-
#60 by Rodal on 11 Feb, 2016 00:31
-
....
Well, you have chosen to ignore Dr. White's and Dr. Minotti's theory for the EM Drive regarding negative energy-mass, I have chosen to address its implications. Regarding
My considerations of energy conservation, since I made them under the specific situation of stationary thrust and stationary velocity, quite don't care about the sign of energy or mass terms. They care about some lack of equality, the fact that so far we are not aware of a frame agnostic definition of total Energy for the whole system, in a deep space context (not reacting on lab's walls), such that we don't see Etot(t+Δt)≠Etot(t). While it's true I spoke mainly about apparent excess of energy, the same idea (stationary thrust at stationary velocity) can be applied to make apparent complete wipeout of some energy content, that is as problematic.Quote from: Frobnicatso far we are not aware of a frame agnostic definition of total Energy for the whole system, in a deep space context (not reacting on lab's walls), such that we don't see Etot(t+Δt)≠Etot(t)
, see this discussion of a more familiar concept, dark energy, and how this is addressed:
http://www.preposterousuniverse.com/blog/2010/02/22/energy-is-not-conserved/
I have not seen these issues addressed in Prof. Woodward's discussions of negative energy-mass, please let me know if you are aware of any such discussion. -
#61 by frobnicat on 11 Feb, 2016 00:57
-
...This is a lot of trouble just for asserting that it could conceivably be a closed system when, given the phenomenology of the considered equations, all indicates that it is open....
My God, I posted with bold red signs and a moving banner, something titled "Under construction" and now you are stating that I am asserting that the EM Drive is a closed system that works on spontaneous and continuous creation of negative mass-energy
Considering what are the implications if the EM Drive would be a closed-system does not at all mean that one is "asserting" that it is
Such considerations are par for the course for anybody involved in R&D !
I'm not stating that you are asserting that the EM Drive is a closed system, I say that this assertion (that your thought experiment and derivation imply, as an hypothesis) would be a lot of trouble... You were clear enough about the speculative nature of all that. Sorry if I'm cluttering this thread, hope this can be of use as possibly representative of the kind of indignation you'll have to face if you were to present that to a wider audience, and help toward a more explicit exposition of the premises. I'll be back after construction, to haunt the basement
-
#62 by Rodal on 11 Feb, 2016 01:09
-
...This is a lot of trouble just for asserting that it could conceivably be a closed system when, given the phenomenology of the considered equations, all indicates that it is open....
My God, I posted with bold red signs and a moving banner, something titled "Under construction" and now you are stating that I am asserting that the EM Drive is a closed system that works on spontaneous and continuous creation of negative mass-energy
Considering what are the implications if the EM Drive would be a closed-system does not at all mean that one is "asserting" that it is
Such considerations are par for the course for anybody involved in R&D !
I'm not stating that you are asserting that the EM Drive is a closed system, I say that this assertion (that your thought experiment and derivation imply, as an hypothesis) would be a lot of trouble... You were clear enough about the speculative nature of all that. Sorry if I'm cluttering this thread, hope this can be of use as possibly representative of the kind of indignation you'll have to face if you were to present that to a wider audience, and help toward a more explicit exposition of the premises. I'll be back after construction, to haunt the basement
I'm glad we understand each other that this is a speculative analysis to examine what are the implications of the EM Drive being a closed system. Who knows, sometimes useful things come out of such speculations
Thanks to you and everybody dedicating their valuable time to post comments: they have been, are, and always will be most welcome, and appreciated
-
#63 by X_RaY on 14 Feb, 2016 20:34
-
I have tried to handle feko lite but I run into several problems due to the strong limitation of the trial version..
Till now the only working model is a TE011 version excited from the large base. For all later modifications the program stops with limitation notification or solving problems because of coarse port-mesh and similar issue.
Nice try but less useful in this configuration. For sure a full version of the program would be very nice.
-
#64 by X_RaY on 18 Feb, 2016 19:40
-
OK just spending a little more time on FEKO LITE.
