-
#3380 by FattyLumpkin on 01 Jul, 2016 04:25
-
This is the sim of Dave's 2nd gen build ...if memory serves TM013. Hats off to Monomorphic for this!!!!
-
#3381 by Chrochne on 01 Jul, 2016 05:15
-
Criticism is never healthy.
One is certainly free to express and hold this uniquely binary viewpoint about experimentation. After all, they designed 'safe spaces' in order to maximize the echo chamber effect for those people who prefer no criticism at all.
Either the experimental device works or it doesn't. Either the experimenter understands why, or else the experimenter doesn't understand why.
So far, the reported experiments haven't floated a device across the conference table at the investor meeting.
It is generally 'safe' to ignore trolls who denigrate one's work in uncharitable terms. Do not feed the trolls, and the site bandwidth will be saved from unnecessary personal nonsense.
Again, one is free to ignore helpful criticism. good luck with that approach.
Finally, many thanks above for the "learn something new every day" moment regarding Lilith. The John Collier painting of her shown on the oracle is inspiring.
I understand what you said and I accept your criticism of me. Thank you.
But how can you continue your healthy criticism when you chase away people that put their effort to construct and test the devices that you and others for various resons can not and will not build and test?
The criticism than lose on its purpose.
-
#3382 by Slyver on 01 Jul, 2016 06:52
-
I understand what you said and I accept your criticism of me. Thank you.
But how can you continue your healthy criticism when you chase away people that put their effort to construct and test the devices that you and others for various resons can not and will not build and test?
The criticism than lose on its purpose.
Analysis of an experiment, with questions, suggestions, or suppositions is different than disregard for an experiment because the criticizer believes a priori that it can not work, though both are often considered "criticism".
In the latter case, it is extremely rare that anyone has a criticism of this type on this forum. If they have it on other forums, who cares? That is not this community, so go address it there instead of here.
In the former case, if the analysis is coupled with specifics of physics and/or math that can be addressed in future experiments, or disproven by further theoretical analysis than it is of the greatest boon to both science and the experimenter. Despite this, it is often taken as the latter form of criticism for a few possible reasons. These reasons include:
The experimenter
-- feels that insufficient attention was given to the analysis that was performed (his work has been ignored)
-- has insufficient knowledge to address the analysis theoretically
-- has insufficient funds to address the issue experimentally and doesn't want everything he has done to be ignored
-- feels that a "burden of proof" has been set upon his shoulders without even an order of magnitude calculation provided to show that the criticism is even a real issue, even if the physics of the criticism has merit.
For the first three, I suggest to the experimenter:
-- Don't worry if your work is not acknowledged, it truly doesn't matter. You are doing what you are doing for yourself, and potentially for all of mankind, not for the individual that is not giving you the props you feel you deserve.
-- Admit when you don't know. Ask for help, ask for proof, look things up, try to understand the criticism so it can be addressed in future experiments. This is how science works. Its how we human beings do truly great things.
-- Say you don't have the money, ask if there is another way to address the issue within the scope of what you can do.
For the last, I see this one as a real problem. People all too often give a criticism (analysis/known physics protest), which has merit, but they don't take the time to even guess at the order of magnitude of the contribution. They put it on the experimenter, that's not fair. These are people doing us all a service. Work with them, don't put everything on them.
This wonderful community has both experienced experimenters and analysts. Both sides can occasionally "take things personal" or not appreciate the work that has been done by the other side. If I have a criticism, it is that both sides do not often enough appreciate the other in every post.
We can do great things with genuine mutual respect. -
#3383 by TheTraveller on 01 Jul, 2016 11:15
-
Here is Daves latest data with work underway to improve the torsion beam's return dynamics.
Also attached is the data for the 1st chart and a photo of his test rig.
I have suggested to Dave that the power wire feed may be causing the issues with the return as it also twists and untwists as the torsion beam rotates and will alter the return dynamics.
As the frustum has a high Q and narrow bandwidth plus the maggie freq drops as it warms up, there is only a limited time period when the maggie freq splatter is close enough to be pulled to the frustum resonance freq, compounded by the power delivered during a freq pull being proportional to the amount of pull. So even when the maggie freq locks to the frustum, the power delivered is not constant.
Such is the cost of powering frustums with non cooled / non temperature regulated maggies. Please note Roger's Demonstrator EmDrive maggie was cooled and temperature plus freq controlled to eliminate this maggie freq pulling variable power effect.
-
#3384 by TheTraveller on 01 Jul, 2016 11:46
-
This is the sim of Dave's 2nd gen build ...if memory serves TM013. Hats off to Monomorphic for this!!!!
As far as I know, Dave's current frustum was designed for TE013 mode but excited in TM113 mode because of the antenna postion being such it only excites TM modes. Both modes resonant at the same freq with the same dimensions. Mode excited depends on the antenna placement.
These 2 modes are the only 2 I know of that exhibit this duality. -
#3385 by TheTraveller on 01 Jul, 2016 11:55
-
...And it is evident from rfmguy's picture that at the moment he can have no idea of what mode shape (if any) is being excited...
Dave's VNA S11 rtn loss scan shows a very high Q dip where TE013 resonance would be expected. But his antenna placement can't excite TE013 but can excite TM113 which exists at the same resonant freq as TE013.
He also knows the cold maggie freq is above this resonant freq and that his maggie needs to warm up a bit, to drop it's freq to get pull lock with the high Q frustum, which can be seen on his 1st test run as attached.
Please note in this zoomed in image, AC filament current and DC anode current flowed over the feed wires for approx 30 sec, until the maggie warmed, freq dropped enough to get lock and force was generated. If you look close you can see a very slight drop in the cold base line as DC anode current flowed and probably generated a very small Lorentz force.
Should also add that over those 1st 30 seconds, there was 900Wrf being injected into the frustum by a maggie antenna INSIDE the frustum, yet there is no sign of any movement of the base line as the frustum heated up prior to achieving freq lock and generating a "Shawyer Effect" thrust output.
At lease in those 1st 30 secs, starting with a ambient temp maggie,, there doesn't appear to be any significant Lorentz or thermal forces active as the base line is flat.
-
#3386 by Monomorphic on 01 Jul, 2016 12:47
-
Dave's VNA S11 rtn loss scan shows a very high Q dip where TE013 resonance would be expected. But his antenna placement can't excite TE013 but can excite TM113 which exists at the same resonant freq as TE013.
The COMSOL sim to the right does not indicate the antenna location. In the FEKO sim to the left, I did simulate the antenna. With Dave's antenna placement, to me it looks like TM013 with a slight distortion in the immediate vicinity of the antenna. TM113 is 33Mhz away and looks very different.
-
#3387 by TheTraveller on 01 Jul, 2016 12:57
-
Dave's VNA S11 rtn loss scan shows a very high Q dip where TE013 resonance would be expected. But his antenna placement can't excite TE013 but can excite TM113 which exists at the same resonant freq as TE013.
The COMSOL sim to the right does not indicate the antenna location. In the FEKO sim to the left, I did simulate the antenna. With Dave's antenna placement, to me it looks like TM013 with a slight distortion in the immediate vicinity of the antenna. TM113 is 33Mhz away and looks very different.
Jamie,
As far as I know and my research has shown, you can't excite TE mode with the antenna positioned as Dave has it. Sure TM looks very different but it should excite at or very close to the TE013 freq as both have 3 x 1/2 guide waves between the end plates and the Bessel function is the same. Plus I have seen experimental results from others that TE013 and TM113 excite at the same freq, with TM113 being referred to as the degenerative shadow of TE013 and stated that a lot of care needs to go into the design of the TE013 antenna to not excite the degenerative TM113 shadow.
As I have said before, I have leaned FEKO does not properly handle what happens at the small end. It shows you can achieve resonance with a small end plate much smaller than Roger's limit, which I verified experimentally as being correct. Roger did tell me that even COMSOL gets the small end wrong and was they needed to develop their own in house simulation software to get frustum designs done right, the 1st time.
BTW do a FEKO end plate eddy current analysis as it will tell you what the mode is as every mode has a unique end plate eddy current pattern.
You must share how you do the 1/4 wave stub excitement in FEKO. -
#3388 by meberbs on 01 Jul, 2016 13:14
-
As far as I know and my research has shown, you can't excite TE mode with the antenna positioned as Dave has it. Sure TM looks very different but it should excite at or very close to the TE013 freq as both have 3 x 1/2 guide waves between the end plates and the Bessel function is the same. Plus I have seen experimental results from others that TE013 and TM113 excite at the same freq, with TM113 being referred to as the degenerative shadow of TE013 and stated that a lot of care needs to go into the design of the TE013 antenna to not excite the degenerative TM113 shadow.
(emphasis mine)
As I have said before, I have leaned FEKO does not properly handle what happens at the small end. It shows you can achieve resonance with a small end plate much smaller than Roger's limit, which I verified experimentally as being correct. Roger did tell me that even COMSOL gets the small end wrong and was they needed to develop their own in house simulation software to get frustum designs done right, the 1st time.
BTW do a FEKO end plate eddy current analysis as it will tell you what the mode is as every mode has a unique end plate eddy current pattern.
You must share how you do the 1/4 wave stub excitement in FEKO.
Yes, you have said this all before, and all of the errors with your statements have been explored in detail before. I am not going to waste time rehashing facts that you have repeatedly refused to listen to.
I will ask you again to provide a single example where an experiment shows a lack of resonance due to "small end cutoff" where Maxwell's equations (what FEKO et al) predict resonance. This should be easy to do if it is true, and would be comparably revolutionary to a working EMDrive. You claim to have done this experiment, just provide the dimensions and frequency so it can be validated. -
#3389 by Monomorphic on 01 Jul, 2016 13:22
-
BTW do a FEKO end plate eddy current analysis as it will tell you what the mode is as every mode has a unique end plate eddy current pattern.
You must share how you do the 1/4 wave stub excitement in FEKO.
Here are the surface currents. For the FEKO sim, I did a couple of tutorials on monopole antenna first, and then just attached a ~3cm monopole to the inside surface of the frustum. It involves using a wire port and then assigning a voltage source.
-
#3390 by TheTraveller on 01 Jul, 2016 13:53
-
BTW do a FEKO end plate eddy current analysis as it will tell you what the mode is as every mode has a unique end plate eddy current pattern.
You must share how you do the 1/4 wave stub excitement in FEKO.
Here are the surface currents. For the FEKO sim, I did a couple of tutorials on monopole antenna first, and then just attached a ~3cm monopole to the inside surface of the frustum. It involves using a wire port and then assigning a voltage source.
May I see the big end plate as well?
Attached is an oldie but a goodie that discusses end plate current patterns versus their modes.
-
#3391 by Rodal on 01 Jul, 2016 13:53
-
Are you saying that Roger Shawyer claims to have greater expertise in numerically solving Maxwell's partial differential equations than the people that wrote FEKO and COMSOL, who are Professors at universities and have plenty of publications in specialized peer-reviewed journals on numerical analysis?Dave's VNA S11 rtn loss scan shows a very high Q dip where TE013 resonance would be expected. But his antenna placement can't excite TE013 but can excite TM113 which exists at the same resonant freq as TE013.
The COMSOL sim to the right does not indicate the antenna location. In the FEKO sim to the left, I did simulate the antenna. With Dave's antenna placement, to me it looks like TM013 with a slight distortion in the immediate vicinity of the antenna. TM113 is 33Mhz away and looks very different.
...
As I have said before, I have leaned FEKO does not properly handle what happens at the small end. It shows you can achieve resonance with a small end plate much smaller than Roger's limit, which I verified experimentally as being correct. Roger did tell me that even COMSOL gets the small end wrong and was they needed to develop their own in house simulation software to get frustum designs done right, the 1st time.
...
Are you saying that Roger Shawyer (*) claims to have developed a better numerical method to solve partial differential equations than the Finite Element method used in COMSOL? The Finite Element method which was created by famous scientists like Richard Courant, who together with David Hilbert authored the influential textbook Methoden der mathematischen Physik ? And that Roger Shawyer claims to have developed a better numerical solution than the Boundary Integral Method used in FEKO?
That's very interesting, because these solutions to partial differential equations (Finite Element Method) have been used for decades now by NASA (see: NASTRAN), and all aerospace contractors. They got us to the Moon and back.
The only such numerical method you have disclosed from Roger Shawyer is your Excel spreadsheet that uses a rudimentary method to solve the equations in a very approximate way. Surely you are not claiming that Roger Shawyer bases his claims about having a superior numerical solution to Maxwell's equations than FEKO and COMSOL based on that spreadsheet.
If Roger Shawyer has created a numerical method superior to the Finite Element method and the Boundary Element method, that achievement by itself would be wasted if not shared with the rest of humanity, as everyone else (NASA, aerospace companies and universities) are using the Finite Element Method and the Boundary Element Method to design just about everything nowadays.
Would you be so kind as to disclose any reports by Roger Shawyer on his solution to partial differential equations that is superior to FEKO and COMSOL? I have found nothing regarding superior numerical methods of solutions to Maxwell's equation in Shawyer's publications.
Thanks
(*) The inventor of the EM Drive, who claims that a closed microwave cavity can self-accelerate all by itself without violating conservation of momentum and conservation of energy, purely using classical physics: Maxwell's equations and Special Relativity. He claims that no new physics are necessary to explain such self-acceleration. -
#3392 by Monomorphic on 01 Jul, 2016 13:58
-
#3393 by TheTraveller on 01 Jul, 2016 14:04
-
Are you saying that Roger Shawyer claims to have greater expertise in numerically solving Maxwell's partial differential equations than the people that wrote FEKO and COMSOL, Ph.D. who are Professors at universities and have plenty of publications in specialized peer-reviewed journals on numerical analysis?
I have stated what Roger told me. If you decide to ignore it, then so be it.
I built a frustum to test Roger's small end cutoff rule, which still showed FEKO resonance. I did not produce any force as Roger predicted. So as far as that FEKO model, it was wrong as there was no force producing resonance occurring in that frustum as the real world small end guide wavelength was beyond cutoff and not reflecting any significant amounts of photons back to the big end. Again you can accept that or not. I will not debate it.
-
#3394 by TheTraveller on 01 Jul, 2016 14:05
-
May I see the big end plate as well?
Ok as I expected the freq is not at resonance, it is close, as both end plate patterns should be the same.
It is clearly not a TE01x resonance as end plate currents flow onto the side walls. In TE01x mode, the end plate eddy currents never leave the end plate, with is another good feature of exciting in TE01x as there is no need of a 5 X skin depth continuance of the surface resistance across the end plate to side wall joint. Which is why I think Roger uses it. -
#3395 by Rodal on 01 Jul, 2016 14:23
-
Yes, we had discussed this before. I didn't understand why rfmwguy went ahead with an EM Drive test that computer simulations (FEKO, COMSOL by NASA, and my exact solution using Wolfram Mathematica) all showed ahead of time that it would not resonate in any of the mode shapes (TE012 and TE013) used by Shawyer or by Yang, or by NASA (TE012 and TM212).May I see the big end plate as well?
Ok as I expected the freq is not at resonance, it is close, as both end plate patterns should be the same.
It is clearly not a TE01x resonance as end plate currents flow onto the side walls. In TE01x mode, the end plate eddy currents never leave the end plate, with is another good feature of exciting in TE01x as there is no need of a 5 X skin depth continuance of the surface resistance across the end plate to side wall joint. Which is why I think Roger uses it.
My recollection is that rfmwguy answered in the forum (when confronted with Monomorphic's simulations) that he was following "his intuition" and he advised other DIY to do what they wanted, since these were their experiments (which they certainly are).
However, as his EM Drive was known ahead of time that would not resonate in any of the previously reported mode shapes, now a comparison is difficult to make. Even a comparison to rfmwguy's previous test is difficult to make (since he changed the frustum dimensions).
Particularly when his own infrared camera does not show any clear mode shape and he has no means at the present time to ascertain what mode shape (or more than one mode shape if there is mode participation) are being excited in his test.
Monomorphic did a great job at helping him by running a number of FEKO models and showing what was going to be happen. -
#3396 by TheTraveller on 01 Jul, 2016 14:43
-
May I see the big end plate as well?
BTW I get resonance in TM013 at 2.469GHz which would be a good fit as it will match what his antenna placement can excite.
What is your FEKO resonance in that mode? -
#3397 by SeeShells on 01 Jul, 2016 14:49
-
Reviewing the thermal video of Dave's frustum it looks like he somehow excited a TE01 where both endplates exhibit thermal increases from induced currents and the center section of the frustum remains cooler. At just over a minute into his run you can see where the thermal splits start to show in the video. I find this a red flag video as to what maybe happening in his frustum.
Shell
-
#3398 by Rodal on 01 Jul, 2016 14:51
-
Here is Daves latest data with work underway to improve the torsion beam's return dynamics.
I see these oscillations, even when powered off, at a regular period.
...
1) What is the period of oscillation of rfmwguy's torsional pendulum ?
2) rfmwguy at some point in time put "stoppers" to stop the rotational motion of the torsional pendulum after exceeding a given angle.
I think that these stoppers are a terrible idea: they decelerate the motion by adding mass after the rotation exceeds a given angle. By installing these stoppers rfmwguy has made his test nonlinear and hence very difficult to analyze.
The oil bath dampens the motion linearly. The stoppers added by rfmwguy introduce a nonlinearity into the problem.
Hopefully other DIY (Monomorphic. SeeShells, etc.) do not imitate this very bad idea of using these added mass stoppers when the motion exceeds a given angle.
See "stoppers" hanging from a tripod after 14:44 in this video, at the right hand side end of the beam:
https://www.youtube.com/watch?v=zkKNXJEzs7w&feature=youtu.be
Did those stoppers come into action during this test?
He should get rid of the stoppers. If he needs more damping he should use heavier oil and/or even better have another oil bath dampening motion symmetrically placed from the center of torsion .
Knowing the answers to these questions would make it easier to understand what is going on.
Thanks -
#3399 by TheTraveller on 01 Jul, 2016 15:04
-
Reviewing the thermal video of Dave's frustum it looks like he somehow excited a TE01 where both endplates exhibit thermal increases from induced currents and the center section of the frustum remains cooler. At just over a minute into his run you can see where the thermal splits start to show in the video. I find this a red flag video as to what maybe happening in his frustum.
Shell
Hi Shell,
In TE01x mode the edges of the end plates would stay cool as in that mode all end plate eddy currents circulate around the end plate.
From that I can see of Jamie's sim and the video, the mode was TM01x as in that mode there are circular end plate currents but they flow onto the side walls.
See attached TE and TM end plate currents. Note the TE modes have no currents around the outer edge and TM modes do.
