Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1798068 times)

Offline apoc2021

  • Member
  • Posts: 18
  • Liked: 37
  • Likes Given: 27

Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251
....I was writing to say about the same thing in what you edited on the mode of TM212, good observation Jose ...
For everybody to notice what Shell just pointed out, I am attaching the image for Hz - x

This means: the magnetic field (H) component in the z direction (perpendicular to the longitudinal coordinate x) viewed on the circular cross-section of the cone (the cross section perpendicular to the longitudinal coordinate x).

Please observe:

1) This shows the magnetic quadrupole we expect for mode TM212 at 2.45 GHz.  (Quadrupole: mode "m"=2 means two full wave patterns along the circumference, since every wave pattern has a hill and valley, there are 2 hills and two valleys along the circumference: 4 poles)

2) The cartesian coordinates y and z are in the horizontal and vertical directions

3) The magnetic field Hz is zero along the y and z coordinates.  To see Hz one has to have a view at 45 degrees to y or z.  Since only x, y and z views are available the 45 degrees view is not available.  When one looks at the views of Hz along y  (Hz -y) or (Hz- z) [not shown] one should see zero.

4)   The reason we don't see zero when looking at the views of Hz along y  (Hz -y) or (Hz- z) [not shown] is two-fold:

a) views are shown to different numerical magnitudes.  Even something close to zero  (0.00000000000001) will be displayed at maximum contour color.  No numbers are displayed.

b) MOST IMPORTANT:  a big advantage to Meep's analysis [that my eigenvalue analysis or COMSOL's eigenvalue analysis do not have] is that Meep is a time-marching scheme that takes into account the participation of ALL MODES.  Thus, adjacent modes, like TM114 that physically occur in the real EM Drive, (albeit with lower participation) will be shown by Meep.  This is a big advantage to Meep.  Unfortunately we don't get the display of numbers, so we don't get to see how big (or negligible) is the participation of adjacent modes like TM114.
« Last Edit: 06/28/2015 03:03 PM by Rodal »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251


Thanks so much for these movies !

The effect of the antenna is beautifully displayed.

highest field intensity is clearly at the SMALL BASE of the truncated cone.

Action is NOT symmetric left to right (which many critics do NOT take into account).

Wave to the SMALL BASE is travelling progressively and well formed towards the SMALL BASE.

Recoil wave originating at small base to the BIG BASE occurs sharply and all of a sudden.  Not progressive and occurs more like a shock.

Asymmetries like this (left to right) are of paramount importance to explain momentum to one side (just like the difference between the ejecting jet of air and the vacuuming intake of air in acoustic propulsion of bottles).



« Last Edit: 06/28/2015 03:08 PM by Rodal »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587


Thanks so much for these movies !

The effect of the antenna is beautifully displayed.

highest field intensity is clearly at the SMALL BASE of the truncated cone.

Action is NOT symmetric left to right (which many critics do NOT take into account).

Wave to the SMALL BASE is travelling progressively and well formed towards the SMALL BASE.

Recoil wave originating at small base to the BIG BASE occurs sharply and all of a sudden.  Not progressive and occurs more like a shock.


Thanks apoc2021! I just finished getting one done and ready to post and boom, you're fast! Thanks again.

This shock wave is a key I was looking for and I suspected it and now makes sense to see it happening.

This is one full cycle, right?

Shell

Offline b0nafide

  • Member
  • Posts: 4
  • Liked: 1
  • Likes Given: 0
Hi, is it safe to assume that kml will not be performing any more tests or is he considering another method of weighing his setup?

Offline fasmax

  • Member
  • Posts: 4
  • Liked: 0
  • Likes Given: 0
257.5 g for frustum exoskeleton, simply tacked together. Will need stiffening, should come in well under 750 g. Weight budget for everything at 1.5 kg looks fine. Circle cutting 11.01 and 6.25 was a hassle. Probably less stiffening needed with thicker copper clad pcb, oh well. Mr. Whiskers supervising...

Edit...corrected gram weight...sheesh...sniffed too much solder smoke.
This excites me to no end!!! Good for you and it looks like you do great work, clean edges and I like the way you supported the axis with the 4 standoffs. Smart man.

cute kitty!

Shell
The cat is curious wondering if rfmwguy will measure any thrust from the experiment.
I canít wait to see the results from all the various experiments.
Thanks  to everyone for sharing your experiments.



Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251


Thanks so much for these movies !

The effect of the antenna is beautifully displayed.

highest field intensity is clearly at the SMALL BASE of the truncated cone.

Action is NOT symmetric left to right (which many critics do NOT take into account).

Wave to the SMALL BASE is travelling progressively and well formed towards the SMALL BASE.

Recoil wave originating at small base to the BIG BASE occurs sharply and all of a sudden.  Not progressive and occurs more like a shock.


Thanks apoc2021! I just finished getting one done and ready to post and boom, you're fast! Thanks again.

This shock wave is a key I was looking for and I suspected it and now makes sense to see it happening.

This is one full cycle, right?

Shell
So people that proclaim left to right symmetry fail to take into account time.

Greg Egan's analysis assumes a sinusoidal change with time.  Clearly this is not the case. There is TIME-ASYMMETRY left to right.  The origin of the asymmetry is the RF feed, that Greg Egan does not take into account.  There is an interaction between standing waves and the travelling waves from the RF feed.

As Notsosureofit said:  steady state standing waves by themselves never occurs as long as the RF feed is on.


We still have to show that this asymmetry can result in space propulsion.

But, I think this is great progress in understanding the field inside the EM Drive


Congratulations aero: this is nowhere in the literature except in this thread !!!

Not in:

1) Shawyer's analysis

2) Yang's analysis

3) COMSOL's Eagleworks analysis

4) Greg Egan's analysis

« Last Edit: 06/28/2015 03:21 PM by Rodal »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587


Thanks so much for these movies !

The effect of the antenna is beautifully displayed.

highest field intensity is clearly at the SMALL BASE of the truncated cone.

Action is NOT symmetric left to right (which many critics do NOT take into account).

Wave to the SMALL BASE is travelling progressively and well formed towards the SMALL BASE.

Recoil wave originating at small base to the BIG BASE occurs sharply and all of a sudden.  Not progressive and occurs more like a shock.

Asymmetries like this (left to right) are of paramount importance to explain momentum to one side (just like the difference between the ejecting jet of air and the vacuuming intake of air in acoustic propulsion of bottles).


The difference in the asymmetries becomes more apparent in grayscale to me with fewer flashing colors. This is a good find and some nice data, good job!
Shell

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
A delta function excitation at the antenna might be instructive, since then the time evolution would be only due to cavity modes and not to a mixture of antenna and cavity. However, I fear that a delta function will not propagate successfully down a waveguide. Perhaps the best one could do in this regard, then, would be to excite the antenna with a single cycle and record the cavity fields one cycle later.

Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251
This also shows that the longitudinal field (electric Ex for a  transverse magnetic TM mode as in this case, or magnetic  Hx for a transverse electric TE mode) is what mostly we should be looking at.

In cavities used in particle accelerators (CERN) it is also the the longitudinal field that matters the most.

Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251
A delta function excitation at the antenna might be instructive, since then the time evolution would be only due to cavity modes and not to a mixture of antenna and cavity. However, I fear that a delta function will not propagate successfully down a waveguide. Perhaps the best one could do in this regard, then, would be to excite the antenna with a single cycle and record the cavity fields one cycle later.

Great observation, deltaMass.

QUESTIONS:

1) Since the RF feed is so important, how representative is this of the RF feeds used in experiments?

2) Yang did not use an antenna.  She achieved the highest reported force/powerInput by using a waveguide RF feed instead  [This is also what is done in particle accelerators: they use waveguides to couple: they are better than antennas ]

3) how does this antenna in the Meep model compare with the one used by NASA Eagleworks ?

As Notsosureofit said: there are so many kinds of "antennas"...
« Last Edit: 06/28/2015 03:51 PM by Rodal »

Offline deuteragenie

  • Member
  • Posts: 71
  • Germany
  • Liked: 22
  • Likes Given: 0


Thanks so much for these movies !

The effect of the antenna is beautifully displayed.

highest field intensity is clearly at the SMALL BASE of the truncated cone.

Action is NOT symmetric left to right (which many critics do NOT take into account).

Wave to the SMALL BASE is travelling progressively and well formed towards the SMALL BASE.

Recoil wave originating at small base to the BIG BASE occurs sharply and all of a sudden.  Not progressive and occurs more like a shock.

Asymmetries like this (left to right) are of paramount importance to explain momentum to one side (just like the difference between the ejecting jet of air and the vacuuming intake of air in acoustic propulsion of bottles).



Althouth it looks like the field strength is stronger at the small end, we have no data.  What about averaging the field values for each frame we see in this video and produce the resulting "averaged" field strength ?

Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251
...

Althouth it looks like the field strength is stronger at the small end, we have no data.  What about averaging the field values for each frame we see in this video and produce the resulting "averaged" field strength ?
We need NUMBERS to quantify this, there is no other way around it.  Without numbers one cannot compare.

At least now we have everything scaled to the same Max Min numbers, unfortunately we don't know the numbers.

Can you tell aero how to output NUMBERS in Meep?
« Last Edit: 06/28/2015 03:52 PM by Rodal »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
This is a 14 frame animation of a full wave function and looking at frames 6,7 and 8 right in the center of the full cycle. I also can see what @notsosureofit was saying about not standing waves.
Frame 6
You see the energy built up in the small end of the cavity, squeezed into the small end of the cavity and held there by the antenna actions.

Frame 7
The very fast release of all the energy built up in the small end "almost instantly" appear at the large end as the waveform switches phases through 0 degrees.

Frame 8
The process begins again with the reverse phase

Note: I would find it interesting to see finer time slices of the phase shift through the 0 degree phase shift to see if a time component could be derived of the engery pulse from the small end to the large.

Shell

Offline tchernik

  • Full Member
  • **
  • Posts: 247
  • Liked: 297
  • Likes Given: 577
Outstanding results, you people rock!

Question: do these new results contradict the calculations made by people like Greg Egan, telling there is no asymmetry of forces in the cavity?

Because visually at least, it seems there is some asymmetry of field strength and maybe forces involved through time. Something that doesn't seem to be included in the existing previous analysis.

I concur with Rodal numeric results would be much better for weighting the asymmetries.

Offline dustinthewind

  • Full Member
  • ****
  • Posts: 583
  • U.S. of A.
  • Liked: 226
  • Likes Given: 256
I was thinking of a few modifications.  One idea was to fill the cavity with barium titanate which I think slows down the propagation speed and allows us to move into the radio frequency range.  ...

...

- Barium titanate at MHz has an extraordinarily high relative pemittivity. This translates (via sqrt) into a high refractive index.

...

Ok thanks deltaMass I guess I was thinking of this patent here for a propellantless EM propulsion device "https://scholar.google.com/scholar?cluster=980245622792438027&hl=en&as_sdt=0,48" on column 8 and 4th paragraph down.  I am sure I have seen other areas suggest an unusually large index of refraction for Barium Titanate also.  It is hard to find anything that is clear about it however. 

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
McClean reinvents a photon rocket; film at 11.
Sorry about that.
The USPTO isn't as well administered as it once was.

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1223
  • Do it!
  • Vista, CA
  • Liked: 1290
  • Likes Given: 1740
...

Althouth it looks like the field strength is stronger at the small end, we have no data.  What about averaging the field values for each frame we see in this video and produce the resulting "averaged" field strength ?
We need NUMBERS to quantify this, there is no other way around it.  Without numbers one cannot compare.

At least now we have everything scaled to the same Max Min numbers, unfortunately we don't know the numbers.

Can you tell aero how to output NUMBERS in Meep?

Outputting numbers is usually just write to a CSV file. The problem is to include some corresponding time stamp so we can associate which numbers go to which frame and which color. I.e., the CSV file will need an interpreter.
Todd


Online Rodal

  • Senior Member
  • *****
  • Posts: 5834
  • USA
  • Liked: 5902
  • Likes Given: 5251
...

Althouth it looks like the field strength is stronger at the small end, we have no data.  What about averaging the field values for each frame we see in this video and produce the resulting "averaged" field strength ?
We need NUMBERS to quantify this, there is no other way around it.  Without numbers one cannot compare.

At least now we have everything scaled to the same Max Min numbers, unfortunately we don't know the numbers.

Can you tell aero how to output NUMBERS in Meep?

Outputting numbers is usually just write to a CSV file. The problem is to include some corresponding time stamp so we can associate which numbers go to which frame and which color. I.e., the CSV file will need an interpreter.
Todd
Meep uses Finite Difference in time, given a uniform deltaT increment.  One can only output results (numbers, the images are built from numbers) at every time step: deltaT, 2 deltaT, 3 deltaT,.....NdeltaT.  The frame times in Meep are clearly associated with the finite difference deltaT, and so are any numbers.  The deltaT is dis-associated from any time stamp attached to the computing time. 

This intrinsic time (N deltaT where deltaT is the Finite Difference time step identifier) is available, providing the best identification for any numbers, as it is associated with the finite difference scheme and completely divorced from the vicissitudes of the machine time.

Please don't use machine time stamps to identify the numbers, instead use the available finite difference time step as an identifier.
« Last Edit: 06/28/2015 05:27 PM by Rodal »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
I'm in hog tech heaven and I never knew I'd be talking about splitting nanoseconds into smaller slices. Very special to me. I still have a ~11 inch piece of wire that I got from Grace Hopper (nanosecond lady)(if you don't know her please look her up) from the early 70's and here we are slicing it into a 1000 more pieces to detail fire.

I was thinking whoever unlocks this world class problem I'll be glad to send that piece of wire to them, they will have earned it. But this isn't a 1 man/woman team so I guess I'll have to divide it equally.

Been said before but you all rock!

Shell

Tags: