Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1873431 times)

Offline VAXHeadroom

  • Full Member
  • *
  • Posts: 195
  • Whereever you go, there you are. -- BB
  • Baltimore MD
  • Liked: 256
  • Likes Given: 151
The higher density views are uploaded.
https://drive.google.com/folderview?id=0B1XizxEfB23tfkVzeXVub2NpTm5fanZTTTdrLXNiT3VHaV9FYnB6TVpDUmJsWjRQbEUwdE0&usp=sharing

These Meep runs were made at resolution = 250 which is 2.5 times higher than previously uploaded views. These are the 14 final time slices of a 32 cycle run. 14 images for each view, separated by 0.1 cycle of the drive frequency 2.45 GHz. The Gaussian noise bandwidth of 2.45 GHz * .025 was used which seems reasonable for a magnetron. The 58 mm dipole antenna was located parallel to and 1/4 wavelength from the small end plate, excited with the Ez field component.

Included are two models of the 10.2 inch NSF-1701 cavity, one using copper and the other using Perfect metal.

I expect questions.

Looking forward to seeing somebody make movies from these higher density runs !

On it.
Emory Stagmer
  Executive Producer, Public Speaker UnTied Music - www.untiedmusic.com

Offline ElizabethGreene

  • Member
  • Posts: 69
  • Nashville, Tennessee
  • Liked: 138
  • Likes Given: 3
How hot can a magnetron get before it is 'damaged'?
I imagine all internal parts are expanding, I wonder how long it takes for parts to plastically deform.
I don't know how magnetrons are constructed.  They can probably withstand a temperature above 100 C without getting damaged

Data point: My donor microwave has a thermal cutout mounted 5mm offset from the exterior wall of the heatsink.  I searched for those and found they come in 3 temperatures.  190, 200, and 230.

Can you look at your pile of scrap parts and see if you have a bit that looks like this?


This is not a 100% answer, as the same vendor also sells Thermal Fuses that go up significantly higher.


Parts source:
http://www.electronix.com/advanced_search_result.php?keywords=microwave+oven+thermostat
 
The way my donor was wired the cutout kills all power to the device until it cools down.  The oven has a mechanical timer, not electronic, so cutting the power to cool down is acceptable.

Here is my wiring diagram.  The thermal cutout is visible in the upper right trace.

Moderator note: I resized this down significantly, please let me know if it needs to be smaller to keep the thread width reasonable.
Copyright note: I assume this device diagram is someone's intellectual property.  If you are the owner and wish it removed, please let me know.

Offline aero

  • Senior Member
  • *****
  • Posts: 2785
  • 92129
  • Liked: 724
  • Likes Given: 249
...
Yes, a very big help from zxcvb, Thank you again, and welcome!

Aero, what is the output power of the antenna in the simulation? The numbers are much larger than they were, but they are still much, much lower than I would expect under a resonant condition.
Todd

Todd, are you taking into account that these fields are in Meep dimensionless units ( http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#Units_in_Meep )? (not normal SI dimensional units)

Also, the value of the field at the antenna is higher than shown, I had to clip the values of the antenna so that the electric fields can be seen.

Sorry, I must've missed that detail. How do I scale it to V/m and A/m then? I'm looking for some peak and average values so I can estimate the copper losses.
Todd

Here: http://meepunits.wikia.com/wiki/Meep_unit_transformation_Wiki

With that and knowing that my scale factor, a, is 0.3, take your best shot.
And by comparing the data from the two runs, you may have a shot. The Perfect metal is lossless as I understand it. Everything that I set in the control file were identical so, except for adjustments that Meep may make internally (I don't know about anything in particular) the runs data should be comparable.
« Last Edit: 07/02/2015 01:55 AM by aero »
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
...
Yes, a very big help from zxcvb, Thank you again, and welcome!

Aero, what is the output power of the antenna in the simulation? The numbers are much larger than they were, but they are still much, much lower than I would expect under a resonant condition.
Todd

Todd, are you taking into account that these fields are in Meep dimensionless units ( http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#Units_in_Meep )? (not normal SI dimensional units)

Also, the value of the field at the antenna is higher than shown, I had to clip the values of the antenna so that the electric fields can be seen.

Sorry, I must've missed that detail. How do I scale it to V/m and A/m then? I'm looking for some peak and average values so I can estimate the copper losses.
Todd

Here: http://meepunits.wikia.com/wiki/Meep_unit_transformation_Wiki

With that and knowing that my scale factor, a, is 0.3, take your best shot.

So the electric field has to be multiplied by to get Volts/meter

and H is already in Amps/meter  (H does not need any transformation)

where does the scale factor, a=0.3, enter the picture?

Into the denominator , which is in meters?

So really to get the Electric field one has to multiply by 3767.3/3 = 1255.77  Volts/meter

and the Magnetic field H has to be multiplied by 1/0.3 = 3.33333   Amps/meter
« Last Edit: 07/02/2015 02:07 AM by Rodal »

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1301
  • Do it!
  • Vista, CA
  • Liked: 1350
  • Likes Given: 1813
Todd, are you taking into account that these fields are in Meep dimensionless units ( http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#Units_in_Meep )? (not normal SI dimensional units)

Also, the value of the field at the antenna is higher than shown, I had to clip the values of the antenna so that the electric fields can be seen.

Sorry, I must've missed that detail. How do I scale it to V/m and A/m then? I'm looking for some peak and average values so I can estimate the copper losses.
Todd

Here: http://meepunits.wikia.com/wiki/Meep_unit_transformation_Wiki

With that and knowing that my scale factor, a, is 0.3, take your best shot.
And by comparing the data from the two runs, you may have a shot. The Perfect metal is lossless as I understand it. Everything that I set in the control file were identical so, except for adjustments that Meep may make internally (I don't know about anything in particular) the runs data should be comparable.

Okay, so if "a = 0.3", then the numbers shown should be divided by 0.3 to give 1/m, or one unit per meter. Then we need to compare runs as;

1. Set an input signal level, make a run.
2. Increase signal level by 5%, make a run and see how the amplitudes were affected. Did they increase 5%, or did resonance amplify the signal?

It also says there is another free variable, the current "I" has to be defined, or the mass "m". One determines the other in this case. If we define I = 1, then we're all good. Easier to scale than if we set m = 1. Can you confirm that you have I = 1?
Todd

Offline VAXHeadroom

  • Full Member
  • *
  • Posts: 195
  • Whereever you go, there you are. -- BB
  • Baltimore MD
  • Liked: 256
  • Likes Given: 151
The higher density views are uploaded.
https://drive.google.com/folderview?id=0B1XizxEfB23tfkVzeXVub2NpTm5fanZTTTdrLXNiT3VHaV9FYnB6TVpDUmJsWjRQbEUwdE0&usp=sharing

These Meep runs were made at resolution = 250 which is 2.5 times higher than previously uploaded views. These are the 14 final time slices of a 32 cycle run. 14 images for each view, separated by 0.1 cycle of the drive frequency 2.45 GHz. The Gaussian noise bandwidth of 2.45 GHz * .025 was used which seems reasonable for a magnetron. The 58 mm dipole antenna was located parallel to and 1/4 wavelength from the small end plate, excited with the Ez field component.

Included are two models of the 10.2 inch NSF-1701 cavity, one using copper and the other using Perfect metal.

I expect questions.

Looking forward to seeing somebody make movies from these higher density runs !

On it.

EY-EZ video -  others coming ... tonight if I don't fall asleep waiting for the videos to render  ::)



Edit: Added 11:30PM EDT

EX-HZ video - HY/HZ will have to wait til the morning



« Last Edit: 07/02/2015 03:30 AM by VAXHeadroom »
Emory Stagmer
  Executive Producer, Public Speaker UnTied Music - www.untiedmusic.com

Offline OttO

  • Member
  • Posts: 81
  • France
  • Liked: 91
  • Likes Given: 11
Something interesting:

Surface plasmon hurdles leading to a strongly localized giant field enhancement on two-dimensional (2D) metallic diffraction gratings
https://www.osapublishing.org/oe/fulltext.cfm?uri=oe-23-7-9167&id=314369

"An extensive numerical study of diffraction of a plane monochromatic wave by a single gold cone on a plane gold substrate and by a periodical array of such cones shows formation of curls in the map of the Poynting vector....
Arranging the cones in a two-dimensional subwavelength periodic array (diffraction grating), supporting a specular reflected order only, resonantly strengthens the field intensity at the tip of cones and leads to a field intensity enhancement of the order of 10 000 with respect to the incident wave intensity."
« Last Edit: 07/02/2015 01:18 PM by OttO »

Offline deuteragenie

  • Member
  • Posts: 71
  • Germany
  • Liked: 22
  • Likes Given: 0
...
Yes, a very big help from zxcvb, Thank you again, and welcome!

Aero, what is the output power of the antenna in the simulation? The numbers are much larger than they were, but they are still much, much lower than I would expect under a resonant condition.
Todd

Todd, are you taking into account that these fields are in Meep dimensionless units ( http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#Units_in_Meep )? (not normal SI dimensional units)

Also, the value of the field at the antenna is higher than shown, I had to clip the values of the antenna so that the electric fields can be seen.

Sorry, I must've missed that detail. How do I scale it to V/m and A/m then? I'm looking for some peak and average values so I can estimate the copper losses.
Todd

Here: http://meepunits.wikia.com/wiki/Meep_unit_transformation_Wiki

With that and knowing that my scale factor, a, is 0.3, take your best shot.

So the electric field has to be multiplied by to get Volts/meter

and H is already in Amps/meter  (H does not need any transformation)

where does the scale factor, a=0.3, enter the picture?

Into the denominator , which is in meters?

So really to get the Electric field one has to multiply by 3767.3/3 = 1255.77  Volts/meter

and the Magnetic field H has to be multiplied by 1/0.3 = 3.33333   Amps/meter


I think aero adjusted the time, the distances and the frequency with the scale factor (0.3). 
One therefore needs to be very careful with conversions to SI.

Offline arc

  • Member
  • Posts: 20
  • In port for a few weeks
  • Liked: 10
  • Likes Given: 32
Confirmation for @Rodal's previous graphs .  I also have decoded the csv files and the results are the same.  I wont bore people here with duplicating the images.

I also see what you meant regarding antenna height (values) sticking up in the wave plot images. Its height values overrode the wave size about 20:1 so excluding antenna plot values {Z axis} above the wave tops was necessary to increase wave signals proportional to background.

Also Thank you and well done to Aero for all his hard work, it is appreciated.

Only other comment is... I do like that Alcubierre like "imaginary space warp" graph. Can I ask you Mr Rodal to create a "Large version" of it that can be accessed/downloaded somewhere, Your program produces much better coloring than the one I use here, the image would be a good talking point for some of the students.
« Last Edit: 07/03/2015 03:24 AM by arc »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
...
I think aero adjusted the time, the distances and the frequency with the scale factor (0.3). 
One therefore needs to be very careful with conversions to SI.

Quote from: Mark Twain
Be careful about reading health books. You may die of a misprint.

Taking everything into account, the transformation factor for the Electric Field becomes so that, as remarked by Todd, what one needs to know is whether aero used Io=1 or some other value.  The value of a is automatically taken into account in the denominator of

Ditto for the magnetizing field H, where the transformation factor is : 

We need aero to disclose what value he used for Io.
« Last Edit: 07/02/2015 11:43 AM by Rodal »

Offline deuteragenie

  • Member
  • Posts: 71
  • Germany
  • Liked: 22
  • Likes Given: 0
...
I think aero adjusted the time, the distances and the frequency with the scale factor (0.3). 
One therefore needs to be very careful with conversions to SI.

Quote from: Mark Twain
Be careful about reading health books. You may die of a misprint.

Taking everything into account, the transformation factor for the Electric Field becomes so that, as remarked by Todd, what one needs to know is whether aero used Io=1 or some other value.  The value of a is automatically taken into account in the denominator of

Ditto for the magnetizing field H, where the transformation factor is : 

We need aero to disclose what value he used for Io.

From what I understand it should be 1 indeed, but aero will for sure confirm after he finishes his beers (plural !) because by now he has already updated his Bash script to pack all CSVs and PNG files into one zip which is subsequently automagically uploaded.
 
No progress yesterday evening on the HDF5 compilation/linking, which is the last remaining bit before I am able to recompile the full latest Meep version on Ubuntu.  Unfortunately, the HDF5 libs seem to be badly designed (require separate compile for MPIs etc.) and to make things worse they have changed the APIs in recent versions. I'll try with a chainsaw this evening to see what gives.

