EM Drive Developments - related to space flight applications - Thread 7

[Rodal]

Dr. Rodal, I lack the expertise to non-concur with anything in your above statement. That said, it should also be impossible to argue with CW's assertion that filtering the RF source to provide consistent frustum input is one key to a repeatable experiment.

Presumably if I have a source from a Panasonic NE-1064 and rfmwguy has an LG LMHM2237ST and Shells is using a Frigidaire FGMV175QF at different ambient temperatures, pressures, AC line voltages, & etc., & etc. - whether or not the effect is due to QV interaction or Unruh radiation, or CU++ spalling, or _________ the experiments are not exactly repeatable (or less so) without knowing the details of the input frequency, bandwidth, power, and so on.

 Are we wrong here?

Your statement that

it should also be impossible to argue ... that filtering the RF source to provide consistent frustum input is one key to a repeatable experiment

presumes that the process is deterministic and linear.  Yes, it is possible to argue otherwise:

Quantum mechanics does not specify the outcome of individual experiments but only the probabilities.  For example, what a single photon is going to do in a slit experiment is not deterministic: it involves probabilities.  The photon is both a particle and a wave.

There are experiments that are not exactly repeatable: they involve a probability distribution, because they involve random variables.

Besides random variables, the process may also involve nonlinearities (not present in elementary Quantum Mechanics problems, but common in macroscopic phenomena in Nature, like in fluid mechanics).

There are experiments whose results are very sensitive to the input variables: they involve nonlinear processes.


Although Shawyer and McCulloch's equations model the process as being linear: the force being a linear function of power input and the quality of resonance Q, previous reports by Star-Drive indicate that the outcome of the plasma QV code is not linearly dependent on Q and not linearly dependent on power input.


"Cleaning" the input to a nonlinear process may ensure that the outcome is null or much reduced.


So, one possible outcome of "filtering the RF souce" is ensuring that the experiment will result in lower force/PowerInput or not a significant level of force.

Yes, having a NULL or low force/powerInput result will be repeatable, but is that the aim of the experimenter? To have a repeatable NULL experimental result?

Further testing will tell the story: what experiments will result in the maximum force/InputPower ?


If the anomalous force is real, is it a linear function of Q and power input as assumed by Shawyer and McCulloch?

Of is it a nonlinear function of Q and power input as shown by Dr.White's QV code?

Is frequency modulation and phase modulation of the input detrimental to the anomalous force (as assumed by those "cleaning" the magnetron input), or are the results of Yang's experiments the real thing, showing that maximum anomalous force is achieved with a magnetron having Amplitude, Frequency and Phase modulation?

[rfmwguy]

I think we need to take a moment to reflect on how fortunate we are to have our hosts provide us a place to debate and discuss this controversial topic. To give you an idea how this and other New Physics topics are treated on another (unnamed) forum. Here's how their hosts treated a student (starting a thread) wanting to experiment with and perhaps debunk the emdrive:

"This thread will remain closed. The EM Drive is on the list of (snip) Forbidden Topics (see the Rules link at the top of the page under INFO). It is crackpot stuff, and not worth the time to debunk. From the Forbidden Topics list:  EMDrive and other reactionless drives Articles suggesting that NASA, the Chinese government, or some other governmental actor is working on such a technology frequently appear in the popular press. These claims have been extensively debunked and are not acceptable references under the (snip) Forums rules."

(snips added to allow this forum to remain anonymous)

Be sure to drop them a note sometime or better yet, sign up for L2 or visit the NSF store.

[Rodal]

I think we need to take a moment to reflect on how fortunate we are to have our hosts provide us a place to debate and discuss this controversial topic. To give you an idea how this and other New Physics topics are treated on another (unnamed) forum. Here's how their hosts treated a student (starting a thread) wanting to experiment with and perhaps debunk the emdrive:

"This thread will remain closed. The EM Drive is on the list of (snip) Forbidden Topics (see the Rules link at the top of the page under INFO). It is crackpot stuff, and not worth the time to debunk. From the Forbidden Topics list:  EMDrive and other reactionless drives Articles suggesting that NASA, the Chinese government, or some other governmental actor is working on such a technology frequently appear in the popular press. These claims have been extensively debunked and are not acceptable references under the (snip) Forums rules."

(snips added to allow this forum to remain anonymous)

Be sure to drop them a note sometime or better yet, sign up for L2 or visit the NSF store.

A topic like the EM Drive is problematic for a forum:

1) If the forum is not moderated or lightly moderated, discussion may involve insults, name calling, etc., as it happens in another forum that you have previously addressed

2) If the forum is strongly moderated, such moderation is very time consuming and problematic: Chris had to shut down the first EM Drive thread and has warned about how problematic is to moderate such threads.

Re: EM Drive Developments Thread
« Reply #3641 on: 12/12/2014 11:37 AM »

Thread back on after at least a trim of bad posts over the last 10 pages or so. I honestly couldn't read back further than that without my brain turning to mush.

Pointless posts (including one post that was basically "LOL") removed. Personal attack posts removed - member removed. Stupid posts removed - member asked not to post on here again.

However, nothing that I read (at least over the pages I looked at) feels like this site's subject matter, so I'm locking it (but putting it back on view) and we'll start a new thread in an attempt to make it relevant to this site. That's the best solution.

Thus the decision by the forum you quote above is understandable, because it is much easier to maintain forum discussion on topics that are well-defined and will not invite people distinguishing between "believers" and "unbelievers", conspiracy theories, or name calling.  It is much easier to have a forum discussion on mathematical problems, for example, on the Poincaré conjecture, on P versus NP, or on the Riemann hypothesis, or on Navier–Stokes solution existence and smoothness.   

[rfmwguy]

Yes, its more difficult to deal with controversial topics, which requires more effort. All the more reason we should thank the NSF staff when we get the chance. In return, they have a broader readership around the planet and perhaps could someday be on the leading edge of a potential new discovery; emdrive or not.

The forum I quoted is not the one you might be thinking, this one is simply dedicated to classical physics. New physics is not a good fit for their business model. No problem, its the reason NSF and others are around...to fulfill a need.

A thought did occur to me, I've known a few people from school who went on to study physics...they all wanted to study and discover something new, not prove/retest theories a century or more old. Seems like a disconnect, but who am I to judge? 

[JaimeZX]

You are presuming that the process is deterministic and linear.

Further testing will tell the story: what experiments will result in the maximum force/InputPower ?

I'm not presuming anything other than "any result should be easier to reproduce if one can understand and reproduce the initial conditions."

If the effect is less (or null) as a result of "cleaning," then that is data. If the effect appears to be strongest with a given test setup, then it behooves us to learn as much as possible about that setup so others can reproduce it and make modifications from that starting point.

[Tetrakis]

Or the effect is a mirage of poor error analysis. But the people arguing that have largely left this thread.

[Rodal]

Or the effect is a mirage of poor error analysis. But the people arguing that have largely left this thread.

I would say that

1) most discussions here involve NO statistical error analysis, rather than "poor error analysis".  Comparisons are often made without any statistical error analysis. "Good error analysis" is not feasible (from the data I have seen so far) because the EM Drive experimental sample population is too small to even properly define what is the actual statistical distribution.  Notice the absence of histograms in experimental reports. It would be just as much unwarranted to naively assume a Gaussian statistical distribution (as naively proposed in another forum) without any experimental basis for such assumption.  To assume that this is a question of P values or Chi-Square analysis, when there is no data showing that the actual distribution is Gaussian. Or to make a P value analysis on such a small sample population.

2) Regarding experimental artifacts, this EM Drive thread has the best discussions in the Internet, for example: Prof. Frobnicat discussion of energy conservation, mathematical modeling of torsion pendulum coupling modes, shifting of the axis, etc., also the discussions involving thermal (convection, conduction and radiation) effects, discussions about cable Lorentz effects,  exact solution and numerical (FEM and BEM) solution of Maxwell's equations, discussions about different experimental methods: torsional pendulum vs. teeter-totter, linear-air-bearing, force measurements by scales, etc.  And those people are still involved.

[A_M_Swallow]

{snip}
I think that there is a profund lack of controlled conditions in respect to signal spectrum integrity, which might produce random results. For some people, it may work - for some time. And for some, it never does. I think, that tightly controlling the HF spectrum is key to repeatable experiments. Otherwise, I think it's not an experiment, but throwing the dishes on the floor and expecting a fine orchestra piece to be playing out. Somewhere in that random spectrum, a nice tune may be hidden. Most of the times, it isn't.
{snip}

You are making the hidden assumption that it is a single frequency that makes the EMDrive work. (Just like the rest of us.) I am reminded of an opera singer breaking a wine glass. He has to use a note containing two tones simultaneously. A treble tone that actually breaks the top of the glass and a base tone to saturate the bass of the glass. By itself each tone will vibrate the glass but not break it.

Generating two or more radio frequencies is not hard but may require extra hardware.

[Tetrakis]

Or the effect is a mirage of poor error analysis. But the people arguing that have largely left this thread.

I would say that

1) most discussions here involve NO statistical error analysis, rather than "poor error analysis".  Comparisons are often made without any statistical error analysis.

2) "Good error analysis" is not possible (from the data I have seen so far) because the EM Drive experimental sample population is too small to even properly define what is the actual statistical distribution.  Notice the absence of histograms in discussions.

3) It would be just as much unwarranted to naively assume a Gaussian statistical distribution (as naively proposed in another forum) without any experimental basis for such assumption.  To naively assume that this is a question of P values or Chi-Square, analysis, when there is no data showing that the actual distribution is Gaussian.

Agreed. I guess I would just say that I've seen many such mirages in my own experimental science (chemistry). When your data tells you something thats "impossible", it almost always is. The right thing to do is to carefully repeat the experiment, perform control experiments, and systematically modify variables. Analyze all aspects of the experiment. Try to collect irrefutable evidence. The height of the "bar" for evidence gets higher when the result is more impossible. In the end these mirages almost always disappear. This sort of thing, exactly like the EMDrive, happens every day in practical science. You don't publish these results because they are almost certainly the result of simple or hidden errors. There are probably 10 of these wild error goosechases for every significant result in the lab. If the literature was full of them, the literature wouldn't be very useful. 

A simple example:

Chemical yield of 1000%? Which is the more likely answer: mass has popped mysteriously into existence from the quantum vacuum, or the compound is hygroscopic and has absorbed water from the atmosphere? Which is more worth your time as a chemist?

I think that the EMDrive gets so much attention from non-scientists for a few reasons. Engineers, especially software engineers, don't often appreciate this error-goosechase side of practical science. The EMDrive is just "complicated" enough that a non-expert can let their imagination (preferred reference frame, free energy, etc.) get the best of them. And finally it lends itself perfectly to the conspiracy theory crowd perfectly: the "man" (sometimes called "Deniers" on this NSF thread) is trying to keep a revolutionary result secret for personal gain.

[Tetrakis]

2) Contrary to what you state concerning "people leaving", regarding experimental artifacts, this EM Drive thread has the best discussions in the Internet, for example: Prof. Frobnicat discussion of energy conservation, mathematical modeling of torsion pendulum coupling modes, shifting of the axis, etc., also the discussions involving thermal (convection, conduction and radiation) effects, discussions about cable Lorentz effects,  exact solution and numerical (FEM and BEM) solution of Maxwell's equations, discussions about different experimental methods: torsional pendulum vs. teeter-totter, linear-air-bearing, force measurements by scales, etc.[/b]  And those people are still involved.

I do apologize. I realize that there are still people posting critical analysis here but most participants act as though Shawyer, Yang, etc.'s data stands on its own merit. Most people say that they are just out to prove them right or wrong, but even that level of attention elevates their frankly wild claims over other wild claims.

[rfmwguy]

2) Contrary to what you state concerning "people leaving", regarding experimental artifacts, this EM Drive thread has the best discussions in the Internet, for example: Prof. Frobnicat discussion of energy conservation, mathematical modeling of torsion pendulum coupling modes, shifting of the axis, etc., also the discussions involving thermal (convection, conduction and radiation) effects, discussions about cable Lorentz effects,  exact solution and numerical (FEM and BEM) solution of Maxwell's equations, discussions about different experimental methods: torsional pendulum vs. teeter-totter, linear-air-bearing, force measurements by scales, etc.[/b]  And those people are still involved.

I do apologize. I realize that there are still people posting critical analysis here but most participants act as though Shawyer, Yang, etc.'s data stands on its own merit. Most people say that they are just out to prove them right or wrong, but even that level of attention elevates their frankly wild claims over other wild claims.

Dr Rodal is one of the leading critics AND supporters of experimentation and theory on this device. Its a rare situation where someone doesn't offhandedly dismiss efforts yet keeps the conversation as grounded as it can be.

I noticed you consider this topic pseudoscience. Its possible many agree here, but they do not use the unnecessary lingo.

<edit - multiple typos...I'm busy  >

[JaimeZX]

<edit - multiple typos...I'm busy  >
True success is never achieved without the mastery of communication skills.

I love the irony here.   

[MaxIsp]

Sorry, It was impossible to resist 

http://www.barnlightelectric.com/industrial-decor/industrial-home-utility/the-moscow-mule-copper-cop-8oz.html



Product Details:
Measurements: 4"W x 3 ¼"H
Material: Solid Copper

I found this a couple of months ago in searching for copper fabricators but did not think to post here. Thank you for doing so. This looks like it would possibly be a 5.8GHZ band resonator and almost fit in a CubeSat.

[tchernik]

2) Contrary to what you state concerning "people leaving", regarding experimental artifacts, this EM Drive thread has the best discussions in the Internet, for example: Prof. Frobnicat discussion of energy conservation, mathematical modeling of torsion pendulum coupling modes, shifting of the axis, etc., also the discussions involving thermal (convection, conduction and radiation) effects, discussions about cable Lorentz effects,  exact solution and numerical (FEM and BEM) solution of Maxwell's equations, discussions about different experimental methods: torsional pendulum vs. teeter-totter, linear-air-bearing, force measurements by scales, etc.[/b]  And those people are still involved.

I do apologize. I realize that there are still people posting critical analysis here but most participants act as though Shawyer, Yang, etc.'s data stands on its own merit. Most people say that they are just out to prove them right or wrong, but even that level of attention elevates their frankly wild claims over other wild claims.

This is quite precisely what I like the most about this forum: you can come, say this is all pseudoscience and exactly the very many reasons why you think it is, and nobody will engage in name calling you because of your thinking (and be tolerated by the moderators). Contrarian opinions can come and do the same too. Ideas here have to live because of their merit, which in this case comes from experimental results, not just words; and specially not because of any social manipulation tricks we humans are so apt at playing.

The reference Dave made to forums where there are Forbidden Ideas not to be discussed because they obviously are pseudo-science is the kind of groupthink, echo-chamber mentality that makes me cringe.

This possibility of having contrarian views so expressed in relative safety and civility (just not the safety of not finding any disagreement with your ideas) is a rare gift indeed.

[rfmwguy]

<edit - multiple typos...I'm busy  >
True success is never achieved without the mastery of communication skills.

I love the irony here.   

You got that right! Back to work...its builder's season!

[rfmwguy]

EMDrive talk on "new age" science show starts about 29 minutes in. NSF and Chris Bergin get a plug @ about 33 minutes in.

[SeeShells]

Or the effect is a mirage of poor error analysis. But the people arguing that have largely left this thread.

I would say that

1) most discussions here involve NO statistical error analysis, rather than "poor error analysis".  Comparisons are often made without any statistical error analysis.

2) "Good error analysis" is not possible (from the data I have seen so far) because the EM Drive experimental sample population is too small to even properly define what is the actual statistical distribution.  Notice the absence of histograms in discussions.

3) It would be just as much unwarranted to naively assume a Gaussian statistical distribution (as naively proposed in another forum) without any experimental basis for such assumption.  To naively assume that this is a question of P values or Chi-Square, analysis, when there is no data showing that the actual distribution is Gaussian.

Agreed. I guess I would just say that I've seen many such mirages in my own experimental science (chemistry). When your data tells you something thats "impossible", it almost always is. The right thing to do is to carefully repeat the experiment, perform control experiments, and systematically modify variables. Analyze all aspects of the experiment. Try to collect irrefutable evidence. The height of the "bar" for evidence gets higher when the result is more impossible. In the end these mirages almost always disappear. This sort of thing, exactly like the EMDrive, happens every day in practical science. You don't publish these results because they are almost certainly the result of simple or hidden errors. There are probably 10 of these wild error goosechases for every significant result in the lab. If the literature was full of them, the literature wouldn't be very useful. 

A simple example:

Chemical yield of 1000%? Which is the more likely answer: mass has popped mysteriously into existence from the quantum vacuum, or the compound is hygroscopic and has absorbed water from the atmosphere? Which is more worth your time as a chemist?

I think that the EMDrive gets so much attention from non-scientists for a few reasons. Engineers, especially software engineers, don't often appreciate this error-goosechase side of practical science. The EMDrive is just "complicated" enough that a non-expert can let their imagination (preferred reference frame, free energy, etc.) get the best of them. And finally it lends itself perfectly to the conspiracy theory crowd perfectly: the "man" (sometimes called "Deniers" on this NSF thread) is trying to keep a revolutionary result secret for personal gain.

Well said!

I will not publish any data until I've felt I've exhausted the device, configurations, test beds and narrowed down the reason(s) that the thrust still persists.

I'm really not supergirl or can see anything others can't, but I am determined to pick this apart bit by bit.

I'm currently on my third iteration of incorporating a new testing bed onto the bench.
I will be able to test a teeter todder force generation, acceleration and also a torsion and rotational actions using the same drive and same test bed. Kudos to Dr. Rodal and to rfmwguy for giving me the flash to do this. It has some distinct advantages by being able to combine all the data sets it is capable of.

I had a frame from one of our pieces of semiconductor equipment we designed and never used with re-enforced angle iron construction.  I'm modifying it to be able to keep almost all of the variables the same except measuring.

Using piano wire stretched between two points and be similar to what I posted a month ago. By rotating the entire "box" 90⁰ I can now have a torsion rotational measurement and even thrust measurements.

Sorry this it taking so much time but it's not a simple chemistry problem. 

Back to work.


Shell

[VAXHeadroom]

EMDrive talk on "new age" science show starts about 29 minutes in. NSF and Chris Bergin get a plug @ about 33 minutes in.

I'm very involved with TMRO and have been a guest 6 times.  I was at Hershey Park on Saturday.  I'm quite embarrassed for them, this show was not up to their usual standard - they didn't even pronounce 'magnetron' correctly.

It's not really 'new age' science, it's mostly space and space news (where they do have expertise).

I am, in fact, here on the forum because they asked me to do a short video on the EMDrive last August.  It's taken me until about now to be able to understand most of the physics discussed here   I'm going to respond to their show, probably with a short video which I hope will get played next week...

 

[X_RaY]

Or the effect is a mirage of poor error analysis. But the people arguing that have largely left this thread.

I would say that

1) most discussions here involve NO statistical error analysis, rather than "poor error analysis".  Comparisons are often made without any statistical error analysis.

2) "Good error analysis" is not possible (from the data I have seen so far) because the EM Drive experimental sample population is too small to even properly define what is the actual statistical distribution.  Notice the absence of histograms in discussions.

3) It would be just as much unwarranted to naively assume a Gaussian statistical distribution (as naively proposed in another forum) without any experimental basis for such assumption.  To naively assume that this is a question of P values or Chi-Square, analysis, when there is no data showing that the actual distribution is Gaussian.

Agreed. I guess I would just say that I've seen many such mirages in my own experimental science (chemistry). When your data tells you something thats "impossible", it almost always is. The right thing to do is to carefully repeat the experiment, perform control experiments, and systematically modify variables. Analyze all aspects of the experiment. Try to collect irrefutable evidence. The height of the "bar" for evidence gets higher when the result is more impossible. In the end these mirages almost always disappear. This sort of thing, exactly like the EMDrive, happens every day in practical science. You don't publish these results because they are almost certainly the result of simple or hidden errors. There are probably 10 of these wild error goosechases for every significant result in the lab. If the literature was full of them, the literature wouldn't be very useful. 

A simple example:

Chemical yield of 1000%? Which is the more likely answer: mass has popped mysteriously into existence from the quantum vacuum, or the compound is hygroscopic and has absorbed water from the atmosphere? Which is more worth your time as a chemist?

I think that the EMDrive gets so much attention from non-scientists for a few reasons. Engineers, especially software engineers, don't often appreciate this error-goosechase side of practical science. The EMDrive is just "complicated" enough that a non-expert can let their imagination (preferred reference frame, free energy, etc.) get the best of them. And finally it lends itself perfectly to the conspiracy theory crowd perfectly: the "man" (sometimes called "Deniers" on this NSF thread) is trying to keep a revolutionary result secret for personal gain.

Well said!

I will not publish any data until I've felt I've exhausted the device, configurations, test beds and narrowed down the reason(s) that the thrust still persists.

I'm really not supergirl or can see anything others can't, but I am determined to pick this apart bit by bit.

I'm currently on my third iteration of incorporating a new testing bed onto the bench.
I will be able to test a teeter todder force generation, acceleration and also a torsion and rotational actions using the same drive and same test bed. Kudos to Dr. Rodal and to rfmwguy for giving me the flash to do this. It has some distinct advantages by being able to combine all the data sets it is capable of.

I had a frame from one of our pieces of semiconductor equipment we designed and never used with re-enforced angle iron construction.  I'm modifying it to be able to keep almost all of the variables the same except measuring.

Using piano wire stretched between two points and be similar to what I posted a month ago. By rotating the entire "box" 90⁰ I can now have a torsion rotational measurement and even thrust measurements.

Sorry this it taking so much time but it's not a simple chemistry problem. 

Back to work.


Shell

I am curious to see your future results however positive or negative. As you state, there is no bad data

[jmossman]

(...) the experiments are not exactly repeatable (or less so) without knowing the details of the input frequency, bandwidth, power, and so on.

(...)
There are experiments that are not exactly repeatable: they involve a probability distribution, because they involve random variables.

Besides random variables, the process may also involve nonlinearities (not present in elementary Quantum Mechanics problems, but common in macroscopic phenomena in Nature, like in fluid mechanics).

There are experiments whose results are very sensitive to the input variables: they involve nonlinear processes.

(...)

So, one possible outcome of "filtering the RF souce" is ensuring that the experiment will result in lower force/PowerInput or not a significant level of force.[/b]

Yes, having a NULL or low force/powerInput result will be repeatable, but is that the aim of the experimenter? To have a repeatable NULL experimental result?

From my perspective, I can understand and appreciate both concerns:  (a) simplify experimental inputs versus (b) avoid assumptions which may nullify or attenuate the signal being sought in said experiment.

While it's been mentioned many times in previous threads, perhaps now is a good time to reiterate a mantra that will help improve the quality of any experiment:  repetition, lots and lots of repetition.   

Unfortunately, DIY budgets make true automation of all of these experiments difficult...  but the differences in approach of (a) versus (b) start to blur if sufficient automation and data logging are possible.  Approach (a) can systematically alter the inputs and look for results, while (b) can let the magnetron do-its-thing and rely on post-processing to potentially help provide and/or improve "repeatability".

(and to be clear, the data logging would ideally need to include some kind of real-time signal spectrum analysis, along with a multitude of other variables...  all very challenging on DIY budgets)

Best of luck to all of the DIY experimenters...  and my thanks to all of the many people contributing their time to help guide the technical and scientific discussion on this very controversial topic.  Perhaps the eventual conclusion will be mundane, but it's one heck of a ride.