Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1842228 times)

Offline leomillert

  • Member
  • Posts: 34
  • Liked: 21
  • Likes Given: 12
This should also help a bit.
http://ab-initio.mit.edu/wiki/index.php/Meep_Tutorial#Output_tips_and_tricks

I also remember: when in doubt, ask the MEEP mailing list. http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss

And how does that help? Tell us why it helps with the problem we have?

It tells how to output directly to images instead of h5 files.
(run-until 200 (at-every 0.6 (output-png Ez "-Zc bluered")))

From that command, it should be possible to discover how to output to csv files, which was, if I understood correctly, what you wanted to do.
Maybe output-txt instead of output-png? I don't know, but it's a single command, so it shouldn't be too hard for us to discover it.

Offline ThinkerX

  • Full Member
  • ***
  • Posts: 311
  • Alaska
  • Liked: 120
  • Likes Given: 59
Quote
Right now, turning off the source after 0.013 microseconds would be much more informative than running Meep up to 1 microseconds and beyond.

This is something that Todd asked many days ago: just turning off the source and seeing what happens.

Self sustaining action of some sort?  (if only for a fragment of a microsecond)

Offline aero

  • Senior Member
  • *****
  • Posts: 2777
  • 92129
  • Liked: 721
  • Likes Given: 246
This should also help a bit.
http://ab-initio.mit.edu/wiki/index.php/Meep_Tutorial#Output_tips_and_tricks

I also remember: when in doubt, ask the MEEP mailing list. http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss

And how does that help? Tell us why it helps with the problem we have?

It tells how to output directly to images instead of h5 files.
(run-until 200 (at-every 0.6 (output-png Ez "-Zc bluered")))

From that command, it should be possible to discover how to output to csv files, which was, if I understood correctly, what you wanted to do.
Maybe output-txt instead of output-png? I don't know, but it's a single command, so it shouldn't be too hard for us to discover it.

Go for it, and good luck.
Retired, working interesting problems

Offline aero

  • Senior Member
  • *****
  • Posts: 2777
  • 92129
  • Liked: 721
  • Likes Given: 246
Quote
Right now, turning off the source after 0.013 microseconds would be much more informative than running Meep up to 1 microseconds and beyond.

This is something that Todd asked many days ago: just turning off the source and seeing what happens.

Self sustaining action of some sort?  (if only for a fragment of a microsecond)

This is another question that is answered with, "write your own function," when ask of the meep-discuss mailing list.

I sometimes get the idea that searching the mailing list for the key words "write your own function." would be a more fruitful way to find the non-answers to questions I have. It's to bad Stack- Overflow doesn't have a good section on meep, or I haven't found it.
Retired, working interesting problems

Offline Ricvil

  • Member
  • Posts: 67
  • Liked: 66
  • Likes Given: 23
http://arxiv.org/abs/1505.02520

Beyound a source of microwave, could the magnetron coupled to cavity be acting as a amplifier too?
It would be nice to elaborate on that.

Yes. The cavity is a periodic structure under mirror images symmetry of the planar ends.
If the magnetron is acting as a amplifier there is a chance  of a effective active load on the cavity, and perhaps some nonreciprocity too.
The article above talks about alternating gain/loss media produces nonreciprocal scattering associated to a PT symmetry breaking.
« Last Edit: 07/21/2015 02:03 AM by Ricvil »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5894
  • USA
  • Liked: 6040
  • Likes Given: 5315
Yes, "write your own functions" is certainly the Meep way to do things.  Although that carried to the extreme would be why bother with Meep, just write your own Meep :)

At the moment what I see is that (really the only ?) new important thing I see with Meep vs the steady-state standing wave solution is the influence of the antenna RF feed

The stress at the opposite end of the antenna is what one sees in the standing wave solution.  What gets modified is the stress at the base near the antenna and the whole field near the antenna.  (Also the Poynting vector field near the antenna gets modified)

I expect that by turning the RF feed OFF we will see the standing wave solution (just at a lower intensity because it is so early in the process after 0.013 microseconds).

Since the only new thing I see coming from Meep transient analysis is the antenna, it would be most important to model:

1) Different antenna locations (being done)

2) Different antennas (monopoles, loop, etc.)

3) A Magnetron  [all kinds of things here -- just see the post above by Ricvil, besides the obvious one of the Magnetron bandwidth, the Frequency, Amplitude and Phase modulation provided by the magnetron, etc.]

4) A waveguide feed (as used by Yang -- who used a waveguide feed instead, and reported the highest force)

All of this is more important IMHO, than running Meep to steady-state.  I agree with Todd.

Todd, where are you ?  we miss you :)
« Last Edit: 07/21/2015 02:09 AM by Rodal »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5894
  • USA
  • Liked: 6040
  • Likes Given: 5315
...some parameter studies WRT antenna placement and / or geometries could be useful to flesh the EM drive out.  We're all missing something ..

RIGHT ON !

Agreed 100%

Offline mwvp

  • Full Member
  • **
  • Posts: 254
  • Coincidence? I think Not!
  • Liked: 169
  • Likes Given: 30
Beyound a source of microwave, could the magnetron coupled to cavity be acting as a amplifier too?
It would be nice to elaborate on that.
...If the magnetron is acting as a amplifier there is a chance  of a effective active load on the cavity, and perhaps some nonreciprocity too.

Long way back I read a tech journal article about magnetrons, modeled as negative-resistance oscillators/amplifiers. You push on it (EMF), it pushes back harder, if the signal is in its bandwidth. If put in a loop with a circulator, you can indeed use it, as well as other negative-resistance devices (tunnel diodes, gunn diodes,  et) or circuits (transistors) as 2-port amplifiers.

As far as modeling goes, I suspect you'd use 8 or so tuned circuits, in parallel with negative resistances, more or less coupled together and coupled to the I/O E-field probe. What frequency are the resonator cavities tuned to? How much coupling? Good questions. Perhaps they try to drill them all the same, but inevitable tolerances mean differences. And different cavities have different gains. You can see it in the output spectrum of the under-heated tube in the IEEE article posted that I referenced.

The load the magnetron drives will then most definitely affect the frequency it generates, withing BW and reason of course. "Oscillator pulling" is a problem with any oscillator, hence heroic efforts are often required to make stable, non crystal/saw oscillators with shieldeding and buffering.
« Last Edit: 07/21/2015 02:41 AM by mwvp »

Offline mwvp

  • Full Member
  • **
  • Posts: 254
  • Coincidence? I think Not!
  • Liked: 169
  • Likes Given: 30
...some parameter studies WRT antenna placement and / or geometries could be useful to flesh the EM drive out.  We're all missing something ..

RIGHT ON !

Agreed 100%

If I might be pardoned for butting in, the simulations so far, while very important to familiarize with the software, don't inspire my confidence WRT utility.

Why not follow conventional industry wisdom using conventional E-field probe, H-field loop, or aperture methods of excitation, similar to Shawyer, White or Yang? For instance, see:
http://www.scribd.com/doc/111892661/Waveguide-Theory  pg. 10 on coupling methods.
That short document has pretty pictures, the pretty equations are in the MIT Radiation Lab series.

A big 'ol dipole thrown in anywhere probably will trash the Q of the cavity.

Rfmwguy has the right idea, I think, in putting the probe 1/4 wave from the wall, but what is the impedance at that point?

I've been perusing documents lately, but I got mpd installed (it can be called by meep) because I think the methodology to follow is to calculate the modes using a fast harmonic solver (mpd), plot the cavity Z and Vg (E/B), calculate the Q and other nearby modes, then -

Use the eigenvalue mode of interest to excite the cavity in the time-domain (meep) and see what it does when the cavity is accelerated. I believe meep can do this. I read a post on the meep mailing list where the Cerenkov effect was being modeled by moving the charge-source between runs. Why not the cavity? Something I hope to look into.

Because Shawyer somewhat obscurely, in a couple documents states that motion is necessary to produce thrust. Since we're considering his cone, why not his important precondition for its operation?

The cone is an impedance gradient; as Frobnicat points out, equivalent to gravity or an accelerated inertial frame.

If the cone is accelerated, it "straightens out" the cavity, affecting field/energy/momentum distribution, and delta-momentum means a change in force inside the frustrum. The antenna is bolted to the inertial frame of the cone, but with a Q of 10k, there is 10^4 times more stale energy than fresh energy to apply force to an accelerating frustrum.

Its great to have newcomers with hardware. Let's not waste this great opportunity! I wish I had more experience with EM and waveguides. We could really use some expert advice. Microwave circuits have been called a "black art" for a reason.

Offline mwvp

  • Full Member
  • **
  • Posts: 254
  • Coincidence? I think Not!
  • Liked: 169
  • Likes Given: 30
While I have the temerity to vomit-out my ignorant opinion,

Its been my sad experience, too often, computers have been a hindrance rather than an asset. Those working in analog electronics know that whatever they calculate, component tolerances and environment, especially around high-Q and high-gain circuits, will need to be physically tuned, so rfmwguy will probably have a working system while others pound on their keyboards. It's really neat the way he can just slide the probe around to find tune. But that won't help elucidate theory. With a loop coupling, the loop can be simply rotated to vary the coupling, to balance cavity Q vs. power supplied.

However, even a battery powered, low-power oscillator driving a frustrum, with probes at various points, can measure amplitude and phase fluctuations  of a moving frustrum to validate/refute Shawyer's claims about the behavior (although not reaction force or thrust) of his system. And it won't cause near (like say, 10^5) as much thermal detuning, neither oscillator or cavity. Saw and dielectric resonator are available cheap for 900MHz & 2.5 GHz.

That would be a relatively simple and safe experiment to acquaint oneself with the technology and art before attempting to use a deadly magnetron or expensive and fragile semiconductor amplifier, which could easily cause big trouble if not shielded adequately.
« Last Edit: 07/21/2015 03:34 AM by mwvp »

Offline DeWeave

  • Member
  • Posts: 2
  • Liked: 0
  • Likes Given: 0
So I won't pretend that I'm an expert on this subject, I've just been following this for a while as I find it interesting to watch this all develop in realtime.

One thing I haven't noticed being discussed is any possibility that the emdrive could be generating a Mach effect.  Reading Woodwards description of transient mass fluctuation, makes it sound very similar to what is being discussed here.  Is there a possibility that all of standing waves and power being pumped into the copper fulstrum is making it mimic one of Woodwards capacitors? 

Would provide a (relatively) clean way to try to tie the effect back into the realm of accepted physics.  But again, I may just be oversimplifying it.

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
The Woodward Effect requires something to be accelerated while its energy density changes in time. Make of that what you will.

And whatever makes you think that the Woodward Effect is "accepted physics"?
« Last Edit: 07/21/2015 03:58 AM by deltaMass »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2335
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2982
  • Likes Given: 2601
While I have the temerity to vomit-out my ignorant opinion,

Its been my sad experience, too often, computers have been a hindrance rather than an asset. Those working in analog electronics know that whatever they calculate, component tolerances and environment, especially around high-Q and high-gain circuits, will need to be physically tuned, so rfmwguy will probably have a working system while others pound on their keyboards. It's really neat the way he can just slide the probe around to find tune. But that won't help elucidate theory. With a loop coupling, the loop can be simply rotated to vary the coupling, to balance cavity Q vs. power supplied.

However, even a battery powered, low-power oscillator driving a frustrum, with probes at various points, can measure amplitude and phase fluctuations  of a moving frustrum to validate/refute Shawyer's claims about the behavior (although not reaction force or thrust) of his system. And it won't cause near (like say, 10^5) as much thermal detuning, neither oscillator or cavity. Saw and dielectric resonator are available cheap for 900MHz & 2.5 GHz.

That would be a relatively simple and safe experiment to acquaint oneself with the technology and art before attempting to use a deadly magnetron or expensive and fragile semiconductor amplifier, which could easily cause big trouble if not shielded adequately.
Got his on order.
http://www.ebay.com/itm/New-USB-138M-4-4G-Signal-Source-Signal-Generator-Simple-Spectrum-Analyzer-/251600378012?pt=LH_DefaultDomain_0&hash=item3a948d189c

Should allow a nice static test and coupled with the USB spectrum analyser.

http://www.ebay.com/itm/281744660852?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Give should give me a fine real world low power test. Data that we can use to fine tune meep and at least compare.

Offline TheTraveller

Should allow a nice static test and coupled with the USB spectrum analyser.

http://www.ebay.com/itm/281744660852?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Give should give me a fine real world low power test. Data that we can use to fine tune meep and at least compare.

Good to see they send a 30dB 2 Watt attenuator to protect the Rf input stage from overload.
« Last Edit: 07/21/2015 04:50 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Offline TheTraveller

While I have the temerity to vomit-out my ignorant opinion,

Its been my sad experience, too often, computers have been a hindrance rather than an asset. Those working in analog electronics know that whatever they calculate, component tolerances and environment, especially around high-Q and high-gain circuits, will need to be physically tuned, so rfmwguy will probably have a working system while others pound on their keyboards. It's really neat the way he can just slide the probe around to find tune. But that won't help elucidate theory. With a loop coupling, the loop can be simply rotated to vary the coupling, to balance cavity Q vs. power supplied.

However, even a battery powered, low-power oscillator driving a frustrum, with probes at various points, can measure amplitude and phase fluctuations  of a moving frustrum to validate/refute Shawyer's claims about the behavior (although not reaction force or thrust) of his system. And it won't cause near (like say, 10^5) as much thermal detuning, neither oscillator or cavity. Saw and dielectric resonator are available cheap for 900MHz & 2.5 GHz.

That would be a relatively simple and safe experiment to acquaint oneself with the technology and art before attempting to use a deadly magnetron or expensive and fragile semiconductor amplifier, which could easily cause big trouble if not shielded adequately.

My 100W Rf amp can be throttled down to 79mWs for testing / lowest VSWR searching. Has a 5 stage inbuilt attenuator as well as auto shutdown at VSWR > 3.0 and on thermal over temperature. So maybe not so fragile?

Experience I have gained from others says it is critical to be able to tune the Rf amp frequency to the highest Return Loss dB (lowest VSWR) from the cavity as otherwise little or no Force generation will happen.

With respect, trying to play the EMDrive Poker Machine and hope a Rf gen frequency matches what the cavity needs is a good recipe to waste a lot of time and money. Or you could get lucky but then as the cavity warms up, you lose the sweet spot and have no way to readjust the frequency to regain the sweet spot.

The data from Prof Yang, that I published earlier, shows what happens if your magnetron output bandwidth fails to match that of the cavity.
http://forum.nasaspaceflight.com/index.php?topic=37642.msg1406672#msg1406672

Can't see much point in doing a EMDrive build if you can't get the Rf energy inside the cavity at it's resonant frequency at the desired excitation mode.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1266
  • Do it!
  • Vista, CA
  • Liked: 1333
  • Likes Given: 1798
Yes, "write your own functions" is certainly the Meep way to do things.  Although that carried to the extreme would be why bother with Meep, just write your own Meep :)

At the moment what I see is that (really the only ?) new important thing I see with Meep vs the steady-state standing wave solution is the influence of the antenna RF feed

The stress at the opposite end of the antenna is what one sees in the standing wave solution.  What gets modified is the stress at the base near the antenna and the whole field near the antenna.  (Also the Poynting vector field near the antenna gets modified)

I expect that by turning the RF feed OFF we will see the standing wave solution (just at a lower intensity because it is so early in the process after 0.013 microseconds).

Since the only new thing I see coming from Meep transient analysis is the antenna, it would be most important to model:

1) Different antenna locations (being done)

2) Different antennas (monopoles, loop, etc.)

3) A Magnetron  [all kinds of things here -- just see the post above by Ricvil, besides the obvious one of the Magnetron bandwidth, the Frequency, Amplitude and Phase modulation provided by the magnetron, etc.]

4) A waveguide feed (as used by Yang -- who used a waveguide feed instead, and reported the highest force)

All of this is more important IMHO, than running Meep to steady-state.  I agree with Todd.

Todd, where are you ?  we miss you :)

Thanks! I've had my nose buried in dispersion equations and metric line elements. Too much I don't remember from way back when, lots of re-learning to do to try and get this right. So far, I've found 3 different thrust-to-power ratios. One of which is consistent with a photon rocket, one of which is consistent with over-unity, and one I haven't quite figured out yet because my significant other thinks I'm spending way too much time on this and I'm going to end up living in a FEMA trailer park. :-/

I for one would like to know how the wave progresses after the source is shut off. How long does it persist and how far back up the frustum does it travel. So far, all of my equations point to the idea that the input frequency matches the small end Xmm resonant mode, with near zero group velocity. From there, the mass-energy falls to the big end and lifts the frustum forward. A trumpet shape is better than a frustum because the long, slow group velocity will allow the wave-train to pile up before it falls over the cliff.

I'm on break...
Todd


Offline TheTraveller

...some parameter studies WRT antenna placement and / or geometries could be useful to flesh the EM drive out.  We're all missing something ..

RIGHT ON !

Agreed 100%

If I might be pardoned for butting in, the simulations so far, while very important to familiarize with the software, don't inspire my confidence WRT utility.

Why not follow conventional industry wisdom using conventional E-field probe, H-field loop, or aperture methods of excitation, similar to Shawyer, White or Yang? For instance, see:
http://www.scribd.com/doc/111892661/Waveguide-Theory  pg. 10 on coupling methods.
That short document has pretty pictures, the pretty equations are in the MIT Radiation Lab series.

A big 'ol dipole thrown in anywhere probably will trash the Q of the cavity.

Rfmwguy has the right idea, I think, in putting the probe 1/4 wave from the wall, but what is the impedance at that point?

I've been perusing documents lately, but I got mpd installed (it can be called by meep) because I think the methodology to follow is to calculate the modes using a fast harmonic solver (mpd), plot the cavity Z and Vg (E/B), calculate the Q and other nearby modes, then -

Use the eigenvalue mode of interest to excite the cavity in the time-domain (meep) and see what it does when the cavity is accelerated. I believe meep can do this. I read a post on the meep mailing list where the Cerenkov effect was being modeled by moving the charge-source between runs. Why not the cavity? Something I hope to look into.

Because Shawyer somewhat obscurely, in a couple documents states that motion is necessary to produce thrust. Since we're considering his cone, why not his important precondition for its operation?

The cone is an impedance gradient; as Frobnicat points out, equivalent to gravity or an accelerated inertial frame.

If the cone is accelerated, it "straightens out" the cavity, affecting field/energy/momentum distribution, and delta-momentum means a change in force inside the frustrum. The antenna is bolted to the inertial frame of the cone, but with a Q of 10k, there is 10^4 times more stale energy than fresh energy to apply force to an accelerating frustrum.

Its great to have newcomers with hardware. Let's not waste this great opportunity! I wish I had more experience with EM and waveguides. We could really use some expert advice. Microwave circuits have been called a "black art" for a reason.

The antenna inside an EMDrive is there to synergistically add energy to, couple to, an existing resonant standing EM wave which has a spherical wavefront. It is not there to radiate EM waves into free space. Different job to do, different design.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Offline TheTraveller

"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Offline CW

  • Full Member
  • *
  • Posts: 187
  • Germany
  • Liked: 141
  • Likes Given: 51
If it's about accelerating simulations, couldn't we theoretically try and make something like the SETI client and distribute prepared packages to those willing to donate CPU time to simulate a ton of EM drive configurations - even over longer periods of time and operation?
Reality is weirder than fiction

Offline mwvp

  • Full Member
  • **
  • Posts: 254
  • Coincidence? I think Not!
  • Liked: 169
  • Likes Given: 30
...Reading Woodwards description of transient mass fluctuation, makes it sound very similar to what is being discussed here.  Is there a possibility that all of standing waves and power being pumped into the copper fulstrum is making it mimic one of Woodwards capacitors? 

Would provide a (relatively) clean way to try to tie the effect back into the realm of accepted physics.  But again, I may just be oversimplifying it.

My take on it, if it's for real, is that it behaves like an animation I saw on (the only?) youtube lecture by Woodward; a rocket with a spring and brick on back bouncing its way forward, sort of like a squid.

Unlike the Woodward effect which relies on the charged capacitor having more inertia than an uncharged one, the "Shawyer Effect" I call a "Sagnac Ratchet"; the force in the forward, and impedance in the reverse directions are the result the frustrum's asymmetrical dispersion, group velocity, and sum/difference frequency filtering characteristics.

Maybe Shawyer would have gotten a different response if he made it clear it was acting as a ratchet, so CoM wouldn't have been the issue.
« Last Edit: 07/21/2015 06:48 AM by mwvp »

Tags: