Author Topic: EM Drive Developments - related to space flight applications - Thread 8  (Read 1004661 times)

Offline TheTraveller

The question I have is why TM010 vs say TM013?

TE013 eliminates eddy currents crossing over the end plate to side wall joint. If that joint is not VERY perfect you can get arcing and bad mode formation in other than TE01x modes. Plus Roger advised all DIYers to use TE013 and it seems to work well.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline CraigPichach

  • Member
  • Posts: 59
  • Calgary Alberta
  • Liked: 42
  • Likes Given: 4

Craig:

The one thing you need to understand is that the off-the-shelf 929 MHz XXXkW magnetrons won't work very well with a high-Q copper frustum due to its nominal very wide bandwidth, which means that most of its RF input power would be wasted heating the cavity walls instead of maximizing the generated E&M fields.  At room temperature, a copper frustum driving a TE013 resonant response has a resonant -3dB bandwidth of say 30 kHz where as the magnetron's normal output is smeared over 30 to 60 MHz dependent on its design and operating temperature.  However, what one can do is to modify the magnetron along the lines in the attached slide so that it becomes a controllable frequency source with similar -3dB bandwidth that can be CONTROLLED with an S11 resonant frequency tracker with force feedback as I was talking about with The Traveler (Phil).

BTW, we still don't know what resonant mode will generate the best thruster efficiency other than high Q-factors, to a point, are better.  The reason I had Jerry Vera run these COMSOL/QV-Plasma code simulations at the TM010 resonant  mode frequency was that we already had the COMSOL results for the EW copper frustum and it had a TM010 resonant frequency close to the 929 MHz output of this L-3 Comm 100kW magnetron with 88% efficiency.  From the testing experience of Shawyer and others, the best frustum resonant mode may be to run the cavity at its TE013 resonance, which means purpose building a copper frustum with its as-built TE013 resonance at 929 MHz and one that can handle the 12+kW of power dissipation. 

Best,

PM and the Traveller,

Can I permission to forward this diagram?

My understanding is that we can adjust +/- 5MHz with the test setup they have at 915MHz.

We were aiming to do a TE013 mode design (Q=133526 at 914.85MHz) with an a S11 of approximately 15dB.

We were going to actually use compress to ensure high quality fabrication of the frustrum (and rate it as an ASME vessel). Building the frustrums with an industrial fabricator isn't as expensive as you would think with the right equipment that we could try to tune by building more frustrums.

Offline mwvp

  • Full Member
  • **
  • Posts: 254
  • Coincidence? I think Not!
  • Liked: 169
  • Likes Given: 30
... L-3 was to send me dimensions (I used the Travellers spreadsheet to come up with some rough dimensions to confirm) but then they stopped emailing me after sending pictures

Perhaps L-3 became aware of the potential fire hazard of igniting a plasma in a high Q cavity with high power microwaves? IIRC, ~100 MW is the max for an outstanding vacuum in accelerator cavities. If the Q is 10K, and you put 10KW in, there's your 100 MW.

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1373
  • Do it!
  • Vista, CA
  • Liked: 1424
  • Likes Given: 1893
I have considered the MHD model for many, many years. My conclusion has been that, the temperature at which electron-positron pairs annihilate each other is in excess of 108 Kelvin. If e-p pairs in the frustum had a density ~1x1012 kg/m3, and a life time of ~10-22 s, the frustum would be vaporized from the heat, faster than dropping it onto the surface of the sun.

The paper in JBIS is saying that it would require a mass of e-p pairs in excess of 105 kg. So the heat and the mass would not go undetected, therefore that's not it.

Regarding the paper you attached, I love this paper! However, at the scale of electrons and quarks, they are constantly undergoing exchange scattering with their counterparts in the QV. At this scale it is possible because the E field exceeds the Schwinger limit, but in the frustum the E field is no where near that limit. So the expectation of producing so many pairs is unreasonable, and I offer that it can't be happening if the frustum is not melting instantly upon their creation.

Todd


Todd:

"My conclusion has been that, the temperature at which electron-positron pairs annihilate each other is in excess of 108 Kelvin."

Again you are making the assumption that the e/p pairs are fully fleshed out in our universe, which does require 0.511 MeV per particle and that would indeed melt the frustum if fully developed.  What Dr. White's QV conjecture posits is that these virtual force carriers can be expressed in our reality with a variable effective mass/energy density that goes from just barely here to fully here at the Schwinger limit energy densities.  Of course the only way to prove this QV conjecture is to test a given frustum design over a broad input power range of four orders of magnitude or greater to see if it generates the COMSOL/QV Plasma code's EW copper frustum's TM010 thrust predictions I posted at NSF.com earlier, or not.

Best, Paul M.

I've read this idea a very long time ago, but I thought it was discredited because I never heard of it again.

If that is the case, a much simpler experiment would be to measure the "linearity" of vacuum permittivity and permeability up to fields strengths equivalent to those in the frustum. Because, any such creation of virtual pairs, or voltage tension between two virtual masses, will change the permittivity and permeability of the vacuum in a measurable way.  How non-linear are vacuum filled capacitors as the field strength approaches 10^7 V/m?

See the attached paper by Urban, which derives these values from the polarizable vacuum. We could probably extend this to apply directly to such an experiment.


Todd:

Great idea and thanks much for the pointer to the Urban paper!  I will read and consider how one might do this on the cheap in my home lab once it is built.

Best, Paul M.

A simple calculation for a parallel plate capacitor with a vacuum dielectric turns out that an electric field of 2.5 x 107 V/m, is not difficult to achieve. For instance, a 1 nF capacitor, is only an area of 10 mm2 with a separation of 8.84 x 10-8 m. This E field strength occurs at only 2.21V!!!! Obviously, if there were significant non-linearity in the vacuum permittivity due to e-p pairs at these field strengths, surely they would've been noticed by now in capacitor manufacturing.

In my view, and my opinion is based on P. W. Milonni's "The Quantum Vacuum", the vacuum "IS" an electromagnetic field. Superimposing a stronger EM field on top of the ZPF, is just raising the energy state of the QV by the number of photons in each state.  The field superimposed on the ZPF, and the ZPF are the same thing, except with a much narrower bandwidth and non-random polarization. If the QV had anything to do with this, then IMO the Casimir effect alone would propel it, but it doesn't. We need to superimpose a stronger field, which we can push against, and that field must be asymmetrically annihilated (dissipated) in order to make it move.

Offline Augmentor

  • Member
  • Posts: 61
  • Liked: 48
  • Likes Given: 53
Tangent to the present discussion, it's curious that the Eagleworks paper shows effectively no linear response in forward thrust between the 60 to 80 watt runs. Thoughts?

I believe doing tuning in vac involves a bit of what someone I know once called "Kentucky Windage"?

Would suggest the force vs power results could have been tighter if they had the lowest reflected freq tuner instead of the PLL.

Phil:

"I believe doing tuning in vac involves a bit of what someone I know once called "Kentucky Windage"?"

With 20/20 hindsight in place some of us now know that PLL frequency trackers are NOT the best way to go.  What is needed is a digital S11 resonant frequency tracker that can accommodate a variable permitivity in the frustum due to pressure changes, but with a force feedback input that can skew the drive frequency off peak resonant power just a few Hz to kHz to maximize the dynamic force output.  It also needs to accommodate vehicle acceleration in the manner that Shawyer has talked about in his patents.

Best, Paul M.

Paul,

For any space drive, tracking changes in momentum is preferred to tracking just a force. Force is a short hand notation.

F = m a = d(mv)/dt

Now we

So one has to not only track acceleration but also any change in mass. The difficulty is that permittivity and permeability may also change. Any pinning of the magnetic field to the walls or to the flow will also create issues to address.

So we have four cases,

1. Inertia mass is constant, acceleration is constant
2. Inertia mass is variable (transient), acceleration is constant
3. Inertia mass is constant, acceleration is variable
4. Inertia mass is variable (transient), acceleration is variable

In the later two cases, variable acceleration suggests da/dt which is the jerk or jolt. When acceleration varies, then one may use differential or delta step functions depending on the continuity desired.

Note that continuity may be challenging in an environment where both mass and acceleration are uncoupled variables.

The Poynting vector, E X B, may be an oversimplification by not containing total angular momentum of the photons.Poynting vector contributions may not be accurate if there is energy stored or released in the total angular momentum stored or released in the RF/microwave photons. In a photon, both spin and orbit angular momentum need to be accounted for.

There is also reference frame issue for acceleration which Woodward addresses in his book but I have not seen White address reference frames.

dm

Offline rq3

  • Full Member
  • **
  • Posts: 231
  • USA
  • Liked: 266
  • Likes Given: 42
Tangent to the present discussion, it's curious that the Eagleworks paper shows effectively no linear response in forward thrust between the 60 to 80 watt runs. Thoughts?

I believe doing tuning in vac involves a bit of what someone I know once called "Kentucky Windage"?

Would suggest the force vs power results could have been tighter if they had the lowest reflected freq tuner instead of the PLL.

Phil:

"I believe doing tuning in vac involves a bit of what someone I know once called "Kentucky Windage"?"

With 20/20 hindsight in place some of us now know that PLL frequency trackers are NOT the best way to go.  What is needed is a digital S11 resonant frequency tracker that can accommodate a variable permitivity in the frustum due to pressure changes, but with a force feedback input that can skew the drive frequency off peak resonant power just a few Hz to kHz to maximize the dynamic force output.  It also needs to accommodate vehicle acceleration in the manner that Shawyer has talked about in his patents.

Best, Paul M.

Yup. Sounds like a force locked loop to me! Glad to help out!
« Last Edit: 11/23/2016 10:34 PM by rq3 »

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2166
  • Liked: 2684
  • Likes Given: 1124
Hey @rfmwguy,

I'm interested in doing spectrum analysis of my magnetrons and was curious of what equipment you used.
I remember EW measured the spectrum of leakage from their microwave button panel with a water load inside. Did you do something similar? What kind of analyzer did you use?

Thank you
Very basic USB spec an. Be sure to isolate with 30dB attenuator and use unity gain dipole antenna. Locate pickup several feet from Sig source and laptop even further. Buy the best spec an you can afford. For general Sig strength and spectral display, a quality USB spec an worked just fine.

Offline otlski

  • Member
  • Posts: 69
  • Liked: 86
  • Likes Given: 64
Simple question to the Forum

If you theory guys had a working EmDrive, on a rotary test rig, at your disposal, what would be the process to develop an acceptable theory to explain what you are observing?

What data would you need from the test rig?

Please try to be specific so I can ensure that data is available.

TT, specific to the test apparatus part of the question. 
First, I would ensure that the air bearing had clean dry air supplied to it.  You would definitely need an oil separator if you were not using an oiless compressor.  You would follow up the output with a 50 foot length of copper tubing coiled inside a tub of water at room temperature.  This is for heat exchanging purposes to ensure the heat of compression was partly eliminated.  Lastly, I would follow with two air regulators.  The first being an inexpensive one to filter out pressure changes as the compressor cycles on and off.  The second being an expensive precision regulator to hold to a fraction of a psi.  If you are using bottled nitrogen I would still use the heat exchanger to compensate for the cooling from expansion.

The air bearing would be placed on a purpose built three legged mounting stand (no cobbled 80/20 or optical breadboard components for this part).  This would all rest on a concrete floor of reasonable thickness and good soil underlayment.  If using a flat bearing combined with a cylindrical bearing, the top of the air bearing would need to be flat enough to support leveling to 0.0005”/ft and that is also the target value for the final level.  Walking around on the concrete floor while leveling should not affect the bubble’s position. If using a hemisphere. You would want to maintain level but not as stringent.  For the hemisphere, maintaining the vertical CG below the spherical center is required.  For either bearing keeping the horizontal CG coincident with the axis of rotation is strongly advised.

The room would be temperature controlled and free of drafts.  A way to shut the HVAC off during tests is important.  A waiting period for the HVAC convection to settle is advised.

A 30 frame per second camera mounted directly above and looking straight down on the experiment is wholly necessary.  30 fps has been proven fast enough for this kind of work although a CCD is preferred over a CMOS sensor that has a rolling shutter.  A rolling shutter would be useless.  Camera and video capture system must not lose its time base by dropping frames as this would corrupt the calculations.  Having angular markings every 10 degree around the bearing and a stationary pointer would allow us to measure position often enough to be useful.  Three full rotations during a testing run is the minimum for good analysis.  Less than one rotation appropriately brings on questions as to the test's validity; in fact, it negates the apparent validity in my mind.  A full rotation would help us evaluate/eliminate level vs. CG errors, interaction with Earth’s magnetic field, and other experimental problems.  Several rotations lets us see if we’ve reached a terminal angular velocity where thrust torque matches profile drag.

A reasonable estimate of the mass moment of inertia of the entire rotating section would allow us to calculate torque and thus force.  I can help with this estimate when the time comes.  As for data, the angular position verses time stamp, and MOI is all that is needed to do the major math.  A side view FLIR, other cameras, room temperature might be useful to analyze if things got weird.

However, before running actual tests we would want to characterize the system with the camera running.  First, with just the bare bearing (nothing mounted) we would want to measure both motoring (Paul called it swirl) torque and coulomb friction.  These let us know that you have a good bearing or need to compensate.  Basically, from stopped, the bearing is allowed to accelerate on its own; it might take hours.  It will accelerate if it is not a perfect bearing and if motoring torque is greater than coulomb friction.  The second test of the still bare bearing would feature you inducing a CW spin by hand and letting it decay on its own, then repeating this CCW.  This might take 10 minutes to an hour for data in each direction.  From this we can confirm the motoring and calculate coulomb.  I have done this for twenty air bearings when required by our customers.  I will scrub a spreadsheet and make it and myself available when the time comes.  Finally, with the full apparatus mounted, we would repeat the hand induced CW and CCW spins and process in the same spreadsheet to get the profile drag components, coulomb, viscous, and turbulence of the all-up experiment.

I have used all these techniques before so none are new to me.  What you require (including MOI and CG) knowledge and control are all part of what we have done for customers and for building our own corporate knowledgebase.  www.space-electronics.com  Count me as someone who wants dearly to see this work while maintaining a healthy skepticism; mainly because I see experiments that simply have not measured up.  That said I am more than willing to help where I can. 
« Last Edit: 11/24/2016 03:12 AM by otlski »

Offline FattyLumpkin

  • Full Member
  • **
  • Posts: 266
  • Boise ID
  • Liked: 69
  • Likes Given: 0
TT, re cavity fabrication ..... if memory serves no more than 4/100s" margins? yes?   thnx , FL
« Last Edit: 11/24/2016 06:04 PM by FattyLumpkin »

Offline EDGEOFTHECOSMOS

  • Member
  • Posts: 1
  • Los Angeles, CA
  • Liked: 0
  • Likes Given: 0
I found this thread on 4chan of all places in the political section, I think what you guys are doing is really cool. My question though is people talk about this being used for long space journey, but what about stuff that can be used on the ground here on earth? Is there any hope it will lets say, change everything? Btw, I never knew this thing costed so much, here I thought you could make these things with using stuff from shelf usage, like microwave magnetrons for example, but the ones you guys are talking about costs few thousand dollars, pretty crazy. Well, anyways, enough of me, goodluck and I will be watching this thread from now on.

Offline spupeng7

......
I'm asking for more insight on the relation between momentum and reflectance, because it would have profound implications :

If i may simplify the situation a bit to explain what was perhaps a bit too poorly elaborated :

I recall the discussions about the tennis balls bouncing back and forth in a huge frustum space station.

As long you consider the reflectance of the walls to be uniform in every direction, it is only the angle of incidence that will determine the size of the momentum. And in the end, when you add up all bounces, the final sum of all forces will be zero. That much I understood...

But I see a problem if the reflectance of the walls would vary according the direction you throw the balls. It would mean that for the direction (small end > big end) the transfer of momentum/force would be smaller then what the angle of incidence would predict.

So, it puzzles me on how the relation is then, between the reflectance and the momentum transfer ?

To put it simply, physics as we know it requires a force to THROW THE BALL! If you're in a funnel shaped copper cavity headed towards the moon, and you throw a tennis ball towards the moon, you and your cavity are accelerated away from the moon due to the energy required to THROW THE BALL. The tennis ball can then hit a steel plate, a mattress, or glance off the walls. It doesn't matter. When all of the energy has dissipated and the ball has come to rest, the net result is zero gain in momentum. At the most basic level, this is the fundamental issue with the Emdrive, and the issue every one here is trying to explain.

EDIT: Of course, an Emdrive is constantly stuffing more and more tennis balls into your copper cavity, and insisting that you throw them in a collimated fashion so that as one tennis ball is leaving your hand, the last one you threw is just exactly bouncing off the wall of your choice (hence all of the discussion about modes and Q). The more tennis balls you can deal with, the greater the Q. The end result is the same. You're going nowhere at a great rate, surrounded by tennis balls. Unless, of course, current physics is incomplete. And I'm sure it is  ;)
Sigh.. I knew this would happen the moment I started writing about tennis balls...
I thought it would be more clear the tennis ball analogy is rather flawed, but apparently, it isn't... :-[

It is not about the total sum of all the action/reactions I'm writing about, but I'm trying to get the understanding about reflection right.
Reflection is the most essential feature/process of what happens inside an EMdrive.
If you take 1 complete cycle, going from big end to small en and back, the total sum of the reflectance from small to big end is < then the total sum, especially when you consider it to be a wave, instead of a particle.
The problem with particles is that we associate the too much with real life observations where hard kinetic impacts have an apparent linear relation with the angle of incidence.

I'm not trying to explain the EMdrive with this, I don't have a comprehensive theory. I only have questions.
In this case about reflection and its relation with force/momentum transfer, because it is the most essential part of what happens inside an EMdrive.
The non linearity of reflection is really counter-intuitive and understanding it "might" give a clue to why an EMdrive functions. (IF it really functions).

My idea is that - IF the EMdrive does indeed work - we need to see and rethink what misconceptions we have, because maybe we then can find clues that help solve these apparent conflicts.

The EMdrive (probably) does not violate CoM. The press likes sensation but totally reverses the logic...
 
The question should be : what did we miss in our understanding of CoM, that doesn't need "new physics", but only a more accurate understanding of what's happening...
This means retracing, step by step, what we know and investigating any of the shortcuts we took for granted for so long...

Because it is a resonance cavity, any minuscule variation can get potentially amplified (Q 100k ? or more) to a point it becomes no longer negligible...

So questions like : what brings us the non-linearity of reflection, or the extraordinary momentum/spin in evanescent waves, become important because they're not really considered in our daily understanding of the world around us.

As far as comprehensive theories goes, I do have a preference for Todd's asymmetric dissipation/attenuation based upon the wave properties of RF waves...It's way above my head how he does it, but it looks like a logical and elegant solution to the apparent CoM/CoE conflict...

Flyby,
we know that reflections within a waveguide are attenuated by electrical resistance, there is no mystery there. We know that there is anomalous thrust and it seems to be proportional to the number of reflections so your question is well asked.

What I visualize as the mechanism of reflection is absorption which induces current in the surface of the reflector, followed by emission when that current is diverted or reversed as it reaches an edge or a discontinuity. This is not a large mirror reflecting light though, the visualization is rapidly complicated by the refractive index and other properties and dimensions of the surface.

Non of this is very relevant until you ask how that constitutes a connection with the universe outside of the frustum, for that I hope other readers will forgive me posting my own yet undeveloped surmise once more. Attached was published at my own expense in the IJEK in 2015.
Optimism equals opportunity.

Offline rq3

  • Full Member
  • **
  • Posts: 231
  • USA
  • Liked: 266
  • Likes Given: 42
TT, re cavity fabrication ..... if memory serves no more than 4/100s margins? yes?   thnx , FL

??? What does this mean? What are 4/100 margins? Margins of what? Have you a reference to previous posts? Perhaps keep in mind you appear to be asking questions on a public forum of a member who has accomplished a lot of arm waving, prematurely released results without permission to do so, and otherwise posted nothing but noise, unless it was ideas he's slowly absorbed from other's input on this very site. "Magic Happens Inside" strongly implies that some folks don't really understand how to impliment the hardware they fantasize, yet would like to appear to be able to do so. Claims of fabrication, with ball point pen sketches on napkins of multi-thousand dollar hardware just smells "odd", in my not so humble opinion.

If, perchance, you are asking about frustum fabrication tolerances, you might want to include the measurement system involved (English, metric, Klingon). In any case, no published results have determined that fabrication tolerances have any effect whatsoever, except as thay may influence cavity Q, which is well established microwave engineering.

As a microwave engineer, please don't get into the "how do I fabricate a cavity to resonate with my crappy source" discussion again. Even the latest NASA paper is past this one. You need to build the highest Q cavity possible, and then tune the driving microwave source to it, with a control loop that optimizes force. I addressed this almost 2 years ago. Since absolutely no one knows whether this effect exists at all, optimizing for reflection coefficient or any other effect other than the desired one (thrust versus input microwave power) is completely pointless. Maximize the effect until it is out of all conceivable noise, and develop the theory once the effect is proven. To date, the effect is polywater. All results are "down in the grass" (the baseline noise you see on a spectrum analyser due to it's own thermal and phase noise signature).

« Last Edit: 11/24/2016 02:35 AM by rq3 »

Offline cvbn

  • Member
  • Posts: 4
  • Liked: 2
  • Likes Given: 10
Have you seen this recent interview with Shawyer for IBTimes UK where he claims that he had EmDrives having 8 g and 18 g of thrust (the latter one Boeing has been testing):

What do you think?

Online RotoSequence

  • Full Member
  • ****
  • Posts: 888
  • Liked: 635
  • Likes Given: 821
Have you seen this recent interview with Shawyer for IBTimes UK where he claims that he had EmDrives having 8 g and 18 g of thrust (the latter one Boeing has been testing):

What do you think?

We have, some number of pages back - its a bit hard to dig through, but I do remember this being posted. My opinion was, and is, that I'll believe those extraordinary thrust numbers when I see them demonstrated.
« Last Edit: 11/24/2016 03:22 AM by RotoSequence »

Offline zen-in

  • Full Member
  • ****
  • Posts: 532
  • California
  • Liked: 469
  • Likes Given: 365

...

"Shouldn't the time process for a degradable Quantum Vacuum be many orders of magnitude faster?"

Agreed, but IMO the time delays observed in the fall 2015 EW lab's in-vacuum testing are specific to the ICFTA design interactions with the EW torque pendulum and are not inherent in the degradable QV interactions.  The EW lab's Dec 2014 split system in-vacuum test had much more prompt force turn-on and turn-off responses as demonstrated in figure 12 in the AIAA/JPP paper and my spring 2015 postings here a NSF.com.  I've attached a couple of slides from this spring 2015 time period as a reminder to all.

Best, Paul M.

I remember those graphs because the reverse response was so much less than the forward response.   If I'm not mistaken that was also when your group started to mitigate the Lorentz error.    My conclusion then and now of the reverse graph is that it is due to magnetic interaction.   It is the only "thrust" waveform that has the characteristics of a second order step response.  There is a fast rise/fall time and an overshoot with ringing.   That type of response curve has not been seen again in your results.   The second graph is a first order step response; ie:thermal.   In a vacuum or low pressure atmosphere the time constant is longer.   It would be useful to see a lot more data, even families of plots.   For example what would the plots look like if the pressure was stepped down by powers of ten (logrithmically), with everything else the same?   My guess is the time constant of the step response would increase as the pressure decreased.   Another experiment I have been asking all the em-drive dy people to do is to heat up the Copper cone with resistive heaters and collect data as if it was an RF input.   No one has done that yet.    I'm sure the EW team has the resources to do this.   

The more things you change in an experimental setup the more information you get from your experiments.   For example what would happen if the EW team redesigned the mounting hardware that holds the Copper cone on the TP arm.   If they could design it so that the CG vs displacement test using the 10 gram weight no longer caused a significant change in displacement how would that affect later tests?   It would be interesting to see.   I know the EW team has done excellent work and are only interested in finding the truth so I assume they will eventually do some of the things I have suggested.
« Last Edit: 11/24/2016 04:36 AM by zen-in »

Offline TheTraveller

Bad data, bad assumptions.

As per the attached, the right side data, which I have posted before, is badly wrong. The vertical scale is in Ohm and not in mOhm plus the Cu Rs curve at 300K (room temp) is about as wrong as it can get.

The left side image shows a truer picture and as the room temp and 77K Rs copper curves are very close together, there is little to be gained by cooling a Cu or Al with Ag coating frustum to 77K.

The original of the bad data screenshot is attached.

Apologies.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline Peter Lauwer

  • Full Member
  • **
  • Posts: 269
  • Setting up an exp with torsion balance
  • Netherlands
  • Liked: 301
  • Likes Given: 431
Quote
As will be discussed in more detail at the end of the section on slope filtering, in order to run the test article in a fully integrated configuration, the torsion pendulum is operated in a highly loaded configuration, which results in slower displacement rates for the torsion pendulum when an impulsive force is applied.
Perhaps nudging the pendulum with a known force could let us get a time to use in place of "slower"?

If there only were some kind of, I don't know, calibration pulse.
In all seriousness... this is pretty bad. I use the calibration pulse to determine the time constant of the pendulum as 4 seconds, and get no thrust in their latest graphs. What ever was responding to power rapidly is now gone.

Yes, there's something odd at 17..20 s , but then, there's great many components being heated non-uniformly. Some may even be undergoing quasi-phase changes (plastics have a glass transition temperature). Some may be warping until they mechanically come in contact with another component.

All:

This will be my last post of the day.  The EW Integrated Copper Frustum Test Article (ICFTA) had metallic and plastic components with competing and non-linear thermal expansions and contractions when heated, see previous posted slides on this topic, that when driving the torque pendulum's center of gravity shifts, blurred the impulsive response of this test article in time, dependent on the magnitude of the impulsive force.  For me, it is fully explained in the text of the JPP report, so please go back and read it this section again until it hopefully makes sense to you.

Best,  Paul M.

Dear Paul,

Do you have additional data for us about the dynamic behavior of the torque pendulum?
Without this, the observed responses are difficult to judge. The torsional constant of the flexure bearings must be quite big, given the rather fast response to a step function. The moment of inertia must be more than 6 kg.m^2, I guess (cooling block 5 kg, frustum, amp, counter weight...).
Has the frequency response been measured? And what is the damping factor?
I think it is really necessary to know this for a system like this.

Thanks,
Peter
Science is a way of trying not to fool yourself. The first principle is that you must not fool yourself, and you are the easiest person to fool.   — Richard Feynman

Offline CraigPichach

  • Member
  • Posts: 59
  • Calgary Alberta
  • Liked: 42
  • Likes Given: 4
Perhaps L-3 became aware of the potential fire hazard of igniting a plasma in a high Q cavity with high power microwaves? IIRC, ~100 MW is the max for an outstanding vacuum in accelerator cavities. If the Q is 10K, and you put 10KW in, there's your 100 MW.

This is a problem that has to be overcome regardless no? What solution would you recommend? Argon environment? Seems like we have to work this out if this unit is to, quite literally, fly.
« Last Edit: 11/24/2016 03:22 PM by CraigPichach »

Offline Bob012345

  • Full Member
  • ****
  • Posts: 582
  • Liked: 125
  • Likes Given: 219
TT, re cavity fabrication ..... if memory serves no more than 4/100s margins? yes?   thnx , FL

??? What does this mean? What are 4/100 margins? Margins of what? Have you a reference to previous posts? Perhaps keep in mind you appear to be asking questions on a public forum of a member who has accomplished a lot of arm waving, prematurely released results without permission to do so, and otherwise posted nothing but noise, unless it was ideas he's slowly absorbed from other's input on this very site. "Magic Happens Inside" strongly implies that some folks don't really understand how to impliment the hardware they fantasize, yet would like to appear to be able to do so. Claims of fabrication, with ball point pen sketches on napkins of multi-thousand dollar hardware just smells "odd", in my not so humble opinion.

If, perchance, you are asking about frustum fabrication tolerances, you might want to include the measurement system involved (English, metric, Klingon). In any case, no published results have determined that fabrication tolerances have any effect whatsoever, except as thay may influence cavity Q, which is well established microwave engineering.

As a microwave engineer, please don't get into the "how do I fabricate a cavity to resonate with my crappy source" discussion again. Even the latest NASA paper is past this one. You need to build the highest Q cavity possible, and then tune the driving microwave source to it, with a control loop that optimizes force. I addressed this almost 2 years ago. Since absolutely no one knows whether this effect exists at all, optimizing for reflection coefficient or any other effect other than the desired one (thrust versus input microwave power) is completely pointless. Maximize the effect until it is out of all conceivable noise, and develop the theory once the effect is proven. To date, the effect is polywater. All results are "down in the grass" (the baseline noise you see on a spectrum analyser due to it's own thermal and phase noise signature).

Shawyer is a microwave engineer too and his results are not in the noise nor are they "polywater"

Offline WarpTech

  • Full Member
  • ****
  • Posts: 1373
  • Do it!
  • Vista, CA
  • Liked: 1424
  • Likes Given: 1893
The 3 boxes are sized such that the diagonal lines are ~ equal length. This illustrates how the dispersion is different for wavelengths along the z axis versus those perpendicular to it.

Tags: