-
#1300
by
rfmwguy
on 15 Apr, 2016 18:51
-
New poll for emdrive readership at top of each page. Just for fun, pick a timeframe when EW's paper will be released. Poll active for 90 days. One guess per user.
-
#1301
by
rfmwguy
on 15 Apr, 2016 19:11
-
-
#1302
by
Monomorphic
on 15 Apr, 2016 19:14
-
What is the deg F resolution possible with this camera?
At the moment, the color scheme is showing the hot magnetron which is much hotter, over 200 deg F, and therefore the mode shape small delta T will not show. Have to hide the magnetron from view
Sensitivity: ability to detect temperature differences as small as 0.18° F (0.1° C)
I'll do a few more runs tonight and try and hide the magnetron. I might also add another coat of paint beforehand as I am still seeing some IR reflections through the one coat of paint.
-
#1303
by
X_RaY
on 15 Apr, 2016 19:27
-
Just done the verification of the NASA-frustum with HDPE-disc.
This are the results for TE012
Looks good but I am still fighting with the limitations of feko light
@FL
I will downscale the whole thing by a factor of 2.794 now...
-
#1304
by
Rodal
on 15 Apr, 2016 19:46
-
Just done the verification of the NASA-frustum with HDPE-disc.
This are the results for TE012
Looks good but I am still fighting with the limitations of feko light
@FL
Now I can downscale the whole thing now by a factor of 2.794 
.
What value of relative electric permittivity did you use for your FEKO run?
If you used
2.26 @ 3 GHz
and you got resonance TE012 at 1.884 GHz close to the same experimental value of natural frequency found by NASA this shows that the HDPE dielectric did not get hot enough (nowhere near melting temperature) for the permittivity to be much affected by temperature.
This fact (that the experimental natural frequency was close to the frequency predicted for a relative electric permittivity of HDPE at room temperature, 3 GHz), effectively nullifies rfmwguy's hypothesis that the HDPE dielectric may have been close to the melting temperature in NASA's tests.
-
#1305
by
X_RaY
on 15 Apr, 2016 20:01
-
Just done the verification of the NASA-frustum with HDPE-disc.
This are the results for TE012
Looks good but I am still fighting with the limitations of feko light
@FL
Now I can downscale the whole thing now by a factor of 2.794 .
What value of relative electric permittivity did you use for your FEKO run?
If you used
2.26 @ 3 GHz
and you got resonance TE012 at 1.884 GHz close to the same experimental value of natural frequency found by NASA this shows that the HDPE dielectric did not get hot enough (nowhere near melting temperature) for the permittivity to be much affected by temperature.
This fact (that the experimental natural frequency was close to the frequency predicted for a relative electric permittivity of HDPE at room temperature, 3 GHz), effectively nullifies rfmwguy's hypothesis that the HDPE dielectric may have been close to the melting temperature in NASA's tests.
Great question!!!
You are very very good!!
I tryed to use the ε
r=2.26 but the resuting frequency was sligtly under 1.86GHz, after that I tried to use PTFE(ε
r=2.08) but the frequency was over 1.9GHz.
At the end I used ε
r=2.15 tanδ=0.00031 to get the same frequency as EW. The difference is tiny, only dε
r≈0.1 lower than expected
Please remember the coarse mesh! I tried to submesh the dielectric but get limitation errors. Therefor do not interpret too much into this number please.
-
#1306
by
Rodal
on 15 Apr, 2016 20:15
-
Just done the verification of the NASA-frustum with HDPE-disc.
This are the results for TE012
Looks good but I am still fighting with the limitations of feko light
@FL
Now I can downscale the whole thing now by a factor of 2.794 .
What value of relative electric permittivity did you use for your FEKO run?
If you used
2.26 @ 3 GHz
and you got resonance TE012 at 1.884 GHz close to the same experimental value of natural frequency found by NASA this shows that the HDPE dielectric did not get hot enough (nowhere near melting temperature) for the permittivity to be much affected by temperature.
This fact (that the experimental natural frequency was close to the frequency predicted for a relative electric permittivity of HDPE at room temperature, 3 GHz), effectively nullifies rfmwguy's hypothesis that the HDPE dielectric may have been close to the melting temperature in NASA's tests.
Great question!!!
You are very very good!!
I tryed to use the εr=2.26 but the resuting frequency was sligtly under 1.86GHz, after that I tried to use PTFE(εr=2.08) but the frequency was over 1.9GHz.
At the end I used εr=2.15 tanδ=0.00031 to get the same frequency. The difference is tiny, only dεr≈0.1 lower than expected
Please remember the coarse mesh! I tried to submesh the dielectric but get limitation errors. Therefor do not interpret too much into this number please.
This is great information !
Thanks for your run!
From the great work of Krupka,
Complex permittivity measurements of common plastics over variable temperaturesJerzy Antoni Krupka
Warsaw University of Technology
Article in
IEEE Transactions on Microwave Theory and Techniques 51(3):727 - 733 · March 2003
Impact Factor: 2.24 · DOI: 10.1109/TMTT.2003.808730 · Source: IEEE Xplore
https://www.researchgate.net/publication/3130152_Complex_permittivity_measurements_of_common_plastics_over_variable_temperatureswe have these experimental values of HDPE vs temperature in the GHz range
It looks like to have ε
r=2.15 one has to have a temperature close to the Tg of HDPE (110 deg C)
This is very important because there are very interesting things happening to the group velocity (*) and the index of refraction, and reflectivity near the glass transition temperature of polymers
PS: Concerning everything that people wrote concerning dielectrics and the group velocity in previous threads: it needs to be revised, as the permittivity (both real part and imaginary) are not constant at the Tg !!!(*) I'll write a mathematical discussion of these effects when I have the time
-
#1307
by
X_RaY
on 15 Apr, 2016 20:29
-
X_Ray, the only information I have re the nature of the dielectric is this:
Congratulations FattyLumpkin Maxwell works very well.
Suggestion 1.8804 GHz * 2.794 = 5.254 GHz
The small difference is a result of rounding errors(I used two digits) and the coarse mesh.
-
#1308
by
X_RaY
on 15 Apr, 2016 21:23
-
How about giving us a thermal view of the big end, with the camera focus perpendicular to the center of the big end, showing only the big end, when you have a chance?
Here you go, but as the magnetron gets hot it swamps everything out. Will need to orient the camera so the magnetron can't be seen.
I'm not sure if this is what rfmwguy is referring to, but I also would like to have the magnetron covered so that the thermal camera cannot see the magnetron or see anything hotter than the big end. As you say, the magnetron is so much hotter that the small difference in temperature ( of the order of a few deg F: 72 deg F is only 4 deg F above room temperature of 68 deg F ?) will not be visible
What is the deg F resolution possible with this camera?
At the moment, the color scheme is showing the hot magnetron which is much hotter, over 200 deg F, and therefore the mode shape small delta T will not show. Have to hide the magnetron from view
I am not sure that the sensitivity of this type or camera is in the range of what is needed. Cooled bolometer cam could be a better opinion! Its much more sensetive and delivers a faithful representation in the sub Kelvin resulutions (of course its extensive, several 10000$
)
-
#1309
by
Rodal
on 15 Apr, 2016 22:19
-
How about giving us a thermal view of the big end, with the camera focus perpendicular to the center of the big end, showing only the big end, when you have a chance?
Here you go, but as the magnetron gets hot it swamps everything out. Will need to orient the camera so the magnetron can't be seen.
I'm not sure if this is what rfmwguy is referring to, but I also would like to have the magnetron covered so that the thermal camera cannot see the magnetron or see anything hotter than the big end. As you say, the magnetron is so much hotter that the small difference in temperature ( of the order of a few deg F: 72 deg F is only 4 deg F above room temperature of 68 deg F ?) will not be visible
What is the deg F resolution possible with this camera?
At the moment, the color scheme is showing the hot magnetron which is much hotter, over 200 deg F, and therefore the mode shape small delta T will not show. Have to hide the magnetron from view
I am not sure that the sensitivity of this type or camera is in the range of what is needed. Cooled bolometer cam could be a better opinion! Its much more sensetive and delivers a faithful representation in the sub Kelvin resulutions (of course its extensive, several 10000$ )
Bolometer?
I had to look this up
Aha, it works like a calorimeter
http://www.photonics.com/Article.aspx?AID=56863I love how much one always learns from these threads
-
#1310
by
Rodal
on 15 Apr, 2016 22:50
-
It is also appropriate to consider the following experimental results and patents in such a review:
How should we interpret/analyze the reported experimental result that NASA only measured anomalous thrust for the Cannae non-superconducting device when introducing a dielectric insert in the Cannae pipe ?
And the fact that Cannae, after NASA's 2014 measurement, incorporated a dielectric in Cannae's 2015 patent claims?
I do not think we should take seriously NASA's measurements with/without dielectric inserts. Was in one measurement the cannie drive grounded, and in another measurement not grounded? If in one measurement the drive was grounded, what was the shape of the RF cable in that measurement? These were factors that could have dramatically changed the Lorentz forces in their measurements, which had the same size as their measured thrust. Since they did not consider the Lorentz forces, their measurements were not valid and no conclusion should be drawn from their measurements.
Thank you for further clarification (not clear whether you are dismissing all Cannae's measurements or just NASA's measurements of Cannae's device):
Do you take more seriously Cannae's claimed force measurements with the grooves? while NASA's measurements showed no anomalous force effect with Cannae's grooves?
Do you take more seriously Cannae's claimed force measurements with the Cannae non-superconducting and superconducting devices rather than NASA's non-superconducting measurements using a torque pendulum?
If so, what is the basis for taking Cannae's measurements more seriously than NASA's measurements?
I was dismissing NASA's measurements because I think I know what they did incorrectly. I do not know Cannie's experiments as they are not published.
No, I do not take Cannie's claims seriously because I do not believe the EmDrive concept is valid. I think if I can have the chance to observe Cannie's experiment closely, I can spot what is done incorrectly.
I fully understand that in a thread (EM Drive) where some have made proclamations that unbelievers "are going to have to eat their hats in 2015" (this goal post now moved to 2016) and extrapolations of trips to the stars using the EM Drive have been made, skeptics are tempted to match the hyperbole, but I think that it would be better to tone down the remarks as the evidence for a remark to the effect of "not taking seriously NASA's data" is just not there. On the contrary:
NASA reported measuring Cannae experiment’s power supply DC Lorentz force interactions with the Earth’s magnetic field (and any other stray magnetic fields) in the vacuum chamber via a RF dummy load test and found them to be less than a uN per the NASA's published 2014 ""Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum" paper on page 10, part D.
The DC grounding system used for all these Cannae tests were the same in all cases.
Since the thrust signature from the PTFE dielectric loaded Cannae flying-saucer-shaped resonator during this test series was on the order of 22 to 50 uN dependent on direction, one can reasonably conclude that an unaccounted for anomalous force was detected during this test series.
The strongest scholarly statement you could make regarding the effect of Lorentz forces pertains to the difference between 22 to 50 uN dependent on direction, but to state that "NASA's experiments cannot be taken seriously" and that NASA's measurements are due entirely to Lorentz forces, and that therefore NASA measured no anomalous force whatsoever (not even 22 uN), is just ... hyperbole that detracts from scholarship IMHO.
-
#1311
by
CuriousDreamer
on 15 Apr, 2016 23:04
-
You know, I had an idea about why this thing worked last thread that nobody liked but might be applicable. The rf bounces between two points taking the shortest 3 dimensional path. The acceleration though has to follow the shell of the device. Both propagate at the speed of light. A pancake shape means a very quick transit between the two reflective points while maximizing the area that the acceleration has to propagate through. The acceleration is still light lagged (isn't in the light cone of the reflector) when the reflection takes place. If KE=1/2mv^2 then, if both endplates are at a constant speed (say 10m/s and 10m/s) the endplate in the direction of motion will gain more KE then the opposite endplate will lose (unless the acceleration has reached the opposite point of reflection in which case the equation will balance).
I suppose I could be very wrong in treating acceleration in the same way I would an electrical charge propagating through the shell of the device. Still, if correct it would explain why a superconducting pancake with the bulk of its energy in the very center works (assuming it actually does work).
If this were the case, would I be right in assuming then that the drive would produce thrust in what ever direction the drive were moving?
If so, this would be very easy to test if experiments begin to show reliable thrust from the drive. turn it on, nudge it one way, record changes in velocity. reset, turn it on, nudge the other way, record changes in velocity again, compare. If the drive accelerates in both directions, then this could be the cause and further testing could be done. If the drive accelerates one direction and decelerates in the other direction then you can be fairly sure this is not the droid we are looking for.
Curious
-
#1312
by
SteveD
on 16 Apr, 2016 00:32
-
You know, I had an idea about why this thing worked last thread that nobody liked but might be applicable. The rf bounces between two points taking the shortest 3 dimensional path. The acceleration though has to follow the shell of the device. Both propagate at the speed of light. A pancake shape means a very quick transit between the two reflective points while maximizing the area that the acceleration has to propagate through. The acceleration is still light lagged (isn't in the light cone of the reflector) when the reflection takes place. If KE=1/2mv^2 then, if both endplates are at a constant speed (say 10m/s and 10m/s) the endplate in the direction of motion will gain more KE then the opposite endplate will lose (unless the acceleration has reached the opposite point of reflection in which case the equation will balance).
I suppose I could be very wrong in treating acceleration in the same way I would an electrical charge propagating through the shell of the device. Still, if correct it would explain why a superconducting pancake with the bulk of its energy in the very center works (assuming it actually does work).
If this were the case, would I be right in assuming then that the drive would produce thrust in what ever direction the drive were moving?
If so, this would be very easy to test if experiments begin to show reliable thrust from the drive. turn it on, nudge it one way, record changes in velocity. reset, turn it on, nudge the other way, record changes in velocity again, compare. If the drive accelerates in both directions, then this could be the cause and further testing could be done. If the drive accelerates one direction and decelerates in the other direction then you can be fairly sure this is not the droid we are looking for.
Curious
My working assumption is that, for some reason, the small end has greater reflectivity than the large base. This means that more reflections take place at one end than the other (generating 2W/C forces instead of simply W/C force for an absorption -- meaning that there is double the acceleration which hasn't yet propagated to the other side of the drive), putting the device into motion towards the small end. I suspect that this is somehow related to the changes in wavelength that occur in a tapered cavity / when use a dielectric insert. This would explain why a similar effect has yet to be detected in symmetrical cavities (though see the reports of capacitors that measure lighter than they should be when charged). If I remember correctly the last iteration of the Cannae drive had only one superconducting endplate in a non-tapered design. That would seem to argue that a difference in reflectivity between the two sides is required. Unfortunately, I'm lacking for a good idea of how to test this. Also, we might be talking about a very small difference to produced the desired motion.
-
#1313
by
CriX
on 16 Apr, 2016 03:30
-
Regarding the theory that thrust is dependent on motion, or acceleration as I interpret it, isn't this problematic as both acceleration and being in a gravity field are the same thing via general relativity? This is not my field so I could be completely mistaken.
-
#1314
by
FattyLumpkin
on 16 Apr, 2016 09:26
-
Monomorphic, as busy as you are, is there any chance you could run your stacked frustums sim perhaps twice or three times as long (I'm not sure if FEKO accounts for the "loading" concept I mentioned earlier). Would you also comment on the sim it self: what is your opinion... did you have to increase the amount of RF energy in order to fill all three vessels, and finally assuming the EM effect is produced by each, would you expect a tripling of thrust?
Thank you for your time, FattyLumpkin
PS: looking forward to your FLIR without the outside "hotspots"
-
#1315
by
TheTraveller
on 16 Apr, 2016 10:33
-
The only reason people are here is because of Roger Shawyer's inventive work and publically shared experimental data.
I find it amazing that so many people here go in very different direction than those suggested by Roger and follow advice from posters that have never tried to build an EmDrive.
Best of luck to you because, mathematics theory that can not correctly / properly model thrust generation will never open the doorway so many seek.
All successful EmDrive thrusters that I know of have incorporated a tuning element of some sort at the input. Also no successful design used COMSOL without correction, as the software does not seem to cope with conditions close to cut-off, as NASA should have realised.
-
#1316
by
X_RaY
on 16 Apr, 2016 11:30
-
... as the software does not seem to cope with conditions close to cut-off...
What? Since when? This software solves field equations and for sure it will work near cutoff.
-
#1317
by
FattyLumpkin
on 16 Apr, 2016 11:49
-
Traveller, I have every intention of moving forward with a build (see attached) However your statement puts me into a bit of a quandary where specs are concerned. I don't know anything about your build in progress save that you are using a shop, as you are having difficulties with your health (I'm sorry to hear about that). IF COMSOL and/or FEKO simulations are not good predictors for coming up with a build that produces real thrust, then where and how is one to "get" the appropriate geometry/dimensions that will/would assure a thrust producing device. I've read everything there is to read on Roger's site.
Do you have precise specs for me for my trapezoidal prism build? I'd like to go with a bigger build (as would others here on NSF)..L band or even something below that. Your input would be greatly appreciated. , FattyLumpkin PS: are you going to publishing you results once testing is completed (predicted and actual thrust etc.)? FL
-
#1318
by
FattyLumpkin
on 16 Apr, 2016 12:02
-
X_Ray , thank you for all of your hard work related to the frustum downscaling. Please contact me (use my email) and include your email address. Thank you FL PS: Mein Deutsch ist furchtbar geworden. K
-
#1319
by
Rodal
on 16 Apr, 2016 12:25
-
...Also no successful design used COMSOL without correction, as the software does not seem to cope with conditions close to cut-off, as NASA should have realised.
What? Since when? This software solves field equations and for sure it will work near cutoff.
There is a contradiction between Shawyer stating in several reports available to the public that EM Drive thrust can be explained in terms of Maxwell's equations and people claiming in this thread that Shawyer e-mailed them that COMSOL cannot cope with conditions close to cut-off. Since it is evident that COMSOL solves Maxwell's partial differential equations, and that COMSOL is used by many companies making waveguides with conditions close to cut-off, this contradiction shows that either
1) Shawyer does not understand that COMSOL solves Maxwell's differential equations and appropriately handles cut-off conditions
or that
2) Shawyer never stated that "COMSOL without correction, does not seem to cope with conditions close to cut-off"
Since there is no public report of Shawyer himself stating that "COMSOL without correction, does not seem to cope with conditions close to cut-off" and this statement cannot be independently verified as having being said by Shawyer, the preferable alternative is that Shaywer is being misquoted and/or misunderstood in some private communication out of context.
In any case, it does not seem right to quote Shawyer criticizing a commercial code like COMSOL quoting a fragment of a private e-mail that cannot be independently verified as having been authored by him. And supposing that even if it would be the case that Shawyer stated this in a private e-mail, we don't know whether Shawyer presently wants such private e-mail criticism of COMSOL to be repeated in public to this forum, long after that private e-mail was written.
.
)