-
#540
by
Mulletron
on 12 Feb, 2015 20:10
-
-
#541
by
Rodal
on 12 Feb, 2015 20:13
-
-
#542
by
Giovanni DS
on 12 Feb, 2015 20:18
-
I have a question about how the thrust is usually measured. Is it a pendulum? if so, could a physical cavity deformation simply move the barycenter? that could be mistaken as thrust.
Far from being an expert, I am still enjoying the thread
-
#543
by
Mulletron
on 12 Feb, 2015 20:25
-
What I don't understand still is how that thermal buckling could appear as a sustained false thrust signature over 30-45 second runs, instead of just a quick impulse. Is it because the CG moved or something?
-
#544
by
Rodal
on 12 Feb, 2015 20:26
-
I have a question about how the thrust is usually measured. Is it a pendulum? if so, could a physical cavity deformation simply move the barycenter? that could be mistaken as thrust.
Far from being an expert, I am still enjoying the thread
It is a horizontal torsional pendulum, with the torsional axis of rotation in the vertical direction (call it "z"), paralel to the direction of the force of gravity and hence perpendicular to the floor. The EM Drive's force produces a rotation around that vertical axis "z", such that the motion of the EM Drive occurs mainly in the x-y plane parallel to the floor, perpendicular to "z". There is a small amount of coupling between the rotations around z, rotation around x and rotation around y due to the moments of inertia. I wrote a program in Mathematica taking into account the nonlinear coupled equations of motion, thinking that the forces could be an artifact due to nonlinear coupling of the equations of motion but I after comparing with the data I rejected that possibility.
-
#545
by
Rodal
on 12 Feb, 2015 20:34
-
What I don't understand still is how that thermal buckling could appear as a sustained false thrust signature over 30-45 second runs, instead of just a quick impulse. Is it because the CG moved or something?
Yes, good question. For a perfectly elastic pendulum it should restore itself, as you say. For it to persist for 45 seconds, one needs to posit some nonlinear non-conservative mechanism, perhaps in the nonlinear coupled equations of motion or perhaps also involving parasitic electromagnetic effects.
EDIT: There is another explanation that was offered by Oak Ridge involving
thermal expansion, shifting the center of mass, we should discuss that sometime... (Thermal expansion is slower, but it could be that the answer is due to 1) thermal buckling (fast) + 2) thermal expansion (slower))
-
#546
by
SleeperService
on 12 Feb, 2015 20:43
-
I have a question about how the thrust is usually measured. Is it a pendulum? if so, could a physical cavity deformation simply move the barycenter? that could be mistaken as thrust.
Far from being an expert, I am still enjoying the thread
It is a horizontal torsional pendulum, with the torsional axis of rotation in the vertical direction (call it "z"), paralel to the direction of the force of gravity and hence perpendicular to the floor. The EM Drive's force produces a rotation around that vertical axis "z", such that the motion of the EM Drive occurs mainly in the x-y plane parallel to the floor, perpendicular to "z". There is a small amount of coupling between the rotations around z, rotation around x and rotation around y due to the moments of inertia. I wrote a program in Mathematica taking into account the nonlinear coupled equations of motion, thinking that the forces could be an artifact due to nonlinear coupling of the equations of motion but I after comparing with the data I rejected that possibility.
Could you share your Mathmatica code nevertheless?
-
#547
by
Star One
on 12 Feb, 2015 20:46
-
-
#548
by
SleeperService
on 12 Feb, 2015 20:50
-
I think we are getting close to a truth here...
-
#549
by
aero
on 12 Feb, 2015 21:01
-
Attached find the fields from the Copper Kettle at the 32-nd half-period of the drive frequency, 1.937115E+009 Hz. This would compare directly to the situation presented yesterday, except the dielectric has been removed.
I note that by watching the development of the fields from the first half-period, the cavity resonance does not seem to be as strong. That is, the fields drift with time. The strong blue point moves.
-
#550
by
aero
on 12 Feb, 2015 21:08
-
Something I should mention. In watching the development of the evanescent waves and the propagating waves within the cavity, I can see no evidence that one moves any faster than the other.
I think what that means is that as soon as the evanescent wave exits the gap in the end plates, its velocity reverts to the classic solution. The equations needed to determine this are in the paper I referenced yesterday,
http://wwwsis.lnf.infn.it/pub/INFN-FM-00-04.pdf though it would take some tinkering to formulate them to correctly make this determination.
Anyone want to tackle that problem?
-
#551
by
aero
on 12 Feb, 2015 21:36
-
@ Rodal
For a perfectly cylindrical resonator we have a simple exact solution (see: http://en.wikipedia.org/wiki/Microwave_cavity#Cylindrical_cavity ) all that one needs to do is to compare the frequency calculated by MEEP for a cylindrical cavity with the exact solution.
Yes, that fact was used in Dr. Dominic's analysis. However, he also wanted to verify that the meep fields were as they were supposed to be and not the cause of the discrepancy. They are OK and peaks match the cylindrical cavity resonance frequencies nicely.
-
#552
by
Rodal
on 12 Feb, 2015 21:36
-
Attached find the fields from the Copper Kettle at the 32-nd half-period of the drive frequency, 1.937115E+009 Hz. This would compare directly to the situation presented yesterday, except the dielectric has been removed.
I note that by watching the development of the fields from the first half-period, the cavity resonance does not seem to be as strong. That is, the fields drift with time. The strong blue point moves.
You calculated a force with the dielectric in the cavity,
what force do you calculate
without the dielectric?
How does the force compare?
-
#553
by
aero
on 12 Feb, 2015 21:41
-
Attached find the fields from the Copper Kettle at the 32-nd half-period of the drive frequency, 1.937115E+009 Hz. This would compare directly to the situation presented yesterday, except the dielectric has been removed.
I note that by watching the development of the fields from the first half-period, the cavity resonance does not seem to be as strong. That is, the fields drift with time. The strong blue point moves.
You calculated a force with the dielectric in the cavity, what force do you calculate without the dielectric?
How does the force compare?
Interesting question.
That is another time consuming computer job. I might have answers for you tonight, if not, tomorrow.
-
#554
by
Mulletron
on 12 Feb, 2015 22:14
-
I accidentally stumbled upon this report:
https://hal.archives-ouvertes.fr/hal-00551421v1/document
I'm still trying to find which reference it was where any of this implied non-reciprocity.
It does imply non-reciprocity:
http://arxiv.org/pdf/1101.1174v1.pdfSection 2.1.
Magneto-electric directional anisotropy can be induced in all media, including centrosymmetric
ones. It was first predicted by G. E. Stedman and coworkers [11] and observed for the first time in
a crystal by G. L. J. A. Rikken and coworkers [12]. This non-reciprocal effect is supposed to be independent on light polarization.
I'm sure any optics expert would tell me, well duh!
-
#555
by
Rodal
on 12 Feb, 2015 22:42
-
I accidentally stumbled upon this report:
https://hal.archives-ouvertes.fr/hal-00551421v1/document
I'm still trying to find which reference it was where any of this implied non-reciprocity.
It does imply non-reciprocity:
http://arxiv.org/pdf/1101.1174v1.pdf
Section 2.1.
Magneto-electric directional anisotropy can be induced in all media, including centrosymmetric
ones. It was first predicted by G. E. Stedman and coworkers [11] and observed for the first time in
a crystal by G. L. J. A. Rikken and coworkers [12]. This non-reciprocal effect is supposed to be independent on light polarization.
I'm sure any optics expert would tell me, well duh!
Yes.
I also read the (C. Robilliard and G. Bailly) paper (Table 1) as stating that the calculated magneto-electric linear birefringence MELB = nB − nE for the Quantum Vacuum is 3*10^(-10) of the value for Nitrogen:
ten orders of magnitude smaller for the Quantum Vacuum than for Air.
-
#556
by
aero
on 12 Feb, 2015 23:38
-
Attached find the fields from the Copper Kettle at the 32-nd half-period of the drive frequency, 2.168E+009Hz. In this case, dielectric has been removed and the drive frequency has been increased. Increased drive frequency means that the actual simulated time is less than in the previous cases because the period is shorter.
-
#557
by
Rodal
on 12 Feb, 2015 23:53
-
Attached find the fields from the Copper Kettle at the 32-nd half-period of the drive frequency, 2.168E+009Hz. In this case, dielectric has been removed and the drive frequency has been increased. Increased drive frequency means that the actual simulated time is less than in the previous cases because the period is shorter.
Thanks. It appears then that the evanescent fields are strong and unsymmetric (fore and aft of the EM Drive) with a resonant cavity with the dielectric at 1.9 GHz
The evanescent fields are weaker and their intensity
much more symmetric (fore and aft of the EM Drive) with the
dielectric removed at the same frequency 1.9 GHz. There is no strong resonance now because with the dielectric removed the natural frequencies for the same mode shapes occur at higher frequencies.
With the dielectric removed, at the higher frequency of 2.1 GHz, the evanescent field intensities are unsymmetric again (fore and aft) like with the dielectric at 1.9 GHz but the fields don't seem to be as strong
1) You need to give us a contour field mapping rule: what do the colors mean in numerical terms, to further understand the contourplot. For example: what is the intensity of the white areas? of the red areas? of the orange areas? of the blue areas?
2) It would be great to get the forces on the center of mass of the EM Drive you compute for the three different cases
-
#558
by
Mulletron
on 12 Feb, 2015 23:56
-
-
#559
by
Mulletron
on 13 Feb, 2015 00:12
-
Is there anything useful right now, today..... that can be done with an always on 50uN of thrust in space?
