Author Topic: EM Drive Developments - related to space flight applications - Thread 2  (Read 3314755 times)

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
And while on the subject of simulation: Guido Fetta of Cannae tells me that his COMSOL(?unsure) sim predicts a nonzero net Lorentz force for his device. Now, we all learned that no closed system of currents can produce such a net force. There's a paradox. He insists that there are no significant cumulative rounding errors.

Anyone have insight into this?

There cannot be a paradox.

Let's remember that the COMSOL FEA solution is the steady state solution showing the spatial distribution of the field.  Remember that the steady state solution of Maxwell's differential equations can be accomplished by separation of variables.

 The harmonic (time varying) part of the field is assumed. So, for example, the Magnetic Field shown on the COMSOL output is the spatial distribution of the magnetic field.  Now, what is shown as a maximum and what is shown as a minimum is arbitrary, since depending at what time one arbitrarily chooses to display the magnetic field, as the magnetic field varies with time like a harmonic function.

Similarly, the Poynting vector is a harmonic function of time, and this is, as you point out, well known in the literature, with a frequency which is twice the frequency of the magnetic and the electric field.

Although the spatial distribution of the Poynting vector is non-zero at arbitrary points in time, over a whole cycle the Poynting vector (and the Lorentz force) for a cavity sums up to exactly zero, just like the mean of the magnetic and electric fields is also zero.  COMSOL is an excellent package.

The Poynting vector solution of Maxwell's equations points towards the Big Base half of the time, and points towards the Small Base half of the time.

COMSOL will not tell that to the analyst obtaining a steady solution where the harmonic function of time is implicit. 
It is recommended that COMSOL and any other FEA  packages (ANSYS MultiPhysics, etc., ABAQUS, ADINA, NASTRAN) should be run by experienced FEA analysts, to prevent errors.  (Ditto for FD, control volume , and any other numerical packages).
« Last Edit: 05/09/2015 11:25 pm by Rodal »

...
BOTTOM LINE: I would suggest for experimenters to try 4 different kinds of ends:

1) A solid reflecting end (copper or aluminum)

2) A conductive wire mesh

3) A transparent glass (transparent to microwaves)

4) An open end 


And compare the results.  Such tests would be very valuable both for scientific and engineering purposes to understand what is being measured.

For completeness, should we also consider:

1a) shaped end plates with narrow RF modulation and

1b) plane end plates with wide modulation?

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
And while on the subject of simulation: Guido Fetta of Cannae tells me that his COMSOL(?unsure) sim predicts a nonzero net Lorentz force for his device. Now, we all learned that no closed system of currents can produce such a net force. There's a paradox. He insists that there are no significant cumulative rounding errors.

Anyone have insight into this?
Yes.  Let's remember that the COMSOL FEA solution is the steady state solution showing the spatial distribution of the field.  The harmonic (time varying) part of the field is assumed. So, for example, the Magnetic Field shown on the COMSOL output is the spatial distribution of the magnetic field.  Now, what is shown as a maximum and what is shown as a minimum is arbitrary, since depending at what time one arbitrarily chooses to display the magnetic field, as the magnetic field varies with time like a harmonic function.

Similarly, the Poynting vector is a harmonic function of time, as you point out, well known in the literature, with a frequency which is twice the frequency of the magnetic and the electric field.

Although the spatial distribution of the Poynting vector is non-zero at arbitrary points in time, over a whole cycle it sums up to exactly zero, just like the mean of the magnetic and electric fields is also zero.

The Poynting vector solution of Maxwell's equations points towards the Big Base half of the time, and points towards the Small Base half of the time.

COMSOL will not tell that to the analyst.  The  FEA analyst is supposed to know that.

It is recommended that COMSOL and any other packages (ANSYS MultiPhysics, etc., ABAQUS, ADINA, NASTRAN) should be run by experience FEA analysts.
Do you know for a fact that Fetta made this elementary blunder, or are you just hypothesising?

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
...
Do you know for a fact that Fetta made this elementary blunder, or are you just hypothesising?
I don't know the gentleman, and I have no idea as to what he may or may not have done, may have said or may not have said.

I was only and innocently answering your question based solely on the basis of my knowledge and experience of FEA and COMSOL and the fact that the correct solution of Maxwell's equations for a cavity of any shape precludes a non-zero force over a whole period.

This follows logically from the fact that a Finite Element solution performs a finite element discretization in space, not a finite element discretization in time.  For a steady state solution, the harmonic variation in time is assumed and not displayed.  For a transient solution different packages use different methods, usually a fiinite difference discretization in time for example (as the time variation is one-dimensional).

COMSOL is an excellent Finite Element package.

If you know the gentleman you may ask him (if interested) exactly what he did and what was the time variation he obtained for the calculated force, and if the variation is not harmonic, how did he obtain a non-harmonic variation using COMSOL.  Obviously, I can't answer for him.
« Last Edit: 05/10/2015 12:07 am by Rodal »

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Thanks. And we are in violent agreement about the consequences of applying Maxwell.

Offline LasJayhawk

Thanks. And we are in violent agreement about the consequences of applying Maxwell.

How about his friend, the demon?

If I pick my waveleghth, design my cavity, and antenna placement, then adjust the phase of the signal such that one end is in a node and the other is riding a crest, both all the time all the time, I would have an unbalance between the two ends???

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Thanks. And we are in violent agreement about the consequences of applying Maxwell.

How about his friend, the demon?

If I pick my waveleghth, design my cavity, and antenna placement, then adjust the phase of the signal such that one end is in a node and the other is riding a crest, both all the time all the time, I would have an unbalance between the two ends???
Yes, but not the sort that is supposed to matter according to Maxwell.

I like the idea of an antenna to avoid the potential measurement artifacts of hooking up a coaxial cable feed. Unfortunately, the practicality of beaming sufficient far field power to an antenna is not great. And even if this sufficed, there'd be RF flying everywhere messing things up.

Really and truly, this experiment cries out for a space test.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
....

Really and truly, this experiment cries out for a space test.
There are several people that think the same way (that it cries out for a space test), while there are several other ones that think it would be premature.  Since you made a powerful argument for the energy paradox, it would be useful if you could list all the reasons (and what and how should be tested) why a space test should be the next step, as a powerful argument in that direction may help to push the ball rolling...upwards  :)
« Last Edit: 05/10/2015 12:31 am by Rodal »

Offline PaulF

  • Member
  • Posts: 52
  • Netherlands
  • Liked: 19
  • Likes Given: 9
....

Really and truly, this experiment cries out for a space test.
There are several people that think the same way (that it cries out for a space test), while there are several other ones that think it would be premature.  Since you made a powerful argument for the energy paradox, it would be useful if you could list all the reasons (and what and how should be tested) why a space test should be the next step, as a powerful argument in that direction may help to push the ball rolling...upwards  :)
If enough people here are solidary, we could petition directly to Elon Musk for a ride on a dragon slated for the ISS, and see what he has to say. As collateral we could offer him all the research that has been done and collected by the fine gentlemen and ladies here. That is, IF the device actually works in space. But Elon Musk likes taking chances, that's what made him a billionaire.

Offline Notsosureofit

  • Full Member
  • ****
  • Posts: 691
  • Liked: 747
  • Likes Given: 1729
FYI   Arghhhh!

Wasted this rare free day chasing constant acceleration transforms til I remembered my own hypothesis is based on negative feedback of the acceleration.  (shows what can happen once the bit is in the mouth)

Time for a hot tub...

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
....

Really and truly, this experiment cries out for a space test.
There are several people that think the same way (that it cries out for a space test), while there are several other ones that think it would be premature.  Since you made a powerful argument for the energy paradox, it would be useful if you could list all the reasons (and what and how should be tested) why a space test should be the next step, as a powerful argument in that direction may help to push the ball rolling...upwards  :)
I am happy that you found my position on the energy paradox powerful. But speaking of premature, some further mulling and reading - ongoing - tells me that there may well be a third possibility for the behaviour of the thrust over time. If this "third thrust scenario" turns out to be the correct one, then the EmDrive may not be useful for any kind of propulsion, and would at best remain as an interesting test framework for various physics variants.

A third thrust scenario
Again we assume that the measured thrust is actual. We note its two characteristics:
a) It is measured as a static force.
b) It remains constant as long as input power is applied to this static configuration.
'a)' implies that we can only theorise as to its dynamic behaviour - i.e. when the EmDrive moves over time.

I take together two further pieces of information:
1. Shawyer's video of the moving EmDrive
2. Mike McCulloch's MiHsC theory of operation.
Now of course, there are a variety of interpretations possible for both of these. For example, that the video is flawed because all the angular momentum is being supplied by imperfections in the air bearing. Or that Mike's theory is nonsense because it violates GR. And so forth. So here I have to decide what to assume so as to justify this 3rd scenario. So here I choose:
1. The vid shows that an impulse is produced which results in a real and constant momentum of the EmDrive.
2. The theory predicts the same thing - constant forward EmDrive momentum.

And so this third thrust scenario is this:
As soon as the EmDrive is free to move, a definite and constant momentum is established and its thrust falls to zero.

So far, so good. But given this third thrust scenario is the correct one, what then is expected to happen when we switch off the power?
« Last Edit: 05/10/2015 01:26 am by deltaMass »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
...
And so this third thrust scenario is this:
As soon as the EmDrive is free to move, a definite and constant momentum is established and its thrust falls to zero.

So far, so good. But given this third thrust scenario is the correct one, what then is expected to happen when we switch off the power?
How do you reconcile this with the NASA Tests showing an impulsive force, in the same direction as the movement, which once it reaches the knee of the uprise, after ~2 sec it either stays fairly constant or continues increasing at a much smaller rate?.  None of the NASA tests showed a Dirac delta function type of response. None of the NASA tests showed the force decreasing to zero once the EM Drive started to move, on the contrary, the force stayed constant or it increased.
« Last Edit: 05/10/2015 01:36 am by Rodal »

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
I'll attempt to answer my own question, using the same set of assumptions as before. Momentum is a signed quantity, and we have to respect the symmetry of the situation. Therefore, when the power is switched off, an equal and opposite momentum is newly established, resulting in the EmDrive coming to rest.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
FYI   Arghhhh!

Wasted this rare free day chasing constant acceleration transforms til I remembered my own hypothesis is based on negative feedback of the acceleration.  (shows what can happen once the bit is in the mouth)

Time for a hot tub...

“Einstein box” Gedanken experiment first proposed by Balazs

N. L. Balazs, “The energy-momentum tensor of the electromagnetic field inside matter,” Phys. Rev. 91, 408-411
(1953).

for the system’s center of mass to be in the same place in both experiments, it is necessary for the slab in the latter case to have shifted to the right. The difference between the free-space momentum of the pulse and its electromagnetic (or Abraham) momentum is thus transferred to the slab in the form of mechanical momentum, pM, causing the slab’s eventual displacement in a manner consistent with the demands of the Einstein box experiment.

http://bit.ly/1DZl2z6

Resolution of the Abraham-Minkowski Controversy
Masud Mansuripur
College of Optical Sciences, The University of Arizona, Tucson, Arizona 85721
[Published in Optics Communications 283, 1997-2005 (2010)]

The dielectric insert may be acting as the slab.  That produces a solid movement towards the small end according to Balazs, which is what NASA observes.  However, the small end of the EM Drive is not open, so doesn't the wave reflect on the small end's copper surface and therefore enters the dielectric again now heading in the direction towards the big end which produces a movement towards the big base? back and forth?

« Last Edit: 05/10/2015 02:42 am by Rodal »

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
...
And so this third thrust scenario is this:
As soon as the EmDrive is free to move, a definite and constant momentum is established and its thrust falls to zero.

So far, so good. But given this third thrust scenario is the correct one, what then is expected to happen when we switch off the power?
How do you reconcile this with the NASA Tests showing an impulsive force, in the same direction as the movement, which once it reaches the knee of the uprise, after ~2 sec it either stays fairly constant or continues increasing at a much smaller rate?.  None of the NASA tests showed a Dirac delta function type of response. None of the NASA tests showed the force decreasing to zero once the EM Drive started to move, on the contrary, the force stayed constant or it increased.
I'm not aware of any NASA tests on a moving EmDrive. Refs?
As for the static force profile, the small residual increase appears to have a slope that matches, to within experimental accuracy, the prevailing thermal drift.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
..
I'm not aware of any NASA tests on a moving EmDrive. Refs?
As for the static force profile, the small residual increase appears to have a slope that matches, to within experimental accuracy, the prevailing thermal drift.

NASA is measuring a displacement of the EM Drive vs. time, which means that the EM Drive is moving.  This is true for all the NASA experiments of the truncated cone EM Drive.

The motion is a rigid body rotation of the EM Drive around the center of the torque pendulum, and the force on the EM Drive is reacted by the torsional stiffness of the torque pendulum.

References:

1) Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum (Brady et al)
AIAA 2014-4029

2) NSF EM Drive threads 1 and 2
« Last Edit: 05/10/2015 01:56 am by Rodal »

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Briefly moving, yes, but constrained from free dynamics - which is what I'm attempting to address in all three thrust scenarios. If the torque balance shows a rectangular pulse (which it largely does, up to thermal drift), this indicates the prompt onset of a constant force. But this arrangement by no means guarantees that the appropriate momentum, per the third thrust scenario, has been established. In fact it guarantees that it has not, since as soon as the torque balances the thrust, the balance ceases to turn and the momentum is zero. And we're back to a situation tantamount to the EmDrive pushing against an immovable wall - i.e. fully static.

Recall the third thrust scenario says that the thrust only falls to zero after the appropriate momentum is established.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Briefly moving, yes, but constrained from free dynamics - which is what I'm attempting to address in all three thrust scenarios. If the torque balance shows a rectangular pulse (which it largely does, up to thermal drift), this indicates the prompt onset of a constant force. But this arrangement by no means guarantees that the appropriate momentum, per the third thrust scenario, has been established. In fact it guarantees that it has not, since as soon as the torque balances the thrust, the balance ceases to turn and the momentum is zero. And we're back to a situation tantamount to the EmDrive pushing against an immovable wall - i.e. fully static.

Recall the third thrust scenario says that the thrust only falls to zero after the appropriate momentum is established.
free dynamics ?

Both Shawyer's rotating EM Drive (in his video) and NASA's EM Drive are under the action of the Earth's gravity, a force which is balanced by the support in both cases.
Both encounter an opposing force: the stiffness of the torsional pendulum in NASA's case, and the friction of the air bearing in Shawyer's case.


How do you define what freely moves and what doesn't freely move?
Is there a velocity threshold? Does it have a preferred frame of reference?
Is there an acceleration threshold?
Is there a threshold for what the opposing, resisting force should be ? (torsional stiffness of the pendulum or frictional resistance of the bearing)
« Last Edit: 05/10/2015 02:47 am by Rodal »

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
I really don't know where you're going with that lot. All I am saying is that the torque balance does not allow a constant momentum of the EmDrive to be established and retained.

And in that sense its dynamics are not free.
« Last Edit: 05/10/2015 06:47 am by deltaMass »

Offline Notsosureofit

  • Full Member
  • ****
  • Posts: 691
  • Liked: 747
  • Likes Given: 1729
FYI   Arghhhh!

Wasted this rare free day chasing constant acceleration transforms til I remembered my own hypothesis is based on negative feedback of the acceleration.  (shows what can happen once the bit is in the mouth)

Time for a hot tub...

“Einstein box” Gedanken experiment first proposed by Balazs

N. L. Balazs, “The energy-momentum tensor of the electromagnetic field inside matter,” Phys. Rev. 91, 408-411
(1953).

for the system’s center of mass to be in the same place in both experiments, it is necessary for the slab in the latter case to have shifted to the right. The difference between the free-space momentum of the pulse and its electromagnetic (or Abraham) momentum is thus transferred to the slab in the form of mechanical momentum, pM, causing the slab’s eventual displacement in a manner consistent with the demands of the Einstein box experiment.

http://bit.ly/1DZl2z6

Resolution of the Abraham-Minkowski Controversy
Masud Mansuripur
College of Optical Sciences, The University of Arizona, Tucson, Arizona 85721
[Published in Optics Communications 283, 1997-2005 (2010)]

The dielectric insert may be acting as the slab.  That produces a solid movement towards the small end according to Balazs, which is what NASA observes.  However, the small end of the EM Drive is not open, so doesn't the wave reflect on the small end's copper surface and therefore enters the dielectric again now heading in the direction towards the big end which produces a movement towards the big base? back and forth?



Yes, I am aware of those arguments, and as you can see, there is no internal arrangement of dielectric and/or absorbers that will not integrate to zero w/ those forces.

The only common thing I see between the photon directions is that they both see the gradient.  So, is the photon smart enough to distinguish between the cavity induced gradient and an acceleration induced one ?

Maybe, maybe not, still an open question.
« Last Edit: 05/10/2015 03:08 am by Notsosureofit »

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1