-
#2060 by TheTraveller on 09 Oct, 2017 20:56
-
...
BTW judging from the pictures, your side wall sheet is very shiny & smooth. Good sheet choice.
If I have to make thermograms in order to be sure what mode is excited, I have to paint the outside of the cavity.
Bt it is probably also possible to use watercolour paint, so I can wash it off later. 
Peter,
For your target TE013 mode, you should be able to thermally "see" the 3 eddy current rings that form around the side walls, plus the single eddy current ring that forms on the end plates. -
#2061 by TheTraveller on 09 Oct, 2017 21:03
-
Thank you
!
Would you mind trying different curves? e.g. top different from bottom or both the same but deeper... and the like ?
And with the bell shape please.Thank you
!
Would you mind trying different curves? e.g. top different from bottom or both the same but deeper... and the like ?
The Division Bell
Hi Kenjee,
The heavy eddy current ring at the small end says the bell is operating beyond cutoff (small end diameter is too small) and very few photons are making it to the small end plate.
Experience says that condition is not a good design for an EmDrive.
-
#2062 by Mulletron on 09 Oct, 2017 21:19
-
TT, where do you see an eddy current ring in that picture?
-
#2063 by ThatOtherGuy on 09 Oct, 2017 21:21
-
Thank you
!
Would you mind trying different curves? e.g. top different from bottom or both the same but deeper... and the like ?
And with the bell shape please.Thank you
!
Would you mind trying different curves? e.g. top different from bottom or both the same but deeper... and the like ?
The Division Bell
Hi Kenjee,
The heavy eddy current ring at the small end says the bell is operating beyond cutoff (small end diameter is too small) and very few photons are making it to the small end plate.
Experience says that condition is not a good design for an EmDrive.
That was just an "experiment"; to say it all, I wonder what may happen by using a half ellipse for the "bell" while keeping a parabolic (curve oriented inside as for the bell design) shape for the end plate and placing the emitter at the ellipse focus; I think that such a design may be "interesting" as well, especially if considering the current "pilot wave" theory and the totally different shape suggested
-
#2064 by Monomorphic on 09 Oct, 2017 21:26
-
...
BTW judging from the pictures, your side wall sheet is very shiny & smooth. Good sheet choice.
If I have to make thermograms in order to be sure what mode is excited, I have to paint the outside of the cavity. Bt it is probably also possible to use watercolour paint, so I can wash it off later.
Peter,
For your target TE013 mode, you should be able to thermally "see" the 3 eddy current rings that form around the side walls, plus the single eddy current ring that forms on the end plates.
I wonder if thermochromatic pigment paint would work?
https://www.amazon.com/GloMania-THRM-10G-BLU-BLUE-Thermochromic-Pigment/dp/B00SZ7X99Y -
#2065 by Monomorphic on 09 Oct, 2017 21:43
-
I ran S11 with the same signal generator/vna Peter has. Similar results for Delta, but since he is 1GHz higher, Q is higher using the -3dB method. Interesting thing is Q was ~9% lower with the spherical end-plates (~6,200) vs flat end-plates (~6,800). Whether that is a product of the foil remains to be seen. Return loss was also significantly better using the flat end-plates.
-
#2066 by TheTraveller on 09 Oct, 2017 22:35
-
I ran S11 with the same signal generator/vna Peter has. Similar results for Delta, but since he is 1GHz higher, Q is higher using the -3dB method. Interesting thing is Q was ~9% lower with the spherical end-plates (~6,200) vs flat end-plates (~6,800). Whether that is a product of the foil remains to be seen. Return loss was also significantly better using the flat end-plates.
Jamie,
For the same mode and cavity geometry, ie same Bd/Sd, Q decreases as freq increases.
Roger's suggestion to use spherical end plates was based on a conventional cavity build. Your foil introduces a new element which should cause the side wall eddy current ring to be broken into smaller isolated sections, even though the foil has a conductive adhesive. Remember that at 2.45 GHz, the eddy currents only penetrate 5x the skin depth or around 6 um into the foil and thus never reach the other surface of the foil nor reach the conductive adhesive. -
#2067 by TheTraveller on 09 Oct, 2017 22:42
-
TT, where do you see an eddy current ring in that picture?
Hi Mulletron,
There are single eddy current rings on each end plate and 2 around the side walls.
Easier to see in the attached, where there are 3 eddy current rings around the side walls and 1 on each of the end plates.
BTW the 1st cavity image is also working beyond cutoff as the side wall eddy current ring nearest the small is much stronger than the eddy current ring on the small end plate. In a cavity that is not working beyond cutoff, the highest current density eddy current ring is on the small end plate.
The wavy rings are caused by the input coupler not being correctly positioned or the freq being a little off resonance.
This also shows how I use a side wall mounted loop, actually a 1/2 loop, in the highest H field density of the central lobe, to excite by cavities. Needs to be position adjusted along the side wall and then rotated for best coupler match, so not the easiest of couplers to build and get working.
-
#2068 by flux_capacitor on 09 Oct, 2017 22:51
-
New article on the EMDrive about modelling using pilot-wave theory and it claims the results are compatible with thrust generation:
https://www.sciencealert.com/physicists-have-a-weird-new-idea-about-how-the-impossible-em-drive-could-produce-thrust
The paper referenced in the article is behind a paywall.
Has anyone just clicked on the paywall link? I tried. Maybe there is a bug in their system, maybe the paper is free: in any event it downloaded straightforward.
This is an absurdly benign traversable paywall.
http://www.ikpress.org/abstract/6485
-
#2069 by OnlyMe on 09 Oct, 2017 22:54
-
I ran S11 with the same signal generator/vna Peter has. Similar results for Delta, but since he is 1GHz higher, Q is higher using the -3dB method. Interesting thing is Q was ~9% lower with the spherical end-plates (~6,200) vs flat end-plates (~6,800). Whether that is a product of the foil remains to be seen. Return loss was also significantly better using the flat end-plates.
And you are in a perfect position to compare the two, flat vs spherical. It will be interesting no matter the results. -
#2070 by TheTraveller on 09 Oct, 2017 23:07
-
I`m going back in shadow and play with simulation of shapes and rainbow elastic balls. If someone find it interesting. Cool
Hi Kenjee,
You need to spend more time with your simulators to further understand how the small end of an asymmetric resonant cavity changes as the small end alters diameter above and below.
The TE012 image (2x side wall eddy current rings) you shared showed the small end was in cutoff. I suggest if you increased the freq to obtain TE013 resonance (3x side wall eddy current rings) you would see the highest eddy current density has shifted from the small end eddy current ring to the small end plate, meaning the small end is no longer operating beyond cutoff.
Basically what you will observe is when the small end plate is well above cutoff, the highest current density eddy current ring is formed on the small end plate. As the diameter of the small end plate decreases beyond a certain diameter, the eddy current density on the small end plate decreases to be less than the eddy current density on the side wall eddy current ring at the small end. When that happens, the small end plate is said to be operating beyond cutoff. This cutoff is not a black and white line but the small end plate eddy current does drop off very quickly as it's diameter is reduced.
Cutoff is a microwave waveguide effect where the waveguide diameter is smaller than required for efficient propagation of microwave energy. There are equations to define the cutoff diameter which involve the excited mode, freq and waveguide geometry.
Microwave engineers hate propagation loss and are not happy if there is more than a small fraction of a dB loss, so they use waveguides that are much larger than cutoff.
BTW the eddy currents are induced by the time varying H fields of the photons and the eddy current heating energy is sourced or taken from the photon's energy, which causes them to decrease their energy and increase their wavelength, until finally almost all their energy is thermalised and they have very long wavelengths. Of course the thermalised microwave photon energy heats the cavity and is radiated away by much higher frequency IR photons.
-
#2071 by Mulletron on 09 Oct, 2017 23:44
-
TT, where do you see an eddy current ring in that picture?
Hi Mulletron,
There are single eddy current rings on each end plate and 2 around the side walls.
Easier to see in the attached, where there are 3 eddy current rings around the side walls and 1 on each of the end plates.
BTW the 1st cavity image is also working beyond cutoff as the side wall eddy current ring nearest the small is much stronger than the eddy current ring on the small end plate. In a cavity that is not working beyond cutoff, the highest current density eddy current ring is on the small end plate.
The wavy rings are caused by the input coupler not being correctly positioned or the freq being a little off resonance.
This also shows how I use a side wall mounted loop, actually a 1/2 loop, in the highest H field density of the central lobe, to excite by cavities. Needs to be position adjusted along the side wall and then rotated for best coupler match, so not the easiest of couplers to build and get working.
This makes sense to me if I treat the eddy current losses as a type of friction doing negative work on the system, like an external force, so it's natural to want the force to be cosine(180 degrees) -1 instead of cosine (90 degrees) 0. -
#2072 by TheTraveller on 10 Oct, 2017 05:29
-
Not sure this makes sense, but trying to come up with something that would allow us to effectively drain the frequency from photons in the cavity. There is a lot of momentum in light but normally very little of that momentum is ever transferred in a single reflection (heat death of the universe - most of the energy gets trapped in light) and in a cavity, normally all that momentum is regained by the photons on the next reflection from the opposite wall.
Dustin,
Consider that for say a Qu of 62,800, it will take 10,000 cycles of the exciting Rf to charge or discharge the cavity as Qu = 2 * Pi * (stored energy / energy loss to wall currents per cycle).
Resonance is not needed to generate force but it is needed to fill a cavity as without resonance the cavity coupler looks to the Rf amp to be a short circuit and will reject any Rf sent to the coupler.
However once the photons are injected into the cavity by the coupler, resonance is not involved in force generation. Which says if you pulse the cavity, the trapped photons existing post the pulse will continually increase their wavelength as they lose energy to eddy currents, the coupler and conversion into work to accelerate mass. All of which continually drain photon energy and increase the wavelength.
At some lower freq, the 5x skin depth is greater than wall and end plate thickness and photons start penetrating the cavity enclosure and escaping.
Fairly simple to calc the photon freq which will generate 5x skin depth = min side wall or end plate thickness. As photon freq continues to drop below that value, increasing numbers of long wavelength photons will penetrate the cavity and escape. -
#2073 by Peter Lauwer on 10 Oct, 2017 08:45
-
...
Your foil introduces a new element which should cause the side wall eddy current ring to be broken into smaller isolated sections, even though the foil has a conductive adhesive. Remember that at 2.45 GHz, the eddy currents only penetrate 5x the skin depth or around 6 um into the foil and thus never reach the other surface of the foil nor reach the conductive adhesive.
We're talking GHz's. It is not DC current. Isn't there some capacitive coupling etc.? -
#2074 by Eusa on 10 Oct, 2017 09:49
-
I ran S11 with the same signal generator/vna Peter has. Similar results for Delta, but since he is 1GHz higher, Q is higher using the -3dB method. Interesting thing is Q was ~9% lower with the spherical end-plates (~6,200) vs flat end-plates (~6,800). Whether that is a product of the foil remains to be seen. Return loss was also significantly better using the flat end-plates.
Decoherence delay may be longer with spherical ends I guess. Meaning more optical energy content...? -
#2075 by sanman on 10 Oct, 2017 10:23
-
Thank you for indulging me with your replies, gentlemen
Not a dumb question. If the EM drive produces thrust, it's a reasonable assumption it has a correlation to field strength. Without an accepted working theory, individuals are checking many possibilities.
Is there an ideal, regarding what a plot should look like? Would that ideal be the interior of the cavity being filled with as many "red donuts" (strong toroidal electric fields) as possible? I keep imagining that the centerline that passes through all the donuts is the thrust centerline generated by these toroidal E-fields. So I keep visualizing these toroidal E-fields as "smoke-ring vortices" of Quantum Foam, which are then moving virtual particles along that centerline to produce thrust. Therefore the more "smoke-ring vortices" (toroidal E-fields) you have, and the stronger (more reddish) they are, then the better the propulsive effect. Am I way off?
In a way, that picture is consistent with the Pilot Wave theory, which holds that the virtual particles have a persistent trajectory, even if the virtual particles themselves are only transiently visible to us as quantum fluctuations.Here's some quick answers Sanman:
QuoteWhen I see the shaded zones inside the frustrum area, they seem to show concentrations of electric field strength - the reddish areas seem to be stronger than the bluish-greenish areas - but why is electric field strength indicative of propulsive capability?
It's related to Q and is critical for surpassing threshold levels in gravitational gradient/time lag based theories. The B component has been neglected slightly during the research, mainly because (opinion) TM has been less promising than TE. A popular explanation for this has to do with the direction in which you would want electron and ion pressure to equalize (small to big end), though there is an unresolved question as to whether TM is truly worse than TE. If new TM results equal those of TE then we know that the interaction of the B component with the endplates is the same as that of the E component.
So at this point it's not confirmed that optimizing for full Lorentz effect (E+B) is better than optimizing for just E alone? I've seen various COMSOL plots of E-field posted at times, but has anyone produced any of the B-field? (just curious as to what they look like, and how they compare)QuoteQuoteAnd why do these resonant cavities have to be pumped with microwaves in particular? What's so special about microwaves? Why not UV-light instead, for example? Is it because the wavelength of microwaves makes them more convenient to work with?
Microwaves have multiple unique and useful resonant freqs which are located nearby on the bandwidth. Microwave cavities are a useful size and magnetrons are common. Also, they have tolerable wall skin depths allowing for feasible DIY builds.
But other than the convenience for human hands, there's no natural benefit from using one frequency of EM over another? (Well, I'd assume that higher frequencies are harder to reflect inside the cavity, while lower frequencies are more easily reflected without losses.)
Jim once told me that a single larger rocket engine with a larger thrust chamber is better than many small ones. Likewise, we know that larger tokamaks operate more efficiently than smaller ones, because of cube-square.
I'm wondering if a larger frustrum/cavity would operate more efficiently than a smaller one, producing a better signal-to-noise ratio. On the other hand, I'd imagine that building an accurate measurement device becomes more difficult if your EMdrive gets too large.QuoteQuoteDoes the actual size of the frustrum matter, or just its ratios?
Put roughly: everything is dependent on ratios of 1/2 waves. Of course the size matters.
FYI: here are two cutoff related papers surrounding the EM Drive and two educational sources regarding more general questions about resonating waveguides.
http://whites.sdsmt.edu/classes/ee481/notes/481Lecture10.pdf
Thank you for that! I recognize that the cavity dimensions have to correspond to the frequency of EM being used, but I was just mainly wondering if the EMdrive performance changes with size scaling.
(ie. would a bigger 10m-wide EMdrive be better than a 10cm-wide one?)
I'm reminded of the challenges of Very Long Baseline Interferometry, whereby telescopes positioned at very large distances from each other can more precisely image distant objects, with the caveat that these telescopes have to be aligned at wavelength-precision.
Likewise, I was imagining that a very large EMdrive could operate more efficiently than a smaller one, with the caveat that its dimensions would have to be at a precision equal to the resonant EM wavelength (or half-wavelength). -
#2076 by Peter Lauwer on 10 Oct, 2017 11:58
-
INPUT:
SD=122mm; BD=202mm; LN=165mm (Theta=13,6270°)
Hi X_RaY,
I was not exact in giving the frustum dimensions. After carefully remeasuring:
SD=122.5(5) mm
BD=202.5(7) mm (it is a little oval)
LN=167.5(5) mm
Cheers,
Peter
-
#2077 by OnlyMe on 10 Oct, 2017 15:05
-
I ran S11 with the same signal generator/vna Peter has. Similar results for Delta, but since he is 1GHz higher, Q is higher using the -3dB method. Interesting thing is Q was ~9% lower with the spherical end-plates (~6,200) vs flat end-plates (~6,800). Whether that is a product of the foil remains to be seen. Return loss was also significantly better using the flat end-plates.
Jamie,
For the same mode and cavity geometry, ie same Bd/Sd, Q decreases as freq increases.
Roger's suggestion to use spherical end plates was based on a conventional cavity build. Your foil introduces a new element which should cause the side wall eddy current ring to be broken into smaller isolated sections, even though the foil has a conductive adhesive. Remember that at 2.45 GHz, the eddy currents only penetrate 5x the skin depth or around 6 um into the foil and thus never reach the other surface of the foil nor reach the conductive adhesive.
TT,
The following, "Roger's suggestion to use spherical end plates was based on a conventional cavity build.", doesn't say Roger ever actually built a frustum with spherical end plates. I know there have been drawings, but no actual pictures claiming spherical end plates, that I remember. Does that suggestion come from actual experience or what he expects based on his theory of operation?
BTW It might cut down on some of the criticism if you mention in your posts whether your comments are calculated, based on design theory, or some practical experimental experience. Folks here have no issue discussing theory and what ifs, just problems with what appears to be theory presented as fact... And an extension of design theory to critical "suggestions".
To this, "Your foil introduces a new element which should cause the side wall eddy current ring to be broken into smaller isolated sections, even though the foil has a conductive adhesive."... The skin depth on Jamie's foil tape would not be broken by any over lapping edges, as long as the adhesive is also conductive, the skin depth would just run over the tape edge and through the adhesive, maybe acting like an imperfection in the dimensions, rather than a break in conductivity. The eddy currents don't have to penetrate the tape to reach the adhesive. They would just "flow" over the edge of the tape across the small adhesive gap and into the next section of tape. A tape overlap would not even act like a scratch as far as how the microwaves interact with the foil and the eddy currents would follow any contour. If the current were high enough there might be arching across the adhesive gap between two sections of tape, but so far Jamie is working with only 30 watts?
The foil may not be as effective as a plated surface, but your argument above would seem to require a non conductive adhesive to be of critical effect.
I could be mistaken but I believe Jamie has mentioned in the past that plating the component parts of his frustum would be possible if needed. And the fact that he is printing them (the frustum sections) makes design adjustments far easier than working with other materials, even if there might be some issues with heat at higher power levels.
-
#2078 by Kenjee on 10 Oct, 2017 15:15
-
Someone asked for magnetic plots and for Dr. Rodal energy density.
-
#2079 by Kenjee on 10 Oct, 2017 15:19
-
Same for Bell
Bt it is probably also possible to use watercolour paint, so I can wash it off later. 
