From (2),

(\gamma)dx = dx' + (\gamma)vdt.

Inserting this in (1) yields

dt'=\sqrt(1- v^2)dt - vdx'.

Now setting dx' =0 (along with dy' = dz' =0) and dt' = dT gives (8).No
surprise…
(End of part 1)
Part2: Motion in a Rotating Relativistic Frame
Let us now consider a frame at rest, S, and a second frame S' rotating
about the z axis - a turntable - with a constant angular velocity w.
In cylindrical coordinates the interval in S is

ds^2 = dt^2 - dr^2 - r^{2d}(\phi)^2 - dz^2. (9)

The coordinates of S' are related to those of S by the
transformations

dt' = dt, (10)

d(\phi)' = d(\phi) - wdt, (11)

dr' = dr, dz' = dz, (12)

in terms of which the interval in S' becomes

ds'^2 = (1 - w^{2r}‘^2)dt‘^2 - dr‘^2 - r'^2d(\phi)'^2- 2wr'^2d(\phi)‘dt' - dz'^2, (13)

where S and S' have the same common origin so that r = r' and z = z'.

Equation (9) can be cast into

ds^2 = (1 - dr^2/dt^2 - r^{2d}(\phi)^2/dt^2 - dz^2/dt^2) dt^2,= (1 -- r^{2w}^2 - 0) dt^2 = (1 - w^{2r}^2) dt^2.
Or,

ds = \sqrt(1- w^{2r}^2)dt, (14)

which corresponds to equation (6) above.

Now, for a clock at rest on the turntable dr' = d(\phi)' = dz' =0 so
that

ds'^2 = (1 - w^{2r}^2) dt'^2, (15)

and

ds' = \sqrt(1 - w^{2r}^2) dt' (16)

In the literature the proper time of the clock at rest on the
turntable is defined as

ds = ds' = dT (17)

so that (14) becomes

dT = \sqrt(1 - w^{2r}^2) dt (18)

and (16) becomes

dT =\sqrt(1 - w^{2r}^2) dt' (19)

Equation (18) is the one quoted in the literature. Defining the proper
time this way, dt' looses its meaning that for a clock at rest on the
turntable, dt' is its proper time! Indeed, equation (19) reveals that
for a clock at rest on the turntable dt' is not the proper time. Both
dT and dt‘ refer to the same clock at rest, and they have different
meanings. IS NOT THIS IN CONTRADICTION WITH THE DEFINITION OF PROPER
TIME?

On the other hand, if we stick to the definition of proper time and
define it for a clock at rest on the turntable as dt' = dT, equation
(16) becomes

ds'= \sqrt(1 - w^{2r}^2)dT , (20)

which when compared with (14) gives

dT = dt (21)

In other words, there is no time dilation! This is consistent with the
transformation equation (10) above, which would directly give dT = dt.

John B. Kogut, in his book "Introduction to Relativity" treats motion
in a rotating, relativistic frame in Section 7.2, p.100. Having
eliminated the d(\phi)'dt' cross term in equation (13) above he writes

[itex]ds^2 = (c^2 - w^{2r}^2)dt'^2-(dr^2 + c^{2r}^2d(\phi)'^2/(c^2-w^{2r}^2)+dz^2),

his equation (7.2.6). He then says

" In order to interpret Eq.(7.2.6), we can compare it to the invariant
interval of an inertial frame of reference chosen to approximate the
transverse velocity wr locally. For example, taking a clock at rest in
the rotating frame,

ds^2 = (c^2 - w^{2r}^2)dt'^2,

and comparing it to a clock at rest in a locally inertial frame,

ds^2 = c^{2dT}^2,
we have time dilation again,

dT=\sqrt(1-w^{2r}^2/c^2)dt'. "

Is this a bad phrasing or a bad mistake?
Note that even though the last expression in Kogut's statement has the
same form as my equation (19) above, the meaning of dT in the two
equations are quite different.
Comments on Kogut's statement and my thinking above will be
appreciated very much.

Regards,

Murat Ozer

Bjorn Danielsson
Sep14-04, 12:25 PM
<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nm_h_ozer@hotmail.com (Murat Ozer) wrote:\n&gt;\n&gt; The coordinates of S\' are related to those of S by the\n&gt; transformations\n&gt;\n&gt; dt\' = dt, (10)\n&gt; d(phi)\' = d(phi) - wdt, (11)\n&gt; dr\' = dr, dz\' = dz, (12)\n\nThe above is a local Galilean transform in cylindrical coordinates.\nFor a local Lorentz transform, (10) and (11) must be modified into:\n\ndt\' = gamma(r*w)*(dt - r^2*w*d(phi))\nd(phi)\' = gamma(r*w)*(d(phi) - w*dt)\n\nwhere gamma(v) = 1/sqrt(1-v^2), and assuming that phi\' is defined so\nthat r\'*d(phi)\' is a reasonable measure of infinitesimal length in the\nrotating frame.\n\n--\nBjorn Danielsson &lt;bonus@algonet.se&gt;\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">&nbsp;&nbsp;View this Usenet post in original ASCII form </a></div><P></jabberwocky>m_{h_ozer}@hotmail.com (Murat Ozer) wrote:
>
> The coordinates of S' are related to those of S by the
> transformations
>
> dt' = dt, (10)
> d(\phi)' = d(\phi) - wdt, (11)
> dr' = dr, dz' = dz, (12)

The above is a local Galilean transform in cylindrical coordinates.
For a local Lorentz transform, (10) and (11) must be modified into:

dt' = \gamma(r*w)*(dt - r^2*w*d(\phi))d(\phi)' = \gamma(r*w)*(d(\phi) - w*dt)

where \gamma(v) = 1/\sqrt(1-v^2), and assuming that \phi' is defined so
that r'*d(\phi)' is a reasonable measure of infinitesimal length in the
rotating frame.

--
Bjorn Danielsson <bonus@algonet.se>

Murat Ozer
Sep16-04, 07:09 AM
<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nBjorn Danielsson &lt;bonus@algonet.se&gt; wrote in message news:&lt;yqf1xh6up7u.fsf_-_@knights-who-say-ni&gt;...\n&gt; m_h_ozer@hotmail.com (Murat Ozer) wrote:\n&gt; &gt;\n&gt; &gt; The coordinates of S\' are related to those of S by the\n&gt; &gt; transformations\n&gt; &gt;\n&gt; &gt; dt\' = dt, (10)\n&gt; &gt; d(phi)\' = d(phi) - wdt, (11)\n&gt; &gt; dr\' = dr, dz\' = dz, (12)\n&gt;\n&gt; The above is a local Galilean transform in cylindrical coordinates.\n&gt; For a local Lorentz transform, (10) and (11) must be modified into:\n&gt;\n&gt; dt\' = gamma(r*w)*(dt - r^2*w*d(phi))\n&gt; d(phi)\' = gamma(r*w)*(d(phi) - w*dt)\n&gt;\n&gt; where gamma(v) = 1/sqrt(1-v^2), and assuming that phi\' is defined so\n&gt; that r\'*d(phi)\' is a reasonable measure of infinitesimal length in the\n&gt; rotating frame.\n\n\nThanks Bjorn.\nThe transformation equations\n\ndt\' = dt, (10)\nd(phi)\' = d(phi) - wdt, (11)\ndr\' = dr, dz\' = dz, (12)\n\nare definitely Galilean and immediately imply that\n\ndt\'= dT = dt (from (10) ), and\n\nr\'d(phi)\' = rd(phi) (from (11).\n\nIn other words, there can be neither time dilation nor length\ncontraction under these transformations. As I have argued in my\nprevious message, an alleged time dilation is obtained in the rotating\nframe by altering the definition of proper time, which should be the\ntime recorded by a clock at rest in the rotating frame. Instead of\nthis definition, people simply define\n\nds = dT (i)\n\nas in Special Relativity. Equation (i) works in SR, but as I have\nargued it,sometimes,leads to inconsistancies in GR. The locally\nLorentz transformations\n\ndt\' = gamma(dt - r^2*w*d(phi)) (ii)\nd(phi)\' = gamma(d(phi) - w*dt) (iii),\n\ngive\n\nds\'^2 = dt\'^2 - r\'^2d(phi)\'^2 - dr\'^2 - dz\'^2,\n\nwhich is just what is expected under Lorenz transformations. Equations\n(ii) and (iii) lead to the correct time dilation and length\ncontraction:\n\ndt\'= dT = gamma dt, while d(phi)\'=dr\'=dz\',\nr\'d(phi}\'=gamma rd(phi), while dt\'=0.\n\nEverybody will agree on the correctness of the last two results. How\nis it then possible that the Galilean transformations (10) and (11)\nabove also lead to these very same results, namely time dilation and\nlength contraction?\n\nComments please...\n\n\nRegards,\n\n\nMurat Ozer\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">&nbsp;&nbsp;View this Usenet post in original ASCII form </a></div><P></jabberwocky>Bjorn Danielsson <bonus@algonet.se> wrote in message news:<yqf1xh6up7u.fsf_-_@knights-who-say-ni>...
> m_{h_ozer}@hotmail.com (Murat Ozer) wrote:
> >
> > The coordinates of S' are related to those of S by the
> > transformations
> >
> > dt' = dt, (10)
> > d(\phi)' = d(\phi) - wdt, (11)
> > dr' = dr, dz' = dz, (12)
>
> The above is a local Galilean transform in cylindrical coordinates.
> For a local Lorentz transform, (10) and (11) must be modified into:
>
> dt' = \gamma(r*w)*(dt - r^2*w*d(\phi))
> d(\phi)' = \gamma(r*w)*(d(\phi) - w*dt)
>
> where \gamma(v) = 1/\sqrt(1-v^2), and assuming that \phi' is defined so
> that r'*d(\phi)' is a reasonable measure of infinitesimal length in the
> rotating frame.


Thanks Bjorn.
The transformation equations

dt' = dt, (10)
d(\phi)' = d(\phi) - wdt, (11)
dr' = dr, dz' = dz, (12)

are definitely Galilean and immediately imply that

dt'= dT = dt[/itex] (from (10) ), and

r'd(\phi)' = rd(\phi) (from (11).

In other words, there can be neither time dilation nor length
contraction under these transformations. As I have argued in my
previous message, an alleged time dilation is obtained in the rotating
frame by altering the definition of proper time, which should be the
time recorded by a clock at rest in the rotating frame. Instead of
this definition, people simply define

ds = dT (i)

as in Special Relativity. Equation (i) works in SR, but as I have
argued it,sometimes,leads to inconsistancies in GR. The locally
Lorentz transformations

dt' = \gamma(dt - r^2*w*d(\phi)) (ii)
d(\phi)' = \gamma(d(\phi) - w*dt) (iii),

give

[itex]ds'^2 = dt'^2 - r'^2d(\phi)'^2 - dr'^2 - dz'^2,

which is just what is expected under Lorenz transformations. Equations
(ii) and (iii) lead to the correct time dilation and length
contraction:

dt'= dT = \gamma dt, while d(\phi)'=dr'=dz',r'd(\phi}'=\gamma rd(\phi), while dt'=0.

Everybody will agree on the correctness of the last two results. How
is it then possible that the Galilean transformations (10) and (11)
above also lead to these very same results, namely time dilation and
length contraction?

Comments please...


Regards,


Murat Ozer

Murat Ozer
Sep17-04, 05:32 AM
<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nm_h_ozer@hotmail.com (Murat Ozer) wrote in message news:&lt;33971fb8.0409150920.20c0fb43@posting.google. com&gt;...\n&gt; Bjorn Danielsson &lt;bonus@algonet.se&gt; wrote in message news:&lt;yqf1xh6up7u.fsf_-_@knights-who-say-ni&gt;...\n&gt; &gt; m_h_ozer@hotmail.com (Murat Ozer) wrote:\n&gt; &gt; &gt;\n&gt; &gt; &gt; The coordinates of S\' are related to those of S by the\n&gt; &gt; &gt; transformations\n&gt; &gt; &gt;\n&gt; &gt; &gt; dt\' = dt, (10)\n&gt; &gt; &gt; d(phi)\' = d(phi) - wdt, (11)\n&gt; &gt; &gt; dr\' = dr, dz\' = dz, (12)\n&gt; &gt;\n&gt; &gt; The above is a local Galilean transform in cylindrical coordinates.\n&gt; &gt; For a local Lorentz transform, (10) and (11) must be modified into:\n&gt; &gt;\n&gt; &gt; dt\' = gamma(r*w)*(dt - r^2*w*d(phi))\n&gt; &gt; d(phi)\' = gamma(r*w)*(d(phi) - w*dt)\n&gt; &gt;\n&gt; &gt; where gamma(v) = 1/sqrt(1-v^2), and assuming that phi\' is defined so\n&gt; &gt; that r\'*d(phi)\' is a reasonable measure of infinitesimal length in the\n&gt; &gt; rotating frame.\n&gt;\n&gt;\n&gt; Thanks Bjorn.\n&gt; The transformation equations\n&gt;\n&gt; dt\' = dt, (10)\n&gt; d(phi)\' = d(phi) - wdt, (11)\n&gt; dr\' = dr, dz\' = dz, (12)\n&gt;\n&gt; are definitely Galilean and immediately imply that\n&gt;\n&gt; dt\'= dT = dt (from (10) ), and\n&gt;\n&gt; r\'d(phi)\' = rd(phi) (from (11).\n&gt;\n&gt; In other words, there can be neither time dilation nor length\n&gt; contraction under these transformations. As I have argued in my\n&gt; previous message, an alleged time dilation is obtained in the rotating\n&gt; frame by altering the definition of proper time, which should be the\n&gt; time recorded by a clock at rest in the rotating frame. Instead of\n&gt; this definition, people simply define\n&gt;\n&gt; ds = dT (i)\n&gt;\n&gt; as in Special Relativity. Equation (i) works in SR, but as I have\n&gt; argued it,sometimes,leads to inconsistancies in GR. The locally\n&gt; Lorentz transformations\n&gt;\n&gt; dt\' = gamma(dt - r^2*w*d(phi)) (ii)\n&gt; d(phi)\' = gamma(d(phi) - w*dt) (iii),\n&gt;\n&gt; give\n&gt;\n&gt; ds\'^2 = dt\'^2 - r\'^2d(phi)\'^2 - dr\'^2 - dz\'^2,\n&gt;\n&gt; which is just what is expected under Lorenz transformations. Equations\n&gt; (ii) and (iii) lead to the correct time dilation and length\n&gt; contraction:\n&gt;\n&gt; dt\'= dT = gamma dt, while d(phi)\'=dr\'=dz\',\n&gt; r\'d(phi}\'=gamma rd(phi), while dt\'=0.\n&gt;\n\nThere is a typo error in these equations. The gamma\'s are supposed to be\ngamma^(-1) so that the correct equations are\n\ndt\'= dT = gamma^(-1)dt= sqrt(1-r^2w^2)dt,\nr\'d(phi}\'=gamma^(-1)rd(phi)= sqrt(1-r^2w^2)rd(phi).\n\n&gt; Everybody will agree on the correctness of the last two results. How\n&gt; is it then possible that the Galilean transformations (10) and (11)\n&gt; above also lead to these very same results, namely time dilation and\n&gt; length contraction?\n&gt;\n&gt; Comments please...\n&gt;\n&gt;\n&gt; Regards,\n&gt;\n&gt;\n&gt; Murat Ozer\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">&nbsp;&nbsp;View this Usenet post in original ASCII form </a></div><P></jabberwocky>m_{h_ozer}@hotmail.com (Murat Ozer) wrote in message news:<33971fb8.0409150920.20c0fb43@posting.google.com>...
> Bjorn Danielsson <bonus@algonet.se> wrote in message news:<yqf1xh6up7u.fsf_-_@knights-who-say-ni>...
> > m_{h_ozer}@hotmail.com (Murat Ozer) wrote:
> > >
> > > The coordinates of S' are related to those of S by the
> > > transformations
> > >
> > > dt' = dt, (10)
> > > d(\phi)' = d(\phi) - wdt, (11)
> > > dr' = dr, dz' = dz, (12)
> >
> > The above is a local Galilean transform in cylindrical coordinates.
> > For a local Lorentz transform, (10) and (11) must be modified into:
> >
> > dt' = \gamma(r*w)*(dt - r^2*w*d(\phi))
> > d(\phi)' = \gamma(r*w)*(d(\phi) - w*dt)
> >
> > where \gamma(v) = 1/\sqrt(1-v^2), and assuming that \phi' is defined so
> > that r'*d(\phi)' is a reasonable measure of infinitesimal length in the
> > rotating frame.
>
>
> Thanks Bjorn.
> The transformation equations
>
> dt' = dt, (10)
> d(\phi)' = d(\phi) - wdt, (11)
> dr' = dr, dz' = dz, (12)
>
> are definitely Galilean and immediately imply that
>
> dt'= dT = dt (from (10) ), and
>
> r'd(\phi)' = rd(\phi) (from (11).
>
> In other words, there can be neither time dilation nor length
> contraction under these transformations. As I have argued in my
> previous message, an alleged time dilation is obtained in the rotating
> frame by altering the definition of proper time, which should be the
> time recorded by a clock at rest in the rotating frame. Instead of
> this definition, people simply define
>
> ds = dT (i)
>
> as in Special Relativity. Equation (i) works in SR, but as I have
> argued it,sometimes,leads to inconsistancies in GR. The locally
> Lorentz transformations
>
> dt' = \gamma(dt - r^2*w*d(\phi)) (ii)
> d(\phi)' = \gamma(d(\phi) - w*dt) (iii),
>
> give
>
> ds'^2 = dt'^2 - r'^2d(\phi)'^2 - dr'^2 - dz'^2,
>
> which is just what is expected under Lorenz transformations. Equations
> (ii) and (iii) lead to the correct time dilation and length
> contraction:
>
> dt'= dT = \gamma dt, while d(\phi)'=dr'=dz',
> r'd(\phi}'=\gamma rd(\phi), while dt'=0.
>

There is a typo error in these equations. The \gamma's are supposed to be
\gamma^(-1) so that the correct equations are

dt'= dT = \gamma^(-1)dt= \sqrt(1-r^{2w}^2)dt,r'd(\phi}'=\gamma^(-1)rd(\phi)= \sqrt(1-r^{2w}^2)rd(\phi).

> Everybody will agree on the correctness of the last two results. How
> is it then possible that the Galilean transformations (10) and (11)
> above also lead to these very same results, namely time dilation and
> length contraction?
>
> Comments please...
>
>
> Regards,
>
>
> Murat Ozer