A constant speed of light in all reference frames? Surely you can't be serious.
in [Relativity]
|
From: glird on 24 Feb 2010 13:35 On Feb 13, 8:29 am, Ste <ste_ro...(a)hotmail.com> wrote: > >I've been absolutely racking my brain (to the point of getting a headache) for the last few days about this issue, and it's clear that the speed of light(where light is either considered in the form of a ballistic photon, or a wave-cycle) cannot, physically, be constant in all relative frames, and at the same time be constant when traveling between two objects in two different frames. It's a physical and logical impossibility. It's also clear that velocities cannot be additive (in the form of speed of bullet + speed of gun), and nor can they be subtractive relative to a background medium (in the form of speed of propagation in medium - speed of source). Take an illustration: A C B A and B are atoms that pass infinitesimally close to each other. In the illustration, A and B are separated from C by a distance L. A and C are stationary relative to each other. B is moving, and approaching C at a speed S. A pulse is emitted from both A and B simultaneously towards C, at the point when A and B are equidistant from C. > Presumably, "simultaneously" means "at the same instant, regardless of what the 'time' might be on coinciding clocks A and B". >< Now, clearly, if velocities were additive, then light from B would reach C much quicker than light from A. We don't see that, so we can dismiss that immediately. Next, if velocities were subtractive, like sound, well that seems like a compelling explanation for what we see, which is that light from both A and B travel towards C at the same speed. But the presence of an absolute medium seems to fall down when one considers that, to be consistent with observation, the speed of propagation orthogonal to the direction of travel must be the same as the speed in the direction of travel. A speed (i.e. a measure of distance traversed within a period of time) cannot possibly be measured constant in all directions within a frame, *and* constant between frames, where the frames themselves are moving at a speed relative to each other. So how the hell does one reconcile this physically? > Lorentz did it in 1904 and Einstein modified his solution a huge trifle. Lorentz showed that when a group of atoms constituting a given object is moving through a stationary luminiferous material, it will contract by q = sqrt(c^2-v^2) in its direction of motion and the rate that events in it transpire will slow down by q. Here is how that solved your problem wrt a roundtrip of light, i.e. a beam going from AB to C and back to A and B: 1. If we let L be x' = one unit long, where 1 unit is the distance light travels in a vacuum in one second, then it will take a ray 2/c = 2 seconds to go from A to C and back again. 2. It would take the ray qL/c = qx'/c = q seconds to get from B to C and ?/c to get from C back to B, where ? is less than L because B has moved closer to C while the ray was traveling. 2'. To simplify the math, let A be stationary in the luminiferous material and let B and C be moving at v = .6c on X of co-ordinate system XYZ. It will take the ray 2(L + vt)/c = 2(1 + .6t) seconds to roundtrip to C and back to A. It will take the ray qx'/(c-v) = .8/.4 = 2 seconds to get from B to C and qL/(c+v) = .8/1.6 = .5 seconds to return. As measured by B, the roundtrip will therefore take qt = .8 x 2.5 = 2 seconds. 3. Let C be perpendicular to co-moving B and it will take the ray L/q = 1/.8 = 1.25 seconds to get from B to C and another x'/q = 1.25 seconds to return, thus t = 2.5 seconds for the roundtrip. Running q- slow, clocks of B will measure this as qt = .8 x 2.5 = 2 seconds. In order to let it take 1 second each way as plotted by B, Lorentz introduced the concept of "local time"; i.e. in order to let a moving system measure c as a constant in any and all directions, successive clocks of a moving system have to be offset relative to each other by vx/c^2 seconds, where v is the velocity of the system through the local material and x is the distance between two such clocks as measured by the moving system itself. Here, then, is his solution to your problem: 1. A and C are stationary and L = x' = 1 unit long, then AC/c = L/c = 1 each way so c = x'/t = 1/1 = 1 unit/second. 2. If B and C are co-moving at v, then the one way time from B to C, as measured by the stationary system A, will be t = qL/(c-v) = [1/(c-v)]q = (1/.4).8 = 2 seconds and the roundtrip time will be t_r = qx'/(c-v) + qx'/(c+v) = [1/(c-v) + 1/(c+v)]q = 2.5 seconds, in which x' is the distance between B and C as measured by A. 3. As measured by the moving system B (xi, tau) this becomes tau = (dtau/dt)t + dtau/dxi = (dtau/dt)[1/(c-v)](dxi/dx)] + (-vxi/c^2) = (dtau/dt)[2 - .6(1.25)] = .8 x 1.25 = 1 second, and the roundtrip time is tau_r = (dtau/dt)t_r = (dtau/dt)[qx'/(c-v) + qx'/(c+v)] = [(dxi/dx)x'/(c-v) + (dxi/dx)x'/(c+v)]dtau/dt = [1/(c-v) + 1/(c+v)] (dxi/dx)( dtau/dt) = [2/(c^2 - v^2)] x .8 x .8 = 2/.64 x .64 = 2 seconds. Q. E. D. What was the "huge trifle" set forth by Einstein? It was this: In his 1904 paper, Lorentz stipulated that the deformation of lengths and rates of a moving system were by ql in the direction of motion and by l in the perpendicular directions. He later let l be equal to one, and -- as in my above treatment -- it disappeared from hisequations. Calling it phi(v) instead of l, Einstein put it back again in the "Einstein transformation equations": tau = phi(v)(t - vx/c^2)/q, xi = phi(v)(x - vt)/q, eta = phi(v)y, zeta = phi(v)z. So what's so "huge" about this trifle? This: There is only ONE group of transformations (the LTE) in which l = phi(v) = 1. But there is an INFINITE number of groups in which l = phi(v) =/= 1; each of which solves the problem equally well. please don't blame meif phi9v) = 1 is experimentally better than all but those in which the difference between 1 and 1 +/1 f is too small to have a measurable impact. glird
From: Ste on 24 Feb 2010 21:32 On 24 Feb, 08:30, mpalenik <markpale...(a)gmail.com> wrote: > On Feb 24, 2:15 am, Ste <ste_ro...(a)hotmail.com> wrote: > > > > > That's not true. A further 100 tosses would not discern definitively > > > > between the two theories, for the outcome would still be technically > > > > consistent with either theory. > > > > In fact, though, you can calculate the probability that either > > > hypothesis is correct if the coin toss comes out a certain way after > > > 100 tosses. 100% heads, for example, would put you well past a 95% > > > confidence interval. And in fact, when experimentalists publish their > > > data, they do also publish such confidence intervals. > > > But even a weighted coin would be unlikely to give 100% heads. In any > > event, the point is that there is still no definitive test, and unlike > > a simple coin-toss outcome theory, the truth of real-world scientific > > theories and experiments are not nearly as easily reducible to this > > kind mathematical probability. > > I don't think you understand how statistics works. We can say "a > weighted coin should come up heads > 50% of the time." We can say > "given that we threw the coin n times and got x number of heads, what > is the probability that this is a fair coin." We can then compare the > null hypothesis (that it is a fair coin) to the "heads comes up more" > hypothesis. I think you're probably arguing this at a different level. I understand the statistics of this scenario perfectly. > > And indeed, if the confidence level is 95%, then that doesn't mean > > 100% of people should hold the weighing-theory to be true (i.e. on the > > basis that the most likely explanation is likely to be the correct > > one). On the contrary, it would be desirable to have, say, only 95% of > > people working within the assumption that the weighting-theory is > > true, and the rest working with the assumption that it is not true. > > First of all, after 100 flips you'd be well above 95%--probably well > above 99%. Whatever the probability is, it doesn't matter, except that it is not 100%. > But regardless, a 95% confidence interval doesn't mean > that 5% of all people should believe that the coin isn't weighted. It > means that everybody should believe that there is a 95% chance that > the coin is weighted. There's a big difference. Ever heard of "putting all your eggs in one basket"? > > > > > Notice how different this approach is from sitting back and trying to > > > > > decide whether mathematicians (or gangsters) are groups of people who > > > > > self-select themselves into delusions, and therefore their models are > > > > > not to be trusted. Why do that kind of nonsense, when you can simply > > > > > ask the coin to show its colors? > > > > > Because in the real world it is not simply a case of flipping the coin > > > > an infinite number of times. Let's face it, we both submit to evidence > > > > - that cannot be the difference between us. > > > > Actually, you tend to hand-wave away any evidence you don't like as > > > 100 years of bad experiments, as in the case of the Michaelson Morely > > > experiment and others like it with similar, albiet more refined, > > > steups, which you were insisting had problems for a long time. > > > Perhaps the difference is, we understand the experimental evidence and > > > you don't. > > > I haven'thand-wavedthe evidence away at all. If you mean that I've > > just dismissed evidence out of hand (presumably because you think I > > find it undesirable), then I would challenge you to identify where I > > have done this. If you mean something else, then I would ask you to > > clarify what you mean when you say I've "hand-wavedevidence away". > > Ok, let's look at your whole thread about measuring the speed of > light. > > You came up with a bunch of hypotheses that have no basis in physical > reality--about how brightness could affect the location of > interference fringes (it doesn't, and I proved it doesn't), about the > speed of light and measured brightness being able to "compensate" for > each other--again none of which have any basis in physical reality-- > all because you wanted to dismiss the Michaelson-Morley type > experiments. Again, there was nothing sensible about any of it, > except that it was a way for you to attempt to justify not having to > believe the speed of light isotropy measurements. And indeed, I said openly that I couldn't make that work, once I'd been able to construct a graphic where I could see the proof with my own eyes, and clearly the physical understanding on which that was based is untenable. As it stands, I don't really have any workable physical model for light under relativity. That said, I don't think anyone can accuse me of being insensible or lacking integrity in the way you suggest that it was a contrivance to avoid believing the obvious. > > And before you jump in and say that this newsgroup is not > > representative of physicists, let me be clear that I've read pretty > > widely already and this utterly lack of conceptual clarity is by no > > means confined to this newsgroup. > > I wasn't going to say that, but what extensive interaction have you > actually had with physicists? And reading "A Breif History of Time" > doesn't count. Oh come on Mark. I first read A Brief History years ago, and just happened to mention that I'd read it again recently. > > For example, consider this illustration: > > > A------B---C > > > Basically you can express the distance AB with the value x, the > > distance BC as value x/2, and the distance AC as value 3x/2. Or you > > can express it as AB = 2x, BC = x, and AC = 3x. But this form of > > expression always relies on comparison, and if you change the > > reference value of X then all the other values change numerically (but > > not physically - there is still some essential relationship that is > > physically invariant). > > First of all, physically, if you double the distance between two > objects, that does make a difference. For example, if you bring two > molecules close engouh together, they will start repelling instead of > attracting (this is the principle behind atomic force microscopy). If > you shrank the sun down into a small enough region, it would become a > black hole. > > The invariant quantity is the ratio of lengths: 2:1. And yes, this > can be expressed as 1:2. Yes, but you've subtly introduced a comparison again, of expressing one length as a ratio of another. The question is how to express the distance of BC in such a way that, no matter how the distance AB changes, and without reference to any other standard, the expressed distance of BC does not change. By your ratio method, if we double the distance of AB (or halve the distance BC), then the ratios change to 4:1. > > But how do you iron out the reliance on > > comparison? How do you describe something with reference only to the > > things in question, and not to a reference standard? > > The distance between B and C = 1/2 the distance between A and B. Thankyou for stating the obvious while the real question apparently whooshed over your head. The question is how to express the distance BC by reference only to B and C (and without reference to A or any other thing).
From: Ste on 24 Feb 2010 21:35 On 24 Feb, 17:37, PD <thedraperfam...(a)gmail.com> wrote: > On Feb 24, 2:56 am, Ste <ste_ro...(a)hotmail.com> wrote: > > > Of course, I don't pretend to even have a > > speculative account for what this variable may be. I will finish > > however by saying that surely you accept that clocks that actually > > exist must have some common principles at a fundamental level, and > > that if the time dilation phenomenon operates at that level then it's > > quite plausible that they would all react in the same way to time > > dilation. > > On this last point, I will simply conclude that, yes, indeed all > clocks operate with a common principle at some level, and that time > dilation operates at that level. That principle and that operation of > dilation are PRECISELY what is described by special relativity, as far > as we can tell from the evidence we have in hand. Congratulations. > > Now, at this point, I imagine you might say, "But I don't BELIEVE in > special relativity and hold faith that there is some OTHER principle > and operation of dilation that is responsible for what is going on." > That is, of course, a possibility. Anything is possible. However, > among those models that have been tested and which do purport to > account for the principle that drives dilation, special relativity is > the demonstrated winner. You are free to put forward a new candidate > to add to the race. There are others who are doing exactly the same > thing this very day. As I say, I don't necessarily disbelieve SR. I just don't think it is conceptually very clear.
From: artful on 24 Feb 2010 21:40 On Feb 25, 1:32 pm, Ste <ste_ro...(a)hotmail.com> wrote: > On 24 Feb, 08:30, mpalenik <markpale...(a)gmail.com> wrote: > > On Feb 24, 2:15 am, Ste <ste_ro...(a)hotmail.com> wrote: > > > But how do you iron out the reliance on > > > comparison? How do you describe something with reference only to the > > > things in question, and not to a reference standard? > > > The distance between B and C = 1/2 the distance between A and B. > > Thankyou for stating the obvious while the real question apparently > whooshed over your head. The question is how to express the distance > BC by reference only to B and C (and without reference to A or any > other thing). Then you need to pose your questions more carefully.. you only mentioned 'reference standard'. He gave an answer without a reference standard. How about this .. the length BC is double the length from B to the point halfway between B and C. That only references B and C (and a point you can derive from those two points).
From: Sam Wormley on 24 Feb 2010 22:02
On 2/24/10 8:35 PM, Ste wrote: > On 24 Feb, 17:37, PD<thedraperfam...(a)gmail.com> wrote: >> On Feb 24, 2:56 am, Ste<ste_ro...(a)hotmail.com> wrote: >> >>> Of course, I don't pretend to even have a >>> speculative account for what this variable may be. I will finish >>> however by saying that surely you accept that clocks that actually >>> exist must have some common principles at a fundamental level, and >>> that if the time dilation phenomenon operates at that level then it's >>> quite plausible that they would all react in the same way to time >>> dilation. >> >> On this last point, I will simply conclude that, yes, indeed all >> clocks operate with a common principle at some level, and that time >> dilation operates at that level. That principle and that operation of >> dilation are PRECISELY what is described by special relativity, as far >> as we can tell from the evidence we have in hand. Congratulations. >> >> Now, at this point, I imagine you might say, "But I don't BELIEVE in >> special relativity and hold faith that there is some OTHER principle >> and operation of dilation that is responsible for what is going on." >> That is, of course, a possibility. Anything is possible. However, >> among those models that have been tested and which do purport to >> account for the principle that drives dilation, special relativity is >> the demonstrated winner. You are free to put forward a new candidate >> to add to the race. There are others who are doing exactly the same >> thing this very day. > > As I say, I don't necessarily disbelieve SR. I just don't think it is > conceptually very clear. Try The Mechanical Universe series. http://www.learner.org/resources/series42.html 42. The Lorentz Transformation If the speed of light is to be the same for all observers, then the length of a meter stick, or the rate of a ticking clock, depends on who measures it. |