-
#240 by meberbs on 02 Aug, 2018 21:20
-
Talk is cheap. The language of physics is mathematics. Equations would give those that are versed in the subject (probably not me - but others) plenty of clarification.
As I said, point me to where I haven't provided detailed enough mathematics and I can correct. I have provided plenty of math, but there is no reason I should have to re-type the math every time someone makes a baseless claim that has already been disproven.
No it doesn't go away. The cavity accelerating increases the blue shift, since by the time the light hits the back wall, the cavity is moving faster. This makes the new force be larger so the direction of cavity motion reverses, so you get nothing but vibration. If you think otherwise do the math and try to show it.But you may say, the red-shift at the small-end is undone by the blue shift at the large-end? No, I think not!
The second statement literally does nothing to support your "No, I think not." It doesn't make the blue shift go away by some kind of magic. Entropy is not relevant to the discussion, since energy and momentum won't just appear out of nowhere just to make entropy increase. Nothing you are describing increases the density of states anyway.
A multi-mode cavity has more modes, density of states (in that foul tongue of QM which I abhor uttering) at the large end. More degrees of freedom. An increase in Entropy! Just like a heat engine.
Literally, I think so. The blue shift goes away either by cavity dissipation or by being red-shifted at the cavity apex if the cavity accelerates. IF...
Your statement about "dissipation" is literally just handwaving.
You don't need years of experience to answer the questions you have. Anyway I don't provide my credentials deliberately, in part because my statements should stand on their own. Experts are in this thread already, more than enough to say that your claims are incorrect.What is frustrating is when people like you who apparently don't even know how to calculate momentum of an EM wave think they are somehow smarter than people who actually know what they are talking about. Try to recognize that maybe the reason that qualified people don't agree with you is because you are wrong.
I have never found a highly qualified microwave engineer here, that is one that has years of EM simulation experience with waveguides and antennas, R&D. Shell, Dave maybe the best, one Nasa test engineer showed up for maybe one post, and Dr. Rodal's math expertise seems to be with gravity and fluid dynamics. One other guy with military radar.I have alluded to a bit of uncertainty regarding my conjectures and reasoning, which are based on references I link. Of course I concede I may be wrong.
This is simple then. You are wrong. You have been told repeatedly that no matter how you add it up, the momentum is going to balance, and the device will not go anywhere. If you start with a wave in the cavity, and ignore the initial momentum imparted by the antenna inputting the wave, the total net momentum imparted to the cavity by the wave will be no more than the initial momentum of the wave. It doesn't matter whether it is absorbed immediately, or bounces back and forth a million times, possibly gradually losing energy. You can now stop wasting your time here.
I think not.
I wouldn't spend time here if I didn't think I was right.
If you still think you are anything but wrong, please provide some math to support your statements so that specific errors can be pointed out to you. (For example, that means not statements like the one below, which has no relevant meaning.)QuoteNow, a stationary electron can be (whether accurately or not) thought of (modeled) as a standing electromagnetic standing wave in a nonlinear Kerr-effect media (quantum vacuum). A moving electron, as a marching wave. If similar conditions are created in an appropriate waveguide (frustrum), will it too move, similar to how particles move? Is the EM drive system acting like a macroscopic particle?
-
#241 by mwvp on 02 Aug, 2018 21:47
-
...
do the math and try to show it.
"Try and show it". LOL. I save my pearly nuggest for the biggest and best
-
#242 by rq3 on 02 Aug, 2018 22:09
-
[quote}
I wouldn't spend time here if I didn't think I was right.
[/quote]
That's funny! I wouldn't spend time here if there wasn't the infinitisimal chance that I was wrong! -
#243 by mwvp on 03 Aug, 2018 06:17
-
[quote}
I wouldn't spend time here if I didn't think I was right.
That's funny! I wouldn't spend time here if there wasn't the infinitisimal chance that I was wrong!
[/quote]
Your motive's probably more popular here than mine. -
#244 by flux_capacitor on 03 Aug, 2018 14:50
-
I may again make some mistakes below (apologies if this is the case to the less indulgent among you all) but the current discussion about redshift and blueshift of wavelengths makes me think about several things:
1) We have already discussed in the past here a particular blue/redshift effect, due to the presence of a gravitational potential (or time dilatation as some like to consider) with the Pound-Rebka experiment. Which is not apparently directly related to the EmDrive (or is it?) bu still a very interesting effect.
2) An EM wave is redshifted when losing its energy upon each collision with a reflecting wall (side wall, small and big ends).
3) Dynamic Doppler shifts occur with the acceleration of the cavity.
What I call a "dynamic" Doppler shift is:
How do all this (except point #1) combine together?
Shawyer says that dynamic Doppler shifts limit the acceleration of the EmDrive.
According to point #3, he specifically talks about the frequency of the travelling wave moving outside the narrow resonant bandwidth of the cavity, leading to a reduction in stored energy, thus a reduction in Q, and a reduction in thrust.
Let's put this claim aside for now and focus on blue vs red shifts only. The acceleration of the cavity (small end leading) indeed introduces Doppler shifts: EM waves are being redshifted (their wavelength becomes longer and the EM wave looses energy) after their reflection on the "receding" small end, while they are being blueshifted (their wavelength getslongershorter, and their energy increases) after reflection on the "approaching" big end. Such effect being amplified by the acceleration, as the velocity of the cavity between two bounces keeps increasing. Shawyer explains this effect:Quote from: Roger ShawyerAssume the EM wavefront propagates initially from the large end plate towards the small end plate. At the end of this forward transit, the wavefront is reflected at the small end plate. At this time, due to cavity acceleration, the cavity velocity has increased to Vr whereas the wavefront has a constant guide velocity of Vg2. The relative addition of these velocities, gives the reflected wavefront a Doppler Shift, resulting in a reduced frequency Fr for the reverse transit.
On reaching the large end plate, the wavefront is again reflected and subjected to a second Doppler shift, resulting in the forward frequency Ff. The increase in frequency is calculated from the relative addition of the guide velocity Vg1 and the new cavity velocity Vf."
Some may say that the velocity increase of the cavity during such rapid transit is infinitesimally small with respect to the speed of light, so such effect would be negligible and undetectable. This would be forgetting that when small end diameter is made just above the cut-off diameter of a cylindrical open waveguide at that same frequency, then the guide velocity Vg2 would become very small.*
As mwvp said:But you may say, the red-shift at the small-end is undone by the blue shift at the large-end? No, I think not!
The blue shift goes away either by cavity dissipation or by being red-shifted at the cavity apex if the cavity accelerates. IF...
Maybe I see what you are thinking. When combining points #2 and #3 above, blueshift at big end is somewhat "dissipated" or "neutralized" wrt to redshift throughout the ringing time of the cavity, since just upon each bounce/reflection (on sidewalls, big and small ends) the new emitted photon has lost some of its energy (via momentum transfer to the cavity, photo-electric effect with electrons, IR radiation through the larger heatsink). So the redshift due to photon reflection substracts to the dynamic (i.e. due to the acceleration) Doppler blueshift at big end, while it adds to the redshift dynamically produced at small end. Finally, re-emitted photons are more redshifted there than they are blueshifted at the other end, when the cavity accelerates. Therefore "redshift eventually wins" after complete decay, and a new pulse of fresh input RF energy is needed.
* Even more fringe hypothesis: Some people asked many time TheTraveller to quantify Shawyer's claims about the importance of the ability of the cavity to "vibrate" and being "free to accelerate", in order for the thrust to appear and induce a dynamic movement of the cavity, as well as to justify so-called "motor" and "generator" modes. So maybe such "vibration" along the axis just need to trigger enough Doppler shifts at the small end to produce a velocity increase exceeding the locally very slow value of the guide velocity. This way one could understand that some "kick" in the positive direction (small end leading) would trigger "motor mode"; while a kick in the negative direction (big end leading) would conversely produce a blueshifted wave upon reflection on small end, triggering "generator mode". This would only work with a properly designed EmDrive, i.e. with a small end diameter being made "just above cut-off". On the contrary, if the small end diameter is made too large, well above cut-off, the value of the minimum vibration (its amplitude and acceleration) mandatory to trigger either mode increases with the guide velocity, maybe up to an impracticable value. -
#245 by flux_capacitor on 03 Aug, 2018 18:20
-
To X_Ray and others.
About the EM lobes (antinodes) in a frustum cavity that look (in FEKO/COMSOL/etc) at the same time:
- axially stretched (and radially squeezed) near small end
- axially squeezed (and radially stretched) near big end
Is the following animated representation of this effect correct, from the point of view of the spatial shape and temporal evolution of the standing wave in a cylindrical vs frustum resonant cavities?
-
#246 by meberbs on 03 Aug, 2018 18:53
-
According to point #3, he specifically talks about the frequency of the travelling wave moving outside the narrow resonant bandwidth of the cavity, leading to a reduction in stored energy, thus a reduction in Q, and a reduction in thrust.
Which is nonsensical because the blueshift restores energy to the wave. The cavity Q is unaffected, because the waves all originate form the cavity so the velocity can't matter.Let's put this claim aside for now and focus on blue vs red shifts only. The acceleration of the cavity (small end leading) indeed introduces Doppler shifts: EM waves are being redshifted (their wavelength becomes longer and the EM wave looses energy) after their reflection on the "receding" small end, while they are being blueshifted (their wavelength gets longer, and their energy increases) after reflection on the "approaching" big end. Such effect being amplified by the acceleration, as the velocity of the cavity between two bounces keeps increasing. Shawyer explains this effect:
But the more blueshift there is, the more cancelling force there will be that will produce acceleration in the opposite direction. This always happens in the direction opposite the original acceleration, so can never be the cause of initial acceleration.
How many times do Shawyer's statements have to be proven self contradictory before people stop using them as a basis for thinking? Starting at 1=0 you can prove anything you want, but it is meaningless. -
#247 by flux_capacitor on 03 Aug, 2018 20:11
-
According to point #3, he specifically talks about the frequency of the travelling wave moving outside the narrow resonant bandwidth of the cavity, leading to a reduction in stored energy, thus a reduction in Q, and a reduction in thrust.
Which is nonsensical because the blueshift restores energy to the wave. The cavity Q is unaffected, because the waves all originate form the cavity so the velocity can't matter.Let's put this claim aside for now and focus on blue vs red shifts only. The acceleration of the cavity (small end leading) indeed introduces Doppler shifts: EM waves are being redshifted (their wavelength becomes longer and the EM wave looses energy) after their reflection on the "receding" small end, while they are being blueshifted (their wavelength gets longer, and their energy increases) after reflection on the "approaching" big end. Such effect being amplified by the acceleration, as the velocity of the cavity between two bounces keeps increasing. Shawyer explains this effect:
But the more blueshift there is, the more cancelling force there will be that will produce acceleration in the opposite direction. This always happens in the direction opposite the original acceleration, so can never be the cause of initial acceleration.
How many times do Shawyer's statements have to be proven self contradictory before people stop using them as a basis for thinking? Starting at 1=0 you can prove anything you want, but it is meaningless.
I'm not starting at 1=0, just want to check where are the flaws if any.
So you seem in agreement with the fact that there is some redshift occurring at the front and blueshift at the rear when a cavity accelerates. Then for a cylindrical cavity your explanation is obvious, but a subtlety due to the taper is introduced in the IAC 2014 conference paper published in 2015 in Acta Astronautica:Quote from: Roger ShawyerBecause the guide velocity is different at each end, the Doppler shifts are different, even for a constant rate of acceleration.
Do you disagree here with the claim that the guide velocities Vg1 > Vg2 and subsequently, that they induce unequal Doppler shifts during the acceleration of a tapered cavity?Quote from: Roger ShawyerThis build-up of net frequency shift causes a widening of the spectrum of the standing wave pattern, and causes much of the power spectrum to fall outside the narrow bandwidth of the resonant cavity. Clearly this effect will increase with increasing cavity Q, as the number of reflections increase, together with the reduction in bandwidth.
which has the same negative effect as a broadband microwave oven magnetron used instead of a narrowband solid state RF generator, the magnetron spattering among a wide unusable range of frequencies, especially for high-Q cavities with spherically-shaped ends, as shown many threads ago by SeeShells, Star-Drive and others. -
#248 by RonM on 03 Aug, 2018 20:36
-
Quote from: Roger Shawyer
Because the guide velocity is different at each end, the Doppler shifts are different, even for a constant rate of acceleration.
Do you disagree here with the claim that the guide velocities Vg1 > Vg2 and subsequently, that they induce unequal Doppler shifts during the acceleration of a tapered cavity?Quote from: Roger ShawyerNote that this configuration ensures that there is no orthogonal component of the guide velocity reflected from the side wall, thus ensuring a zero side wall force component in the axial plane.
I thought we determined that this statement is incorrect. So, taking into account all Doppler shifts from both ends and the side wall, the net force should be zero. If you disagree, please show the math. -
#249 by X_RaY on 03 Aug, 2018 21:18
-
To X_Ray and others.
It is a oversimplification.
About the EM lobes (antinodes) in a frustum cavity that look (in FEKO/COMSOL/etc) at the same time:
- axially stretched (and radially squeezed) near small end
- axially squeezed (and radially stretched) near big end
Is the following animated representation of this effect correct, from the point of view of the spatial shape and temporal evolution of the standing wave in a cylindrical vs frustum resonant cavities?
The animation is correct as long as it should illustrate the polarity(vector E into phi direction) of the E-field component of the EM-field within the dielectric* only over time. The Field intensity at the very locations depend on the cavity shape. Please see the attached paper from Dr. Rodal for more details.
The following FEKO animation below shows what happens in detail over a full 360 deg cycle (for TE012 in this case) for both E & H.
FEKO EM animation of a TE012 cavity resonator
https://forum.nasaspaceflight.com/index.php?topic=39214.msg1646352#msg1646352
*vacuum or whatever dielectric material
-
#250 by meberbs on 03 Aug, 2018 22:39
-
I'm not starting at 1=0, just want to check where are the flaws if any.
Shawyer's "theory" is self contradictory. It is actually kind of impressive that he has managed to say so many different things that are equivalent to 1=0, but if you take a random quote from him about the emDrive, it will likely be equivalent to 1=0.So you seem in agreement with the fact that there is some redshift occurring at the front and blueshift at the rear when a cavity accelerates.
To describe this situation clearly, it helps to just assume there is an outside force accelerating the cavity. This would clearly cause shifts. Intuitively this would produce an unbalanced force, but not a propulsive one. It would be an inertial force. The EM waves represent part of the mass-energy and therefore the inertia of the overall system. This would manifest as more radiation pressure on the back than the front, even for a perfectly cylindrical cavity. This is not some strange or useful effect. If you had a rubber ball bouncing between 2 ends of a cavity that you were accelerating, it would bounce harder off the back than the front in order to keep up with the overall acceleration of the cavity.
Since this boils down to nothing more than Newton's law of inertia, there is nothing truly new or useful here.
I disagree with the claim that anyone has ever come up with a precise definition of guide wavelength in a resonating cavity, so there is no meaningful direct answer to your question as stated. To the extent that the concept is useful, there is less force on the small end than the large end, which is due to some of the momentum change happening at the sidewalls. Any unbalanced red/blueshift from acceleration (which again I assume to be caused externally, since there is no internal cause), is just the inertial mass equivalent of the energy stored in the cavity in the form of EM waves. Whether the cavity is tapered has no effect on this.Quote from: Roger ShawyerBecause the guide velocity is different at each end, the Doppler shifts are different, even for a constant rate of acceleration.
Do you disagree here with the claim that the guide velocities Vg1 > Vg2 and subsequently, that they induce unequal Doppler shifts during the acceleration of a tapered cavity?Quote from: Roger ShawyerThis build-up of net frequency shift causes a widening of the spectrum of the standing wave pattern, and causes much of the power spectrum to fall outside the narrow bandwidth of the resonant cavity. Clearly this effect will increase with increasing cavity Q, as the number of reflections increase, together with the reduction in bandwidth.
It is possible that a sufficiently accelerating cavity could have Q drop a little, but for the relevant accelerations in any emDrive experiment, this should be at an impossible to measure level. Regardless of the level, this does not explain anything Shawyer tries to use it to explain. It is in the wrong direction to ever provide useful acceleration by definition. It would be proportional to acceleration, not velocity or how long the drive is on, so it does nothing to resolve conservation of energy (or momentum). -
#251 by flux_capacitor on 03 Aug, 2018 23:35
-
Quote from: Roger Shawyer
Because the guide velocity is different at each end, the Doppler shifts are different, even for a constant rate of acceleration.
Do you disagree here with the claim that the guide velocities Vg1 > Vg2 and subsequently, that they induce unequal Doppler shifts during the acceleration of a tapered cavity?Quote from: Roger ShawyerNote that this configuration ensures that there is no orthogonal component of the guide velocity reflected from the side wall, thus ensuring a zero side wall force component in the axial plane.
I thought we determined that this statement is incorrect. So, taking into account all Doppler shifts from both ends and the side wall, the net force should be zero. If you disagree, please show the math.
Why do you resort to an obvious wrong claim ("no force on side walls") to disprove the one being discussed (Vg1 > Vg2) which has nothing to do with it? I am sorry, but this is a suggested irrelevant conclusion using an association fallacy.
Therefore I do not feel the obligation to agree or disagree (providing maths!) on the veracity of such unrelated sentence about THRUST, that I have not quoted nor broached, as my questions are about Doppler shifts, and Doppler shifts only… especially as I cared to precise in my first message about this:… and a reduction in thrust.
Let's put this claim aside for now and focus on blue vs red shifts only.
If I wasn't clear enough from the beginning, let't write it extensively: I am not trying in the last couple of messages to prove the origin or even the reality of any anomalous thrust in relation with possible Doppler shifts. I just ask if we can discuss properly about these Doppler shifts, their origin and their behavior, in various cavity shapes accelerating (or not) by an arbitrary external force. Not any possible or impossible "thrust" as this is another story. That's all. -
#252 by flux_capacitor on 04 Aug, 2018 00:23
-
To X_Ray and others.
It is a oversimplification.
About the EM lobes (antinodes) in a frustum cavity that look (in FEKO/COMSOL/etc) at the same time:
- axially stretched (and radially squeezed) near small end
- axially squeezed (and radially stretched) near big end
Is the following animated representation of this effect correct, from the point of view of the spatial shape and temporal evolution of the standing wave in a cylindrical vs frustum resonant cavities?
The animation is correct as long as it should illustrate the polarity(vector E into phi direction) of the E-field component of the EM-field within the dielectric* only over time. The Field intensity at the very locations depend on the cavity shape. Please see the attached paper from Dr. Rodal for more details.
The following FEKO animation below shows what happens in detail over a full 360 deg cycle (for TE012 in this case) for both E & H.
FEKO EM animation of a TE012 cavity resonator
https://forum.nasaspaceflight.com/index.php?topic=39214.msg1646352#msg1646352
*vacuum or whatever dielectric material
Yes I know this is an oversimplification, with only one component of the EM wave. It's to be clear enough.
The paper from Rodal you pointed out answers my questions about the stretching/squeezing of the EM waves, as he says in it that both the amplitude and the wavelength of the EM wave get longer when approaching the small end.
I was not sure and annoyed about the question of the amplitude variation, as I forgot it was studied in the paper. Makes sense as there is much greater energy density near small end, and the amplitude of an EM wave is directly proportional to its energy.
The amplitude of the EM wave being greater at the small end is something that was not shown at all in my simplistic 2D diagrams (on the contrary!).
Thanks for the monthly Rodal reminder
PS: I had something in mind about this axial wavelength stretching at small end and squeezing at big end that may IMHO interfere negatively with Doppler shifts when the cavity is accelerating (by an arbitrary force, yes)… maybe it's too fringe and silly. Thank you for your appreciated help to separate the wheat from the chaff! -
#253 by meberbs on 04 Aug, 2018 00:28
-
If I wasn't clear enough from the beginning, let't write it extensively: I am not trying in the last couple of messages to prove the origin or even the reality of any anomalous thrust in relation with possible Doppler shifts. I just ask if we can discuss properly about these Doppler shifts, their origin and their behavior, in various cavity shapes accelerating (or not) by an arbitrary external force. Not any possible or impossible "thrust" as this is another story. That's all.
Maybe you didn't see my last post before you posted, but my response convers these things, using an external force to provide the acceleration to avoid confusion with any anomalous force.
Here is a back of the envelope calculation of how much Doppler shift to expect:
Round trip length = 1 m (longer than 2*cavity to account for not moving in straight lines)
number of trips = 10^6
speed of light = 3e8 m/s
total time = 0.0033 s
acceleration = 1m/s^2 (way more than typical)
velocity change of cavity in photon lifetime = 0.0033 m/s
ratio of this velocity to speed of light = 1.11e-11.
That last ratio tells you the frequency shift to expect as a fraction of the original frequency as a sum of all individual Doppler shifts. You can pick different numbers, or do a more detailed calculation, but the effect will remain negligible in realistic cases.
-
#254 by RonM on 04 Aug, 2018 00:36
-
Why do you resort to an obvious wrong claim ("no force on side walls") to disprove the one being discussed (Vg1 > Vg2) which has nothing to do with it? I am sorry, but this is a suggested irrelevant conclusion using an association fallacy.
The quote I used is from the same page in the same paper you quoted. Since these statements from Shawyer are only a few paragraphs apart, I think that's relevant.
-
#255 by flux_capacitor on 04 Aug, 2018 00:38
-
Why do you resort to an obvious wrong claim ("no force on side walls") to disprove the one being discussed (Vg1 > Vg2) which has nothing to do with it? I am sorry, but this is a suggested irrelevant conclusion using an association fallacy.
The quote I used is from the same page in the same paper you quoted. Since these statements from Shawyer are only a few paragraphs apart, I think that's relevant.
Relevant to Shawyer's own broken explanation for thrust, for sure. Bur not for what I asked. -
#256 by flux_capacitor on 04 Aug, 2018 00:48
-
If I wasn't clear enough from the beginning, let't write it extensively: I am not trying in the last couple of messages to prove the origin or even the reality of any anomalous thrust in relation with possible Doppler shifts. I just ask if we can discuss properly about these Doppler shifts, their origin and their behavior, in various cavity shapes accelerating (or not) by an arbitrary external force. Not any possible or impossible "thrust" as this is another story. That's all.
Maybe you didn't see my last post before you posted, but my response convers these things, using an external force to provide the acceleration to avoid confusion with any anomalous force.
Here is a back of the envelope calculation of how much Doppler shift to expect:
Round trip length = 1 m (longer than 2*cavity to account for not moving in straight lines)
number of trips = 10^6
speed of light = 3e8 m/s
total time = 0.0033 s
acceleration = 1m/s^2 (way more than typical)
velocity change of cavity in photon lifetime = 0.0033 m/s
ratio of this velocity to speed of light = 1.11e-11.
That last ratio tells you the frequency shift to expect as a fraction of the original frequency as a sum of all individual Doppler shifts. You can pick different numbers, or do a more detailed calculation, but the effect will remain negligible in realistic cases.
Thanks for the numbers. I think the velocity to take into account for the EM wave during such Doppler shifts is not the speed of light though, but the "guide velocity" which is nothing but the local value of the group velocity of the EM wave in the cavity (ie near small end vs big end). But as you already said, nobody ever shown that the group velocity varies (inversely proportionally to the guide wavelength) in a closed tapered resonator like it does in open waveguides of various diameters, so being suspicious topics (both the "guide velocity" and the "guide wavelength") I imagine this will be classified as a cold case and we have to move on for now. -
#257 by Monomorphic on 04 Aug, 2018 21:59
-
It took me a while to get the ~20mm spacer fabricated as I have been away on vacation. This spacer is used to reduce the resonant frequency of the cavity Oyzw sent me, from 2.5Ghz to 2.4Ghz (for mode TE013). The spacer will be compressed one or two mm while aligning the big end parallel with the small end using three bolts. I also polished the end plates as they arrived a little rough around the edges.
Next step is to go back to the simulation and figure out the best place to drill the two holes for the couplers...
-
#258 by Mark7777777 on 05 Aug, 2018 01:58
-
It took me a while to get the ~20mm spacer fabricated as I have been away on vacation. This spacer is used to reduce the resonant frequency of the cavity Oyzw sent me, from 2.5Ghz to 2.4Ghz (for mode TE013). The spacer will be compressed one or two mm while aligning the big end parallel with the small end using three bolts. I also polished the end plates as they arrived a little rough around the edges.
Next step is to go back to the simulation and figure out the best place to drill the two holes for the couplers...
Did you ever publish test results from your 3D printed frustum? Forgive me if you did and I missed it. -
#259 by Monomorphic on 05 Aug, 2018 12:16
-
Did you ever publish test results from your 3D printed frustum? Forgive me if you did and I missed it.
I published one of the first tests with the 3D printed cavity, but nothing since then. Once I have two cavities that I can swap in and out, then I will run full sequences on both and publish those results. It will be interesting to compare the results between the copper foil and solid copper cavity.
