Author Topic: EM Drive Developments - related to space flight applications - Thread 4  (Read 1245626 times)

Offline X_RaY

  • Full Member
  • ****
  • Posts: 852
  • Germany
  • Liked: 1146
  • Likes Given: 2479
Quote
The smaller antenna

Are you referring to the antenna with circumference of 1 wavelength?
No. Circumference Lambda/4 (or Lambda/3 like Shell said).

EDIT: The blue and red regions in your simulations corresponds to the direction of the H field (remember it's the Hz component). The TE01p mode is high symmetrically at each point around a single circumference, around the middle axis. The full loop has to have only one color for a ~half wavelength ( complete red, after that complete blue and so on) for this type of mode.

At least i am not 100 percent sure with that. Please try it. May be i am wrong  :-\ . At this point your help with the simulations is so important!!
« Last Edit: 09/04/2015 09:04 pm by X_RaY »

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360
An artefact of my implementation is that as the loop gets smaller the number of point sources making up the loop are reduced.  It is still one point for every meep time step but it takes a lot fewer time steps to walk around the smaller loop. I don't know if that will have an adverse effect or not. It does affect the amplitude of the signal.
Retired, working interesting problems

Offline X_RaY

  • Full Member
  • ****
  • Posts: 852
  • Germany
  • Liked: 1146
  • Likes Given: 2479
An artefact of my implementation is that as the loop gets smaller the number of point sources making up the loop are reduced.  It is still one point for every meep time step but it takes a lot fewer time steps to walk around the smaller loop. I don't know if that will have an adverse effect or not. It does affect the amplitude of the signal.
Yes smaller loops gives lower amplitudes(also true in the real world, lower coupling...), but the wrong pattern at the loop could lead to a wrong field pattern means wrong mode..

On the other hand (for most geometries of such a resonator) there is only one mode possible, theoretically.

I don't believe that this is an artefact, half the number of points means you will see only one color for each half cycle. (Is that right??)
How it looks like for a lambda/4 circumference of the loop? Can you upload one or two full cycles? Myself or someone else will make a gif based on it :)
« Last Edit: 09/04/2015 09:42 pm by X_RaY »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2442
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 3186
  • Likes Given: 2708
An artefact of my implementation is that as the loop gets smaller the number of point sources making up the loop are reduced.  It is still one point for every meep time step but it takes a lot fewer time steps to walk around the smaller loop. I don't know if that will have an adverse effect or not. It does affect the amplitude of the signal.
Yes smaller loops gives lower amplitudes(also true in the real world, lower coupling...), but the wrong pattern at the loop could lead to a wrong field pattern means wrong mode..

On the other hand (for most geometries of such a resonator) there is only one mode possible, theoretically.

I don't believe that this is an artefact, half the number of points means you will see only one color for each half cycle. (Is that right??)
How it looks like for a lambda/4 circumference of the loop? Can you upload one or two full cycles? Myself or someone else will make a gif based on it :)
In for a minute then have to dash off. A friend needs a lift her car broke down.

The wavelength of 2.5GHz (resonate frequency determined by harmiv and meep) for CE is 120mm. So, you want to try a lambda /4 of 30mm of loop circumference?  Direct multiples of the wavelength  1/2, 1/4, 1/8... will lead to excitement of other TMxx modes I do believe instead of the TExx That's why I was thinking of a 1/3 mode. But, we are in uncharted territory and aero is in the drivers seat. Speaking of driving... got to go. BBL.

Shell

PS: The wire faraday cage looks and tests out very well! I will be working late tonight it seems....

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360
I've been making resonance runs for this antenna with the circumferences mentioned and see two interesting differences in the numbers from the dipole antennas used previously.

First though, the resonant frequency seems to be the same 2.50 GHz calculated before. More exactly, this

2.50021516E+009,
2.49990486E+009,
2.49989601E+009

for three runs.

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?

2 ) And this is interesting. When the drive frequency is less than the resonance frequency, Q does not suffer very much (20%) but the imaginary part of the complex amplitude is huge compared to the real part. This characteristic goes away when driving the cavity at the Harminv calculated resonant frequency. I don't recall what this means but I think I recall that the large imaginary component was related to something important. Dispersion, maybe?

Anyway, I can organize and post those numbers if there is interest.
Retired, working interesting problems

Online Eer

  • Full Member
  • ****
  • Posts: 625
  • Liked: 469
  • Likes Given: 913
New meep simulations source link, below.

I've shared my folder meeper-files, which is laid out as follows:

meeper-files/
   CE2-0009-150904/                                             - all files associated with this run in this folder
      cvs-directory/                                                   - the cvs files from the Continuous run
          400+ files
      CE2-8Ey-dual_dipoles-250-csv.log                     - log file from running CE2-csv.sh
      CE2-dual_dipoles-250res.ctl                              - control file for the Continuous run at 250 resolution
      CE2-dual_dipoles-C-250res.log                          - log file from Continuous run
      CE2-dual_dipoles-G.ctl                                      - control file for the Gaussian run at 100 resolution
      CE2-dual_dipoles-G.log                                     - log file from Gaussian run
      meep-data-description-CE-2r9-32cy.txt             - my take at the data description file
  scripts/
      CE2-csv.sh                                                       - my current version of the parameterized script file

Note that in the future, there will be better agreement between run names (CE2r9), directory names (CE2-0009-150904), control file names (CE2-dual_dipoles-xxx), and log file names.  Sorry about that - it's been an evolution.  I can recreate the files with the CE2-0009 common prefix, if you'd like.

Link to meeper-files folder is https://drive.google.com/folderview?id=0B527OOY4hxdZfldTN2FoVm5SRDZ2MHFJYmhaM2ZFcXVEeklpd3NnTy1RUUtnS3d1YllCWGc&usp=sharing

Folder is shared public for viewing only.

Hope you'll find them useful.

Edit - correct root folder name to meeper-files.
« Last Edit: 09/05/2015 03:43 am by Eer »
From "The Rhetoric of Interstellar Flight", by Paul Gilster, March 10, 2011: We’ll build a future in space one dogged step at a time, and when asked how long humanity will struggle before reaching the stars, we’ll respond, “As long as it takes.”

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
I've been making resonance runs for this antenna with the circumferences mentioned and see two interesting differences in the numbers from the dipole antennas used previously.

First though, the resonant frequency seems to be the same 2.50 GHz calculated before. More exactly, this

2.50021516E+009,
2.49990486E+009,
2.49989601E+009

for three runs.

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?

2 ) And this is interesting. When the drive frequency is less than the resonance frequency, Q does not suffer very much (20%) but the imaginary part of the complex amplitude is huge compared to the real part. This characteristic goes away when driving the cavity at the Harminv calculated resonant frequency. I don't recall what this means but I think I recall that the large imaginary component was related to something important. Dispersion, maybe?

Anyway, I can organize and post those numbers if there is interest.

Several things appear problematic above.  Since I'm short of time, I'll just pick this one:

<<When the drive frequency is less than the resonance frequency, Q does not suffer very much (20%) but the imaginary part of the complex amplitude is huge compared to the real part.>>

This appears contradictory, because Q is calculated by Harminv and only by Harminv.  It is calculated by Harminv as the half ratio of the real part to the imaginary part of the frequency, so if the imaginary part would increase by a "huge" amount with respect to the real real part of the frequency, then the calculated Q by Harminv should decrease in ("huge") inverse proportion to the huge increase in  the imaginary part of the frequency.

How can one divorce the calculated Q from the imaginary frequency, when the Q is calculated based on the ratio of the real to the imaginary parts of the frequency?

The calculated Q is conjoined with the ratio of the real part to the imaginary part of the frequency.  It does not make sense that these conjoined twins would be separated all of a sudden.

Only you have seen the Q and the real and imaginary parts of the frequency you are mentioning, but since you are posting, I guess that you were looking for some comment.
« Last Edit: 09/05/2015 02:25 pm by Rodal »

Offline Rodal

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

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?...
What component(s) were you using for the straight dipole antenna you used to excite TM modes in the initial runs?

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360
...

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?...
What component(s) were you using for the straight dipole antenna you used to excite TM modes in the initial runs?

I used each of them but only one at a time. A single component exciting the cavity for each of 6 runs. All 6 Q values were higher than the Q I'm getting with the loop. One by only a small factor (1.5) and the rest by orders of magnitude. It may be that I am not detecting the right mode, using Hx and keep forgetting to change it. Hx is a TM mode component. But without your question, that possibility wouldn't have occurred to me for days. I'll check now.

Regarding the huge Imaginary component, I'm talking about amplitude, not frequency. I agree, for frequency that couldn't happen.
Retired, working interesting problems

Offline Rodal

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

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?...
What component(s) were you using for the straight dipole antenna you used to excite TM modes in the initial runs?

I used each of them but only one at a time. A single component exciting the cavity for each of 6 runs. All 6 Q values were higher than the Q I'm getting with the loop. One by only a small factor (1.5) and the rest by orders of magnitude. It may be that I am not detecting the right mode, using Hx and keep forgetting to change it. Hx is a TM mode component. But without your question, that possibility wouldn't have occurred to me for days. I'll check now.

Regarding the huge Imaginary component, I'm talking about amplitude, not frequency. I agree, for frequency that couldn't happen.

So, if you used only one component before, why use 3 components now?

I can see using 2 components for a circle instead of 1 component for a straight dipole aligned along one of the Cartesian axes.  To define a component at an angle to the Cartesian axes you need the 2 components in the plane of the circle, but I have no idea as to why you are using a 3rd component, which is perpendicular to that plane.

Offline SteveD

  • Full Member
  • ***
  • Posts: 312
  • United States
  • Liked: 83
  • Likes Given: 10
Purely as a practical matter, what would be the effect of putting an unpowered dipole at the center of the circle so that this could be fired in lower Q but interesting circle mode or higher Q but not TE12 central antenna mode?

Offline Ricvil

  • Full Member
  • *
  • Posts: 171
  • Liked: 110
  • Likes Given: 71
http://arxiv.org/abs/0810.2136

http://arxiv.org/pdf/1501.04586

http://arxiv.org/pdf/1507.01807


Interesting articles.

It's about  "Hall effect of light", berry phases, lorentz boost, and "anomalous velocity drift"
A little of "PT symmetry breaking" to force a prevalence of a determined helical mode, and perhaps a "thrust" may be explained.

« Last Edit: 09/05/2015 02:42 am by Ricvil »

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360
...

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?...
What component(s) were you using for the straight dipole antenna you used to excite TM modes in the initial runs?

I used each of them but only one at a time. A single component exciting the cavity for each of 6 runs. All 6 Q values were higher than the Q I'm getting with the loop. One by only a small factor (1.5) and the rest by orders of magnitude. It may be that I am not detecting the right mode, using Hx and keep forgetting to change it. Hx is a TM mode component. But without your question, that possibility wouldn't have occurred to me for days. I'll check now.

Regarding the huge Imaginary component, I'm talking about amplitude, not frequency. I agree, for frequency that couldn't happen.

So, if you used only one component before, why use 3 components now?

I can see using 2 components for a circle instead of 1 component for a straight dipole aligned along one of the Cartesian axes.  To define a component at an angle to the Cartesian axes you need the 2 components in the plane of the circle, but I have no idea as to why you are using a 3rd component, which is perpendicular to that plane.

Well, no better idea, and anyway that was left over from debugging. Both TE and TM have 3 components, that's why the capability is built in. I have done as you suggested and am now making a run with only Ex and Ey excitation. Those two components are in the TE mode which is what I am hoping to excite.

I did change the Harminv detection location and component to Ey located 1/4 wavelength from the small end and maybe 70% toward the wall from the zero axis. It has been very near the center of the cavity. (Where is the best place to locate the detector?) This change increased the calculated Q to 7800 which is up from 3722.  That is of course another indication of the futility of trying to adjust the copper model to get realistic calculated Q values. The calculated Q depends on several seemingly extraneous parameters not related to the material model.

Ok - run finished -  exciting with only Ex and Ey reduced Q to 6661 and reduced the detected resonant frequency by 10 kHz. And in case I didn't mention it, this is with a 1/4 wave length circumference loop, 1/4 wavelength from the big end and centered.
Retired, working interesting problems

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2205
  • Liked: 2713
  • Likes Given: 1134
Will you guys slow down!  ;) I'm on a short vacation away from the interwebs and man, its a lot of work to catch up with the conversation...actually, I'm happy to see the work-flow continue. - From steamy central Indiana for a couple of more days.

Offline X_RaY

  • Full Member
  • ****
  • Posts: 852
  • Germany
  • Liked: 1146
  • Likes Given: 2479
...

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?...
What component(s) were you using for the straight dipole antenna you used to excite TM modes in the initial runs?

I used each of them but only one at a time. A single component exciting the cavity for each of 6 runs. All 6 Q values were higher than the Q I'm getting with the loop. One by only a small factor (1.5) and the rest by orders of magnitude. It may be that I am not detecting the right mode, using Hx and keep forgetting to change it. Hx is a TM mode component. But without your question, that possibility wouldn't have occurred to me for days. I'll check now.

Regarding the huge Imaginary component, I'm talking about amplitude, not frequency. I agree, for frequency that couldn't happen.

So, if you used only one component before, why use 3 components now?

I can see using 2 components for a circle instead of 1 component for a straight dipole aligned along one of the Cartesian axes.  To define a component at an angle to the Cartesian axes you need the 2 components in the plane of the circle, but I have no idea as to why you are using a 3rd component, which is perpendicular to that plane.

Well, no better idea, and anyway that was left over from debugging. Both TE and TM have 3 components, that's why the capability is built in. I have done as you suggested and am now making a run with only Ex and Ey excitation. Those two components are in the TE mode which is what I am hoping to excite.

I did change the Harminv detection location and component to Ey located 1/4 wavelength from the small end and maybe 70% toward the wall from the zero axis. It has been very near the center of the cavity. (Where is the best place to locate the detector?) This change increased the calculated Q to 7800 which is up from 3722.  That is of course another indication of the futility of trying to adjust the copper model to get realistic calculated Q values. The calculated Q depends on several seemingly extraneous parameters not related to the material model.

Ok - run finished -  exciting with only Ex and Ey reduced Q to 6661 and reduced the detected resonant frequency by 10 kHz. And in case I didn't mention it, this is with a 1/4 wave length circumference loop, 1/4 wavelength from the big end and centered.

Try to use a H(z) component detector (if possible) at or close too the central axis, lambda/4 distance to the smaller end of the frustum. The magnetic field of TE012 would be strongest there.
For that mode Ex,Ey,Ez at the central axis and lambda/4 away from the plate are negligible/ very small.

« Last Edit: 09/05/2015 07:05 am by X_RaY »

Offline Mezzenile

  • Full Member
  • *
  • Posts: 130
  • Liked: 63
  • Likes Given: 24
Quote from: flux_capacitor link=topic=38203.msg1422976#msg1422976 date=1441395215
[hr
The text by Woodward you refer to was cowritten with Paul March:
Woodward, James F.; Mahood, Thomas L.; March, Paul (July 2001). "Rapid Spacetime Transport and Machian Mass Fluctuations: Theory and Experiment". JPC 2001 Proceedings. 37th AIAA/ASME Joint Propulsion Conference, Salt Lake City, Utah. American Institute of Aeronautics and Astronautics. doi:10.2514/6.2001-3907

Most importantly, it was written in 2001, while Puthof's Polarizable Vacuum theory was only a few months old. Nowadays Woodward does not think PV is realistic anymore, nor any ZPF theory (including McCulloch's MiHsC or White's QVF conjecture) to explain Mach effects. He is very clear about that when asked on that matter.
 
Woodward is even currently writing a paper demonstrating why virtual particles of the vacuum can not be used for propulsion.
Sorry for the missing of the other authors of the article but as you have seen in the version retrieved on the WEB and that I have attached in my previous post, Woodward was the only cited author.
The fact that March is a co-author of this article is interesting as he is himself willing to found an explanation of the Shawyer's EMDrive in term of Woodward's work:
http://nextbigfuture.com/2009/09/mach-effect-interview-with-paul-march.html
Quote
Question: "What do you think of the EMdrive work?"
March's answer: The proposed E&M/SRT conjecture IMO is garbage. The experimental results is tantalizing, but it has to be repeated in a vacuum chamber to get rid of possible spurious error sources for the thrust signatures observed. If it still moves in a 1x10-4 Torr vacuum, then we have to explain what is going on in view of Jim's work.

Now I agree with you that the Puthof's Polarizable Vacuum theory cannot be a replacement for General Relativity and its endorsement of Mach (not March ! ;)) intuition on the origin of inertia. It was also; I would say; the view of Puthof himself who clearly wrote in 1999 that the Polarizable Vacuum theory was not yet in a positon to explain both gravitational radiation and frame-dragging effects (base of Mach ideas on inertia incorporation in General Relativity).
« Last Edit: 09/05/2015 07:31 am by Mezzenile »

Offline Mezzenile

  • Full Member
  • *
  • Posts: 130
  • Liked: 63
  • Likes Given: 24
An article of this year on the  Biefeld-Brown effect : "On the Anomalous Weight Losses of High Voltage Symmetrical Capacitors" : http://arxiv.org/pdf/1502.06915.pdf

Offline Rodal

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

1 ) The calculated quality factor is much lower whether excited by a magnetic or an electric source. Now, I am using all three components to excite the cavity (Ex, Ey, Ex) or (Hx, Hy, Hz) and maybe this is wrong. Does anyone know?...
What component(s) were you using for the straight dipole antenna you used to excite TM modes in the initial runs?

I used each of them but only one at a time. A single component exciting the cavity for each of 6 runs. All 6 Q values were higher than the Q I'm getting with the loop. One by only a small factor (1.5) and the rest by orders of magnitude. It may be that I am not detecting the right mode, using Hx and keep forgetting to change it. Hx is a TM mode component. But without your question, that possibility wouldn't have occurred to me for days. I'll check now.

Regarding the huge Imaginary component, I'm talking about amplitude, not frequency. I agree, for frequency that couldn't happen.

So, if you used only one component before, why use 3 components now?

I can see using 2 components for a circle instead of 1 component for a straight dipole aligned along one of the Cartesian axes.  To define a component at an angle to the Cartesian axes you need the 2 components in the plane of the circle, but I have no idea as to why you are using a 3rd component, which is perpendicular to that plane.

Well, no better idea, and anyway that was left over from debugging. Both TE and TM have 3 components, that's why the capability is built in. I have done as you suggested and am now making a run with only Ex and Ey excitation. Those two components are in the TE mode which is what I am hoping to excite.

I did change the Harminv detection location and component to Ey located 1/4 wavelength from the small end and maybe 70% toward the wall from the zero axis. It has been very near the center of the cavity. (Where is the best place to locate the detector?) This change increased the calculated Q to 7800 which is up from 3722.  That is of course another indication of the futility of trying to adjust the copper model to get realistic calculated Q values. The calculated Q depends on several seemingly extraneous parameters not related to the material model.

Ok - run finished -  exciting with only Ex and Ey reduced Q to 6661 and reduced the detected resonant frequency by 10 kHz. And in case I didn't mention it, this is with a 1/4 wave length circumference loop, 1/4 wavelength from the big end and centered.

A Q of 7800 with the Drude model constants you are using means that the cavity is NOT resonating much, as Meep resonance with this Drude model in the past has given Q's of millions (when the actual Q should have been in the tens of thousands).

What is the diameter of the loop you are using (diameter in meters) and what is the diameter of the cone at the location of your loop?
« Last Edit: 09/05/2015 12:56 pm by Rodal »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2442
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 3186
  • Likes Given: 2708
Trying to put more time into the build as I'm getting in the final hardware. Should have the waveguide to coax in today and splitters. Different laptop arriving today also that needs to be setup, the old one never did work and had to fight to get a refund, thank goodness I've been an ebayier for 14 years and have a good record.

Worked on the lasers yesterday getting the pinhole lens to work and tested the Faraday cage with the access ports for the beam, did very well. Laid out the ceramics for drilling for the center quartz rod today. Figured out the mounting for the first surface mirrors and got the hardware around to place them. Plus a bazzilion other things in verifying and profiling the test rig.

A good friend sent me a picture of a eco friendly wood power frustum, aero would you like to run this in your meep?

Will have drawings for you in a bit aero, less confusion with them I think.

Shell

Online Eer

  • Full Member
  • ****
  • Posts: 625
  • Liked: 469
  • Likes Given: 913
New meep simulations source link, below.

I've shared my folder meeper-files, which is laid out as follows:

meeper-files/
   CE2-0009-150904/                                             - all files associated with this run in this folder
      cvs-directory/                                                   - the cvs files from the Continuous run
          400+ files
      CE2-8Ey-dual_dipoles-250-csv.log                     - log file from running CE2-csv.sh
      CE2-dual_dipoles-250res.ctl                              - control file for the Continuous run at 250 resolution
      CE2-dual_dipoles-C-250res.log                          - log file from Continuous run
      CE2-dual_dipoles-G.ctl                                      - control file for the Gaussian run at 100 resolution
      CE2-dual_dipoles-G.log                                     - log file from Gaussian run
      meep-data-description-CE-2r9-32cy.txt             - my take at the data description file
  scripts/
      CE2-csv.sh                                                       - my current version of the parameterized script file

Note that in the future, there will be better agreement between run names (CE2r9), directory names (CE2-0009-150904), control file names (CE2-dual_dipoles-xxx), and log file names.  Sorry about that - it's been an evolution.  I can recreate the files with the CE2-0009 common prefix, if you'd like.

Link to meeper-files folder is https://drive.google.com/folderview?id=0B527OOY4hxdZfldTN2FoVm5SRDZ2MHFJYmhaM2ZFcXVEeklpd3NnTy1RUUtnS3d1YllCWGc&usp=sharing

Folder is shared public for viewing only.

Hope you'll find them useful.

Edit - correct root folder name to meeper-files.

I've added a png-directory to meeper-files/CE2-0009-150904 containing contour maps for each of the csv files. 

A word about scaling - these png files are each autoscaled.  I understand that may not be what we want.

I'm willing to work with folks who care to create png files that are scaled the way you want them, and to figure out how to automate the process.  There are two obvious choices - which is preferred?

1) autoscale each file, so differences in levels are clearly visible
2) fixed scale applied to all png files, so they can be readily and easily compared one to another.  This could mean that some files show very faint contours, if their range is much smaller than the fixed scale.

In the fixed scale choice, how would you like me to pick the scale?  The matplotlib library function contourf (which is what I'm using, so the images are filled, as opposed to just lines) can take an array of values for which contour lines should be generated - my question is:  how should I pick or generate that array of values?

Thanks,
From "The Rhetoric of Interstellar Flight", by Paul Gilster, March 10, 2011: We’ll build a future in space one dogged step at a time, and when asked how long humanity will struggle before reaching the stars, we’ll respond, “As long as it takes.”

Tags:
 

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