Quote from:  myself
All mushrooms can be eaten; some of them will kill you.
« Last Edit: 07/02/2015 12:18 PM by deuteragenie »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
The higher density views are uploaded.
https://drive.google.com/folderview?id=0B1XizxEfB23tfkVzeXVub2NpTm5fanZTTTdrLXNiT3VHaV9FYnB6TVpDUmJsWjRQbEUwdE0&usp=sharing

These Meep runs were made at resolution = 250 which is 2.5 times higher than previously uploaded views. These are the 14 final time slices of a 32 cycle run. 14 images for each view, separated by 0.1 cycle of the drive frequency 2.45 GHz. The Gaussian noise bandwidth of 2.45 GHz * .025 was used which seems reasonable for a magnetron. The 58 mm dipole antenna was located parallel to and 1/4 wavelength from the small end plate, excited with the Ez field component.

Included are two models of the 10.2 inch NSF-1701 cavity, one using copper and the other using Perfect metal.

I expect questions.

Trying to understand the terminology being used.  In traditional Finite Difference terminology (von Neumann, Ricthmyer, Lax, Hildebrand, and other classic authors, etc.) the terms "finite difference gridpoint" or  "finite difference meshpoint" and "finite difference grid" or "finite difference mesh" is used to characterize the geometrical points at which the Finite Difference scheme is implemented.  Von Neumann, Lax, etc., did not use the terminology "pixel" or "resolution" to refer to the Finite Difference mesh.

In some of the Meep literature (http://ab-initio.mit.edu/wiki/index.php/Meep_Reference) I see graphic display terms being used like "pixel" instead of "grid point".  And now the term "resolution" as in "display resolution".

I would rather use the traditional, mathematical terms rather than these graphic terms, when referring to the Finite Difference grid (or "mesh") because using graphic terms is apt for confusion.  Confusion between the graphic display post-processing of Finite Difference solutions and the Finite Difference mesh used to obtain the solution.

For example, let's say that we have a Finite Difference mesh in the x y plane with 100 FD grid points in both the x and y directions, equally spaced.

For graphic displays one has many alternatives to display the FD solution, for example:

1) display the field variables at every FD grid point (giving 100x100 pixels)
2) display the field variables at every other FD gridpoint (giving 50x50 pixels)
3) use spline (or other form of interpolaton) to display variables in between FD gridpoints in addition to displaying field variables at the gridpoints (giving much more than >>100x100 pixels)

which shows that one should distinguish between the graphic display pixels and resolution and the Finite Difference mesh, because they are not necessarily the same.  And in my experience with codes they have not been the same.

And this distinction is particularly important when the discussion, like here, gyrates around graphic displays.

pixel = the smallest element of an image that can be individually processed in a video display system.

QUESTION:  is what is being displayed in the Meep graphs obtained at every FD grid point, and that's why display terminology is used instead of FD classic terminology?
« Last Edit: 07/02/2015 02:49 PM by Rodal »

Offline deuteragenie

  • Member
  • Posts: 71
  • Germany
  • Liked: 22
  • Likes Given: 0
The higher density views are uploaded.
https://drive.google.com/folderview?id=0B1XizxEfB23tfkVzeXVub2NpTm5fanZTTTdrLXNiT3VHaV9FYnB6TVpDUmJsWjRQbEUwdE0&usp=sharing

These Meep runs were made at resolution = 250 which is 2.5 times higher than previously uploaded views. These are the 14 final time slices of a 32 cycle run. 14 images for each view, separated by 0.1 cycle of the drive frequency 2.45 GHz. The Gaussian noise bandwidth of 2.45 GHz * .025 was used which seems reasonable for a magnetron. The 58 mm dipole antenna was located parallel to and 1/4 wavelength from the small end plate, excited with the Ez field component.

Included are two models of the 10.2 inch NSF-1701 cavity, one using copper and the other using Perfect metal.

I expect questions.

Trying to understand the terminology being used.  In traditional Finite Difference terminology (von Neumann, Ricthmyer, Lax, Hildebrand, and other classic authors, etc.) the terms "finite difference gridpoint" or  "finite difference meshpoint" and "finite difference grid" or "finite difference mesh" is used to characterize the geometrical points at which the Finite Difference scheme is implemented.  Von Neumann, Lax, etc., did not use the terminology "pixel" or "resolution" to refer to the Finite Difference mesh.

In some of the Meep literature I see graphic display terms being used like "pixel" instead of "grid point".  And now the term "resolution" as in "display resolution".

I would rather use the traditional, mathematical terms rather than these graphic terms, because using graphic terms is apt for confusion.  Confusion between the graphic display post-processing of Finite Difference solutions and the Finite Difference mesh used to obtain the solution.

For example, let's say that we have a Finite Difference mesh in the x y plane with 100 FD grid points in both the x and y directions, equally spaced.

For graphic displays one has many alternatives to display the FD solution, for example:

1) display the field variables at every FD grid point (giving 100x100 pixels)
2) display the field variables at every other FD gridpoint (giving 50x50 pixels)
3) use spline (or other form of interpolaton) to display variables in between FD gridpoints in addition to displaying field variables at the gridpoints (giving much more than >>100x100 pixels)

which shows that one should distinguish between the graphic display pixels and resolution and the Finite Difference mesh, because they are not necessarily the same.  And in my experience with codes they have not been the same.

QUESTION:  is what is being displayed in the Meep graphs obtained at every FD grid point, and that's why display terminology is used instead of FD classic terminology?

Lattice (http://ab-initio.mit.edu/wiki/index.php/Meep_Reference#lattice) specifies the size of the computational cell, whereas resolution specifies the computational grid resolution, in pixels per distance unit. 
If the lattice is 100 x 100 and the resolution 10, it results in 1000 x 1000 pixel images.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
...

Lattice (http://ab-initio.mit.edu/wiki/index.php/Meep_Reference#lattice) specifies the size of the computational cell, whereas resolution specifies the computational grid resolution, in pixels per distance unit. 
If the lattice is 100 x 100 and the resolution 10, it results in 1000 x 1000 pixel images.

They are using different terminology than used by von Neuman (https://en.wikipedia.org/wiki/John_von_Neumann), Lax (https://en.wikipedia.org/wiki/Peter_Lax), Courant (https://en.wikipedia.org/wiki/Richard_Courant), Richtmyer (https://en.wikipedia.org/wiki/Robert_D._Richtmyer ), Hildebrand (https://en.wikipedia.org/wiki/Francis_B._Hildebrand) and other giants in Finite Difference Method.

Is pixel used only in its proper form (graphic display):

pixel = the smallest element of an image that can be individually processed in a video display system.

Or is pixel being used in some of the Meep literature to mean "Finite Difference Grid Point (or Mesh point)" ?
which would be the wrong use of the word pixel, since pixel is a graphic term that should not be confused with the FD gridpoint, since they are not necessarily the same
« Last Edit: 07/02/2015 03:01 PM by Rodal »

Offline deuteragenie

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

Lattice (http://ab-initio.mit.edu/wiki/index.php/Meep_Reference#lattice) specifies the size of the computational cell, whereas resolution specifies the computational grid resolution, in pixels per distance unit. 
If the lattice is 100 x 100 and the resolution 10, it results in 1000 x 1000 pixel images.

But is pixel used only in its proper form:

pixel = the smallest element of an image that can be individually processed in a video display system.

Or is it being used in some of the Meep literature to mean "Finite Difference Grid Point (or Mesh point)" ?
which would be the wrong use of the word pixel, since pixel is a graphic term that should not be confused with the FD gridpoint, since they are not necessarily the same

From reading the manuals, it appears that it is used adequately id est to represent one graphical unit.  Maybe what is causing confusion is that the lattice could be 1 x 1, and resolution 100, giving images 100 x 100 pixels, which depending on the distance of 1 lattice unit can represent 1 mm, 1 meter or 1 km.


Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
..

From reading the manuals, it appears that it is used adequately id est to represent one graphical unit.  Maybe what is causing confusion is that the lattice could be 1 x 1, and resolution 100, giving images 100 x 100 pixels, which depending on the distance of 1 lattice unit can represent 1 mm, 1 meter or 1 km.

Don't agree.  For example, a flagrant bad use of the word pixel in the Meep ab initio:

Quote
For CYLINDRICAL simulations with |m| > 1, compute more accurate fields near the origin r = 0 at the expense of requiring a smaller Courant factor. Empirically, when this option is set to true, a Courant factor of roughly min[0.5,1 / ( | m | + 0.5)] (or smaller) seems to be needed. The default is false, in which case the Dr, Dz, and Br fields within |m| pixels of the origin are forced to zero, which usually ensures stability with the default Courant factor of 0.5, at the expense of slowing convergence of the fields near r = 0. 

The use of the word pixel there, in conjunction with the Courant factor is wrong, it should have read Finite Difference grid point.

Courant is turning on his grave, unhappy at this use: 

but (Courant's student) Anneli Lax (https://en.wikipedia.org/wiki/Anneli_Cahn_Lax) does not seem to worry about it (probably because "pixel" is a convenient -if double-meaning- short for the much longer "Finite Difference Grid Point") :-)

She said:  <<''There is a misconception among people and schoolchildren about the nature of mathematics,'' she said in a 1979 interview. ''They consider it a matter of rules and regulations instead of thinking.''>>


« Last Edit: 07/02/2015 03:24 PM by Rodal »

Offline aero

  • Senior Member
  • *****
  • Posts: 2785
  • 92129
  • Liked: 724
  • Likes Given: 249

snip ...

Here: http://meepunits.wikia.com/wiki/Meep_unit_transformation_Wiki

With that and knowing that my scale factor, a, is 0.3, take your best shot.

So the electric field has to be multiplied by to get Volts/meter

and H is already in Amps/meter  (H does not need any transformation)

where does the scale factor, a=0.3, enter the picture?

Into the denominator , which is in meters?

So really to get the Electric field one has to multiply by 3767.3/3 = 1255.77  Volts/meter

and the Magnetic field H has to be multiplied by 1/0.3 = 3.33333   Amps/meter


I think aero adjusted the time, the distances and the frequency with the scale factor (0.3). 
One therefore needs to be very careful with conversions to SI.

Well - not the time, Meep somehow knows what the time is. Length and frequency, yes. In SI units, the lengths and frequency are as given by the NSF-1701 cavity.
Retired, working interesting problems

Offline aero

  • Senior Member
  • *****
  • Posts: 2785
  • 92129
  • Liked: 724
  • Likes Given: 249
As I understand and use the term "pixel" it means Finite Difference Grid Point when referring to Meep calculations and data. Meep stores an array of data at each pixel. It outputs one number from that array to define the field component requested.  I don't know, but I expect the usage was accepted during the interminable meetings at MIT that are part and parcel of DoD, Navy and DARPA funded developments.

H5topng converts that single data item defining the output field component at each Finite Difference Grid Point into an actual image pixel. It took finding the word "pixel" strangely used in the documentation a few times before I came to this understanding. People who use Meep use the term as above. I don't know the usage of people who use other FDTD programs or FD schemes but I do know that "Finite Difference Grid Point" is safe usage. And pixel is safe usage among Meep users.

I don't know what the value of Io is but I speculate that it is 1. I do not adjust the Meep default currents in any way.

And by the way, I don't drink... and don't miss it (very much).
« Last Edit: 07/02/2015 04:02 PM by aero »
Retired, working interesting problems

Offline VAXHeadroom

  • Full Member
  • *
  • Posts: 195
  • Whereever you go, there you are. -- BB
  • Baltimore MD
  • Liked: 256
  • Likes Given: 151
The higher density views are uploaded.
https://drive.google.com/folderview?id=0B1XizxEfB23tfkVzeXVub2NpTm5fanZTTTdrLXNiT3VHaV9FYnB6TVpDUmJsWjRQbEUwdE0&usp=sharing

These Meep runs were made at resolution = 250 which is 2.5 times higher than previously uploaded views. These are the 14 final time slices of a 32 cycle run. 14 images for each view, separated by 0.1 cycle of the drive frequency 2.45 GHz. The Gaussian noise bandwidth of 2.45 GHz * .025 was used which seems reasonable for a magnetron. The 58 mm dipole antenna was located parallel to and 1/4 wavelength from the small end plate, excited with the Ez field component.

Included are two models of the 10.2 inch NSF-1701 cavity, one using copper and the other using Perfect metal.

I expect questions.

Looking forward to seeing somebody make movies from these higher density runs !

On it.

EY-EZ video



EX-HX video



HY-HZ video


« Last Edit: 07/02/2015 04:43 PM by VAXHeadroom »
Emory Stagmer
  Executive Producer, Public Speaker UnTied Music - www.untiedmusic.com

Offline kitsuac

  • Member
  • Posts: 10
  • NW Ohio
  • Liked: 9
  • Likes Given: 4
I took a quick look through the MEEP source code and didn't see any assembly optimizations. By identifying bottlenecks and hand writing SIMD assembly routines, you can often improve performance on the order of several hundred percent (it's my day job). If you folks have a set of representative input data, I'll try to take a look with a profiler in search of low-hanging optimization fruit.

Tags: