Author Topic: EM Drive Developments - related to space flight applications - Thread 2  (Read 2163225 times)

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
@ Mulletron

["As far as the difference in vacuum energy goes, we've discussed the possibility that there exists a "more negative" energy condition at the small end of the cavity WRT the large end. Less modes fit small end vs large end. No calculations were made."]

This is the dispersion relation calculation.  The evaluation is made as the difference from one end of the cavity to the other.  A "boost" is made to an accelerated frame of reference which eliminates that difference, ie. "v is everywhere close to c"  This is just the "trivial" approximation as in:

* Hydrodynamics of the Vacuum_0409292v2.pdf

" However, the vacuum is a Lorentz invariant medium; it has no rest frame. The appropriate frame for the NFA is determined solely by the initial conditions. If in some frame the NFA conditions are satisfied at t = 0 then they will remain satisfied at all later times. One may trivially take a NFA solution and boost it by a large Lorentz boost to obtain an approximate solution to the original relativistic equations in which v is everywhere close to 1. Only when the range of v values is a significant fraction of unity is it necessary to abandon the NFA and return to the relativistic equations, (4.26, 4.27)."

The (static) force then appears as the equivalent "weight" of the photons in the AFR.

Good point, that's in page 8 of Hydrodynamics of the Vacuum_0409292v2.pdf

Also in page 9:

Quote from: page 9 of Hydrodynamics of the Vacuum_0409292v2.pdf
Although the flow velocity is nonrelativistic (v ≪ 1), disturbances tend to “propagate” superluminally, at 1/v. Hence, the NFA here is not a normal nonrelativistic reduction. The resulting equations are “anti-Galilean” invariant...This is certainly strange, and takes some getting used to, but one should simply view it
as an approximation to the full Lorentz transformations, valid in the stated context. One
is used to dealing with small objects that move slowly, so that their density distributions
vary rapidly in space, but slowly in time. In the present case one is dealing with large
objects, slowly varying in space, but relatively rapidly varying in time. This is related to
the fact that the Higgs vacuum, as a spontaneous Bose-Einstein condensate, has almost
all its particles in the same quantum state. Small disturbances of this state involve vast
numbers of particles, spread over long distances, all moving nearly in lockstep, so that
the disturbance varies only slowly with position while the whole collective has the same,
relatively rapid time dependence.
(Bold added for emphasis)
« Last Edit: 02/08/2015 02:27 PM by Rodal »

Offline Notsosureofit

  • Full Member
  • ****
  • Posts: 656
  • Liked: 704
  • Likes Given: 1367
Yes, I have to look at that in the self-acceleration.  Seems I thought they were using instantaneous, but I'll have to check.  It's pretty heavy going for an old guy.

Offline Mulletron

  • Full Member
  • ****
  • Posts: 1116
  • Liked: 779
  • Likes Given: 1030
Paul, so last summer it was reported that TE012 was a top performer yet difficult to work with:
Quote
The measured power applied to the test article was measured to be 2.6 watts, and the (net) measured thrust was 55.4 micronewtons. With an input power of 2.6 watts, correcting for the quality factor, the predicted thrust is 50 micronewtons. However, since the TE012 mode had numerous other RF modes in very close proximity, it was impractical to repeatedly operate the system in this mode, so the decision was made to evaluate the TM211 modes instead.

The landscape appears to have changed somewhat:
Quote
approx. +50 micro-Newton (uN) with 50W at 1,937.115 MHz
Quote
BTW, we have found that both the TE and TM E&M modes of this copper frustum can produce a thrust signature, but so far the TM modes appear to be the better performer, at least for the few modes we have been able to study to date. 

So a couple questions from this. It appears that TM modes are the top dogs now and at the same time performance has gone down significantly since vacuum testing began. See table below for what I mean. TM212 reported now vs TM211 reported then? Do you have any insight about this? Did the vacuum serve to eliminate artifact thrust signals significantly?

Also is TE012 still a good q-thruster performer, but it is just a dog to work with? It showed promising results with only 2.6 watts input producing 55.4uN of thrust (only 1 run though). Is TE012 not so good after all? If so, what made the difference?

Finally, quoting from last summer's paper:
Quote
The tapered thruster has a mechanical design such that it will be able to hold pressure at 14.7 pounds per square inch (psi) inside of the thruster body while the thruster is tested at vacuum to preclude glow discharge within the thruster body while it is being operated at high power.



So it appears the cavity contains atmospheric pressure air inside. Given the thin walls of the cavity and the end sections, did you experience any issues with the resonant cavity expanding under suction of vacuum? And did it affect your testing significantly?

Thanks in advance. There's likely to be a lot of questions coming. We really appreciate all that you have done so far.

On a separate note @Notsosureofit made a calculation based off the dispersion relation inside the cavity that yielded pretty tight results starting here: http://forum.nasaspaceflight.com/index.php?topic=36313.msg1317866#msg1317866


« Last Edit: 02/08/2015 03:30 PM by Mulletron »
Challenge your preconceptions, or they will challenge you. - Velik

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
Paul, so last summer it was reported that TE012 was a top performer yet difficult to work with:
...
The landscape appears to have changed somewhat:
Quote
approx. +50 micro-Newton (uN) with 50W at 1,937.115 MHz
Quote
BTW, we have found that both the TE and TM E&M modes of this copper frustum can produce a thrust signature, but so far the TM modes appear to be the better performer, at least for the few modes we have been able to study to date. 

So a couple questions from this. It appears that TM modes are the top dogs now and at the same time performance has gone down significantly since vacuum testing began. See table below for what I mean. TM212 reported now vs TM211 reported then? Do you have any insight about this? Did the vacuum serve to eliminate artifact thrust signals significantly?

Some caveats here:

1) When discussing mode shapes in experiments, the first question is whether the mode shape has been actually verified with experimental measurements.  My reading of the information is that the answer is NO.   There has been no experimental verification of what the actual mode shape was in the NASA experiments (either in the original "Anomalous ..." Brady et.al. report or in the new experiments).  Is my reading correct? Has anyone actually measured the electric and magnetic fields inside the EM Drive during the experiments to experimentally verify what the actual mode shape was?  Are there experimental measurements showing the orientation, gradient, and magnitude of the electrical and magnetic fields inside the EM Drive that can enable NASA to plot contour plots of the mode shapes to asses what particular mode shapes have taken place?

2) The mode shapes discussed in the original "Anomalous ..." Brady et.al. report or in the new experiments are based on COMSOL Finite Element modeling of the experiments.  Having written Finite Element computer programs, having been involved in their theoretical formulation for very nonlinear problems at MIT and elsewhere, as well as their numerical implementation, and having used (and written constitutive equation subroutines for) commercial codes like ADINA, ANSYS, and others, I'm quite aware of issues dealing with a) theoretical formulation, b) numerical implementation and c) convergence of the Finite Element mesh.  By no means one can accept a Finite Element solution as a correct modeling of reality.  As a minimum one needs to a) show convergence of the finite element solution and preferably b) comparison of the Finite Element solver to exact solutions.  In this case, there is a readily available solution for cylindrical cavities.  I would like to see a comparison of COMSOL Finite Element (using a similar space and time domain discretization) to the exact solution. 

3) For the problem at hand (truncated cone cavity) we know that exact solutions show that the truncated cone has mode shapes that are similar to those in a cylindrical cavity, but that the mode shapes in a truncated cone  have a transition to an evanescent region. A strict distinction between pure propagating and pure evanescent modes in a truncated cone can not be achieved.  Hence from a rigorous point of view, it is not correct to use the same terminology for the modes in a truncated cone than as used for the modes in a cylindrical cavity (TM212 or TE012, for example).  If we are going to use the same terminology, we need to better define our convention: is this terminology used in the sense that only real solutions to the eigenvalue problem in a truncated cone are taken into account?

Therefore, there are very important issues discussing what the actual mode shape in these experiments are.

I suggest to proceed as follows.  Let's first start with addressing whether the actual mode shapes were measured in the experiments and if so how they were measured. 

If the answer is no, that they were not measured, then let's then proceed to validate a numerical solution.  Can NASA show a comparison of a COMSOL Finite Element solution for a cylindrical cavity (with similar discretization as used for the truncated cone) vs. the exact solution?

Once that has been done, we should proceed to the next question: is the COMSOL Finite Element solution solving the eigenvalue problem for the truncated cone taking into account evanescent modes, that is, is it considering complex value solutions to the eigenvalue problem? or is it only considering real solutions to the eigenvalue problem?

« Last Edit: 02/08/2015 04:13 PM by Rodal »

Offline Star-Drive

  • Member
  • Full Member
  • ****
  • Posts: 829
  • TX/USA
  • Liked: 876
  • Likes Given: 9
Rodal:

Look at the copper frustum part-2 COMSOL & IR thermal study that I submitted to this group yesterday for an answer to your "Have you experimentally verified that we are using the TM212 mode as predicted by our COMSOL simulations?  The answer BTW is yes for the TM212 mode, but no for the TE012 mode, but since COMSOL predicted the right PE loaded resonant frequency for the TE212 mode as verified by my IR camera studies 1 & 2 of the copper frustum, I would assume that it got it right for the TE012 mode as well.  In fact I should have provided you my IR study-1 first, so find it attached.

Best, Paul M. 
Star-Drive

Offline Star One

  • Senior Member
  • *****
  • Posts: 8249
  • UK
  • Liked: 1333
  • Likes Given: 168

Folks:

If the quantum vacuum is degradable and malleable as we think it should be, then to conserve momentum a QV wake has to be generated in the QV media as a Q-Thruster goes by just like a ship's propeller leaves a disturbance in the water as it goes by.  We think that the density of the QV is normally around its cosmological average of 9.1x10^-27 kg/m^3, but its density can be greatly increased by the presence of E&M fields and especially very strong and fast time-varying E&M fields that occur is microwave resonant cavities with large Quality Factors greater than say 1,000, or around elementary charged particles like electrons or protons where the QV density goes up to nuclear mass density as you approach the surface of the particle.  Suggest anything?  However in the paper we are now trying to get published with no takers so far, we find that the QV density should drop off very rapidly from a high density volume like a proton and in fact it follows the same drop off in density with distance as the Casimir effect does, i.e., 1 / r^4 where r = the distance from the resonant cavity boundary.  With that being the case it would be near impossible to detect the QV wake behind a Q-Thruster only generating milliNewtons or Newtons or even in tens of Newtons. 

So what's to do?  To detect a QV wake from a Q-thruster at even short distances from the source we think we will have to use another RF excited resonant cavity in a form of QV parametric amplification that is designed to produce a high density QV state just like in a Q-Thruster, but not to produce thrust.  Instead it will be optimized to monitor its time varying QV density as various very weak QV wake fields come in, are amplified and detected, then pass out of it again to go back to the low density QV state once again.  This has some interesting implications especially when you finish reading the attached paper from a PhD from Rice University here in Houston.

Last topic for the night for me.  Someone on this list asked if one could extract energy from the QV.  If the QV is GRT space-time, and space-time is the cosmological gravitational field that is created by all the causally connected mass/energy in our section of the universe, then we live in a high pressure sea of gravitational energy.  Now if the QV energy state is degradable and locally changeable, then one can posit the possibility of a thermodynamic energy conversion cycle that can extract energy from a pressure difference created in this QV media relative to the QV background average pressure, with a net decrease in this universal gravitational pressure or temperature reflective of the amount of energy so extracted.  And try to remember that gravitational energy is negative energy.  I'll leave the rest to you folks to draw your own conclusions from what this might mean...

Best, Paul March

Thank you for participating in the  forum Paul. As far as the paper goes, why not publish publicly and let your peers see it and validate it without the "Star Chamber" reviewers?

Regarding the QV wake, does measuring it really matter in terms of validity if tens of Newtons of thrust (or more) are predictably being measured?
 [Serious question]

As to your last few sentences. Woah!!!.......

As too publishing publicly, where publicly would be the best place to publish to get the most eyes on it from the type of people you want to see it?

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
Rodal:

Look at the copper frustum part-2 COMSOL & IR thermal study that I submitted to this group yesterday for an answer to your "Have you experimentally verified that we are using the TM212 mode as predicted by our COMSOL simulations?  The answer BTW is yes for the TM212 mode, but no for the TE012 mode, but since COMSOL predicted the right PE loaded resonant frequency for the TE212 mode as verified by my IR camera studies 1 & 2 of the copper frustum, I would assume that it got it right for the TE012 mode as well.  In fact I should have provided you my IR study-1 first, so find it attached.

Best, Paul M.

Paul, thank you.  This is the first time I see the attached IR study  ("Comparison of COMSOL Predictions of Copper Frustrum Heat Dissipation with Dec 30 IR Data") .  It was an excellent idea for NASA to conduct this  experimental study to verify the TM212 mode.  I congratulate you for that because it is of the utmost importance to understand the actual mode shapes being excited.

I attach the TM21 mode for a cylindrical cavity for comparison with TM21 in the truncated cone

Magnetic field: - - - - - - dashed lines
Electric field:   _______solid  lines

PS: Concerning whether COMSOL's discretization predicting TE012 was correct, my attitude (based on conducting experiments and numerical analysis) is always I'm from Missouri "show me"  :), so I would rather also have experimental verification for that experiment as well.  But considering your tight budget constraints, you deserve congratulations for what has been done.
« Last Edit: 02/08/2015 05:22 PM by Rodal »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
Rodal:

Look at the copper frustum part-2 COMSOL & IR thermal study that I submitted to this group yesterday for an answer to your "Have you experimentally verified that we are using the TM212 mode as predicted by our COMSOL simulations?  The answer BTW is yes for the TM212 mode, but no for the TE012 mode, but since COMSOL predicted the right PE loaded resonant frequency for the TE212 mode as verified by my IR camera studies 1 & 2 of the copper frustum, I would assume that it got it right for the TE012 mode as well.  In fact I should have provided you my IR study-1 first, so find it attached.

Best, Paul M.

Paul, thank you.  This is the first time I see the attached IR study  ("Comparison of COMSOL Predictions of Copper Frustrum Heat Dissipation with Dec 30 IR Data") .  It was an excellent idea for NASA to conduct this  experimental study to verify the TM212 mode.  I congratulate you for that because it is of the utmost importance to understand the actual mode shapes being excited.

I attach the TM21 mode for a cylindrical cavity for comparison with TM21 in the truncated cone

Magnetic field: - - - - - - dashed lines
Electric field:   _______solid  lines

PS: Concerning whether COMSOL's discretization predicting TE012 was correct, my attitude (based on conducting experiments and numerical analysis) is always I'm from Missouri "show me"  :), so I would rather also have experimental verification for that experiment as well.  But considering your tight budget constraints, you deserve congratulations for what has been done.

Notice from:

"Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum" ( http://www.libertariannews.org/wp-content/uploads/2014/07/AnomalousThrustProductionFromanRFTestDevice-BradyEtAl.pdf )



TE012  mode shape at 1.8804 GHz Table 2. Tapered Cavity Testing Summary (with the dielectric)
TM211 mode shape at 1.9326 GHz Table 2. Tapered Cavity Testing Summary (with the dielectric)
TM211 mode shape at 1.9367 GHz Table 2. Tapered Cavity Testing Summary (with the dielectric)
TM212 mode shape at 1.937188 GHz (Jan 2015 data (with the dielectric))


(TM212 mode shape at 1.937188 GHz from here http://forum.nasaspaceflight.com/index.php?topic=36313.msg1327177#msg1327177 )

(or an inconsequential difference 0.5%: TM212 mode shape at 1.946647 GHz http://forum.nasaspaceflight.com/index.php?action=dlattach;topic=36313.0;attach=634723 )

TM211 and TM212 have the same number of wave patterns in the circumferential and radial directions, but

TM211 has one half-wave pattern in the longitudinal direction
TM212 has two half-wave patterns in the longitudinal direction  (twice as many as TM211)

yet they are reported to take place at practically the same frequency ?

Now, mode shapes also depend on:

a)geometry (I understand that the same tapered cone was used for all the above experiments, so that was not a variable)
b) relative electric permittivity and relative magnetic permeability (I understand that the same dielectric  was used for all the above experiments, so that was not a variable)

So, if the geometry and the dielectric were the same in the above experiments:

QUESTION: 1.937188 GHz is only 0.025% different from  1.9367 GHZ. 
It is not possible to have TM211 at 1.9367 GHz and TM212 at just 0.025% higher frequency, because TM211 has 1 half-wave in the longitudinal direction, while TM212 has 2 half-waves in the longitudinal direction. Twice the number of half-waves in the longitudinal direction imply a significantly higher frequency keeping the number of wave-patterns constant in the circumferential and radial directions (the same m=2 and n=1). Therefore doubling the number of half-waves in the longitudinal direction with just a  0.025%  increase in frequency cannot be justified.

Therefore it appears that the labeling of mode TM211 in the Brady et.al. NASA report was an error, and it really should have been TM212 based on the Jan 2015 IR data
If so, was there was an error in the COMSOL analysis for TE012 as well as for TM211  in the report since they may have used the same COMSOL analysis and mesh?






Also, in "Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum" ( http://www.libertariannews.org/wp-content/uploads/2014/07/AnomalousThrustProductionFromanRFTestDevice-BradyEtAl.pdf )

page 18 (bottom) states (without the dielectric)

TE012 mode shape at  2.168 GHz . 

So that the same mode shape (TE012) is stated to take place with the dielectric at 1.8804 GHz and without the dielectric at (15% higher frequency) 2.168 GHz, but there was no measurable response without the dielectric.

However,

Quote from: page 18 of Brady et.al."Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum"

Numerous COMSOL® analysis runs also indicated a strong dependency between thrust magnitude and antenna type, location, orientation, and number of antenna feeds. Slight changes in antenna design and number of feeds changed the COMSOL® thrust prediction by a factor of three which forced our team to implement tighter configuration control protocols during testing to ensure close representation of as built hardware to the analyzed configuration.

Finally, our experience with the TE012 mode indicated that it is important to design the RF prototype such that any target mode of operation is as isolated as possible in the frequency domain to help ensure that the system can be effectively tuned manually. This also protects for the ability to implement and use a phase lock loop (PLL) automated frequency control circuit. Due to the slow process commensurate with manual tuning, our future test articles will make use of a PLL whenever practical in order to increase the amount of data that can be collected for a given test article configuration and operating condition during a given amount of test time
(Bold added for emphasis)

So it is not clear whether the "no appreciable response without the dielectric at 2.168 GHz" may have been due to "slight changes in antenna location and orientation" in the  2.168 GHz experiment without the dielectric compared to the experiment at 1.8804 GHz with the dielectric.

Also the above-mentioned report states:

Quote from: page 17 of Brady et.al."Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum"
the TE012 mode had numerous other RF modes in very close proximity, it was impractical to repeatedly operate the system in this mode, so the decision was made to evaluate the TM211 modes instead.
(Bold added for emphasis)

perhaps throwing further uncertainty as to what resonance took place at  1.8804 GHz
« Last Edit: 02/08/2015 07:03 PM by Rodal »

Offline Mulletron

  • Full Member
  • ****
  • Posts: 1116
  • Liked: 779
  • Likes Given: 1030
Quote
TM212 mode shape at 1.937188 GHz (Jan 2015 data (with the dielectric))

That comsol plot above says 1946.647. Where's the disconnect between numbers reported I wonder? Also unless comsol is taking into account all the little intricacies like heat expansion, bowing and buckling, simulation won't yield an exact result in reality. The thermal image pretty much nails whether or not the excited mode is actually being excited though.
« Last Edit: 02/08/2015 06:28 PM by Mulletron »
Challenge your preconceptions, or they will challenge you. - Velik

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
Quote
TM212 mode shape at 1.937188 GHz (Jan 2015 data (with the dielectric))

That comsol plot above says 1946.647. Where's the disconnect between numbers reported I wonder? Also unless comsol is taking into account all the little intricacies like heat expansion, bowing and buckling, simulation won't yield an exact result in reality.

1) I took the 1.937188 GHz figure from here:  http://forum.nasaspaceflight.com/index.php?topic=36313.msg1327177#msg1327177

Quote from: Star-Drive http://forum.nasaspaceflight.com/index.php?topic=36313.msg1327177#msg1327177
In that slide which is based on the copper frustum cavity running in its TM212 mode with 50W of 1,937.188 MHz RF power applied

But 1946.647 makes practically NO difference.  One cannot have TM212 at only 0.5% higher frequency than TM211

2) My reading is that the COMSOL Multiphysics simulation did not take into account buckling.  Buckling is an instability that requires a nonlinear stability analysis.

3) "simulation won't yield an exact result in reality".  No way this can justify  TM212 at only 0.5% higher frequency than TM211 for the same known truncated cone geometry and identical dielectric material .  A Finite Element calculation converges from below, it may produce too low a frequency for a given mode shape if the finite element mesh is too crude, but it is not going to give TM212 at only 0.5% higher frequency than TM211 for a cavity with the known dimensions of the NASA truncated cone, since TM212 has twice the number of longitudinal half-wave patterns as TM211
« Last Edit: 02/08/2015 07:06 PM by Rodal »

Offline Mulletron

  • Full Member
  • ****
  • Posts: 1116
  • Liked: 779
  • Likes Given: 1030
I get that, but with a cavity loaded with a dielectric, you can't add up wavelengths trying to satisfy E field requirements like we do with empty waveguides. With the dielectric inserted, there is a dielectric resonator inside the cavity resonator which means the solution is more complicated to figure out by hand.
« Last Edit: 02/08/2015 06:36 PM by Mulletron »
Challenge your preconceptions, or they will challenge you. - Velik

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
I get that, but with a cavity loaded with a dielectric, you can't add up wavelengths trying to satisfy E field requirements like we do with empty waveguides. With the dielectric inserted, there is a dielectric resonator inside the cavity resonator which means the solution is more complicated to figure out by hand.
I am not figuring out the solution by hand.  I am using Mathematica.  Also, I have based my above statements on my extensive experience developing Finite Element formulations for nonlinear analysis,  writing code for Finite Element programs and using commercial Finite Element codes to solve practical problems. 

Concerning the dielectric, COMSOL uses the dielectric constant: the electric permittivity.  My understanding is that the same dielectric was used for both the experiments for which COMSOL is stated as giving TM211 and TM212

If NASA used substantially-different truncated cone geometry or  substantially-different dielectric geometry or materials for the experiments reportedly giving TM211 and TM212, it is useful to find out.

If I have made an error somewhere, I would appreciate it being pointed out. 

If there is an error in the NASA report, it is useful to find out. It is not uncommon at all for articles to have errata, and in my experience, authors (certainly in the R&D field), are always appreciative when such errors are pointed out.
« Last Edit: 02/08/2015 07:08 PM by Rodal »

Offline Star-Drive

  • Member
  • Full Member
  • ****
  • Posts: 829
  • TX/USA
  • Liked: 876
  • Likes Given: 9
All:

Please note that our first COMSOL analyst who was volunteering his time for this activity while holding down his NASA day job, transcribed the dimensions for the copper frustum PE discs incorrectly when he did the analysis for the TE012 mode.  He used 6.0" OD by 1.0" thick whereas the actual dimensions for the PE discs was 6.13" OD by 1.062" thick.  The extra volume in the two PE discs lowered the actual observed resonant frequencies for all the resonant modes in the cavity down by about 8-to-10 MHz from COMSOL calculated.   When you have to beg for help, one can't be too critical of the results.  As to the TM212 mode analysis it was performed by another volunteer, so again I'm not going to complain that he didn't get these calculated frequencies spot on to what was measured with our Agilent Field-Fox Vector Network Analyzer (VNA) measured.

Rodal:

The next time we look at the TE012 mode I will perform the same IR camera survey I did for the TM212 mode.  The reason I didn't do the IR camera survey of the TE012 mode the first time around was that we didn't have that capability during March of 2014 when we ran that test series, since the IR camera didn't come along until the summer of 2014.

Best, Paul M.
Star-Drive

Offline Mulletron

  • Full Member
  • ****
  • Posts: 1116
  • Liked: 779
  • Likes Given: 1030
OMG that is the best thing ever!!!! :) :D  Thanks!





Wow, what a day. Thanks Paul March. Seriously thanks.
And there ya go, ISM band dimensions. Any competent tinkerer should be able to try it for themselves if they are so inclined. Be safe.

Let's not forget that if these Emdrives do work (looking pretty likely), Shawyer had the vision to make them happen. The folks at NASA had the courage to take a look.
« Last Edit: 02/08/2015 08:06 PM by Mulletron »
Challenge your preconceptions, or they will challenge you. - Velik

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
All:

Please note that our first COMSOL analyst who was volunteering his time for this activity while holding down his NASA day job, transcribed the dimensions for the copper frustum PE discs incorrectly when he did the analysis for the TE012 mode.  He used 6.0" OD by 1.0" thick whereas the actual dimensions for the PE discs was 6.13" OD by 1.062" thick.  The extra volume in the two PE discs lowered the actual observed resonant frequencies for all the resonant modes in the cavity down by about 8-to-10 MHz from COMSOL calculated.   When you have to beg for help, one can't be too critical of the results.  As to the TM212 mode analysis it was performed by another volunteer, so again I'm not going to complain that he didn't get these calculated frequencies spot on to what was measured with our Agilent Field-Fox Vector Network Analyzer (VNA) measured.

Rodal:

The next time we look at the TE012 mode I will perform the same IR camera survey I did for the TM212 mode.  The reason I didn't do the IR camera survey of the TE012 mode the first time around was that we didn't have that capability during March of 2014 when we ran that test series, since the IR camera didn't come along until the summer of 2014.

Best, Paul M.

Paul,

Thanks for the response.  It is a pleasure communicating with you. 

It is therefore my understanding (based on the IR data) that the NASA report should have read TM212 instead of TM211

"Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum" ( http://www.libertariannews.org/wp-content/uploads/2014/07/AnomalousThrustProductionFromanRFTestDevice-BradyEtAl.pdf )


TM211 mode shape at 1.9326 GHz Table 2. Tapered Cavity Testing Summary (with the dielectric)

should read:

TM212 mode shape at 1.9326 GHz Table 2. Tapered Cavity Testing Summary (with the dielectric)
« Last Edit: 02/08/2015 07:29 PM by Rodal »

Offline ThinkerX

  • Full Member
  • ***
  • Posts: 311
  • Alaska
  • Liked: 120
  • Likes Given: 59
If I am following the last couple pages correctly at all, it seems that the cavity dimensions relative to the frequency is of supreme importance to making this device work. 

I am reminded of the precision machining required to properly re-bore cylinders in car engines: get that wrong, even by a very tiny fraction, and the engine won't run properly.  I've known a number of people who didn't get that right.  It seems the same degree of precision is required here. 

Offline Star One

  • Senior Member
  • *****
  • Posts: 8249
  • UK
  • Liked: 1333
  • Likes Given: 168
If I am following the last couple pages correctly at all, it seems that the cavity dimensions relative to the frequency is of supreme importance to making this device work. 

I am reminded of the precision machining required to properly re-bore cylinders in car engines: get that wrong, even by a very tiny fraction, and the engine won't run properly.  I've known a number of people who didn't get that right.  It seems the same degree of precision is required here.
Sounds like it may make them tricky to reproduce at least initially.

Offline Mulletron

  • Full Member
  • ****
  • Posts: 1116
  • Liked: 779
  • Likes Given: 1030
If I am following the last couple pages correctly at all, it seems that the cavity dimensions relative to the frequency is of supreme importance to making this device work. 

I am reminded of the precision machining required to properly re-bore cylinders in car engines: get that wrong, even by a very tiny fraction, and the engine won't run properly.  I've known a number of people who didn't get that right.  It seems the same degree of precision is required here.
Sounds like it may make them tricky to reproduce at least initially.

Yes and here is how you tune a resonant cavity:
http://www.navymars.org/national/training/nmo_courses/nmo1/module11/14183_ch1.pdf


The cylinder part of the thruster above is for volume tuning.
If your frequency source is fixed, you have to tune your cavity. If you can change your frequency source, the cavity can remain fixed.
« Last Edit: 02/08/2015 08:52 PM by Mulletron »
Challenge your preconceptions, or they will challenge you. - Velik

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
If I am following the last couple pages correctly at all, it seems that the cavity dimensions relative to the frequency is of supreme importance to making this device work. 

I am reminded of the precision machining required to properly re-bore cylinders in car engines: get that wrong, even by a very tiny fraction, and the engine won't run properly.  I've known a number of people who didn't get that right.  It seems the same degree of precision is required here.

Sorry, but in order to have clarity, I am obliged to point out that I find it difficult to believe that a marginal change like this one:

" 6.0" OD by 1.0" thick whereas the actual dimensions for the PE discs was 6.13" OD by 1.062" thick"

can produce a change from TM211 into TM212 mode.

I think that what happened is that in a truncated cone the (electric and magnetic field) strengths attenuate   (the fields become weaker) considerably as one gets closer to the apex of the cone (going from the Big Diameter towards the Small Diameter).  This is shown in the literature of exact solutions for the truncated cone.  It is a feature of the truncated cone, and I think it maybe a reason why a truncated cone performs better as an EMDrive because this field attenuation towards the apex plays an important role.

Therefore, due to this field attenuation, it becomes difficult to distinguish TM211 from TM212 (and even from TM213) for a truncated cone.



I have to think hard, taking a look at the pictures below, whether the mode is TM211 or whether it is TM212.

This is very different from a cylindrical cavity, because in a cylindrical cavity there is no such attenuation of field strength toward a conical apex.

Therefore the error is a human error of distinguishing what happens near the small diameter (near the apex of the cone), and the error from the contour plot imaging (enough contours have to be available to tell the small gradient near the apex of the cone).

What really matters in a truncated cone, then is A) whether the mode is transverse electric (TE) or transverse magnetic (TM) annd B) what are the number of wave patterns in the circumferential and radial directions.

I would advise to look at the contour plots on the Big Diameter, to establish whether the mode is TM or TE, and what is the number of wave patterns in the circumferential and radial directions at the Big Diameter location, and not to put too much emphasis on the number of wavepatterns in the longitudinal direction for a truncated cone.

Please take a gander at the field strengths towards the small diameter: do you see the very small magnitude gradient of the half-wave mode (deep blue compared to lighter blue) in the longitudinal direction near the small diameter?

Compare that to the much stronger gradient near the Big Diameter (from red to blue).
« Last Edit: 02/08/2015 09:40 PM by Rodal »

Offline Notsosureofit

  • Full Member
  • ****
  • Posts: 656
  • Liked: 704
  • Likes Given: 1367
@ RODAL

Did you say you had an exact analytic solution for the truncated cavity ? 

["This is shown in the literature of exact solutions for the truncated cone. "]
« Last Edit: 02/08/2015 09:42 PM by Notsosureofit »

Tags: