-
#2740 by Rodal on 21 Feb, 2017 19:56
-
There's an awesome amount of experimental physics learning happening in this thread. Should be a lecture series.
The amazing thing with the Internet and the testing person (Monomorphic), is that instead of Monomorphic's testing being peer-reviewed by a few editors if he would have submitted the test results to a Journal at the end of the testing (editors which usually do not have the time or the complete information to perform a thorough review to inspect every little detail of testing submitted to a journal), in this case we have real-time complete information of Monomorphic's test and that it is being reviewed by thousands of viewers in real-time ! plus Monomorphic has the professional dedication not to quit and declare premature victory -or defeat- but to continue refining his test and accommodate the suggestions from the audience
It is almost like watching Edison, Goddard or Tesla in real-time and having them interact with the audience
-
#2741 by X_RaY on 21 Feb, 2017 20:26
-
I hope I did this right
I interpret this as just stating that Notsosureofit's hypothesis predicts a lower force than measured. (If measured 5 μN, for example, using 1 watt, with Q=16,000, Notsosureofit would have predicted 5 μN *(16,000/70,000)= 1.14 μN instead)Reference: For 5 micronewtons I would expect PQ to be around 70,000.
Is this a backward calculation using your set of equations?
The theoretically QU for Monomorphic's cavity is 90k...100k (for copper at 20°C, no dielectric, TE013, 2.45GHz).
His measurement of QL shows ~8k.
https://forum.nasaspaceflight.com/index.php?topic=41732.msg1642070#msg1642070
Because he reached near impedance matched conditions the approximated QU (aka Q0) should be ~16k.
PQ=1W*QU
The question is where this discrepancy comes from.
Measurement errors of QL, nonlinear relation of thrust to P*Q, too large extrapolation or other issues with/of the equations (somewhat else?)
There is nothing wrong with this, as if the measured force is
F = Fr + Ft + FL
for example
where
Fr= real EM Drive force
Ft=thermal convection artifact force due to conducting the test in air and no flat end insulation
FL =Lorentz forces
that
Fr = F - Ft - FL
Fr < F
meaning that the real EM Drive force is smaller than the one measured (after subtracting thermal convection artifacts, Lorentz forces and other artifacts).
If the coupling is small, then that is another reason why Notsosureofit could predict a smaller force.
The only problem would be for theories that predict a larger force for Monomorphic, not a smaller force
How large a force does Shawyer predict ?
How large a force does McCulloch predict ?
-
#2742 by Rodal on 21 Feb, 2017 23:38
-
Please see http://forum.nasaspaceflight.com/index.php?topic=41732.msg1645390#msg1645390
What is the reason to have the EM Drive on for 1.5 minutes ?
Given the fact that the period of oscillation is 48 seconds, should you have the EM Drive on for 1/2 a period (24 seconds) or 1/4 of a period (12 seconds) ?
Having the EM Drive on for more than 1/2 period of the pendulum (more than 24 seconds) will be getting the EM drive into what Shawyer calls generator mode (EM Drive being moved in reverse). Whether one agrees or not with Shawyer, would it make sense to first test for 1/2 period (24 seconds) or on for only 1/4 period (12 seconds)?
Should motions with EM Drive ON longer than 1/4 or 1/2 period of pendulum be examined only after one has examined 1/4 and 1/2 period of pendulum ?
And, should the time that the EM Drive is on be an integer multiple of 1/4 period of the pendulum?
-
#2743 by jmossman on 22 Feb, 2017 00:15
-
Please see http://forum.nasaspaceflight.com/index.php?topic=41732.msg1645390#msg1645390
What is the reason to have the EM Drive on for 1.5 minutes ?
Given the fact that the period of oscillation is 48 seconds, should you have the EM Drive on for 1/2 a period (24 seconds) or 1/4 of a period (12 seconds) ?
Having the EM Drive on for more than 1/2 period of the pendulum (more than 24 seconds) will be getting the EM drive into what Shawyer calls generator mode (EM Drive being moved in reverse). Whether one agrees or not with Shawyer, would it make sense to first test for 1/2 period (24 seconds) or on for only 1/4 period (12 seconds)?
Should motions with EM Drive ON longer than 1/4 or 1/2 period of pendulum be examined only after one has examined 1/4 and 1/2 period of pendulum ?
And, should the time that the EM Drive is on be an integer multiple of 1/4 period of the pendulum?
Dr. Rodal,
If one were to use post-processing techniques to analyze the data from the experiment, would it be better to inject a signal at a different frequency than that of the fundamental pendulum oscillation?
I concur that stopping stimulus prior to the wire tension saturating and the pendulum reversing directions would help keep initial analysis simpler. However, I would have thought the time to reach the saturation point would be independent of the pendulum oscillation period and instead be a function of the anomalous force(s) being measured and the buildup of tension in the wire as movement approaches the first "plateau".
Thanks,
James
Edit: Restored full quotation of Dr. Rodal for proper context/continuity. -
#2744 by Rodal on 22 Feb, 2017 00:36
-
Please see http://forum.nasaspaceflight.com/index.php?topic=41732.msg1645390#msg1645390
What is the reason to have the EM Drive on for 1.5 minutes ?
Given the fact that the period of oscillation is 48 seconds, should you have the EM Drive on for 1/2 a period (24 seconds) or 1/4 of a period (12 seconds) ?
Having the EM Drive on for more than 1/2 period of the pendulum (more than 24 seconds) will be getting the EM drive into what Shawyer calls generator mode (EM Drive being moved in reverse). Whether one agrees or not with Shawyer, would it make sense to first test for 1/2 period (24 seconds) or on for only 1/4 period (12 seconds)?
Should motions with EM Drive ON longer than 1/4 or 1/2 period of pendulum be examined only after one has examined 1/4 and 1/2 period of pendulum ?
And, should the time that the EM Drive is on be an integer multiple of 1/4 period of the pendulum?
Dr. Rodal,
If one were to use post-processing techniques to analyze the data from the experiment, would it be better to inject a signal at a different frequency than that of the fundamental pendulum oscillation?
I concur that stopping stimulus prior to the wire tension saturating and the pendulum reversing directions would help keep initial analysis simpler. However, I would have thought the time to reach the saturation point would be independent of the pendulum oscillation period and instead be a function of the anomalous force(s) being measured and the buildup of tension in the wire as movement approaches the first "plateau".
Thanks,
James
Edit: Restored full quotation of Dr. Rodal for proper context/continuity.
This is an idea of Shawyer. I cannot defend it, because I do not find his analysis, mmm how could I say this...let's say it is not "mainstream" physics (if one considers the EM Drive as just an electromagnetically resonant cavity). It has to do with Shawyer's ideas of the EM Drive behaving differently when restrained than unrestrained, and behaving differently when made to move in the opposite direction than it wants to when free. See emdrive.com papers for "motor mode" and "generator mode" for the EM Drive. I don't know how he can make a difference between restrained and unrestrained, but I can see the difference between moving the EM Drive in one direction vs. the opposite/
Although I cannot defend Shawyer's analysis of this, all things being equal, one might as well test his idea, in case they are based on his experiments...which in this case it would mean having the EM Drive on for 1/4 of the pendulum period, another test fo 1/2, and then testing a with the EM Drive on for a full period and see what difference it makes. See the above drawing comments from Shawyer via TheTraveller. -
#2745 by jmossman on 22 Feb, 2017 01:12
-
Please see http://forum.nasaspaceflight.com/index.php?topic=41732.msg1645390#msg1645390
...
I concur that stopping stimulus prior to the wire tension saturating and the pendulum reversing directions would help keep initial analysis simpler.
...
This is an idea of Shawyer. I cannot defend it, because I do not find his analysis, mmm how could I say this...let's say it is not "mainstream" physics (if one considers the EM Drive as just an electromagnetically resonant cavity). It has to do with Shawyer's ideas of the EM Drive behaving differently when restrained than unrestrained, and behaving differently when made to move in the opposite direction than it wants to when free. See emdrive.com papers for "motor mode" and "generator mode" for the EM Drive. I don't know how he can make a difference between restrained and unrestrained, but I can see the difference between moving the EM Drive in one direction vs. the opposite/
Although I cannot defend Shawyer's analysis of this, all things being equal, one might as well test his idea, in case they are based on his experiments...which in this case it would mean having the EM Drive on for 1/4 of the pendulum period, another test fo 1/2, and then testing a with the EM Drive on for a full period and see what difference it makes. See the above drawing comments from Shawyer via TheTraveller.
I commend your willingness to approach the problem using all available information.
Using the early oil based power-on run from Monomorphic [TE013 Maximum Resonance (1W)] annotated by Shawyer/TheTraveler, I would have thought a power-on duration of about 23 seconds to avoid reaching the first plateau might be worthwhile (i.e. "7:56:28pm - 7:56:05pm"). I think we want the power-on duration to be as large as possible to maximize the Signal-to-Noise-Ratio (SNR), but small enough so that the wire tension doesn't saturate and the pendulum motion doesn't stop/reverse during the power-on region.
As Monomorphic can get additional energy into the cavity, the power-on duration to maximize the SNR will likely change... even though his pendulum apparatus and it's characteristics will hopefully remain static.
I'm excited to watch the upcoming testing campaign! The low power levels and the ability to thermally insulate the cavity exterior should finally help provide some compelling data on magnitude of thermally driven "anomalies".
https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=41732.0;attach=1406776;image -
#2746 by Rodal on 22 Feb, 2017 01:42
-
In the case of possible thermal convection ("natural convection") propulsion (an experimental artifact masquerading as an EM Drive force), there could be oscillations produced in the air adjacent to the induction heated end plates of the truncated cone, Rayleigh–Bénard convection for example.
Thermal convection force would then result in greater motion when harmonic of the natural frequency of the pendulum. Maybe this is what Shawyer observed ?
(Otherwise for EM Drive proponents, it seems like the EM Drive force instead of being monotonic would have to have a spectrum with peaks that would excite greater or smaller motion depending on whether a frequency is a harmonic of the spectrum ? ). It looks like Shawyer must have observed something in his experiments to have come with this strange idea of the EM Drive as having "motor" and "generator" modes ...
-
#2747 by Chrochne on 22 Feb, 2017 08:32
-
Not related to the EmDrive, but definitely worth to share if EmDrive can get us there!
NASA will hold conference on some very interesting discovery beyond our solar system today. We know they held similar briefings in the past. This can be bit different. They usually repeat event for few hours. This time it is 12 hours shift.
https://www.nasa.gov/multimedia/nasatv/schedule.html
They got my interest so lets see what they found out. Could it be that 9th planet we heard so much about lately?
-
#2748 by Flyby on 22 Feb, 2017 08:54
-
According "rumors" , this should be the topic of the pending NASA announcement :
http://nasawatch.com/archives/2017/02/spitzer-discove.html
It boils down to the question : how unique is our solar system ? (star+many planets) -
#2749 by Chrochne on 22 Feb, 2017 09:49
-
According "rumors" , this should be the topic of the pending NASA announcement :
http://nasawatch.com/archives/2017/02/spitzer-discove.html
It boils down to the question : how unique is our solar system ? (star+many planets)
Thank you Flyby. That looks as legitimate topic. If our solar system is not unique and Earth-type planets are more common, that would be really great news! -
#2750 by Monomorphic on 22 Feb, 2017 12:15
-
Please see http://forum.nasaspaceflight.com/index.php?topic=41732.msg1645390#msg1645390
What is the reason to have the EM Drive on for 1.5 minutes ?
Given the fact that the period of oscillation is 48 seconds, should you have the EM Drive on for 1/2 a period (24 seconds) or 1/4 of a period (12 seconds) ?
Having the EM Drive on for more than 1/2 period of the pendulum (more than 24 seconds) will be getting the EM drive into what Shawyer calls generator mode (EM Drive being moved in reverse). Whether one agrees or not with Shawyer, would it make sense to first test for 1/2 period (24 seconds) or on for only 1/4 period (12 seconds)?
Should motions with EM Drive ON longer than 1/4 or 1/2 period of pendulum be examined only after one has examined 1/4 and 1/2 period of pendulum ?
And, should the time that the EM Drive is on be an integer multiple of 1/4 period of the pendulum?
Dr. Rodal, We must be on the same wavelength this week as I was going to send you a private message today asking your opinion on this very question. The durations I have used in the past were essentially chosen at random. I think I will try the 24 second duration first as I'm concerned that 12 seconds will not be enough time to pull the signal out of the noise. Thanks!
Brief update: I have to move my wireless router from the top (third level) of my town home down to the basement, which is close to the workspace. The wifi signal is a little weak so my refresh rate on remote desktop is slower than I would like. It still works, but I would like for it to be working more smoothly during powered tests. As such, I need to purchase a network switch and move some boxes around. Once that is completed, powered tests will resume - definitely by this weekend. -
#2751 by Rodal on 22 Feb, 2017 12:31
-
Please see http://forum.nasaspaceflight.com/index.php?topic=41732.msg1645390#msg1645390
What is the reason to have the EM Drive on for 1.5 minutes ?
Given the fact that the period of oscillation is 48 seconds, should you have the EM Drive on for 1/2 a period (24 seconds) or 1/4 of a period (12 seconds) ?
Having the EM Drive on for more than 1/2 period of the pendulum (more than 24 seconds) will be getting the EM drive into what Shawyer calls generator mode (EM Drive being moved in reverse). Whether one agrees or not with Shawyer, would it make sense to first test for 1/2 period (24 seconds) or on for only 1/4 period (12 seconds)?
Should motions with EM Drive ON longer than 1/4 or 1/2 period of pendulum be examined only after one has examined 1/4 and 1/2 period of pendulum ?
And, should the time that the EM Drive is on be an integer multiple of 1/4 period of the pendulum?
Dr. Rodal, We must be on the same wavelength this week as I was going to send you a private message today asking your opinion on this very question. The durations I have used in the past were essentially chosen at random. I think I will try the 24 second duration first as I'm concerned that 12 seconds will not be enough time to pull the signal out of the noise. Thanks!
Brief update: I have to move my wireless router from the top (third level) of my town home down to the basement, which is close to the workspace. The wifi signal is a little weak so my refresh rate on remote desktop is slower than I would like. It still works, but I would like for it to be working more smoothly during powered tests. As such, I need to purchase a network switch and move some boxes around. Once that is completed, powered tests will resume - definitely by this weekend.
Any of this would be interesting. Notice that 1 minute exactly gives you 1 1/4 period:
sec
12 1/4 period
24 1/2 period
36 3/4 period
48 1 period
60 1 1/4 period ONE MINUTE
72 1 1/2 period
84 1 3/4 period
96 2 periods ONE MINUTE 36 seconds
Maybe interesting to start with just 6 seconds more than the last time: 96 seconds 2 periods ONE MINUTE 36 seconds
An argument to do 96 seconds is that if there is any delay in the force, this throws all the calculations off by the delay
You can always do other tests later, once you know what the delay is, to examine whether the delay is repeatable, and what difference if any it makes to stop at 1/4 period multiples + delay -
#2752 by RERT on 22 Feb, 2017 13:16
-
I would throw in an 'inifinite' test time - say enough time to leave the room, make a decent pot of coffee, and drink a cup. Say 15 minutes. We might learn something, even if only about thermal forces and holding resonance.
Another point is that, ideally, the frustrum and rig would return to the same initial temperatures/starting positions between tests. Then the proposed battery of tests would actually give us multiple instances of the first e.g. 1/4 period, and there would be scope for statistical averaging.
I remain curious as to what the largely undamped oscillations of the pendulum arm are. They are now the main impediment to much higher accuracy. -
#2753 by sghill on 22 Feb, 2017 13:36
-
There's an awesome amount of experimental physics learning happening in this thread. Should be a lecture series.
It is almost like watching Edison, Goddard or Tesla in real-time and having them interact with the audience
Sorry, couldn't help myself....
-
#2754 by Rodal on 22 Feb, 2017 13:54
-
...
I understand Monomorphic to have said that he has verified that the noise increases with his movement in the room. The noise is most likely due to excitation from ground vibrations, not only from Monomorphic's and others movements but also from nearby vehicles, etc. The "noise" contains a spectrum of frequencies, pendulum will move mostly at its first natural frequency when excited with a wide spectrum.
I remain curious as to what the largely undamped oscillations of the pendulum arm are. They are now the main impediment to much higher accuracy.
Compare the shape of this typical ground motion response (displacement vs time at the bottom left):
with Monomorphic's "noise" trace:
In order to ameliorate this, Monomorphic would need to have an isolated damped spring foundation, as the one used at NASA Eagleworks.
The preferable way to correct this is not to have a critically damped pendulum that critically damps everything (I disagree with somebody that posted that a critically damped pendulum was OK with him). The preferable way to correct this is to have an isolated damped foundation under a very sensitive pendulum set-up. One wants to damp the noise, not the signal, as SeeShell wrote "we are trying to measure a snowflake"l
This is the reason that Paul is spending the bucks to make this foundation for his new laboratory:
http://forum.nasaspaceflight.com/index.php?topic=41732.msg1639119#msg1639119 -
#2755 by Monomorphic on 22 Feb, 2017 14:18
-
In order to ameliorate this, Monomorphic would need to have an isolated damped spring foundation, as the one used at NASA Eagleworks.[/b][/color]
I do have sorbothane pads beneath each leg of the torsional pendulum. So there's at least that. I have four more somewhere. Perhaps I should double up on them.
-
#2756 by Rodal on 22 Feb, 2017 14:24
-
In order to ameliorate this, Monomorphic would need to have an isolated damped spring foundation, as the one used at NASA Eagleworks.[/b][/color]
I do have sorbothane pads beneath each leg of the torsional pendulum. So there's at least that.
For a vibration absorber you need not just a high damping material but you need a separate tuned mass-spring-dashpot
Just using a high damping material like Sorbothane (it is a thermoset polyurethane) will only do so much.
That's the reason for isolated foundations:
Look at this vibration absorber in this building:
see:
https://en.wikipedia.org/wiki/Tuned_mass_damper
https://en.wikipedia.org/wiki/Antiresonance
-
#2757 by Monomorphic on 22 Feb, 2017 14:31
-
What about a tuned liquid damper? Seems like something that could be incorporated on the pendulum relatively simply and with commonly available materials.
http://avestia.com/ICMEM2014_Proceedings/papers/68.pdf
-
#2758 by Rodal on 22 Feb, 2017 14:34
-
What about a tuned liquid damper? Seems like something that could be incorporated on the pendulum relatively simply and with commonly available materials.
For an effective vibration absorber, you need a separately tuned mass-spring-damper,, where the separate mass, spring and dashpot are tuned to eliminate the excitation, not just a damper. A damper by itself or a spring-damper will never be as effective.
...
Look at the links I gave.
Think of the examples I gave above in the post (how one isolates a building for example). You need a mass as well to make a vibration absorber. This has been put into effect in numerous Civil Engineering, Mechanical Engineering and Aerospace Engineering applications, during the last century. Prof. Den Hartog at MIT (who I had the privilege to listen to, while he was Prof. Emeritus) was one of the first at applying this.
OK NSF is about spaceflight applications...
Remember the horribly designed ARES rocket "stick"?
It needed a vibration absorber to correct the vibrations..
And remember that there is nothing where mass is more at a premium than a rocket !
A rocket is the last thing where one wants to add mass....But when you have something as horribly designed as the Ares was, it was needed -
#2759 by Monomorphic on 22 Feb, 2017 14:49
-
What about a tuned liquid damper? Seems like something that could be incorporated on the pendulum relatively simply and with commonly available materials.
For an effective vibration absorber, you need a separately tuned mass-spring-damper,, where the separate mass, spring and dashpot are tuned to eliminate the excitation, not just a damper. A damper by itself or a spring-damper will never be as effective.
...
Look at the links I gave.
Think of the examples I gave above in the post (how one isolates a building for example). You need a mass as well to make a vibration absorber.
So for a 48 second oscillation period that is roughly (trough to peak) 5uN, how large of a mass, spring, and absorber would be required? I'm not sure how practical it would be to build a mass-spring-damper on these scales.
