-
#960 by rfmwguy on 30 Dec, 2015 20:15
-
You nailed it once again. Its unfortunate that some of the negative aspects include public ridicule...but that's life in today's interwebs. Its not for the insecure to even delve into this project. Several science authorities have written it off.(snip)
However, if it does compel a new university to seek out proof or explain the measurement error...I can live with that
I think that in the end we agree. There is certainly some amount of risk aversion. Nobody is going to undertake the risk unless:
1) The uncertain payoff is large enough to justify the risk (this is a very subjective choice: some individuals are willing to bet in Vegas, other individuals are very conservative and will not even trade in stocks)
or
2) They love what they are doing and are willing to withstand the negative aspects of undertaking the risk involved -
#961 by rfmwguy on 30 Dec, 2015 20:24
-
rfmwguy,
That's very good insight and something Chris B and I had briefly talked about. This humble topic+ has sprouted multiple sub-topics onto itself.
Now that this subject has its own section, might I suggest letting this topic spread its wings beyond a single thread?
I think it would be useful to separate some of the broad streams in this thread into their own threads.
Lock this thread and restart it as "EM Drive Construction and Experimentation", purely for the technical discussion side.
Start a thread for the more general "EM Drive theory and discussion", and another for the more specific "EM Drive results and updates (UPDATES ONLY)".
And, since it comes up so often, maybe a general "'New Physics' thrusters and the Conservation of Energy issue" topic.
There's probably more (MEEP howto?) but those 3-4 topics would be good to get started.
Where this was left at is if it continues to grow, it will have its own Heading with topics underneath just like what you said. I think Chris believes it has to pan out as a viable/real commodity for spaceflight first. If it does, look out, your 3-4 topics could easily triple.
If an error measurement discovery was made and proven, thread 6, 7 or whatever would be closed which I think is more than reasonable.
Right now, we're in a waiting mode for NASA and Dresden as well as DIY data releases.
-
#962 by graybeardsyseng on 30 Dec, 2015 20:53
-
The Wright Brothers broke a lot of wood and tore a lot of fabric learning to fly. A few dead Maggies will be a small price to pay.Neutrality in these matter (which are still unresolved) has been losing some ground to advocacy over the past couple of pages. Vouch for quality data above all else.
There is no set theory. Simply we are trying to explain something that defies conventional knowledge. There is no solid reason this should work in GR or even QED or any other branch of science we have a grasp of.
When I started my build I figured this will only take a few months a few dollars, some copper and a old microwave surplus oven. That has drastically changed for I've been shown that good quality data is not only needed but demanded. Demanded to make any kind of impact or progress.
This last test showed me that it needs to be taken up another level... again. It wasn't a failure by any means, it showed me where my mistakes in the build were and where I fell short and where I didn't. So, I'm rebuilding it better than before with digital data in the acquisition of pressure and movements of the EMDrive. More cross checks and more controls. This is not a race this is well defined scientific pursuit of something unknown happening. This is for the data.
To the group here and my supporters who remain lurkers or who just know me, thank you for you inputs, your help brain wise and pocket wise. This is getting very very close, I can feel it.
Here I thought I retired. If I was sitting on a beach daydreaming, I'd be dreaming of doing something like this. It doesn't get better then doing what I'm doing.
Shell
There is a quote I vaguely remember, which I read many decades ago in my teenage years, from **I think** Wilber Wright to the effect that each 'failure', each detour down an alley of reasoning or broken wing or strut - was an opportunity to learn a new fact or unlearn a wrong one.
Dream on DIYers.
-
#963 by Prunesquallor on 30 Dec, 2015 20:58
-
...
As I've stated before, in my mind the salient observation would be thrust in excess of a photon rocket when all significant error sources are removed. This would allow 1) serious investigation of underlying physics and 2) experimental variation of design parameters to understand operational characteristics and potential increases in efficiency. At this point, one can only estimate (as some have) the implications to spaceflight given assumptions of efficiency and operational characteristics.
I just wanted to make sure we weren't losing lock on the fact that the rationale for vacuum testing was to eliminate significant error sources, NOT because the EMDrive would need to be in a vacuum environment for spaceflight applications.
That's reasonable, but you have to compete with a large number of alternative options for spacecraft propulsion.
A photon rocket is a very inefficient form of space propulsion: there is no present interest in photon rockets that I know of in any practical application with today's technology because there are better alternatives.
I would say that although it would be scientifically interesting to exceed a photon's rocket (as it would involve a scientific enigma how could that happen), that practical interest would mean a performance many times better than a photon rocket, of the order of what has been claimed by Shawyer and Yang and what has been used for conceptual missions to the outer planets by White (and there would be another puzzle to solve: constant acceleration at constant power input is presently an unresolved energy conservation enigma; if the performance is not scalable and force is not proportional to power input, the interest in the EM Drive may also dissappear.
...
Agree, although the Joosten/White paper examined only crewed outer planet missions to draw conclusions regarding useful EMDrive thrust/power (and, as you say constant thrust/power). It would be interesting if someone examined challenging robotic missions that were beyond any current propulsion capabilities, say a combined Europa, Enceladus, Titan orbiter/sample return mission. It might be that lower EMDrive thrust/power capabilities would be sufficient, given longer allowable mission times. -
#964 by Paul451 on 30 Dec, 2015 21:01
-
Now that this subject has its own section, might I suggest letting this topic spread its wings beyond a single thread?
Where this was left at is if it continues to grow, it will have its own Heading with topics underneath
I think it would be useful to separate some of the broad streams in this thread into their own threads.
To be clear, I mean to have more topics within the New Physics section. Not to create yet-another-Header-section for EM Drive topics.
It seems to me that it would be much easier for people doing builds to have a builds-only thread forcefully limited to just the technical discussion. And for those wanting to discuss the theory side, not having a burst of discussions of MEEP and frustum dielectric in the middle.
I think people would benefit from not having one bogged down with the other.
[Personally, when the thread is "build" heavy, I try to stay out. I don't want to start yet-another-CoE argument while people are talking about bending metal, I want the builders to have priority.] -
#965 by rfmwguy on 30 Dec, 2015 21:20
-
There are some spin-offs already. From what I understand, a member in good standing is free to start a new topic on whatever they desire. Only problem is (lack of) moderation and visibility of that thread.Now that this subject has its own section, might I suggest letting this topic spread its wings beyond a single thread?
Where this was left at is if it continues to grow, it will have its own Heading with topics underneath
I think it would be useful to separate some of the broad streams in this thread into their own threads.
To be clear, I mean to have more topics within the New Physics section. Not to create yet-another-Header-section for EM Drive topics.
It seems to me that it would be much easier for people doing builds to have a builds-only thread forcefully limited to just the technical discussion. And for those wanting to discuss the theory side, not having a burst of discussions of MEEP and frustum dielectric in the middle.
I think people would benefit from not having one bogged down with the other.
[Personally, when the thread is "build" heavy, I try to stay out. I don't want to start yet-another-CoE argument while people are talking about bending metal, I want the builders to have priority.]
I encourage you to start a new thread, whatever your emdrive subtopic interest is, OR I can. Whichever you prefer. Can't promise any moderation though. -
#966 by rq3 on 30 Dec, 2015 21:22
-
Now that this subject has its own section, might I suggest letting this topic spread its wings beyond a single thread?
Where this was left at is if it continues to grow, it will have its own Heading with topics underneath
I think it would be useful to separate some of the broad streams in this thread into their own threads.
To be clear, I mean to have more topics within the New Physics section. Not to create yet-another-Header-section for EM Drive topics.
It seems to me that it would be much easier for people doing builds to have a builds-only thread forcefully limited to just the technical discussion. And for those wanting to discuss the theory side, not having a burst of discussions of MEEP and frustum dielectric in the middle.
I think people would benefit from not having one bogged down with the other.
[Personally, when the thread is "build" heavy, I try to stay out. I don't want to start yet-another-CoE argument while people are talking about bending metal, I want the builders to have priority.]
I strongly disagree. All information is valuable. A good experimentalist is a "synthesist" who gathers a huge spectrum of information from different disciplines, and the oddest bit of data from the oddest source can provide the "AHA" moment that leads to a change in theory, or a change in the experiment that eventually leads to success (or failure). Often, in fact usually, it's the failure that eventually leads to success. -
#967 by ThereIWas3 on 30 Dec, 2015 21:43
-
I was curious about how important the frustrum angle was, so I did a little radical experiment. I wrote a script that ran MEEP in a loop, changing the cone slope by one degree each time, keeping the small-end diameter, height, and frequency constant. height was about 2 wavelengths. So with each run the large end got bigger and bigger. End plates were flat.
To have this complete in a reasonable amount of time I also considerably simplified the model, taking advantage of "cylindrical symmetry", placing the signal source at the enter of the small end. For now it was a linear source, but I intend to do this over again using a circularly polarized source. The source was located on the central axis, one quarter wavelength from the small end.
Using cylindrical symmetry allowed me to generate pictures for Hr, Hz, Er, and Ez fields, for every angle from 0 to 45 degrees in under one hour. Using 3D cartestian coordinates and no symmetry this would have taken nearly two days. Four videos are shown below, showing how each field pattern changes as the angle increases, one degree per second. Use the zoom feature of your video playback software to make them bigger if necessary. The pictures are all trapezoidal in shape, and show only half of the frustrum. This is because MEEP knows that everything is symmetrical and so only computes that one view. The field names with "r" in them mean "radial".
Analysis: 1. the field patterns are extremely sensitive to cone angle. Radical changes were seen sometimes with only one degree of change. This suggests that experimenters should not choose cone angles at random.
2. The curvature of the fields due to the diverging frustrum walls was clearly evident even at this low resolution. And in some ranges of angle the fields were curved in the reverse direction! I expect this to be considerably affected by the flat end plate. I will do it again later using spherical end-plates. And also with exponential ("trombone") wall shapes.
3. At some angles the 'sign' of the fields reversed completely. (Indicated by blue vs red in the pictures.)
-
#968 by zellerium on 30 Dec, 2015 21:43
-
TT, I did notice all that.. but my Question is WHY ? why do you use a coupling window?
The adjusting screws allow the VSWR seen by the Rf gen to be adjusted as low as possible.
The slot also effects VSWR and acts to decouple the 2 sections of the wave guide.
Later designs put the slot into the side wall, so the cavity is highly decoupled from the wave guide feed. See attachment.
To start with, I suppose it isn't for aesthetic reasons ...
euh... VSWR... I'm not an engineer, but might that be "Voltage Standing Wave Ratio" ?QuoteThe parameter VSWR is a measure that numerically describes how well the antenna is impedance matched to the radio or transmission line it is connected to.
added:
nevermind on the VSWR. Found a good explanation : https://www.maximintegrated.com/en/glossary/definitions.mvp/term/VSWR/gpk/815
In case you haven't seen Yang's 2014 paper, I made an effort to translate it awhile back. With some of these types of experiments, it seems the guess and check method using a VNA works quite well. We saw interesting effects when twisting the inner plate of our cylinder resonator because of the two screws protruding and the holes in the end plate. Sometimes we could tighten the resonance to relatively high Q (~10^3) but it was often too sensitive to maintain.
Microwave Engineering by Pozar 3rd ed has a lot of good info on tuning.
-
#969 by Rodal on 30 Dec, 2015 21:59
-
I was curious about how important the frustrum angle was, so I did a little radical experiment. I wrote a script that ran MEEP in a loop, changing the cone slope by one degree each time, keeping the small-end diameter, height, and frequency constant. height was about 2 wavelengths. So with each run the large end got bigger and bigger. End plates were flat.
To have this complete in a reasonable amount of time I also considerably simplified the model, taking advantage of "cylindrical symmetry", placing the signal source at the enter of the small end. For now it was a linear source, but I intend to do this over again using a circularly polarized source. The source was located on the central axis, one quarter wavelength from the small end.
Using cylindrical symmetry allowed me to generate pictures for Hr, Hz, Er, and Ez fields, for every angle from 0 to 45 degrees in under one hour. Using 3D cartestian coordinates and no symmetry this would have taken nearly two days. Four videos are shown below, showing how each field pattern changes as the angle increases, one degree per second. Use the zoom feature of your video playback software to make them bigger if necessary. The pictures are all trapezoidal in shape, and show only half of the frustrum. This is because MEEP knows that everything is symmetrical and so only computes that one view. The field names with "r" in them mean "radial".
Analysis: 1. the field patterns are extremely sensitive to cone angle. Radical changes were seen sometimes with only one degree of change. This suggests that experimenters should not choose cone angles at random.
2. The curvature of the fields due to the diverging frustrum walls was clearly evident even at this low resolution. And in some ranges of angle the fields were curved in the reverse direction! I expect this to be considerably affected by the flat end plate. I will do it again later using spherical end-plates. And also with exponential ("trombone") wall shapes.
3. At some angles the 'sign' of the fields reversed completely. (Indicated by blue vs red in the pictures.)
Having written finite element and finite difference codes for transient response, as well as used commercial codes (ANSYS, ADINA, etc.etc) to solve nonlinear problems (Meep's equations are linear, which is simpler) and compared them to numerous experiments throughout my life:
Congratulations for carrying this study, which is part of the needed verification (a sensitivity analysis) to gain confidence in any numerical model (are the nodes fairly equidistant from each other? (it is known that misshapen distances between nodes leads to ill-conditioning of the matrix for simultaneous solution of equations, and hence to severe inversion problems of the matrix , is the mesh fine enough?, is the central difference time step small enough?, are the number of digits of precision large enough for the amount of time steps desired?, is the final time step late enough to correspond to the solution we are seeking for ? are the constitutive properties fully characterizing the real problem? are the boundary conditions imposed correctly? --Garbage In-Garbage out-)
Please consider whether to interpret the extreme dependence on angle (1 degree ) as possibly due to a Meep numerical model issue and not necessarily as a real physical effect (not a Meep problem, but a Finite Difference Model problem). It may not relate to actual physical experiments. As I discussed previously, the Meep model is a transient marching problem representing a small fraction of a microsecond (1% of a microsecond for several Aero runs) while as previously discussed (including by TheTraveller using a neat simple argument) tens of microseconds (thousands of times longer runs) are needed to come close to describe the formation of standing waves. The very early response may consist of malformed travelling waves that are very dependent on initial conditions and finite-difference mesh. There are questions of:
1) Convergence of the Meep Finite Difference solution ( I have not seen anyone conduct a mesh convergence study)
2) Transient response (only 1% of a microsecond, typically)
exhibiting, as discussed by me and further analysis by Frobnicat:
3) Exponential growth of electromagnetic fields, nowhere close to stability
the reason for the exponential growth has not been adequately addressed.
No comparison of the Meep model fields vs experiments and other models (I had suggested the TM212 thermal experimental results and the COMSOL FEA analyis at NASA for comparison) has been performed, to my knowledge, concerning these issues.
Simply stating that the resonant frequency is some number and that the Q is another number is not enough to validate the model, unfortunately, particularly when looking at the fields (this is known from all numerical solutions, as certain quantities converge much faster than others, typically fields are much more problematic to converge because numerical derivatives are very ill-conditioned, while numerical variables that depend on integration are numerically well-conditioned).
By contrast, I understand that the COMSOL FEA studies by NASA have been eigenvalue solutions, instead of a transient time solution.
-
#970 by Paul451 on 30 Dec, 2015 23:30
-
From what I understand, a member in good standing is free to start a new topic on whatever they desire. Only problem is (lack of) moderation and visibility of that thread.
The issue is pushing topics off of this thread, onto the others.
[Would help more if there was a consensus from the regulars that having a separate thread for theory-debates and another for building/meep-sims/etc is actually in any way desirable to them.]I strongly disagree. All information is valuable. A good experimentalist is a "synthesist" [....]
Not seeing what your objection is. I've not suggested hiding information. -
#971 by rfmwguy on 30 Dec, 2015 23:35
-
From what I understand, a member in good standing is free to start a new topic on whatever they desire. Only problem is (lack of) moderation and visibility of that thread.
The issue is pushing topics off of this thread, onto the others.
[Would help more if there was a consensus from the regulars that having a separate thread for theory-debates and another for building/meep-sims/etc is actually in any way desirable to them.]I strongly disagree. All information is valuable. A good experimentalist is a "synthesist" [....]
Not seeing what your objection is. I've not suggested hiding information.
Sure, we run the show here zo if you gain consensus, I have no issue referring subjects to another thread. I physically cannot move them to my knowledge and wouldn't want to. -
#972 by TheUberOverLord on 31 Dec, 2015 00:18
-
Now that this subject has its own section, might I suggest letting this topic spread its wings beyond a single thread?
Where this was left at is if it continues to grow, it will have its own Heading with topics underneath
I think it would be useful to separate some of the broad streams in this thread into their own threads.
To be clear, I mean to have more topics within the New Physics section. Not to create yet-another-Header-section for EM Drive topics.
It seems to me that it would be much easier for people doing builds to have a builds-only thread forcefully limited to just the technical discussion. And for those wanting to discuss the theory side, not having a burst of discussions of MEEP and frustum dielectric in the middle.
I think people would benefit from not having one bogged down with the other.
[Personally, when the thread is "build" heavy, I try to stay out. I don't want to start yet-another-CoE argument while people are talking about bending metal, I want the builders to have priority.]
I strongly disagree. All information is valuable. A good experimentalist is a "synthesist" who gathers a huge spectrum of information from different disciplines, and the oddest bit of data from the oddest source can provide the "AHA" moment that leads to a change in theory, or a change in the experiment that eventually leads to success (or failure). Often, in fact usually, it's the failure that eventually leads to success.
This is a good read and I think it echoes your statements:
http://www.huffingtonpost.com/david-h-bailey/data-vs-theory-the-mathem_b_8886292.html?ncid=txtlnkusaolp00000592
Don -
#973 by TheTraveller on 31 Dec, 2015 00:31
-
Thank your for confirming this most important information:
* Yang has retired
* Yang can NOT get recognition of the academic committee
* Yang has NO further project funding
* Yang's project grounded to a halt in 2014
* Yang cannot explain the conflict between her "theory" with existing physical theory
Seems you missed listing one item?
* Yang confirmed the results of the experiment
-
#974 by SeeShells on 31 Dec, 2015 00:36
-
One other question I would like to ask if our new builder from China is reading this. What happened to all the research data that Dr. Yang produced?...
As I've stated before, in my mind the salient observation would be thrust in excess of a photon rocket when all significant error sources are removed. This would allow 1) serious investigation of underlying physics and 2) experimental variation of design parameters to understand operational characteristics and potential increases in efficiency. At this point, one can only estimate (as some have) the implications to spaceflight given assumptions of efficiency and operational characteristics.
I just wanted to make sure we weren't losing lock on the fact that the rationale for vacuum testing was to eliminate significant error sources, NOT because the EMDrive would need to be in a vacuum environment for spaceflight applications.
That's reasonable, but you have to compete with a large number of alternative options for spacecraft propulsion.
A photon rocket is a very inefficient form of space propulsion: there is no present interest in photon rockets that I know of in any practical application with today's technology because there are better alternatives.
I would say that although it would be scientifically interesting to exceed a photon's rocket (as it would involve a scientific enigma how could that happen), that practical interest would mean a performance many times better than a photon rocket, of the order of what has been claimed by Shawyer and Yang and what has been used for conceptual missions to the outer planets by White (and there would be another puzzle to solve: constant acceleration at constant power input is presently an unresolved energy conservation enigma; if the performance is not scalable and force is not proportional to power input, the interest in the EM Drive may also dissappear).
At one point, shortly after the article that I co-authored with Mulletron et.al. in NSF, there was so much interest that a known venture capitalist was attracted to our discussions at NSF, to discuss an X-Prize, unfortunately there was a loss of interest in the X-Prize when it came to define on what merits to award it
The:
* poor results reported at TU Dresden, and Tajmar's remarks on the EM Drive
* news now that Yang's project came to a close in 2014 due to lack of confidence from Chinese academicians
acts as a further damper...
we need to hear from NASA Glenn.
Shell -
#975 by TheTraveller on 31 Dec, 2015 00:42
-
Time is a wasting
Have decided to push ahead as rapidly as possible with the build of the spherical high Q 2.45 GHz frustum, powered by the resonance tracking 100W Rf amp, on the rotary test rig. The work on the maggie freq spatter control and the maggie in the big end plate build is on hold for another day.
Believe I have the best change to get a good result following this pathway than splitting my efforts between the maggie unit and the solid state unit.
I need to understand a bad blood test will put me into chemo and on other drugs, which will kill my research. So while I'm reasonable healthy, it is time to engage the pathway with the least unknowns. As Shell is going the maggie pathway, we should be able to compare notes and results from our different approaches.
It is New Years eve in Australia, so to all on NSF, Have a Good New Years celebration and may 2016 see us all in good health and with good EmDrive results. Cheers. -
#976 by TheTraveller on 31 Dec, 2015 00:46
-
One other question I would like to ask if our new builder from China is reading this. What happened to all the research data that Dr. Yang produced?
Shell
Shell,
Have emailed him / her and offered to produce a Prof Yang archive of all her published work and any additional information they can provide. We may get a few pictures and actual frustum dimensions.
We are trusting the information is correct and the post was not just an attempt to sell spun Chinese frustums. Guess time will tell.
Have also asked Roger his opinion on the "News".
Phil -
#977 by VAXHeadroom on 31 Dec, 2015 01:11
-
*hides eyes in shame*
Here's the rendered video. YouTube may still be processing, it should be available at 1080p. 19.5 hours to render the POV-Ray files. This is probably the coolest raytraced animation I've ever done
I'm so embarrassed - I made SUCH a rookie error in rendering the videos I've been doing.
In order to make the development go faster I only output every 10th row and column...and then forgot I had done it. I've only been showing 1% of the data! It really looks like the attached, and will take overnight to render. I can't wait to see this in motion...
P.S. recalculated. At 2.5 min/frame and 560 frames it'll be done in about 23 hours...
Still trying to figure out how to show just the surfaces, and specifically the thermal profile of the end plates. Stay tuned
-
#978 by Space Time Engineer on 31 Dec, 2015 01:28
-
Just scrolling through all the conversations here from the last few days and I am a little disappointed with the philosophical debates of good science, bad science, "my experience is x", "my experience is Y", "Shawyer is crazy", "academics and risk", arguing over semantics and how something was said, etc etc..... Like a massive intellectual pride based pissing contest.
Can I ask a major favor? Please continue to be committed to reporting data, discussing designs, and potential theories of operation. I think this forum needs to tone down the emotions and turn up the science. For the love of Einstein, don't let this become Reddit. All this back and forth detracts from the science.. Leave emotion and personal pride at the door.
Just MHO, I could be wrong, but its hard to continue to follow the thread when we have the multi message debates on she said/he said semantic based arguments.
That said, I am patiently awaiting SHELL's data, TT's data, and NASA reports.
Carry on.
Dr B
-
#979 by ThereIWas3 on 31 Dec, 2015 01:38
-
Quote from: Rodal
No comparison of the Meep model fields vs experiments and other models (I had suggested the TM212 thermal experimental results and the COMSOL FEA analyis at NASA for comparison) has been performed, to my knowledge, concerning these issues.
If you think Meep is inadequate for this, then I will not waste any more time on it. I have no idea how Meep models could be validated to satisfy your questions.