I remove the waveguide excitation (used before) and instead of this I currently use 2 dipoles near the big base. The mesh is still, aaah yes, I dont like to talk about...
The version is restricted to 500 volume elements, these are 450.
Nevertheless I could identify the same mode TE011 again near 2.4GHz. (pic 1&2)
After that I used my spreadsheet and try to use the following dimensions:
SD=150
BD=300
L=270
TE013 has to be nearby ~2.4GHz (using the bad mesh it may be close enought)
I am a little perplex for the field pattern delivered by FEKO LITE for this last run. Looks more like TE012 (the two last pics below) 
(Look to my avatar pic, this looks like TE013(!) created via EMPro full version)
https://forum.nasaspaceflight.com/index.php?topic=37642.msg1412912#msg1412912
-
#65 by X_RaY on 18 Feb, 2016 20:24
-
Spreadsheet results for this dimensions using different calculation methods:
-
#66 by X_RaY on 19 Feb, 2016 18:43
-
I got it
-
#67 by X_RaY on 21 Feb, 2016 17:42
-
At the moment I tried to place dielectric material at the small base.
The excitation is still realized using two dipoles near the large base.
These dipoles are no ports where S-parameter can be extracted by FEKO Lite, only the power level can be displayed and due to the trial version only 10 data points over the chosen frequency range.
The visible mode is still TE013 in both cases shown in the pics below.
EDIT:
Poynting vector fields can also be displayed. (I don't know if this is an average over a full cycle since I am using the dipoles.)*
*I am familiar with the the possibilities of EMPro. There an endless animation can be shown over a full cycle using any fixed frequency at least for the evolution of the E and H fields, but I never try to show the Pointing fields in EMPro. I will look for it if I have some time.
-
#68 by X_RaY on 22 Feb, 2016 13:25
-
Instead of the permittivity of 3.2 I tried to redefine the inlay using the same values like before but now for permeability of the material. The value seems unnatural but this is for comparison with the run before only.
The resulting field pattern is quite different to the last calculation, especially the amplitudes of the single lobes.
-
#69 by X_RaY on 25 Mar, 2016 20:07
-
I think about the "better as large as possible" condition to get very high Q value. Using the linear Q approximation it follows that the Q factor will be larger also the higher the mode indices be. Using well known dimensions, of course at higher frequencies, for the Brady frustum I get a Q value at ~24.5GHz of ~730000 for TE45(34) [the mode number is only used as an example]. The total number may be wrong** but the context still holds. Now based on this I understand the usage of whispering modes gallery in the Cannae thruster much better, it simply leads to higher Q values. It's not only the volume of the cavity, in general frequency volume and mode number are of interest*.
It's simply another solution/expression of the linearity of the maxwell equations while scaling some factors.
* Beside other effects like εr & tanδ of the volume material and the cavity wall conductivity (σ) and so on.
** As discussed in this thread the total Q number for higher than the fundamental modes don't holds using a simple approximation formula. https://forum.nasaspaceflight.com/index.php?topic=39214.msg1476709#msg1476709
-
#70 by Rodal on 25 Mar, 2016 22:44
-
I think about the "better as large as possible" condition to get very high Q value. Using the linear Q approximation it follows that the Q factor will be larger also the higher the mode indices be. Using well known dimensions, of course at higher frequencies, for the Brady frustum I get a Q value at ~24.5GHz of ~730000 for TE45(34) [the mode number is only used as an example]. The total number may be wrong** but the context still holds. Now based on this I understand the usage of whispering modes gallery in the Cannae thruster much better, it simply leads to higher Q values. It's not only the volume of the cavity, in general frequency volume and mode number are of interest*.
It's simply another solution/expression of the linearity of the maxwell equations while scaling some factors.
* Beside other effects like εr & tanδ of the volume material and the cavity wall conductivity (σ) and so on.
** As discussed in this thread the total Q number for higher than the fundamental modes don't holds using a simple approximation formula. https://forum.nasaspaceflight.com/index.php?topic=39214.msg1476709#msg1476709
Very interesting.
How did you calculate the Q for TE45? Is m=4, n=5 ? What is p?
Did you use FEKO?
I would like to have an analytical result showing this to check it and understand it. Essentially one needs a mode shape that maximizes the following quantity:
∫ElectromagneticEnergy dV/ ∫ ElectromagneticEnergy dA
-
#71 by X_RaY on 26 Mar, 2016 08:42
-
I used my spreadsheet and formula discussed in page 2 of this thread. https://forum.nasaspaceflight.com/index.php?topic=39214.msg1476709#msg1476709I think about the "better as large as possible" condition to get very high Q value. Using the linear Q approximation it follows that the Q factor will be larger also the higher the mode indices be. Using well known dimensions, of course at higher frequencies, for the Brady frustum I get a Q value at ~24.5GHz of ~730000 for TE45(34) [the mode number is only used as an example]. The total number may be wrong** but the context still holds. Now based on this I understand the usage of whispering modes gallery in the Cannae thruster much better, it simply leads to higher Q values. It's not only the volume of the cavity, in general frequency volume and mode number are of interest*.
It's simply another solution/expression of the linearity of the maxwell equations while scaling some factors.
* Beside other effects like εr & tanδ of the volume material and the cavity wall conductivity (σ) and so on.
** As discussed in this thread the total Q number for higher than the fundamental modes don't holds using a simple approximation formula. https://forum.nasaspaceflight.com/index.php?topic=39214.msg1476709#msg1476709
Very interesting.
How did you calculate the Q for TE45? Is m=4, n=5 ? What is p?
Did you use FEKO?
I would like to have an analytical result showing this to check it and understand it. Essentially one needs a mode shape that maximizes the following quantity:
∫ElectromagneticEnergy dV/ ∫ ElectromagneticEnergy dA
I simply picked the Bessel value of TE45 and choose the p-number to mach 24.5GHz, for the given geometry its p=34
(34 half wavelength in axial-direction).
I also tried this for TE01p with p=37, fres≈24.3GHz, the resulting Q is 339838.
The cavity is simple larger in relation of the wavelength. So instead to use a low order mode at low frequency which leads to a large cavity one also can store a equivalent high energy using higher frequency and mode number while keeping the cavity volume constant. This describes why its more efficient to use TE013 instead of TE011 for example. The energy desity depends on the frequency/wavelength.
I fully agree to the volume to surface relation!
I can not confirm this with FEKO because I have only the trial version.
Much more than the possible 500 volume elements would be necessary for 10 times higher frequency and this huge mode order.
-
#72 by X_RaY on 26 Mar, 2016 14:50
-
If I have the time I could check this using EMPro within the next week's. Using the eigen resonance solver the resonant frequency and the Q will be displayed directly for each mode.
For now I only have the results of the spreadsheet with the approximation formulas.
-
#73 by Rodal on 26 Mar, 2016 15:01
-
If I have the time I could check this using EMPro within the next week's. Using the eigen resonance solver the resonant frequency and the Q will be displayed directly for each mode.
It would be great if you check this with EMPro. It seems to me that to maximize Q one wants to maximize
For now I only have the results of the spreadsheet with the approximation formulas.
∫ElectromagneticEnergy dV/ ∫ ElectromagneticEnergy dA
this means minimizing ∫ ElectromagneticEnergy dA
while maximizing
∫ElectromagneticEnergy dV
this means a mode shape that would have most of the high electromagnetic energy in the interior volume of the cavity instead of the exterior of the cavity near the metal -
#74 by X_RaY on 31 Mar, 2016 18:35
-
Dr. Rodal is it possible that the displacement of a large number of electrons inside the copper walls could cause a temporal displacement of the cavity/ force on the cavity? Since electrons will driven by thermal currents, they travel away from the heat source into the direction of the heatsink (cooler end of the cavity) due to their statistical velocity**. The center of mass may also change due to this effect. In the case of the (comsol)pic below the electrons will driven in the direction of the small end while the displacement of the cavity itself would be forwards into the direction of the big end diameter. Of course this effect is reversible and holds only as long as a thermal equilibrium is re-established.
https://en.wikipedia.org/wiki/Thermophoresis
http://aerosols.wustl.edu/Education/Thermophoresis/section02.html
**Some years ago I measured this kind of termal driven electron current using a brass rod and a very high sensitive fluxgate sensor system. One end of the rod was heated by a flame for a few seconds the other not. It was easy to measure the magnetic field of this current and its direction/sign dependent on which end of the rod was heated (while the position of the sensor and the rod was static all time) . When no temperatur gradient was present, no field could be detected.
-
#75 by Rodal on 01 Apr, 2016 13:31
-
Dr. Rodal is it possible that the displacement of a large number of electrons inside the copper walls could cause a temporal displacement of the cavity/ force on the cavity? Since electrons will driven by thermal currents, they travel away from the heat source into the direction of the heatsink (cooler end of the cavity) due to their statistical velocity**. The center of mass may also change due to this effect. In the case of the (comsol)pic below the electrons will driven in the direction of the small end while the displacement of the cavity itself would be forwards into the direction of the big end diameter. Of course this effect is reversible and holds only as long as a thermal equilibrium is re-established.
https://en.wikipedia.org/wiki/Thermophoresis
http://aerosols.wustl.edu/Education/Thermophoresis/section02.html
**Some years ago I measured this kind of termal driven electron current using a brass rod and a very high sensitive fluxgate sensor system. One end of the rod was heated by a flame for a few seconds the other not. It was easy to measure the magnetic field of this current and its direction/sign dependent on which end of the rod was heated (while the position of the sensor and the rod was static all time) . When no temperatur gradient was present, no field could be detected.
Interesting. For resonant cavities used in particle accelerators, there is also the phenomenon of multipaction that has long been known to be a problem for operation of resonant cavities. Multipaction is described as an electron resonance effect that occurs when radio frequency fields accelerate electrons in a vacuum and cause them to impact with a surface, which depending on its energy, release one or more electrons, and hence the effect cascades:
https://en.wikipedia.org/wiki/Multipactor_effect
It is not clear to me how these effects would result in self-acceleration of the cavity, since they are all internal effects and no mass or energy is expelled out of the cavity. So, the mass and energy of the cavity doesn't change, does it?
Is there a way to explain self-acceleration ? -
#76 by X_RaY on 01 Apr, 2016 16:47
-
No question the effect described in my last post is not useful to produce thrust in this sense, but it could lead to false interpretation of some displacement of an torsion pendulum for example. Whereas at the moment I have no idea about the amount of involved thermal driven electrons and if the effect would be big enough or just lower order. One could measure how much electrical DC current is needed to generate the same power of the H field as in a well defined thermal driven case to get an idea of the numbers?Dr. Rodal is it possible that the displacement of a large number of electrons inside the copper walls could cause a temporal displacement of the cavity/ force on the cavity? Since electrons will driven by thermal currents, they travel away from the heat source into the direction of the heatsink (cooler end of the cavity) due to their statistical velocity**. The center of mass may also change due to this effect. In the case of the (comsol)pic below the electrons will driven in the direction of the small end while the displacement of the cavity itself would be forwards into the direction of the big end diameter. Of course this effect is reversible and holds only as long as a thermal equilibrium is re-established.
https://en.wikipedia.org/wiki/Thermophoresis
http://aerosols.wustl.edu/Education/Thermophoresis/section02.html
**Some years ago I measured this kind of termal driven electron current using a brass rod and a very high sensitive fluxgate sensor system. One end of the rod was heated by a flame for a few seconds the other not. It was easy to measure the magnetic field of this current and its direction/sign dependent on which end of the rod was heated (while the position of the sensor and the rod was static all time) . When no temperatur gradient was present, no field could be detected.
Interesting. For resonant cavities used in particle accelerators, there is also the phenomenon of multipaction that has long been known to be a problem for operation of resonant cavities. Multipaction is described as an electron resonance effect that occurs when radio frequency fields accelerate electrons in a vacuum and cause them to impact with a surface, which depending on its energy, release one or more electrons, and hence the effect cascades:
https://en.wikipedia.org/wiki/Multipactor_effect
It is not clear to me how these effects would result in self-acceleration of the cavity, since they are all internal effects and no mass or energy is expelled out of the cavity. So, the mass and energy of the cavity doesn't change, does it?
Is there a way to explain self-acceleration ?
Again I dont think there is realy thrust based on such effects without something can escape from the cavity through the walls. I was simply talking about a temporary displacement of electrons from one end of the cavity to the other and the related backreaction nothing else.
BTW this could explain why some experiments (with modes like TE013) generates a force in the opposite direction, for this mode the heating is more effective at the small end due to the larger field strength and wall currents.
Pure speculation.
The effect you describe is also interesting. I would be surprised if the impacted electrons could escape through the relative thick metal walls. -
#77 by X_RaY on 01 Apr, 2016 21:37
-
Q measurement of a resonator with matched antenna impedance
In the orginal EMDrive forum at NSF there where discussions how to measure the Q in general.
This sample measurement shows the difference of the correct Q measurement in contrast to a bad one. The DUT is a conical cavity resonator made of steel with a conductivity of ~1.4e6 S/m, the mode is TE011 @ 24.192318GHz, the antenna is optimized to excite this mode shape and the antenna feed contains adjusting elements for impedance match. The resonator contained a thin dielectric plate at the small base. The following example shows a S11 measurement.
Rather than measure the 3dB above from the peak of the resonance curve which leads to absurd high Q value (pic 3) one has to measure the -3dB loss from the maximal reflected energie level down*. Pic 1 shows this, pic 2 confirms the measurement conditions in the complex plane. This kind of go ahead is consistent as be shown in many textbooks.
(Nasa report also shows how to measure the correct value!)
Marker 1 shows the center of the peak
Marker 5 & 6 marks the -3dB border measured against the reference Marker 4 at -0.9dB.
Q=f/df
Q=fcenter/fhigh - flow
The result is a Q of ~1055 which is a natural value for a steel cavity resonator in this frequency regime with dielectric insert.
(In addition the result is backed by EMPro calculations.)
In contrast to the upper result while using measurements like in the last pic, 3dB above the peak the calculation leads to a Q of ~22000 what ist quite unrealistic for this frequency, mode shape and cavity material.
*Please note that this simple metode is usable under impedance matched conditions, for great over- or under-coupling the situation is more complicated and the calculation is more complex. Here the simple -3dB approximation condition is not longer useful.
For more details see here:
http://www-elsa.physik.uni-bonn.de/Lehrveranstaltungen/FP-E106/E106-Erlaeuterungen.pdf
EDIT
https://forum.nasaspaceflight.com/index.php?topic=41732.msg1635542#msg1635542
-
#78 by TheTraveller on 01 Apr, 2016 21:53
-
Q measurement of a resonator with matched antenna impedance
In the orginal EMDrive forum at NSF there where discussions how to measure the Q in general.
This sample measurement shows the difference of the correct Q measurement in contrast to a bad one. The DUT is a conical cavity resonator made of steel with a conductivity of ~1.4e6 S/m, the mode is TE011 @ 24.192318GHz, the antenna is optimized to excite this mode shape and the antenna feed contains adjusting elements for impedance match. The following example shows a S11 measurement.
Rather than measure the 3dB above from the peak of the resonance curve which leads to absurd high Q value (see last pic) one has to measure the -3dB loss from the maximal reflected energie level down*. Pic 1 shows this, pic 2 confirms the measurement conditions in the complex plane. This kind of go ahead as consistent as be shown in many textbooks.
(Nasa report also shows how to measure the correct value!)
Marker 1 shows the center of the peak
Marker 5 & 6 marks the -3dB border measured from the reference Marker 4 at -0.9dB.
Q=f/df
Q=fcenter/fhigh - flow
The result is a loaded Q of ~1055 which is a natural value for a steel cavity resonator in this frequency regime.
(The result is backed by EMPro calculations also)
In contrast to the upper result while using measurements like in the last pic, 3dB above the peak the calculation leads to a Q of ~22000 what ist quite unrealistic for this frequency, mode and cavity material.
*Please note that this simple metode is usable under impedance matched conditions, for great over- or under-coupling the situation is more complicated and the calculation is more complex. Here the simple -3dB approximation condition is not longer useful.
For more details see here:
http://www-elsa.physik.uni-bonn.de/Lehrveranstaltungen/FP-E106/E106-Erlaeuterungen.pdf
Is simple to directly measure the unloaded Q of a resonant cavity by using TC = Qu / (2 Pi Fres)
With an empty cavity, apply a resonant Rf signal and measure the time it takes forward power to increase from the initial value of 0 to 63.2% of the final value. That is 1 TC time. Then Qu = 1 TC time X 2 Pi Fres. -
#79 by X_RaY on 02 Apr, 2016 20:43
-
This is your interpretation, for me at this single frequencies (-3dB below the max reflection) it's simple the resulting reflection coefficient (of an oscillating RLC circuit) for a cavity near the resonant frequency. It also describes a defined amount of stored RF energy at given frequencies for this circuit..Q measurement of a resonator with matched antenna impedance
In the orginal EMDrive forum at NSF there where discussions how to measure the Q in general.
This sample measurement shows the difference of the correct Q measurement in contrast to a bad one. The DUT is a conical cavity resonator made of steel with a conductivity of ~1.4e6 S/m, the mode is TE011 @ 24.192318GHz, the antenna is optimized to excite this mode shape and the antenna feed contains adjusting elements for impedance match. The following example shows a S11 measurement.
Rather than measure the 3dB above from the peak of the resonance curve which leads to absurd high Q value (see last pic) one has to measure the -3dB loss from the maximal reflected energie level down*. Pic 1 shows this, pic 2 confirms the measurement conditions in the complex plane. This kind of go ahead as consistent as be shown in many textbooks.
(Nasa report also shows how to measure the correct value!)
Marker 1 shows the center of the peak
Marker 5 & 6 marks the -3dB border measured from the reference Marker 4 at -0.9dB.
Q=f/df
Q=fcenter/fhigh - flow
The result is a loaded Q of ~1055 which is a natural value for a steel cavity resonator in this frequency regime.
(The result is backed by EMPro calculations also)
In contrast to the upper result while using measurements like in the last pic, 3dB above the peak the calculation leads to a Q of ~22000 what ist quite unrealistic for this frequency, mode and cavity material.
*Please note that this simple metode is usable under impedance matched conditions, for great over- or under-coupling the situation is more complicated and the calculation is more complex. Here the simple -3dB approximation condition is not longer useful.
For more details see here:
http://www-elsa.physik.uni-bonn.de/Lehrveranstaltungen/FP-E106/E106-Erlaeuterungen.pdf
Is simple to directly measure the unloaded Q of a resonant cavity by using TC = Qu / (2 Pi Fres)
With an empty cavity, apply a resonant Rf signal and measure the time it takes forward power to increase from the initial value of 0 to 63.2% of the final value. That is 1 TC time. Then Qu = 1 TC time X 2 Pi Fres.
It simply defines the bandwidth per definition as well as the dimensionless Q value.
I agree its possible to calculate the time for a given circuit to reach a defined amount of stored energy as you state like for simple a capacitor, when all circuit parameter frequency,R,L&C are known as well as the power of the source**.
I dont know how helpful this definition could be to explain the generation of thrust in the case of a turncated conical cavity. As a first step I would try to understand steady state/CW conditions until try to explain pulse driven emdrive resonators. You may do it using the way you like.
**Or normalized to simply "1".
(Look to my avatar pic, this looks like TE013(!) created via EMPro full version)