What is your EM crankosity?
in [Relativity]
|
From: Benj on 28 Sep 2009 03:59 On Sep 27, 6:12 pm, Jim Black <fmla...(a)organization.edu> wrote: > I'm not aware of any purely electromagnetic clocks in the strictest sense; > they all have some mechanical parts, such as wires and the electrons in > them. If you model those with Newtonian mechanics, of course it's possible > to get an answer different from special relativity. So I would say that > "leading to the conclusion" is the only wrong part of #9. Relativistic > time dilation affects not just electromagnetic clocks, but any > self-contained clock that does not rely on outside parts which may not be > set in motion along with the clock. Examples of clocks failing the "no > outside parts" condition are pendulum clocks and WWV receivers. I'm not going to review jefimenko's calculations for you, but his "clocks" are in essence theoretical calculations not practical clocks for experiments. His examples include things like a point charge oscillating in a ring charge, Point charges oscillating under the influence of one or two line charges and so forth. These are in essence electromagnetic models and are calculated with Maxwell's equations. The fact that they give various answers in moving frames is significant. > On the other questions: > For #5, Susan's answer is best; the statement is only approximate; there is > a small error term. Susan's (and your) assertion that there is a "small error term" obviously implies the statement it false but isn't really good physics. The use of the term "error" implies that a distorted static field is somehow "wrong". It isn't wrong. As was first calculated in 1888 by Heaviside, the electric field from a point charge moving at constant velocity concentrates itself in the direction perpendicular to the the direction of motion and the field decreases along the direction of the motion. See Oliver Heaviside, "The Electromagnetic Effects of a Moving charge", The Electrician, 22, 147--148 (1988) and also see "On the Electromagnetic effects due to motion of Electricity through a Dielectric" Phil. Mag. 27, 324-339 (1889) Hence my statement that the field stays the same as that from a static charge (evenly distributed) is false. > Statement #4, on the other hand, is only valid as an > approximation if E is not too much larger or smaller than cB. Otherwise > the B field from the transformed E field can swamp the original B field, or > vice versa. We are not doing approximations here except for v <<c. The truth is that if you have instruments to measure stationary fields E, D, B, H and get readings. If you mount those meters on a railroad car and again measure the fields you find they are different and NOT THE SAME. The moving meters read fields according to the following transforms: E* = E + v x B H* = H - v x D Where * indicates the moving reading and v is the velocity of the meters. In other words the moving meters measure two additional fields: El = v x B and Hl = -v x D which are termed Lorentz fields. Thus to speak of electric or magnetic fields without specifying a reference frame in which they are measured is meaningless. The above relations hold for v<<c. We are not worried about magnitudes of the various fields at this point so your point about "swamping" though a practical one, really isn't the point. I said that the moving meters always show the same readings regardless of the velocity which is clearly false. See Jefimenko's textbook, "Electricity and Magnetism" Appleton-Century- Crofts, New York, 1966, p388 ff. > #7 is too sloppily posed for me to call true or false, but I will say that > the actual length of the line segment in the observer's frame is different > from the actual length of the line segment in its own frame. Unless the > motion is perpendicular to the line segment, of course. I'm interested in > seeing your reasons for calling it false; then there might be something to > discuss. I seen nothing sloppy here. The reasoning is as follows. Given a moving line charge segment, it is known that due to retardation the apparent length of that segment changes. (Yes, we are talking about motion across the observer and along the direction of the segment not perpendicular to it). That is fact one. Fact two is that by relativity one can calculate the correct transformations of the fields from these moving charges due to motion. These transformations are known to be the correct way fields from moving charges transform from motion. Now (here comes the tricky part) if we calculate fields due to motion of the line charge segments but ALSO include the fact that the charge segments are changing length due to an actual Lorentz change in length of the charges, we find a formula for the field transformations due to motion. The only problem is that this transformation including the shortened charge lines is NOT the same one we found when the charge lines were not actually shortened. Indeed it turns out it is the UNSHORTENED charge lines that give the correct relativistic answer. The conclusion therefore is that the line charges are NOT actually being shortened by the relative motion. Only their apparent length is changed due to that. Hence an actual shortening is not occurring as I stated and the statement is false. For the complete calculations see O. Jefimenko, "Electromagnetic Retardation and the Theory or Relativity", 2004, ISBN 0-917406-24-9 Chapter 9. "Common misconceptions about relativity theory". > #8 is wrong because of "cosmic rays." Correct. Note also that light does not obey classical Electromagnetism. A single photon gives ALL it's energy to a single electron essentially instantaneously. And also light intensity is related to the number of photons not the usual wave energy parameters. Thus it appears that properties of what people used to call the "electromagnetic spectrum" are quite different at it's opposite ends. I agree with your reasons for the > others being wrong. #10 is also wrong on the uniqueness claim. Correct and is also false because in free space electric and magnetic fields DO NOT create each other. Your LC arguments on this matter don't apply because to have an L and a C one needs a geometry of conductor in space. In free space there is none of that to modify and interact with the fields. Since E and B occur simultaneously they simply cannot cause each other. Causality requires that a cause precede an effect. Cheers
From: Jim Black on 29 Sep 2009 02:00 On Mon, 28 Sep 2009 00:59:29 -0700 (PDT), Benj wrote: > On Sep 27, 6:12�pm, Jim Black <fmla...(a)organization.edu> wrote: > >> I'm not aware of any purely electromagnetic clocks in the strictest sense; >> they all have some mechanical parts, such as wires and the electrons in >> them. �If you model those with Newtonian mechanics, of course it's possible >> to get an answer different from special relativity. �So I would say that >> "leading to the conclusion" is the only wrong part of #9. �Relativistic >> time dilation affects not just electromagnetic clocks, but any >> self-contained clock that does not rely on outside parts which may not be >> set in motion along with the clock. �Examples of clocks failing the "no >> outside parts" condition are pendulum clocks and WWV receivers. > > I'm not going to review jefimenko's calculations for you, but his > "clocks" are in essence theoretical calculations not practical clocks > for experiments. His examples include things like a point charge > oscillating in a ring charge, Point charges oscillating under the > influence of one or two line charges and so forth. These are in > essence electromagnetic models and are calculated with Maxwell's > equations. The fact that they give various answers in moving frames is > significant. There is a mechanical component there. To predict the motion of those point and line charges, you need to use either Newton's second law or the corrected relativistic formula. And those line charges will be contracted according to relativity but not according to Newtonian mechanics. So how he models the mechanics matters. I found this in the reviews of the book on Amazon: : In the text, clock#3 (Einsteinian) is the same as clock#7 : (Non-Einsteinian) except that clock#7 is moving in a direction : perpendicular to the direction specified for clock#3. At first when I : looked at this, Jeffimenko's argument and equations looked correct. : However, I noticed that for clock#7 he did not take into effect the : Lorentz contraction of the fixed charges in the system. When I made this : change I got the same results for time dilation as clock#3 and in perfect : agreement with Einstein. >> On the other questions: >> For #5, Susan's answer is best; the statement is only approximate; there is >> a small error term. � > > Susan's (and your) assertion that there is a "small error term" > obviously implies the statement it false but isn't really good > physics. The use of the term "error" implies that a distorted static > field is somehow "wrong". I interpreted that term differently. In my view, the implication was that the spherically symmetric version was slightly wrong. But a better term would be "correction." The reason I would hesitate to call the less accurate statement outright false is that everything we know about physics is probably slightly wrong. If the velocity is low enough, the spherical approximation will be sufficient for some purposes. >> #7 is too sloppily posed for me to call true or false, but I will say that >> the actual length of the line segment in the observer's frame is different >> from the actual length of the line segment in its own frame. �Unless the >> motion is perpendicular to the line segment, of course. �I'm interested in >> seeing your reasons for calling it false; then there might be something to >> discuss. > > I seen nothing sloppy here. The reasoning is as follows. Given a > moving line charge segment, it is known that due to retardation the > apparent length of that segment changes. This apparent length being distinct from the Lorentz contracted length. > (Yes, we are talking about > motion across the observer and along the direction of the segment not > perpendicular to it). That is fact one. Fact two is that by relativity > one can calculate the correct transformations of the fields from these > moving charges due to motion. Yes, you can transform the fields from a line charge at rest to get the fields from a Lorentz-contracted moving line charge. > These transformations are known to be > the correct way fields from moving charges transform from motion. Now > (here comes the tricky part) if we calculate fields due to motion of > the line charge segments but ALSO include the fact that the charge > segments are changing length due to an actual Lorentz change in length > of the charges, we find a formula for the field transformations due to > motion. Where is the change in length being plugged in? Into the original proper length of the line segments? Into an integral over a bunch of moving point charges? Or something else? Or easier yet, is this http://wwwphy.princeton.edu/~kirkmcd/examples/EM/jefimenko_ajp_63_454_95.pdf a similar or the same argument as the one in the book? > > #10 is also wrong on the uniqueness claim. > > Correct and is also false because in free space electric and magnetic > fields DO NOT create each other. Your LC arguments on this matter > don't apply because to have an L and a C one needs a geometry of > conductor in space. In free space there is none of that to modify and > interact with the fields. Since E and B occur simultaneously they > simply cannot cause each other. Causality requires that a cause > precede an effect. The plane electromagnetic wave is even simpler. We can interpret d(Ex)/dz = - d(By)/dt - d(By)/dz = (1/c^2) d(Ex)/dt (d's are partial here) as saying that a *spatial* variation in the electric field causes a *temporal* change in the magnetic field, and vice versa. They're not creating each other; they're altering each other's values in a way that moves the wave forward through space. And here is the same thing in ASCII diagrams: http://groups.google.com/group/sci.physics.electromag/browse_thread/thread/ea5ae80177e26081?fwc=1 -- Jim E. Black (domain in headers) How to filter out stupid arguments in 40tude Dialog: !markread,ignore From "Name" +"<email address>" [X] Watch/Ignore works on subthreads
From: Benj on 29 Sep 2009 17:10 On Sep 29, 2:00 am, Jim Black <fmla...(a)organization.edu> wrote: > Where is the change in length being plugged in? Into the original proper > length of the line segments? Into an integral over a bunch of moving point > charges? Or something else? The idea is that if the lines are actually contracting the total charge must stay the same so therefore the charge density is altered. This changes the transforms so they are not correct. > Or easier yet, is this > > http://wwwphy.princeton.edu/~kirkmcd/examples/EM/jefimenko_ajp_63_454... Which are the same arguments without the actual contraction calculation. Basically it shows that retardation is implicit in relativity (both approaches give the correct transform), but an actual change in length with velocity does not. > a similar or the same argument as the one in the book? Same except for the charge density change with velocity thing... > The plane electromagnetic wave is even simpler. We can interpret > > d(Ex)/dz = - d(By)/dt > - d(By)/dz = (1/c^2) d(Ex)/dt > > (d's are partial here) > > as saying that a *spatial* variation in the electric field causes a > *temporal* change in the magnetic field, and vice versa. They're not > creating each other; they're altering each other's values in a way that > moves the wave forward through space. > > And here is the same thing in ASCII diagrams:http://groups.google.com/group/sci.physics.electromag/browse_thread/t... You are trying hard to use other words to say they create each other but I don't agree. Anything that causes or generates or alters or whatever word you use must precede the event. In this case the fields "measure" each other (to use Maxwell's word) but they are both simultaneous and both created by the source currents that created the wave. Thus they are related to each other and are a "measure" of each other but are not "altering" each other. Your theory would have to take into account the retardation of all points about the given observation location if the spatial waves are to create temporal changes. I don't see this. Electric and magnetic fields each individually travel outward at the speed of light from their current and charge sources. This is known for cases where there are no waves and the fields expand individually. To me it makes sense that the same operations take place with both fields are sent out simultaneously. Your equations above say the two sides are EQUAL but do not say they are casual. In fact, the implication is that both sides are evaluated at the same time.
From: Timo Nieminen on 29 Sep 2009 20:36 On Mon, 28 Sep 2009, Jim Black wrote: > On Mon, 28 Sep 2009 00:59:29 -0700 (PDT), Benj wrote: > > > On Sep 27, 6:12 pm, Jim Black <fmla...(a)organization.edu> wrote: > > > >> I'm not aware of any purely electromagnetic clocks in the strictest sense; > >> they all have some mechanical parts, such as wires and the electrons in > >> them. If you model those with Newtonian mechanics, of course it's possible > >> to get an answer different from special relativity. So I would say that > >> "leading to the conclusion" is the only wrong part of #9. Relativistic > >> time dilation affects not just electromagnetic clocks, but any > >> self-contained clock that does not rely on outside parts which may not be > >> set in motion along with the clock. Examples of clocks failing the "no > >> outside parts" condition are pendulum clocks and WWV receivers. > > > > I'm not going to review jefimenko's calculations for you, but his > > "clocks" are in essence theoretical calculations not practical clocks > > for experiments. His examples include things like a point charge > > oscillating in a ring charge, Point charges oscillating under the > > influence of one or two line charges and so forth. These are in > > essence electromagnetic models and are calculated with Maxwell's > > equations. The fact that they give various answers in moving frames is > > significant. > > There is a mechanical component there. To predict the motion of those > point and line charges, you need to use either Newton's second law or the > corrected relativistic formula. And those line charges will be contracted > according to relativity but not according to Newtonian mechanics. So how > he models the mechanics matters. It's worse than that - classical EM theory can't do point charges properly. See the problems with the Lorentz-Dirac equation of motion of an electron for the most common example. So, even with relativistic mechanics, there are still some deep problems here. -- Timo
From: blackhead on 1 Oct 2009 16:59
On 30 Sep, 01:36, Timo Nieminen <t...(a)physics.uq.edu.au> wrote: > On Mon, 28 Sep 2009, Jim Black wrote: > > On Mon, 28 Sep 2009 00:59:29 -0700 (PDT), Benj wrote: > > > > On Sep 27, 6:12 pm, Jim Black <fmla...(a)organization.edu> wrote: > > > >> I'm not aware of any purely electromagnetic clocks in the strictest sense; > > >> they all have some mechanical parts, such as wires and the electrons in > > >> them. If you model those with Newtonian mechanics, of course it's possible > > >> to get an answer different from special relativity. So I would say that > > >> "leading to the conclusion" is the only wrong part of #9. Relativistic > > >> time dilation affects not just electromagnetic clocks, but any > > >> self-contained clock that does not rely on outside parts which may not be > > >> set in motion along with the clock. Examples of clocks failing the "no > > >> outside parts" condition are pendulum clocks and WWV receivers. > > > > I'm not going to review jefimenko's calculations for you, but his > > > "clocks" are in essence theoretical calculations not practical clocks > > > for experiments. His examples include things like a point charge > > > oscillating in a ring charge, Point charges oscillating under the > > > influence of one or two line charges and so forth. These are in > > > essence electromagnetic models and are calculated with Maxwell's > > > equations. The fact that they give various answers in moving frames is > > > significant. > > > There is a mechanical component there. To predict the motion of those > > point and line charges, you need to use either Newton's second law or the > > corrected relativistic formula. And those line charges will be contracted > > according to relativity but not according to Newtonian mechanics. So how > > he models the mechanics matters. > > It's worse than that - classical EM theory can't do point charges > properly. See the problems with the Lorentz-Dirac equation of motion of an > electron for the most common example. So, even with relativistic > mechanics, there are still some deep problems here. > > -- > Timo- Hide quoted text - > > - Show quoted text - I think Feynman and Wheeler got around the problem by assuming the field of a charge doesn't act upon itself but only on other charges: Wheeler, John Archibald and Feynman, Richard Phillips (1945) Interaction with the absorber as the mechanism of radiation. Reviews of Modern Physics, 17 (2-3). pp. 157-181. ISSN 0034-6861 http://authors.library.caltech.edu/11095/1/WHErmp45.pdf |