-
#2940 by aero on 04 Nov, 2014 20:28
-
Thanks. But we need to explain this video before we can attribute the EM Drive thrust in general to chamber wall/cavity interactions.
http://emdrive.com/dynamictests.html
Video clips are near the bottom of the page. I don't see anything in the video that the EM Drive could interact with.
After all this lengthy discussion I have been tracking for a while, what stays with me is the absolute need of more experimental results to talk about. More data points, more confirmations (or refutations).
And the video you bring, while enticing, is alas not enough. I can imagine several ways to trick a video like that, just requiring enough willingness and lack of a consciousness to do it.
Don't get me wrong. I like visual demonstrations as the next guy, it's that we only have this one so far.
But if more people replicated that... we could start becoming really intrigued.
No question that we need more data and more replication.
Operating on the assumption that there is a real force generated by the EM Drive precludes the need to speculate that the data is somehow faked. If we assume that the data is faked and the force is not real then that's the end of the story. No need to go any further.
We here are operating on the assumption that the measurements are real and that the force is from a real EM Drive effect or else it is from an experimental artifact. We are pursuing both lines of investigation so the story ends when we find either the cause of the EM Drive effect or the experimental artifact that fits all of the data we have. Or when we have exhausted the data available. Without more data the latter seems likely at this point. -
#2941 by Rodal on 04 Nov, 2014 20:31
-
Next batch of scraped data from figure 19 page 15 of "anomalous thrust..." from Brady et al. The top (result1.txt) and middle (result2.txt) graphs are scraped.
Same caveats as previously posted. For first curve (top figure 19) I removed the (non existent) flat last sampled data of the previous version to avoid artefacts when analysing with filters.
Each line of those files is the value in µN at each .1 s interval (linearly interpolated from manual reconstruction). The vertical scale were roughly given by the calibration pulses at about 30µN (expect no more than 5% precision). Absolute values are arbitrary (because of the drifting baseline). Horizontal scale given by the indication of 196 s for the whole display graph window of the pictures.
Will proceed with other graphs when time permits. Will post attempts at original signal reconstruction : thrust(t) while what we see is only balance displacement(t). Since the balance is underdamped, a lot can hide behind those oscillations and drifts in position.
Frobnicated Top of Fig. 19 page 15 of anomalous (Mean and Linear Least Squares Fit)
Autocorrelation of Top of Fig. 19 page 15 (from FFT) on raw data detrended by Mean (Blue)
Autocorrelation of Top of Fig. 19 page 15 (from FFT) on raw data detrended by Linear LS (Red)
Power Spectral Density (from FFT) on raw data detrended by Linear LS (Red)
horizontal scale = frequency(Hz) * 0.1 * (DataLength/2) = frequency(hz)*94.6
Peaks Period (seconds)
3 1/(3/(94.6)) = 31.53 s Pulse period
5 1/(5/(94.6)) = 18.92 s 4*Pendulum Period
7 1/(7/(94.6)) = 13.51 s
10 1/(10/(94.6)) = 9.46 s 2*Pendulum Period
15 1/(15/(94.6)) = 6.31 s <---- This unidentified frequency appears on both Top and Middle
18 1/(18/(94.6)) = 5.26 s
20 1/(20/(94.6)) = 4.73 s Pendulum Period
25 1/(25/(94.6)) = 3.78 s
41 1/(41/(94.6)) = 2.31 s 1/2 Pendulum Period
Frobnicated Middle of Fig. 19 page 15 of anomalous NASA report (Mean, Linear Least Squares Fit and Quadratic Least Squares Fit)
Autocorrelation of Middle of Fig. 19 page 15 (from FFT) on raw data detrended by Mean (Blue), by Linear LS (Red) and by Quadratic LS (Green)
Power Spectral Density of Middle of Fig. 19 page 15 (from FFT) on raw data detrended by Quadratic LS (Red)
horizontal scale = frequency(Hz) * 0.1 * (DataLength/2) = frequency(hz)*98
Peaks Period (seconds)
3 1/(3/(98)) = 32.67 s Pulse period
5 1/(5/(98)) = 19.60 s 4*Pendulum Period
7 1/(7/(98)) = 14..00 s
12 1/(13/(98)) = 7.54 s
16 1/(16/(98)) = 6.13 s <---- This unidentified frequency appears strongly on both Top and Middle
22 1/(22/(98)) = 4.45 s Pendulum Period
29 1/(29/(98)) = 3.38 s
34 1/(34/(98)) = 2.88 s
36 1/(36/(98)) = 2.72 s
40 1/(40/(98)) = 2.45 s
42 1/(42/(98)) = 2.33 s 1/2 Pendulum Period
Based on the 4 runs on Fig. 19 top and middle of NASA's Brady et.al. "Anomalous..." report we can estimate NASA Eagleworks pendulum period, using the Power Spectral Density (from Fast Fourier Transforming the data) as follows:
1) Based on the frequencies corresponding to the 1/2 period:
2*Mean[ 2.31, 2.33] = 4.64
1) Based on the frequencies corresponding to the 1/2 period and the frequencies corresponding to the full period:
Mean[2*2.31,2* 2.33, 4.45,4.73] = 4.62
Both estimates are very close (2.6%) to the period given by Paul March (4.5 seconds) -
#2942 by Rodal on 04 Nov, 2014 20:36
-
Thanks. But we need to explain this video before we can attribute the EM Drive thrust in general to chamber wall/cavity interactions.
http://emdrive.com/dynamictests.html
Video clips are near the bottom of the page. I don't see anything in the video that the EM Drive could interact with.
After all this lengthy discussion I have been tracking for a while, what stays with me is the absolute need of more experimental results to talk about. More data points, more confirmations (or refutations).
And the video you bring, while enticing, is alas not enough. I can imagine several ways to trick a video like that, just requiring enough willingness and lack of a consciousness to do it.
Don't get me wrong. I like visual demonstrations as the next guy, it's that we only have this one so far.
But if more people replicated that... we could start becoming really intrigued.
No question that we need more data and more replication.
Operating on the assumption that there is a real force generated by the EM Drive precludes the need to speculate that the data is somehow faked. If we assume that the data is faked and the force is not real then that's the end of the story. No need to go any further.
We here are operating on the assumption that the measurements are real and that the force is from a real EM Drive effect or else it is from an experimental artifact. We are pursuing both lines of investigation so the story ends when we find either the cause of the EM Drive effect or the experimental artifact that fits all of the data we have. Or when we have exhausted the data available. Without more data the latter seems likely at this point.
I was agreeing with your message up to the point where you wrote "Without more data the latter seems likely at this point". I disagree with your likelihood forecast as it is only recently that we started really independently examining the data (*), with Fourier Transform, Periodogram, Autocorrelation, Power Spectrum Density, a full model of NASA Eagleworks torsional pendulum as well as classical models (Frobnicat) to numerically explain the results.
(*) as opposed to taking for granted the author's written interpretation of the data. To our knowledge the authors themselves have not carried out as complete an examination of their data as we are doing (certainly the authors do not have an analytical model for their pendulum behavior as we do, and the author's did not provide an FFT, Periodogram, Autocorrelation, and Power Spectrum of their data as we are doing).
We have shown that the pendulum acts as a complicated filter of what the actual forcing function is, and that it is incorrect to assume that what was measured was necessarily the actual forcing function of the EM Drive.
Unlike the researchers, we are doing this "as a labor of love" on our spare time, so patience is required, but as stated we have already examined the data to a higher level than the published results. -
#2943 by Mulletron on 04 Nov, 2014 20:39
-
@momerathe.... Momentum is a form of energy. Let's not split hairs here.
You're right about rf cancelling out. That's where simultaneously breaking pt symmetry comes in.
Also if they suck the air out of the conical frustum and remove the dielectric the effect should vanish. -
#2944 by aero on 04 Nov, 2014 20:50
-
@Rodal - Ok. But note that there are two perspective angles.
1. Its real, caused by ...
2. Its an artifact, caused by ...
Perhaps the data is more nearly exhausted from one perspective than the other? Still, the two perspectives do go hand in glove. -
#2945 by frobnicat on 04 Nov, 2014 20:57
-
Ok,
Admittedly, the math for this a\has finally gotten WAY beyond me.
Could someone give me an idea of the power to motion ratio that seems to be being generated with this system, verses, say, a regular chemical rocket?
What I am trying to find out is simple; Is this system somehow generating more motion than should be possible, assuming a direct conversion of energy to motion?
In other words, is 1 calorie of energy somehow rasing 1 cubic centimeter of water's temprature higher than 1 degree celcius, or is the amount of power being used within a reasonable ratio of energy effecient conversion, say, 70% of power applied is being converted to motion, as an example?
For the moment, set aside HOW it appears to be doing what it is doing, and let's see if it violates any of the Laws of Thermodynamics.
Sorry, but you guys have gone so far beyond me mathematically, (plus, I think I may have missed a couple of equations that would have made it simpler to follow) that I am having the devil's own time trying to keep up with this thread.
My take on that : it does very clearly violate the laws of thermodynamic unless :
- It gets energy from an "exterior" source. (1/ or 2/ in post linked)
- It emits tachyons (particles of imaginary mass and negative energy) as a way to get rid of an energetic debt. (3/ in post linked)
That it needs more energy than it takes from its onboard generator for a given result can be shown with relatively basic Newtonian mechanics and thought experiment :
Assume a long "railway" in otherwise empty space and a small 1kg ship accelerating along this path by conventionally pushing with wheels. This is the most favorable situation to get momentum from energy. At some time, the ship is going 1km/s=1000m/s (pretty fast for wheels... use whatever magnetic coupling to "rails") and want to go faster. The power it takes to push against a moving thing (the rails are moving relative to the ship) is power=speed*force. Say we want a force of 1N, then it takes 1000*1 = 1000 Watts (1kW) of power. At 1N, a mass of 1kg experiences an acceleration = force/mass = 1/1 = 1m/s˛ (that is roughly 0.1g since 1g=9.81m/s˛). Ok so, at this 1kW power, at this speed of 1000m/s, 1kW gives an added 1m/s each second.
So after 1s at 1kW we now have a 1kg ship at 1001m/s (speed relative to rails). The energy amount of 1kW during 1s is energy_used = power * time = 1000 * 1 = 1000 J. We are far from c speeds so kinetic energy is accurately given by energy_kinetic = 0.5*mass*speed˛.
Before the acceleration : energy_kinetic_before = 0.5*1*1000˛ = 500000 J
After the acceleration : energy_kinetic_after = 0.5*1*1001˛ = 501000.5 J
energy_gained = energy_kinetic_after - energy_kinetic_before = 1000.5 J
So we have used 1000 J and we have gained 1000.5 J Alles ist gut.
All right, there is an annoying .5 J excess which is due to neglecting the fact that it gets a little bit harder to accelerate when going from 1000m/s to 1001m/s. For instance at 1000.5 m/s it would already take 1000.5*1 = 1000.5 W to get the same 1N force that we had at only 1000W when going 1000m/s. So in average we would have to spend exactly 1000.5 J to augment the kinetic energy by 1000.5 J
Now consider the same experiment with a "propellantless system" that would have a thrust to power ratio of 1N/kW, that is in the ballpark of what are claiming the proponents when they are designing their mission profiles. We have the same force of 1N for the same power of 1kW that will get us from 1000m/s to 1001 m/s in 1 second, at the energetic cost of exactly 1000J we gained exactly 1000.5 J in kinetic energy.
This speed of 1000m/s was not chosen randomly by me: the inverse of thrust to power ratio of a propellantless scheme (that is the power to thrust ratio) is the speed beyond which the thruster gives more energy than it takes. It only gets worse (from a conservation of energy standpoint) or better (from a mission profile point of view) when higher speeds are considered :
Same values as above except we are going from 10000 m/s to 10001 m/s :
Gained kinetic energy = 0.5*1*10001˛ - 0.5*1*10000˛ = 10000.5 J
Used energy by conventional wheels = (average_speed*force) * time = (10000.5*1)*1 = 10000.5 J
Used energy by propellantless scheme = power*time = 1000*1 = 1000W (you read well, ten times less)
So in effect, with any propellantless scheme of fixed thrust/power ratio we have over unit efficiencies in converting power to kinetic energy the over unit factor being speed/(power/thrust). And the mission profiles do exploit this over unit factor to the maximum extent permitted by their law (but forbidden by Usual law). And the attempts at explaining how the thrust/power would not be constant but would depend on "speed" to respect energy conservation fail at showing what is the "railway" : what frame would be used as a reference in the vacuum ? Vacuum has no intrinsic "rest speed".
In conclusion, proponents say that energy could be conserved, but they fail to explain how, and worse from an "ethical" standpoint, all their mission profiles critically depend on the fact that energy is not conserved (or is otherwise tapped from "somewhere", the local vacuum or the walls of the Universe, whatever).
-
#2946 by momerathe on 04 Nov, 2014 21:12
-
@momerathe.... Momentum is a form of energy. Let's not split hairs here.
No, it really isn't. They may be mathematically similar concepts, but they are not the same. This is not splitting hairs; without precision of expression we're not going to get anywhere. -
#2947 by Rodal on 04 Nov, 2014 21:17
-
Momerathe, a warm welcome to this forum !
-
#2948 by Mulletron on 04 Nov, 2014 21:26
-
@momerathe.... Momentum is a form of energy. Let's not split hairs here.
No, it really isn't. They may be mathematically similar concepts, but they are not the same. This is not splitting hairs; without precision of expression we're not going to get anywhere.
Fine let's split hairs. If an object has momentum, then it is moving. If it is moving, then it has kinetic energy. And if an object has kinetic energy, then it definitely has mechanical energy... Kinetic energy is a form of energy.
We're not doing a lick of math here. When we do we can use that all you want. We can throw in vector operators when we do math. I can tell that the momentum in a baseball hitting things sure feels like energy to me. -
#2949 by IslandPlaya on 04 Nov, 2014 21:30
-
You are pre-supposing.@momerathe.... Momentum is a form of energy. Let's not split hairs here.
No, it really isn't. They may be mathematically similar concepts, but they are not the same. This is not splitting hairs; without precision of expression we're not going to get anywhere.
Fine let's split hairs. If an object has momentum, then it is moving. If it is moving, then it has kinetic energy. And if an object has kinetic energy, then it definitely has mechanical energy... Kinetic energy is a form of energy.
Who knows what energy actually is? -
#2950 by Rodal on 04 Nov, 2014 21:34
-
....
More info on the Feigel Effect:
http://rspa.royalsocietypublishing.org/content/early/2011/10/06/rspa.2011.0481.full.pdf (A peer review with surprising results!)
http://physics.aps.org/story/v13/st3
The anomalous thrust production from an RF test device was due to the Feigel–van Tiggelen effect.
Interesting paper. Some thoughts occur:
* the dielectric constant in the can will be very low. it's meant to be evacuated, after all.
* the ExB fields in the Feigel effect are steady, those in an EM wave oscillate sinusoidally. Thus the force would be continually swapping direction, and average to zero.
* If you are right, it spells the death-knell for this device as a form of propulsion. From the paper:QuoteFeigel considers the following situation: a region of a dielectric fluid far from
the boundaries of its container is initially at rest (t = 0). Subsequently, strong
electrical and magnetic fields crossed at right angles to each other are applied
to the region. As the fields reach their constant final values, Eext and Bext for
electrical and magnetic fields, respectively, the fluid is accelerated by the Lorentz
forces (FLorentz ∝ vt(Eext × Bext)) to a final velocity v.
It's a one-off impulse. It can't provide a steady thrust (again, I'm not saying that you're one of the people saying it does) unless you can keep increasing the field forever. When you turn the field off, the dielectric will stop moving. Kind of how the casimir effect provides a one-off energy gain when you bring the two plates together.
The same author has a paper published later, where he concludes that the Feigel effect violates the first law of thermodynamics, as there is work without an energy input:
http://arxiv.org/pdf/1304.3338.pdf -
#2951 by Mulletron on 04 Nov, 2014 21:43
-
Good catch. Gotta see what changed their mind. Good thing emdrive only works when it is switched on.
-
#2952 by Rodal on 04 Nov, 2014 21:46
-
Good catch. Gotta see what changed their mind. Good thing emdrive only works when it is switched on.
Good catch from you in finding this unusual stuff in the first place. The Proceeding of the Royal Society of London is an excellent journal with outstanding peer review. Maxwell published his electromagnetic theory in this Journal, Karl Pearson his famous papers on statistics and so on and on... -
#2953 by aero on 04 Nov, 2014 22:24
-
Question - Has anyone thought of looking for correlations between Shawyer's thrust profile and Brady's? Be very interesting if something showed up.
The attached plots are from Shawyer's CEAS 2009 paper.
-
#2954 by Rodal on 04 Nov, 2014 22:57
-
Question - Has anyone thought of looking for correlations between Shawyer's thrust profile and Brady's? Be very interesting if something showed up.
The attached plots are from Shawyer's CEAS 2009 paper.
Good point.
If Frobnicat can scrape the Shawyer data and provide it in a digital form, I could run Power Cross-Spectral Density and Cross-Correlation analysis between them.
However, to my eyesight, the data from Shawyer is very different in that it has that peaking at the end. -
#2955 by Mulletron on 05 Nov, 2014 09:08
-
As notsosureofit keenly mentioned here: http://forum.nasaspaceflight.com/index.php?topic=29276.msg1282610#msg1282610
You really have to treat the entirety of the cavity as a quantum oscillator to make sense of it. A classical approach doesn't seem to work, but a fully quantum approach yields results.
See slide 15 here:
http://qvg2013.sciencesconf.org/conference/qvg2013/program/Donaire_qvg2013.pdf
I really wish I could find more of this.
This is an interesting throwback to way back (page 96) when I mentioned that I treat EVERYTHING as a wave, particle, and field at the same time, even if something is a macro object. It sounds counter intuitive but that is how I choose to operate and I believe I'm correct in doing so.
Relating to this is that it turns out that even macro objects have a very small hard to measure wave function. -
#2956 by Mulletron on 05 Nov, 2014 09:51
-
Well I have to see if "PT symmetry breaking" is actually happening here or not. I gotta try and break this theory as best I can.
-
#2957 by momerathe on 05 Nov, 2014 10:16
-
Momerathe, a warm welcome to this forum !
Cheers. I usually try to stay away from speculative physics threads as they are usually hair-tearingly frustrating, but sometimes they get the better of me.
You are pre-supposing.
Who knows what energy actually is?
generations of scientists?
The important thing to remember is that energy is a property of a system. It is not "stuff", even though we often talk about it as if it is. This is, IMO, a bad habit among science communicators.Relating to this is that it turns out that even macro objects have a very small hard to measure wave function.
Not always small; the current loop in a superconducting wire can be kilometers long and yet have clear quantum characteristics.
The problem is entanglement; macroscopic objects are generally strongly entangled with their surroundings. That means it no longer makes sense to talk about the wavefunction of the object, but instead of the whole system.
(pedantic aside: you cannot measure a wavefunction; it's not an observable.) -
#2958 by Mulletron on 05 Nov, 2014 10:38
-
Not directly observable. Not measurable, aka "hard to measure." Am I going to have to walk on egg shells with my wording here now? This is getting into a philosophical debate about whether or not for something to be real, it has to be/can be observed. I had a long expose on that philosophy of science way back in the thread concerning Ernst Mach.
I operate as if in order to prove something is "real" you only need to observe it. Be it directly or indirectly. You don't necessarily need to measure it.
What is measurable is limited by technology. Reality goes on being what it is, regardless of it is observable or measurable.
-
#2959 by Mulletron on 05 Nov, 2014 12:29
-
Carl M. Bender and Stefan Boettcher seem to be the heavy hitters on PT symmetry. Since I don't fully understand PT symmetry, I have to learn about it.
Useful resources I've found:
http://www.thphys.uni-heidelberg.de/~hofmann/ICNAAM2012/Carl.pdf
http://arxiv.org/pdf/quant-ph/0501052v1.pdf
http://arxiv.org/pdf/physics/9712001v3.pdf
http://www.physics.emory.edu/faculty/boettcher/Publications/papers/jmp40_2201.pdf
http://ptsymmetry.net/?page_id=10
http://arxiv.org/pdf/1304.6767v2.pdf (more info on Casimir Momentum of a Chiral Molecule in a Magnetic Field)
