Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1799688 times)

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
That looks like useful information, unfortunately it's Greek to me. :)

Need to use a translator I guess.

No, Meep doesn't use GPUs, but of course the source language code is available, written in C++, so someone really, really motivated could make it so.

Same answer re. writing new functions for material characteristics, except Meep does already provide hooks for new user supplied functions in some instances and I think material characteristics is one place where it does.
Digging I found this.
http://www.ajol.info/index.php/ijest/article/viewFile/83885/73892
Simulation and analysis of microwave heating while joining bulk copper
M. S. Srinath 1*, P. Suryanarayana Murthy2
, Apurbba Kumar Sharma3
,
Pradeep Kumar4
, M. V. Kartikeyan5
1*Department of Mechanical Engineering, Malnad College of Engineering, Hassan-573201, INDIA
2,3,4Department of Mechanical and Industrial Engineering, Indian Institute of Technology Roorkee, Roorkee 247 667, INDIA.
5Department of Electronics and Computer Science Engineering, Indian Institute of Technology Roorkee, Roorkee

I believe you will find what you're looking for here.
Shell

That's an interesting paper, Thanks.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5835
  • USA
  • Liked: 5915
  • Likes Given: 5253
That looks like useful information, unfortunately it's Greek to me. :)

Need to use a translator I guess.

No, Meep doesn't use GPUs, but of course the source language code is available, written in C++, so someone really, really motivated could make it so.

Same answer re. writing new functions for material characteristics, except Meep does already provide hooks for new user supplied functions in some instances and I think material characteristics is one place where it does.
Digging I found this.
http://www.ajol.info/index.php/ijest/article/viewFile/83885/73892
Simulation and analysis of microwave heating while joining bulk copper
M. S. Srinath 1*, P. Suryanarayana Murthy2
, Apurbba Kumar Sharma3
,
Pradeep Kumar4
, M. V. Kartikeyan5
1*Department of Mechanical Engineering, Malnad College of Engineering, Hassan-573201, INDIA
2,3,4Department of Mechanical and Industrial Engineering, Indian Institute of Technology Roorkee, Roorkee 247 667, INDIA.
5Department of Electronics and Computer Science Engineering, Indian Institute of Technology Roorkee, Roorkee

I believe you will find what you're looking for here.
Shell

That's an interesting paper, Thanks.

Note"

Quote
Due to the unavailability of permeability values, only the effect of permittivity (εr)  is  taken  in  this  model.  In  the  heat  transfer  module  only  the  effect  of  conduction  is  taken  into  account.  For  the
future study, the other modes of heat transfer can be included.

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
This is what Meep calculates. (Yang model)

Permittivity    frequency      imag. freq.       Q             |amp|      
 Perfect Metal 2.4834659343 -3.79E-007      3,273,813  0.8598093348
amplitude :  -0.5636892860372409+0.6492507073419053i error :  3.663515085704893e-10+0.0i

copper model 2.4834659343 -4.26E-007   2,914,661 0.8598042117
amplitude :  -0.5636862008658441+0.6492466013378828i error :  3.663577500401384e-10+0.0i


Note that the only difference of any significance is the quality factor which was reduced a little over 10% only. So not much difference but that is what we expected.
« Last Edit: 06/19/2015 09:06 PM by aero »
Retired, working interesting problems

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
This is what Meep calculates. (Yang model)

Permittivity    frequency      imag. freq.       Q             |amp|                           amplitude   
Perfect Metal 2.4834659343 -3.79E-007  3,273,813  0.8598093348 -0.5636892860372409+0.6492507073419053i    
                                                                                                     error :  3.663515085704893e-10+0.0i
copper model 2.4834659343 -4.26E-007   2,914,661 0.8598042117 -0.5636862008658441+0.6492466013378828i
                                                                                                     error :  3.663577500401384e-10+0.0i


Note that the only difference of any significance is the quality factor which was reduced a little over 10% only. So not much difference but that is what we expected.
This is gobbledygook to me. What is it that you are calculating and what does "imag. freq." mean, for example? And is it truly the case that the calculated Q value is up in the millions?
« Last Edit: 06/19/2015 09:05 PM by deltaMass »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5835
  • USA
  • Liked: 5915
  • Likes Given: 5253
This is what Meep calculates. (Yang model)

Permittivity    frequency      imag. freq.       Q             |amp|      
 Perfect Metal 2.4834659343 -3.79E-007      3,273,813  0.8598093348
amplitude :  -0.5636892860372409+0.6492507073419053i error :  3.663515085704893e-10+0.0i

copper model 2.4834659343 -4.26E-007   2,914,661 0.8598042117
amplitude :  -0.5636862008658441+0.6492466013378828i error :  3.663577500401384e-10+0.0i


Note that the only difference of any significance is the quality factor which was reduced a little over 10% only. So not much difference but that is what we expected.


What are the computer times for both cases? (is there any advantage in computer time to run the perfect metal case?) [it matters for a time-marching scheme, as every time step is another set of calculations]
« Last Edit: 06/19/2015 09:42 PM by Rodal »

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
This is what Meep calculates. (Yang model)

Permittivity    frequency      imag. freq.       Q             |amp|                           amplitude   
Perfect Metal 2.4834659343 -3.79E-007  3,273,813  0.8598093348 -0.5636892860372409+0.6492507073419053i    
                                                                                                     error :  3.663515085704893e-10+0.0i
copper model 2.4834659343 -4.26E-007   2,914,661 0.8598042117 -0.5636862008658441+0.6492466013378828i
                                                                                                     error :  3.663577500401384e-10+0.0i


Note that the only difference of any significance is the quality factor which was reduced a little over 10% only. So not much difference but that is what we expected.
This is gobbledygook to me. What is it that you are calculating and what does "imag. freq." mean, for example? And is it truly the case that the calculated Q value is up in the millions?

It isn't formatted very well but I dislike taking digets off to fit it onto one line. This data is in Meep units. The first number is the real component of frequency, the second is the imaginary frequency component. Scale to SI units by multiplying by c and dividing by my scale factor, 0.3. (Or approximatly 1x109) Yes, for the Yang model, Meep calculates very high Q values. Following Q is the magnitude of the signal amplitude in meep units followed by the complex amplitude, then an Meep internal error calculation. Small value means "good."
Retired, working interesting problems

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
Quote
What are the computer times for both cases? (is there any advantage in computer time to run the perfect metal case?) [it matters for a time-marching scheme, as every time step is another set of calculations]

100.0% done in 4343.1s   I really didn't keep track of run time, but they were about the same. As it happens, the resonance calculations take significantly more time than calculating the field evolution. Here I used resolution of 200 which is higher than needed but I wanted to be sure that resolution was not a factor in this evaluation. I normally run at resolution of 100 and only go higher to verify an unexpected result.
Retired, working interesting problems

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
This is what Meep calculates. (Yang model)

Permittivity    frequency      imag. freq.       Q             |amp|                           amplitude   
Perfect Metal 2.4834659343 -3.79E-007  3,273,813  0.8598093348 -0.5636892860372409+0.6492507073419053i    
                                                                                                     error :  3.663515085704893e-10+0.0i
copper model 2.4834659343 -4.26E-007   2,914,661 0.8598042117 -0.5636862008658441+0.6492466013378828i
                                                                                                     error :  3.663577500401384e-10+0.0i


Note that the only difference of any significance is the quality factor which was reduced a little over 10% only. So not much difference but that is what we expected.

....This is gobbledygook to me. What is it that you are calculating and what does "imag. freq." mean, for example? And is it truly the case that the calculated Q value is up in the millions?

So, those values for the Drude model gave you a Q in the superconducting range ?

If so, either the values are not a good representation of the metal conditions, or there is something wrong with the model, or MEEP does not handle well those values numerically

Am I missing something ?
I doubt it.  While I didn't expect significant changes, I was thinking that Q would be more strongly affected. The Yang model is the nearest to a cylinder that we have so perhaps we could expect higher Q values with Perfect metal.

Here's what I emplimented, feel free to check it.

(define f2_4GHzmeep (/ (* 2.4E+9 asi) csi))  - asi is scale factor, 0.3, and csi is speed of light, m and m/s.
(define CU-D-conduct (/ (* 2 pi f2_4GHzmeep 3.25E+8) 1))
(material (make medium (epsilon 1) (D-conductivity CU-D-conduct)))
Retired, working interesting problems

Online WarpTech

  • Full Member
  • ****
  • Posts: 1227
  • Do it!
  • Vista, CA
  • Liked: 1293
  • Likes Given: 1744
...
Also pulled out reference 15 as it is of interest:
http://www.asps.it/article2.pdf

...

The paper in bold and underlined, at the end they are also stating that it appears the static part of the current is working against the magnetic part in electromagnetic propulsion.  "The origin of the propulsion is related to the nearfields, such as the static electric field and the inductive electromagnetic field.".  It appears they are talking here about the same thing WarpTech and I were discussing how the static field opposes the magnetic but still leads to some propulsion.  Here they imply that the near fields are a part of the propulsion and they are using two dipole antennas.

Something interesting to note is that in dipole antennas the current alternates between kinetic energy (magnetic) and potential energy (separation of charge) and these two fields are what oppose each other (w.r.t. propulsion) but also provide some propulsion (they don't balance out).  I also suspect the current passing through the magnetic fields of radiation also provide some propulsion and I think they mention this also. 

Now in the case of two circular cavities in TE011 mode but with one cavity out of phase by 90 degrees and taking into account time retardation we notice the current around the axis of the cavity doesn't allow for this charge separation and so we don't have the opposing static fields.  This is because the energy alternates from (kinetic current) to being stored in the light (also kinetic).  In the case that cavities are brought close to each other we only get magnetic interaction I think...  but there may/may-not be a draw back. 
...

Warning Edited Moving mode push.png and changed to imply I think the modes are opposing propulsion

I think the current will have a charge distribution on the surface because the perimeter of the antenna is not small compared to a wavelength. In other words, the charge density is not uniform, so there will be both magnetic and static forces at work between them.

I'm familiar with this uni-directional antenna stuff as I've modeled it before on MathCAD 7, like 15 years ago. My end result was that it is possible to get propulsion. However, it is simply due to leakage flux and is essentially a well focused antenna radiation lobe. It will never give more thrust than an equivalent photon rocket with an equal power output. So I gave up on the idea, lasers are more efficient. ;)
Todd

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5835
  • USA
  • Liked: 5915
  • Likes Given: 5253
This is what Meep calculates. (Yang model)

Permittivity    frequency      imag. freq.       Q             |amp|                           amplitude   
Perfect Metal 2.4834659343 -3.79E-007  3,273,813  0.8598093348 -0.5636892860372409+0.6492507073419053i    
                                                                                                     error :  3.663515085704893e-10+0.0i
copper model 2.4834659343 -4.26E-007   2,914,661 0.8598042117 -0.5636862008658441+0.6492466013378828i
                                                                                                     error :  3.663577500401384e-10+0.0i


Note that the only difference of any significance is the quality factor which was reduced a little over 10% only. So not much difference but that is what we expected.

....This is gobbledygook to me. What is it that you are calculating and what does "imag. freq." mean, for example? And is it truly the case that the calculated Q value is up in the millions?

So, those values for the Drude model gave you a Q in the superconducting range ?

If so, either the values are not a good representation of the metal conditions, or there is something wrong with the model, or MEEP does not handle well those values numerically

Am I missing something ?
I doubt it.  While I didn't expect significant changes, I was thinking that Q would be more strongly affected. The Yang model is the nearest to a cylinder that we have so perhaps we could expect higher Q values with Perfect metal.

Here's what I emplimented, feel free to check it.

(define f2_4GHzmeep (/ (* 2.4E+9 asi) csi))  - asi is scale factor, 0.3, and csi is speed of light, m and m/s.
(define CU-D-conduct (/ (* 2 pi f2_4GHzmeep 3.25E+8) 1))
(material (make medium (epsilon 1) (D-conductivity CU-D-conduct)))

I don't think that Q's in the millions are realistic, so if there is nothing wrong with the model, either the Drude model values are not a good representation of the metal conditions  (either because the values are unrepresentative or there is something wrong with the Drude model in the microwave range), or MEEP does not handle well numerically those Drude model values .   We have to agree that the cavity is not going to resonate with a Q in the millions.
« Last Edit: 06/20/2015 11:44 AM by Rodal »

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
Million's seems excessive, but Perfect metal cavity also resonates in the millions, even 10% higher so ....

I guess I'll run the other models just to see what happens with them.
Retired, working interesting problems

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
What is the meaning of "imaginary frequency"?

Online aero

  • Senior Member
  • *****
  • Posts: 2739
  • 92129
  • Liked: 704
  • Likes Given: 237
What is the meaning of "imaginary frequency"?

Here's one answer.

http://www.researchgate.net/post/What_does_imaginary_frequency_mean_while_optimizing_dimer_geometry_by_DFT_approach

These frequencies are obtained through calculating second derivatives of energy, which constitute the hessian matrix. Diagonalization of hessian matrix gives its eigenvalues and eigenvectors. Eigenvalues are squared harmonic frequencies of 3N-6 normal vibrations, while eigenvectors are 3N-6 normal coordinates. A negative eigenvalue means that one of its square roots (i.e harmonic frequency) contains imaginary number as a factor, and hence the term imaginary frequency. Physically, a negative eigenvalue corresponds to negative curvature of corresponding normal coordinate. In other wrds, the geometry on which this second derivative is yieled is passing through a energy maximum (and geometry corresponds to a saddle point) along this particular normal coordinate. Very low, and close to zero negative frequency means that geometry is already converged to an energy minimum, but the accuracy of energy gradient, and/or electronic scf gradient is not sufficient to eliminate negative frequencies correctly.
Retired, working interesting problems

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Thanks. In other words, it's an artifact of an optimisation process, and as such has no physical significance. Moreover, if it's nonzero it signifies that the solution process is broken in some way and that its results are not to be trusted.

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Million's seems excessive, but Perfect metal cavity also resonates in the millions, even 10% higher so ....

I guess I'll run the other models just to see what happens with them.
"Millions" is not just excessive - it's nonsense. Something isn't right with your simulator. Either the Q will appear to be infinite when the resistivity is identically zero, or the Q will have a physically reasonable value. Does MEEP know the resistivity of copper and is it using it?

Offline dustinthewind

  • Full Member
  • ****
  • Posts: 585
  • U.S. of A.
  • Liked: 226
  • Likes Given: 257
...
Also pulled out reference 15 as it is of interest:
http://www.asps.it/article2.pdf

...

The paper in bold and underlined, at the end they are also stating that it appears the static part of the current is working against the magnetic part in electromagnetic propulsion.  "The origin of the propulsion is related to the nearfields, such as the static electric field and the inductive electromagnetic field.".  It appears they are talking here about the same thing WarpTech and I were discussing how the static field opposes the magnetic but still leads to some propulsion.  Here they imply that the near fields are a part of the propulsion and they are using two dipole antennas.

Something interesting to note is that in dipole antennas the current alternates between kinetic energy (magnetic) and potential energy (separation of charge) and these two fields are what oppose each other (w.r.t. propulsion) but also provide some propulsion (they don't balance out).  I also suspect the current passing through the magnetic fields of radiation also provide some propulsion and I think they mention this also. 

Now in the case of two circular cavities in TE011 mode but with one cavity out of phase by 90 degrees and taking into account time retardation we notice the current around the axis of the cavity doesn't allow for this charge separation and so we don't have the opposing static fields.  This is because the energy alternates from (kinetic current) to being stored in the light (also kinetic).  In the case that cavities are brought close to each other we only get magnetic interaction I think...  but there may/may-not be a draw back. 
...

Warning Edited Moving mode push.png and changed to imply I think the modes are opposing propulsion

I think the current will have a charge distribution on the surface because the perimeter of the antenna is not small compared to a wavelength. In other words, the charge density is not uniform, so there will be both magnetic and static forces at work between them.

I'm familiar with this uni-directional antenna stuff as I've modeled it before on MathCAD 7, like 15 years ago. My end result was that it is possible to get propulsion. However, it is simply due to leakage flux and is essentially a well focused antenna radiation lobe. It will never give more thrust than an equivalent photon rocket with an equal power output. So I gave up on the idea, lasers are more efficient. ;)
Todd

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2736591/  quote: "where the TE011 mode power loss normalized to the peak rf magnetic field is given by (6).  The self-inductance and capacitance of cylindrical cavity modes are derived from the stored electric and magnetic energies" (light) "and the cavity wall currents" (magnetic) "at resonance."  They don't mention charge separation. 

Also here "http://arxiv.org/ftp/physics/papers/0603/0603154.pdf", "Intrinsically, the TE011 mode has closed E and H loops..."  Closed E field loops indicate no separation of charge unless I am miss understanding things.

Offline ThinkerX

  • Full Member
  • ***
  • Posts: 303
  • Alaska
  • Liked: 112
  • Likes Given: 59
Doctor McCulloch has a new blog post up.

http://physicsfromtheedge.blogspot.com/

Seems he is looking into combining his theory with the Zero Point Field (ZPF), which to me seems to be aiming into Doctor White territory.

While disagreements abound, to me it almost seems like the different EM Drive theories are sort of...converging, or attempting to describe the same...force???...from different perspectives in different terminology.

Doctor McCulloch has a new blog post up.

http://physicsfromtheedge.blogspot.com/

Seems he is looking into combining his theory with the Zero Point Field (ZPF), which to me seems to be aiming into Doctor White territory.

While disagreements abound, to me it almost seems like the different EM Drive theories are sort of...converging, or attempting to describe the same...force???...from different perspectives in different terminology.

He is also posting on Reddit with the username memcculloch

http://www.reddit.com/r/EmDrive/comments/3acayb/mihsc_lets_talk_about_this/

Offline Fugudaddy

  • Member
  • Posts: 31
  • Rochester NY
  • Liked: 57
  • Likes Given: 70
Future considerations:
Should the emdrive become a reality, I can envision a next-generation of high-power, miniaturized frustums (from my old filter days - when these were the newest disruptive innovation/technology introduced)

Speaking of disruptive technologies; this was from back in April, hard to tell if this was discussed or not as an option but:

http://www.nasa.gov/marshall/news/nasa-3-D-prints-first-full-scale-copper-rocket-engine-part.html

small frustum models should be easier and faster to 'print' than a whole engine part.

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Baby's twitching again.

That's the good news.

The bad news is that she's twitching when turned off, too.
Oh, Well.

« Last Edit: 06/20/2015 03:13 AM by deltaMass »

Tags: