Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1804616 times)

Offline TheTraveller

You can do this by using stubs/bolts directly into the frustum (near the antenna for example), but well you have to measure for while tuning.

I went for the external 3 stub coax tuner because I don't want anything inside the frustum messing with the phase relationship between the waves, the end plates and the excitation antenna. Plus it would be a guess where they should be best placed and I'm trying to minimise unknowns / guessing.
« Last Edit: 08/01/2015 02:24 PM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Offline TheTraveller

TT did you ever run Yang's model in your spreadsheet as reported in the Wiki site? What did you see, inquiring minds want to know.

Shell

Must have missed that.

Simple to do. What are the dimensions, freq and mode? Or should just I use the dimensions and report on what I find?
« Last Edit: 08/01/2015 02:23 PM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Online Rodal

  • Senior Member
  • *****
  • Posts: 5841
  • USA
  • Liked: 5923
  • Likes Given: 5268
The main thing I'd like to know about Yang is whether anybody can verify her experimental data at her facility. Neutral parties, I mean.
I'd love to be the proverbial fly on the wall, my luck I'd buzz to close to the giant bug zapper called a frustum.

I'm hoping that the frustum I'm putting together matches the best guess of the group here. Nobody has taken the bull by the horns to try to at least validate their claims by building something close to their design. I honestly believe it needs to be addressed and I'm trying to do so as best as I can. So criticism is welcome.

TT did you ever run Yang's model in your spreadsheet as reported in the Wiki site? What did you see, inquiring minds want to know.

Shell

Shell,

Using your Computer Aided Drawing program to assess the following figure:  https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=37642.0;attach=1051616,

what

1) Cone half angle (angle between lateral side and perpendicular to base)

2) Ratio of internal Big Base diameter to internal Small Base diameter

3) Ratio of internal  Big Base diameter to internal Length perpendicular to bases

do you get ?

This would be helpful to better estimate dimensions.

Offline glennfish

  • Full Member
  • ***
  • Posts: 387
  • Liked: 281
  • Likes Given: 168
@zellerium: the "cardboard-like square part" is a thin piece made of mica which covers the end of the waveguide in a microwave oven. Mica (as well as white Teflon) is indeed transparent to microwaves. It lets the EM waves pass through it but protects the food from being impregnated by some undesirable substance that may be emitted from the magnetron cavity, like oil or metallic particles.

This is a thought to EmDrive DIYers: maybe it is a good idea to insert a Teflon or Mica sheet in the waveguide to protect the interior of the cavity from those substances, in the same manner as food is protected in the oven.

@Rodal, flux capacitor and all other experts:
I was thinking about the glimmer plate inside the microwave cooking oven and MW sputtering technique. Without such a plate how many metal ions would enter the cavity? Is it possible that the ions (with high velocity a.k. relativistic mass times rest mass) cause the trust while the EM-field inside the cavity give them preferred direction to a single end plate?

@flux capacitor
Thanks for surprising to Tajmar. ;D
Here one can see how important the peer review process really is!

Is there really no one with ideas about (blue text) ? ???

Yes, I had an idea, which was to "like" your post, which meant that I agree with you that this is another possible source of thrust.  :)
I know about your like :)
But the magnitude of this effect is not really clear for me at the moment, how many mass(particles/s will be emit) using a standard magnetron? That's what i ask for.

Maybe it's a little less or it's not? And with regard to thrust after 'Power off' in several experiments, i am not sure about the relevance.

I disagree with this as a possibility for reported thrust results, not that much on magnitude arguments but because, again, this is a closed device : a given significant thrust (say 20N) can be obtained either by a "high" mass flow at low velocity ( 2g/s*.01m/s=20N ) or by a low mass flow at high velocity typical of MW sputtering ( 20g/s*1km/s=20N ). 20g/s for 100s = 2mg, about .2mm3 worth of removed material at copper density. I must say I don't know exactly how much energetic and where (what material) would be accidentally sputtered by a magnetron, let's say it could go to a few hundreds of eV, velocities above 5km/s, removed material could stay below .1mm3 for some amount of time while still thrusting at 20g/s (notice that microwave oven magnetrons operated at nominal conditions don't wear near that fast, but they are not supposed to be driving a high Q cavity...).

So maybe stretching numbers a little bit, a sputtering magnetron could possibly thrust for many minutes without noticing wear or significantly altered performances. But, back to the comparison high flow low velocity vs low flow high velocity : both can impart same thrust, but this is assuming the "exhaust" is not yet interacting with the limits of the rigid closed system as a whole, since we are exhausting inside a box that is attached to the thruster ! For high velocity this occurs too soon.

Let's take such a reaction thruster attached to (and exhausting reaction mass into) a box of 1m span. With 20g/s mass flow 1km/s we do record an apparent net thrust overall (system wide) of 20N but only for 1ms. Then, assuming constant flow, this net thrust goes to 0 exactly : we are still sputtering and eroding material but the momentum gained at emission is exactly compensated by the opposite momentum lost at the wall where exhaust ends its course. Obviously trying to cheat by playing around with reorientations of path, magnetic mirrors and such, is futile (as any change of path of exhaust must recoil on what is causing the change of path, whatever the mean). Then when the process of emission stops, we would observe an apparent net thrust overall of -20N (opposite direction as for the power on) for 1ms again.

On the other hand, if a magician decided to play a trick, it could be relatively straightforward to send a flow of 2g/s at .01m/s (say, by pumping some fluid), that could impart an apparent net thrust (as observed from the outside) of 20N during 100s. This apparent thrust would be perfectly well behaved, it could accelerate the system if the system is free to accelerate, it could push against a spring and maintain a force while being static. This is Newtonian propellantless thrust in all its glory. But this is only momentum hidden by a container : after 100s, the box will pay back by thrusting -20N for 100s. Spatial extent 1m, secretly transferred mass 200g (offset of CoM will remain unnoticed if box weighs a few kgs). Advice to the magician : 1 minute of demonstration, then say that if it can thrust for 1 minute then surely it could also thrust for years on (only a matter of engineering and $). Make sure the device is no longer on the scale after 1 minute of demo, put it back in an IP box.

In an earlier message, I suggested that oxidation would differentially change the weight of a heated cone.  What I couldn't figure out is how there would be enough oxygen in a partial vacuum to have enough mass exchange to be measured.

The sputtering thought got me thinking about the kinetics of oxydation, which is definitely happening in Tajmar's experiment.

It turns out that you can generate force vectors during oxidation, and there are several studies of designing objects that move as a direct result of redox reactions.

see:  http://rspa.royalsocietypublishing.org/content/467/2130/1645

The narrow end of the cone would have a lower rate of oxidation than the broad end, simply from the differences in surface area, and there would be a definite force because of that difference.  The forces would increase as the temperature of the device increases.

Microwaves might or might not increase the redox rate, but probably would contribute to oxidation with any ambient water in the system.

Offline Flyby

  • Full Member
  • ***
  • Posts: 384
  • Belgium
  • Liked: 447
  • Likes Given: 47
what

1) Cone half angle (angle between lateral side and perpendicular to base)

2) Ratio of internal Big Base diameter to internal Small Base diameter

3) Ratio of internal  Big Base diameter to internal Length perpendicular to bases

do you get ?

This would be helpful to better estimate dimensions.

I thought I uploaded this a while ago, or did I forget? ???

not sure if I understand your questions correctly, but...

As I'm not sure about the actual measurements, the drawing is made with proportional dimensions, with the internal height set to 100%

1) 74.56

2) 100/ 67 (separately rescaled big diameter to 100% to find the % of small diameter)

3) 115.4 / 100

If you find it more convenient to have another dimension as 100% base value, please say so... it's only a few clicks away..
« Last Edit: 08/01/2015 02:46 PM by Flyby »

Offline X_RaY

  • Full Member
  • ****
  • Posts: 774
  • Germany
  • Liked: 993
  • Likes Given: 2150
@SeeShells and @Rodal,
The data you are referring to was something I generated after reading a post/paper by Dr. Rodal in which he discussed extending the Brady frustum, using the same big end diameter and the same taper angle, but extending until the small end diameter equalled 25% of the big end diameter. The views posted are for the start of the cavity, power on until some cycles later. We remarked at the time about the changing range of the signal causing color changes, but no one realized that I had discovered the first step in inventing the "flashy light thingee" used by the Men in Black.

I have used time today to re-run that model generating what has become rather standard upload data, both csv files and png views, here:

https://drive.google.com/folderview?id=0B1XizxEfB23tfkZVbi1MY2RQZmVkeEVHUmVfQkc3UEdlVkdOVXZENmFYbmg4czJUd1lqcDg&usp=sharing

Read the data request file, where I blamed Rodal for asking for the data. I copied some data from my meep run log into the file giving basic cavity and run information. I did not use run logs back in mid June but it is the same model so today's run log data should be the same as was ran back in  June. I don't recall where the 2.14 GHz center drive frequency number came from, perhaps it was Harminv.

In any case, I hope this data tells us something, and Dr. Rodal, the complete set of 14 time slice csv files are there so you now have the data to calculate stresses. I'd be interested in seeing the result. Looks to me like zero force on the small end, but appearances can be deceiving.

A short heads up on this run (NASA Brady et.al. geometry extended to near the apex of the cone):

1) Non-symmetric placement of antenna, very offset from middle

2) Amplitude of response is very low: not a good resonance.  It shows fractal numerical artifacts due to very low amplitude.

3) Forces at both ends are very low.  Force at small end close to zero.  Stress distribution at big end very low and very asymmetric.

Suggestion: may need to run again, at an excitation frequency for which we know (from other calculations) that there is a good resonant response and with an antenna located on the axis axi-symmetry. 

If run again, it may be better to do it based on Yang/Shell geometry extended to near the apex of the cone, to model the geometry she intends to test.
I have almost the same insertion geometry as them except have a monopole rather than a loop. I think my temp tests are confirming a bad match at this locale. Will know more this weekend.
The example was for TE01p and a dipole with different orientation. As far as i know you want to use TM212 correct? And you use a Monopole antenna direction from the magnetron. This is a complete different Situation.
But IMHO your penetration point and the orientation of the antenna is a good choice!
You use an unloaded cavity. If you dont put food in your MW oven, the magnetron will also heats up, thats why people dont do that. The device can be broken caused by overheating the magnetron.

BTW got found an interesting link
https://www.cst.com/Applications/Article/Magnetron-And-Microwave-Oven-Design-To-Solve-Wi-Fi-Interference-Issues

Offline X_RaY

  • Full Member
  • ****
  • Posts: 774
  • Germany
  • Liked: 993
  • Likes Given: 2150
You can do this by using stubs/bolts directly into the frustum (near the antenna for example), but well you have to measure for while tuning.

I went for the external 3 stub coax tuner because I don't want anything inside the frustum messing with the phase relationship between the waves, the end plates and the excitation antenna. Plus it would be a guess where they should be best placed and I'm trying to minimise unknowns / guessing.
You asked for possibilities. That was one ;) but with regard to the thrust measurements i can unterstand that you dont want to it this way
« Last Edit: 08/01/2015 03:18 PM by X_RaY »

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2165
  • Liked: 2681
  • Likes Given: 1124
@SeeShells and @Rodal,
The data you are referring to was something I generated after reading a post/paper by Dr. Rodal in which he discussed extending the Brady frustum, using the same big end diameter and the same taper angle, but extending until the small end diameter equalled 25% of the big end diameter. The views posted are for the start of the cavity, power on until some cycles later. We remarked at the time about the changing range of the signal causing color changes, but no one realized that I had discovered the first step in inventing the "flashy light thingee" used by the Men in Black.

I have used time today to re-run that model generating what has become rather standard upload data, both csv files and png views, here:

https://drive.google.com/folderview?id=0B1XizxEfB23tfkZVbi1MY2RQZmVkeEVHUmVfQkc3UEdlVkdOVXZENmFYbmg4czJUd1lqcDg&usp=sharing

Read the data request file, where I blamed Rodal for asking for the data. I copied some data from my meep run log into the file giving basic cavity and run information. I did not use run logs back in mid June but it is the same model so today's run log data should be the same as was ran back in  June. I don't recall where the 2.14 GHz center drive frequency number came from, perhaps it was Harminv.

In any case, I hope this data tells us something, and Dr. Rodal, the complete set of 14 time slice csv files are there so you now have the data to calculate stresses. I'd be interested in seeing the result. Looks to me like zero force on the small end, but appearances can be deceiving.

A short heads up on this run (NASA Brady et.al. geometry extended to near the apex of the cone):

1) Non-symmetric placement of antenna, very offset from middle

2) Amplitude of response is very low: not a good resonance.  It shows fractal numerical artifacts due to very low amplitude.

3) Forces at both ends are very low.  Force at small end close to zero.  Stress distribution at big end very low and very asymmetric.

Suggestion: may need to run again, at an excitation frequency for which we know (from other calculations) that there is a good resonant response and with an antenna located on the axis axi-symmetry. 

If run again, it may be better to do it based on Yang/Shell geometry extended to near the apex of the cone, to model the geometry she intends to test.
I have almost the same insertion geometry as them except have a monopole rather than a loop. I think my temp tests are confirming a bad match at this locale. Will know more this weekend.
The example was for TE01p and a dipole with different orientation. As far as i know you want to use TM212 correct? And you use a Monopole antenna direction from the magnetron. This is a complete different Situation.
But IMHO your penetration point and the orientation of the antenna is a good choice!
You use an unloaded cavity. If you dont put food in your MW oven, the magnetron will also heats up, thats why people dont do that. The device can be broken caused by overheating the magnetron.

BTW got found an interesting link
https://www.cst.com/Applications/Article/Magnetron-And-Microwave-Oven-Design-To-Solve-Wi-Fi-Interference-Issues
Wonderful article, thanks. Take note of their Q values, a little less that 2000. Another validation for reputable numbers below 10,000.

https://www.cst.com/Content/Articles/article618/Comparison.PNG

Their gif movie demonstrates the circular buildup of energy within the magnetronm which I've seen before. This "spraying" of radiation continues to the radome (monopole antenna):

https://www.cst.com/Content/Articles/article618/Trajectory_movie_3.gif
« Last Edit: 08/01/2015 03:04 PM by rfmwguy »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2326
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2956
  • Likes Given: 2589
TT did you ever run Yang's model in your spreadsheet as reported in the Wiki site? What did you see, inquiring minds want to know.

Shell

Must have missed that.

Simple to do. What are the dimensions, freq and mode? Or should just I use the dimensions and report on what I find?
Details are on the wiki site. http://emdrive.wiki/Experimental_Results Start with them and see what you get with your spreadsheet.

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2326
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2956
  • Likes Given: 2589
The main thing I'd like to know about Yang is whether anybody can verify her experimental data at her facility. Neutral parties, I mean.
I'd love to be the proverbial fly on the wall, my luck I'd buzz to close to the giant bug zapper called a frustum.

I'm hoping that the frustum I'm putting together matches the best guess of the group here. Nobody has taken the bull by the horns to try to at least validate their claims by building something close to their design. I honestly believe it needs to be addressed and I'm trying to do so as best as I can. So criticism is welcome.

TT did you ever run Yang's model in your spreadsheet as reported in the Wiki site? What did you see, inquiring minds want to know.

Shell

Shell,

Using your Computer Aided Drawing program to assess the following figure:  https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=37642.0;attach=1051616,

what

1) Cone half angle (angle between lateral side and perpendicular to base)

2) Ratio of internal Big Base diameter to internal Small Base diameter

3) Ratio of internal  Big Base diameter to internal Length perpendicular to bases

do you get ?

This would be helpful to better estimate dimensions.
Sure give be a bit.

Offline aero

  • Senior Member
  • *****
  • Posts: 2744
  • 92129
  • Liked: 705
  • Likes Given: 239
The antenna modelled in the meep simulation was not a very good approximation to the real thing. The location was correct but it was a dipole antenna. I quote from a private communication from Paul received back in April.

Quote
Steve:  The current loop antenna in the copper frustum has a 14mm OD and is made from #20 AWG magnet wire soldered to an SMA bulkhead connector.  This assembly is then rotated to maximize the S11 return loss for a given resonant frequency while using an RG-142 SMA-to-SMA coaxial cable run, which is typically 2.0 feet long.  Location is on the frustum sidewall 1.35 inch up from the interior flat surface of the large OD end of the frustum.

A proper model of a loop antenna would seem to be a reasonable next step but I would not know how to rotate it to maximize the S11 return loss.
Retired, working interesting problems

Offline flux_capacitor

  • Full Member
  • ****
  • Posts: 552
  • France
  • Liked: 654
  • Likes Given: 909
As I'm not sure about the actual measurements, the drawing is made with proportional dimensions, with the internal height set to 100%

1) 74.56

2) 100/ 67 (separately rescaled big diameter to 100% to find the % of small diameter)

3) 115.4 / 100

If you find it more convenient to have another dimension as 100% base value, please say so... it's only a few clicks away..

Your drawing does not fit the data in the EM drive wiki, where Yang's frustum has the following dimensions:

- cavity length (m): 0.24
- big diameter (m): 0.201
- small diameter (m): 0.1492

In particular, the big end is smaller than the height, not larger.
« Last Edit: 08/01/2015 03:31 PM by flux_capacitor »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5841
  • USA
  • Liked: 5923
  • Likes Given: 5268
what

1) Cone half angle (angle between lateral side and perpendicular to base)

2) Ratio of internal Big Base diameter to internal Small Base diameter

3) Ratio of internal  Big Base diameter to internal Length perpendicular to bases

do you get ?

This would be helpful to better estimate dimensions.

I thought I uploaded this a while ago, or did I forget? ???

not sure if I understand your questions correctly, but...

As I'm not sure about the actual measurements, the drawing is made with proportional dimensions, with the internal height set to 100%

1) 74.56

2) 100/ 67 (separately rescaled big diameter to 100% to find the % of small diameter)

3) 115.4 / 100

If you find it more convenient to have another dimension as 100% base value, please say so... it's only a few clicks away..
Yes, this was intended as an exercise for people to realize on their own, using the Socratic method, that the dimensions of the cone in the EM Drive are completely incompatible with the implied dimensions in Yang's drawing:


(The cone half angle is the angle between the lateral and the perpendicular to the base, not the angle between the lateral and the base)

Cone half-angle as per your drawing =  90 - 74.56
                                                      = 15.44

Cone half-angle in Yang/Shell           = 6.159  (see http://emdrive.wiki/Experimental_Results)

Cone half-angle NASA                      = 14.78  (see http://emdrive.wiki/Experimental_Results)

Thus we show that the cone-half angle in Yang's drawing is close to NASA's cone half-angle.  We also show that the cone-half angle in Yang/Shell (http://emdrive.wiki/Experimental_Results) is less than 40% of the cone half-angle in Yang's drawing.

The cone half-angle in Yang/Shell (the geometry for Yang in the EM Drive wiki) is close to a cylinder, it is very far away from the geometrical dimensions shown in Yang's drawings.

A cylinder has a cone half-angle equal zero.

« Last Edit: 08/01/2015 03:28 PM by Rodal »

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2165
  • Liked: 2681
  • Likes Given: 1124
The antenna modelled in the meep simulation was not a very good approximation to the real thing. The location was correct but it was a dipole antenna. I quote from a private communication from Paul received back in April.

Quote
Steve:  The current loop antenna in the copper frustum has a 14mm OD and is made from #20 AWG magnet wire soldered to an SMA bulkhead connector.  This assembly is then rotated to maximize the S11 return loss for a given resonant frequency while using an RG-142 SMA-to-SMA coaxial cable run, which is typically 2.0 feet long.  Location is on the frustum sidewall 1.35 inch up from the interior flat surface of the large OD end of the frustum.

A proper model of a loop antenna would seem to be a reasonable next step but I would not know how to rotate it to maximize the S11 return loss.

Shell - A loop antenna radiates maximally along its horizontal axis, therefore its match will be dependent on the ralationship to the closest ground plane. Typically the best rl will be perpendicular to the mounting ground plane.

Online Rodal

  • Senior Member
  • *****
  • Posts: 5841
  • USA
  • Liked: 5923
  • Likes Given: 5268
As I'm not sure about the actual measurements, the drawing is made with proportional dimensions, with the internal height set to 100%

1) 74.56

2) 100/ 67 (separately rescaled big diameter to 100% to find the % of small diameter)

3) 115.4 / 100

If you find it more convenient to have another dimension as 100% base value, please say so... it's only a few clicks away..

Your drawing does not fit the data in the EM drive wiki, where Yang's frustum has the following dimensions:

- cavity length (m): 0.24
- big diameter (m): 0.201
- small diameter (m): 0.1492
In particular, the big end is smaller than the height, not larger.

That's the point I have been trying to make for several pages: the dimensions in the EM Drive wiki are incompatible with this Yang drawing  (:  https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=37642.0;attach=1051616,
).  Completely incompatible.  It is not just a little bit off.  It is completely different.  Take a look at Yang's drawing: does it look like a cylinder?

I strongly doubt that Yang's geometry is close to a cylinder, particularly if she didn't use dielectric inserts.  It does not make any sense that her cone half-angle would have been only 6 degrees.

Her drawing shows this.

The Meep runs shows this.

My computer runs shows this.

The only similitude is that the geometry in the EM Drive wiki gives TE012 at 2.45 GHz

That's not enough.  I can also give a perfect cylinder geometry with TE012 at 2.45 GHz, that doesn't mean that Yang's experiment was conducted on a cylinder.

« Last Edit: 08/01/2015 03:27 PM by Rodal »

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1229
  • Do it!
  • Vista, CA
  • Liked: 1296
  • Likes Given: 1753
rfmwguy.

I'm thinking replace the magnetron [see attachment below looks the same] with a RF transistor as used here

http://www.zdnet.com/article/freescales-radio-frequency-oven-the-end-of-the-microwave/

maybe it runs with less heat loss?
and runs cleaner r.f. waves
http://www.freescale.com/webapp/video_vault/videoSummary.sp?code=RF-SAGE-VIDEO

 see 2.37

They still require a heat sink. 3.7 deg. C/Watt is a lot of heat. These are also only good for about 200 to 400 mW per transistor, if you've got a good heat sink. If you want to get back up to 400W, you will need to parallel about a dozen of them on a well designed PC board, that is thermally conductive to the heatsink. At these frequencies, I have no idea how you would do that, or what proper design criteria are necessary for a microwave amplifier of sufficient power.
Todd
I take issue with your use of the arithmetic division operator.

Nick picking again...  oC/W.
Not that - that you say you need "about a dozen" 400 mW transistors to get 400 Watts.

Oops! 4W each, not 400mW. Off by a decimal. It was a long day.
Todd

Offline flux_capacitor

  • Full Member
  • ****
  • Posts: 552
  • France
  • Liked: 654
  • Likes Given: 909
That's the point I have been trying to make for several pages: the dimensions in the EM Drive wiki are incompatible with this Yang drawing  (:  https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=37642.0;attach=1051616,
).  Completely incompatible.  It is not just a little bit off.  It is completely different.  Take a look at Yang's drawing: does it look like a cylinder?

I strongly doubt that Yang's geometry is close to a cylinder, particularly if she didn't use dielectric inserts.  It does not make any sense that her cone half-angle would have been only 6 degrees.

Her drawing shows this.

The Meep runs shows this.

My computer runs shows this.

The only similitude is that the geometry in the EM Drive wiki gives TE012 at 2.45 GHz

That's not enough.  I can also give a perfect cylinder geometry with TE012 at 2.45 GHz, that doesn't mean that Yang's experiment was conducted on a cylinder.

Evidently. But what is the source of those dimensions on the wiki page? I already asked several months ago and you answered they are all over the original Yang 2010 paper. I may be stupid but I only find the (various) cavity lengths she tried at different modes, and not the big end nor the small end dimensions???

EDIT: I see she calculates the end diameters from the cavity length and the cut-off frequency, and shows a diagram (figure 3). But she doesn't provides the results so who calculated those diameters on the wiki?

From the translated paper:

Quote from: Juan YANG
With the three main modes of cylinder waveguide, TE011, TE012, TE111, using finite element to calculated the distribution of electromagnetic field of the conical resonators under the four modes TE011, TE012, TE111 and TM011 with frequency near 2.45GHz. To select the conical resonator diameters, first follow the cylinder resonator operation mode in Figure 3 to select cylinder diameter and height in single mode, that diameter is the conical cavity average diameter, select the diameter of the Small-End as the waveguide cut-off diameter, the diameter of the large end can be obtained according to average diameter. Perform the finite element numerical simulate the distribution of microwave electromagnetic field in resonant for the cavity. Repeat adjusting the Small and Large ends diameters D1, D2 and height H1 of cavity within a limit, to obtain the distribution of the electromagnetic field.


Figure 3
« Last Edit: 08/01/2015 03:58 PM by flux_capacitor »

Offline Flyby

  • Full Member
  • ***
  • Posts: 384
  • Belgium
  • Liked: 447
  • Likes Given: 47

Your drawing does not fit the data in the EM drive wiki, where Yang's frustum has the following dimensions:

- cavity length (m): 0.24
- big diameter (m): 0.201
- small diameter (m): 0.1492

In particular, the big end is smaller than the height, not larger.

I'm aware of that, but I do not know where these values were obtained from, and the question was to get the proportions of the drawing (and not of the Wiki dimensions), while being well aware that they are probably not correct.
But neither can we be sure about the wiki dimensions...

I have no clue at all what are the correct dimensions, so it's poking in the dark, hoping we might get something that can lead us further...
« Last Edit: 08/01/2015 03:40 PM by Flyby »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5841
  • USA
  • Liked: 5923
  • Likes Given: 5268
That's the point I have been trying to make for several pages: the dimensions in the EM Drive wiki are incompatible with this Yang drawing  (:  https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=37642.0;attach=1051616,
).  Completely incompatible.  It is not just a little bit off.  It is completely different.  Take a look at Yang's drawing: does it look like a cylinder?

I strongly doubt that Yang's geometry is close to a cylinder, particularly if she didn't use dielectric inserts.  It does not make any sense that her cone half-angle would have been only 6 degrees.

Her drawing shows this.

The Meep runs shows this.

My computer runs shows this.

The only similitude is that the geometry in the EM Drive wiki gives TE012 at 2.45 GHz

That's not enough.  I can also give a perfect cylinder geometry with TE012 at 2.45 GHz, that doesn't mean that Yang's experiment was conducted on a cylinder.

Evidently. But what is the source of those dimensions on the wiki page? I already asked several months ago and you answered they are all over the original Yang 2010 paper. I may be stupid but I only find the (various) cavity lengths she tried at different modes, and not the big end nor the small end dimensions???

EDIT: I see she calculates the end diameters from the cavity length and the cut-off frequency, and shows a diagram (table 3). But she doesn't provides the results so who calculated those diameters on the wiki?

Quote from: Juan YANG
With the three main modes of cylinder waveguide, TE011, TE012, TE111, using finite element to calculated the distribution of electromagnetic field of the conical resonators under the four modes TE011, TE012, TE111 and TM011 with frequency near 2.45GHz. To select the conical resonator diameters, first follow the cylinder resonator operation mode in Figure 3 to select cylinder diameter and
height in single mode, that diameter is the conical cavity average diameter, select the diameter of the Small-End as the waveguide cut-off diameter, the diameter of the large end can be obtained according to average diameter. Perform the finite element numerical simulate the distribution of microwave electromagnetic field in resonant for the cavity. Repeat adjusting the Small and Large ends diameters D1, D2 and height H1 of cavity within a limit, to obtain the distribution of the electromagnetic field.

As I pointed out several pages ago, and frequently I am the one that is the source of those dimensions.  I carefully explained some time ago how the dimensions were obtained, carefully going through her paper and using her tables, and finally verifying that the dimensions give TE012. 

As far as I know, nobody else attempted to obtain dimensions from her tables. 

Also, as I have been pointing out through several pages, based on her drawings, based on the Meep analysis, and based on my independent computer analysis, something is completely off in those dimensions.  Maybe there is a typo in her paper, maybe there is a misinterpretation of a variable in her tables, something is off by a factor of 2.   Nothing strange about that, as I pointed out Tajmar's dimensions were off by a factor of 2 and he has confirmed this.

At this point in time it is kind of a funny situation: I am the one that went through the trouble to calculate the dimensions that are in the Wiki and I am also the one that has been arguing for the last several pages that the angle in the Wiki must be off at least by a factor of 2, and therefore the Yang dimensions in the wiki canNOT be representative.

It doesn't matter if TheTraveller confirms that the dimensions in the Wiki give TE012 at near 2.45 GHz.  That proves nothing. I calculated them that way.  I can also calculate a perfect cylinder with TE012 at 2.45 GHz that doesn't mean that she used a cylinder.
« Last Edit: 08/01/2015 03:56 PM by Rodal »

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1229
  • Do it!
  • Vista, CA
  • Liked: 1296
  • Likes Given: 1753
As I'm not sure about the actual measurements, the drawing is made with proportional dimensions, with the internal height set to 100%

1) 74.56

2) 100/ 67 (separately rescaled big diameter to 100% to find the % of small diameter)

3) 115.4 / 100

If you find it more convenient to have another dimension as 100% base value, please say so... it's only a few clicks away..

Your drawing does not fit the data in the EM drive wiki, where Yang's frustum has the following dimensions:

- cavity length (m): 0.24
- big diameter (m): 0.201
- small diameter (m): 0.1492
In particular, the big end is smaller than the height, not larger.

That's the point I have been trying to make for several pages: the dimensions in the EM Drive wiki are incompatible with this Yang drawing  (:  https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=37642.0;attach=1051616,
).  Completely incompatible.  It is not just a little bit off.  It is completely different.  Take a look at Yang's drawing: does it look like a cylinder?

I strongly doubt that Yang's geometry is close to a cylinder, particularly if she didn't use dielectric inserts.  It does not make any sense that her cone half-angle would have been only 6 degrees.

Her drawing shows this.

The Meep runs shows this.

My computer runs shows this.

The only similitude is that the geometry in the EM Drive wiki gives TE012 at 2.45 GHz

That's not enough.  I can also give a perfect cylinder geometry with TE012 at 2.45 GHz, that doesn't mean that Yang's experiment was conducted on a cylinder.

However, with the longer cylinder-like frustum, her half-angle is so small that extrapolating from the Impedance diagrams on Zeng and Fan, the part of the impedance curve and slope that cylinder is working on is about the same as it is for the other designs with more angle and shorter length. In other words, although her k*r value is in the 30's, the relative impedance shift is in the same ball park as the shorter designs with a larger angle. It may be that having less angle allows it to reach a much higher Q, like a cylinder but the shift in Z is still there to produce thrust.

@TT, if you look at the impedance graphs in Zeng and Fan, you can estimate what the impedance is as the distance k*r from the apex of the cone. That should eliminate the guessing.
Todd


Tags: