Author Topic: EM Drive Developments Thread 1  (Read 762753 times)

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #420 on: 09/10/2014 01:19 AM »
@Slyver  (CONTINUED)

In the above formula, I neglected the mass of the pendulum's arm "mArm".  If this mass is not negligible, then the angular frequency of swinging oscillation would be:

omega= Sqrt[(k-(g/l) ( M+(mArm/2) ) )  /   ( M+(mArm/3) )]

where:

k= spring constant (about a rotating axis normal to the vertical arm of the pendulum) of the bottom joint that tries to keep the inverted pendulum upright in a vertical position

mArm=mass of the pendulum's arm

M=total supported mass ( total weight of the upper portion: the tested object, its support table,  and any other fixtures and equipment on the table)

g=acceleration of gravity

l=length of inverted pendulum's arm between the lower torsional support and the supported mass

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #421 on: 09/10/2014 01:31 AM »
@Slyver  (CONTINUED)

If the supported mass is small enough, then omega ( the angular frequency of swinging oscillation ) is a real positive number and the pendulum may exhibit a harmonic swinging motion (hopefully of very small amplitude).  If the supported mass is too large, omega is an imaginary number (or in other words, the expression inside the square root becomes negative) then the pendulum will collapse, falling over over towards one side against the centering bearing.

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #422 on: 09/10/2014 01:44 AM »
@Slyver  (CONTINUED)

The swinging pendulum motion is, in principle, around two orthogonal swinging directions.

___________
There are other possible frequencies of swinging oscillation which I did NOT model.  For example if the pendulum arm bends at the upper end, one would also expect a composite pendulum with another frequency. 

I mention this because the NASA authors mention:

<<If needed, ballast is added to the pendulum arm to eliminate moments that affect the neutral position of the pendulum arm.>> (p.5)

In any case it is interesting that they are conscious that their experimental inverted pendulum set-up is capable of having moments affecting the neutral position of the pendulum arm (even if it doesn't bend).  The authors took care of using ballast to eliminate these moments at start-up (however, we have to calculate whether such moments affecting the neutral position of the pendulum arm are significant due to unsymmetric thermal expansion).  It is evident that "k/M" is not infinite !. The bearings allow some off-center motion. Their inverted pendulum is capable (as all inverted pendulums are) of swinging oscillations.
« Last Edit: 09/10/2014 02:25 AM by Rodal »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #423 on: 09/10/2014 02:56 AM »
@Slyver  (CONTINUED)

So (neglecting the mass "mArm" of the arm) , for the swinging motion of the inverted pendulum, we have two forces acting along the swinging arc, in opposite directions:

1) Due to gravity: the force component tangent to the swinging arc,
Fg = - M g theta ~ - M g x / l ,
which is de-stabilizing: it has the opposite sign to the force of gravity in the simple pendulum

and

2) Due to the elastic spring "k" (about a rotating axis normal to the vertical arm of the pendulum)  that tries to keep the inverted pendulum upright in a vertical position,
Fk = + k l theta ~ + k x

where the swinging angle can be approximated as

theta ~ x / l 

where "x" is the horizontal displacement and "l" is the length of the inverted pendulum's arm between the lower  support and the upper supported mass

For the inverted pendulum to work as designed, the elastic spring force (that tries to keep the inverted pendulum upright in a vertical position) must be greater that the de-stabilizing force of gravity, otherwise, as already discussed the pendulum will collapse towards one side:

Fk  > Fg

or

 k l  > M g

or

( (k / M) - (g / l ) ) > 0

« Last Edit: 09/10/2014 03:10 AM by Rodal »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #424 on: 09/10/2014 03:09 AM »
I'll continue this tomorrow...  as well as answering the other questions, from John Fornaro and others

Offline JohnFornaro

  • Not an expert
  • Senior Member
  • *****
  • Posts: 9162
  • Delta-t is the salient metric.
  • Planet Eaarth
    • Design / Program Associates
  • Liked: 610
  • Likes Given: 314
Re: EM Drive Developments
« Reply #425 on: 09/10/2014 12:26 PM »
Just for grins, here is a random linear actuator which gets 4N of "peak force" for 28 watts.
Sometimes I just flat out don't get it.

Offline Star-Drive

  • Member
  • Full Member
  • ****
  • Posts: 823
  • TX/USA
  • Liked: 856
  • Likes Given: 9
Re: EM Drive Developments
« Reply #426 on: 09/10/2014 12:42 PM »
it´s possible Rodal. I already forwarded your previous questions to the Talk Polywell forums, but I also asked Dr Paul March if he could answer your questions directly here, so I can stop playing messenger boy haha ;)

AcesHigh and Dr. Rodal:

Interesting thread, but you would have saved yourselves a lot of wasted effort if you had first obtained the needed details before going so far down this analysis rabbit trail.  That said, yes I'm covered by a number of NDA's that do not allow me to talk about third party test articles.  However since Dr. White has already published several papers on the Eagleworks Lab home-grown torque pendulum thrust measurement system and frustum cavity test articles, I think I can comment at least on those topics when my time permits.

Now in regards to the Eagleworks Lab's torque pendulum, it is a conventional horizontal torque pendulum with two C-flex torsional bearing blocks with one bearing block mounted directly above the torque pendulum arm and the other below it on the same rotational axis.  From memory the distance between the bearing blocks to the torque pendulum arm is around 4.0", but I'll re-measure it today to make sure.  The length of the aluminum pendulum arm is 24.00 inches with the center of rotation being offset from its center of mass by about 4.0" before adding additional masses, but again I'll re-measure it today to get its current dimensions.  The Riverhawk C-flex torsion bearing's spring constant is a nominal 0.007 in-Lb/deg., but that varies with the mass load mounted on the torque pendulum arm and selected balance point of the test article mass and its counterbalance mass on the other end of the pendulum arm relative to the torque pendulum’s center of rotation.  Each bearing block is rated for ~25.0 Lb of vertical mass load, so we nominally restrict ourselves to a 25 pound total load limit on the torque pendulum arm to give ourselves a 100% support mass margin.

More later for I now have to get to work...
Star-Drive

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #427 on: 09/10/2014 10:13 PM »
@Paul March,

What are the units of the Electric Field (shown ranging from 0 to 3000) in Fig. 14 ?

titled "Cross section of test article (left) and close up of fields in RF drive pipe (right)", from the <<Computer modeling of the electric field within the pillbox and beam pipe (using COMSOL Multiphysics® software>> 

on page 10 of your paper (with Brady, White, et.al.)

"Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum" July 28-30, 2014, Cleveland, OH,  50th AIAA/ASME/SAE/ASEE Joint Propulsion Conference
« Last Edit: 09/10/2014 10:14 PM by Rodal »

Offline Star-Drive

  • Member
  • Full Member
  • ****
  • Posts: 823
  • TX/USA
  • Liked: 856
  • Likes Given: 9
Re: EM Drive Developments
« Reply #428 on: 09/10/2014 11:32 PM »
@Paul March,

What are the units of the Electric Field (shown ranging from 0 to 3000) in Fig. 14 ?

titled "Cross section of test article (left) and close up of fields in RF drive pipe (right)", from the <<Computer modeling of the electric field within the pillbox and beam pipe (using COMSOL Multiphysics® software>> 

on page 10 of your paper (with Brady, White, et.al.)

"Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum" July 28-30, 2014, Cleveland, OH,  50th AIAA/ASME/SAE/ASEE Joint Propulsion Conference


The units for all the E-field measurements in our 2014 JPC paper is volts per meter (V/m).

As to the torque pendulum dimensions, the center of the two C-flex bearing blocks is 2.38" above and below the centerline of the 24.00" long by 1.50" square aluminum pendulum arm.  The long end of the pendulum arm is 15.5" from the torque pendulum's center of rotation, which makes the other short-end of the pendulum arm 8.5" from the center of rotation.  And all the pendulum's aluminum structural elements are electrically bonded together and then grounded to the vacuum chamber's 304 alloy stainless steel walls that is in turn grounded to the facility's green wire safety ground system.  This grounding arrangement's function is to preclude the buildup of electrical patch charges on the various parts of the pendulum and vacuum chamber during operations.

BTW, the reason we didn't test in vacuum for these test series was that our 35W RF amplifier, that was mounted on the torque pendulum arm as the counterbalance mass for the test articles, was that it's electrolytic capacitors would pop at the low pressures, thus disabling it.  We have since obtained two ~100W RF amplifiers that are hermetically sealed that will allow us to test down to ~5x10^-6 Torr vacuum pressures in the near future, at least once we sort out our current phase locked loop design issues.

Best,
Star-Drive

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #429 on: 09/11/2014 12:50 AM »
@Paul March,

An elegant way to show that the experimentally measured pulse is NOT due to thermal effects is to make the following calculations based on the electric field shown in Fig. 14, titled "Cross section of test article (left) and close up of fields in RF drive pipe (right)", from the <<Computer modeling of the electric field within the pillbox and beam pipe (using COMSOL Multiphysics® software>>  of page 10 of your 2014 JPC paper co-authored with Brady, White, et.al. and to compute the maximum dissipated power as follows:

DissipatedPower = 2 Pi f (E^2) (permittivity of free space) (epsilon')(tan delta)

Taking the maximum value of the Electric Field shown in Fig. 14, p.10, as 4.7189*10^4 V/m , and the given frequency of 935 MHz, it immediately follows (for the Teflon PTFE dielectric resonator) that:


DissipatedPower [W/m^3] = 2 Pi (935*10^6 1/s) (( 4.7189*10^4)^2) (8.85418782*10^(-12)) (2.1)(0.0003)

DissipatedPower =  72973  W/m^3 = 0.0729 W/cm^3

and then use this value to compute the temperature rise in Teflon (without any heat losses):

dT/dt = DissipatedPower / (HeatCapacity*Density)


dT/dt = ( 0.0729 W/cm^3 ) /( (1.3 J/(g deg C)) (2.22 g /cm^3 ))

dT/dt = 0.0253 deg C /s = 1.517 deg C /min

or in deg F:

dT/dt = 0.0455 deg F /s = 2.731 deg F /min



(Fron Fig. 12), the measured pulse's temperature rise takes about 4 seconds, during which time the maximum possible temperature reached is only 0.18 deg F.

Also, at the end of the whole 35 seconds pulse, the maximum possible temperature reached is just 1.59 deg F


Jose' Rodal


Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #430 on: 09/11/2014 01:04 AM »
Also, since the coefficient of thermal expansion of Teflon PTFE is 135* 10^(−6) 1/degC,

and the maximum possible temperature rise is  dT/dt = 0.0253 deg C /s

during the 4 seconds rise, the maximum possible thermal strain expansion of the Teflon PTFE resonator is:


UnrestrainedThermalStrain = 4 s *( 0.0253 deg C /s ) * ( 135* 10^(−6) 1/degC ) = 0.00137 %

which is extremely small.  For example, if the Teflon dielectric resonator is 1 inch (25.4mm) long, the maximum possible (unrestrained) change in length during the 4 sec pulse rise is only: 0.347 micrometers.

Jose' Rodal

« Last Edit: 09/11/2014 01:32 AM by Rodal »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #431 on: 09/11/2014 02:00 AM »
..the 24.00" long by 1.50" square aluminum pendulum arm...

Thanks Paul for all this information.  I am thinking of another explanation, and I have these further questions:

1) Several places in your report discuss the Magnetic Damper System:

<<Figure 3. Electrostatic Fins Calibration System and Magnetic Damper>> (p.4)

<<Whenever a force is induced upon the pendulum arm, the resultant harmonic motion must be damped. This is accomplished via the use of a magnetic dampening system (MDS) at the back of the test rig. Three Neodymium (NdFeB Grade N42) block magnets interact with the pendulum’s aluminum angle to dampen oscillatory motion.>> (p.4)

<<This current causes the power cable to generate a magnetic field that interacts with the torsion pendulum magnetic damper system>> (p.14)

<<Figure 20. Null Test on Torsion Pendulum – average null force is 9.6 micronewtons due to 5.6A DC current in power cable (routes power from liquid metal contacts to RF amplifier; interacts with magnetic damper system)>> (p.16)

QUESTIONS: 

A) Was the magnetic damper on at all times in the shown traces (for the calibration pulses as well as for the thrust pulses both to the left and right ?

B) What is the nature of the interaction << ....with magnetic damper system)>>  discussed in Fig. 20 (p.16) ?

C) Are you able to apply different levels of magnetic damping (and if so have you tested them, with what results) or all you are able to do is to have this level of magnetic damping either on or off ?

D) There is a range of thrust values that were measured for the same nominal conditions.  Were changes in the total supported mass and location of this mass made, and if so do they have any correlation with the range of thrust measured?

2) Do you know the cross-section of the 1.50" square aluminum pendulum arm ? (Is it solid, or if it is a channel, what are its internal dimensions?)

Jose' Rodal

« Last Edit: 09/11/2014 02:01 AM by Rodal »

Offline frobnicat

  • Full Member
  • ****
  • Posts: 518
  • Liked: 500
  • Likes Given: 151
Re: EM Drive Developments
« Reply #432 on: 09/11/2014 09:14 AM »
If the measured thrust pulse were due to displacement of a mass m relative to the fixation point of the device on the arm, a constant thrust (during the pulse) implies a constant acceleration of this mass (again, relative to the mobile arm) a=F/m, and an integrated displacement at the end of the pulse x = 1/2at˛ = 1/2 F/m t˛ = 0.064/m  (approx. with F=80µN and t=40s, and mass m given in kg). After this phase at constant acceleration, the displacement of mass would have to continue at constant velocity for some more time to mimic a sharp fall in thrust but not an opposite thrust pulse (has would be the case if displacing mass suddenly stopped from its integrated vt at velocity). So 0.064/m is a minimum displacement, and it's 6.4 cm assuming something 1kg is moving, or 6.4mm for 10kg. Even if a thermally expanding part of the device were driving a more massive part (not needing heating the whole mass to move it), the required displacement seems much too huge for a thermal expansion, given the scales and max temperatures overall.

Nice to have some first hand feedback on this thread.

Maybe I missed that but anyone inquired or commented on the apparently systematic slope changes (downward) after the relatively clean square thrust periods ? Is there a higher period (>200s) harmonic mode in the system that gives those overall slopes on the order of 1µN/s, or is this just long period "random drifts" due to sensitivity of system ? The charts show the signals measured relative to this slowly drifting baseline (drawn as dotted yellow curve, like piecewise linear best fits). The statistically small sample makes hypothesis risky, but visually there appear a systematic downward change of slope after thruster's pulses, and (also not quite clearly) no slope changes after the (arguably smaller magnitude) 30 µN calibration periods. What is the relevance or irrelevance of this drifting baseline ?

edit : I meant "has would be the case if displacing mass suddenly stopped from its integrated at velocity"
« Last Edit: 09/11/2014 09:39 PM by frobnicat »

Offline Star-Drive

  • Member
  • Full Member
  • ****
  • Posts: 823
  • TX/USA
  • Liked: 856
  • Likes Given: 9
Re: EM Drive Developments
« Reply #433 on: 09/11/2014 12:39 PM »
..the 24.00" long by 1.50" square aluminum pendulum arm...

Thanks Paul for all this information.  I am thinking of another explanation, and I have these further questions:

1) Several places in your report discuss the Magnetic Damper System:

<<Figure 3. Electrostatic Fins Calibration System and Magnetic Damper>> (p.4)

<<Whenever a force is induced upon the pendulum arm, the resultant harmonic motion must be damped. This is accomplished via the use of a magnetic dampening system (MDS) at the back of the test rig. Three Neodymium (NdFeB Grade N42) block magnets interact with the pendulum’s aluminum angle to dampen oscillatory motion.>> (p.4)

<<This current causes the power cable to generate a magnetic field that interacts with the torsion pendulum magnetic damper system>> (p.14)

<<Figure 20. Null Test on Torsion Pendulum – average null force is 9.6 micronewtons due to 5.6A DC current in power cable (routes power from liquid metal contacts to RF amplifier; interacts with magnetic damper system)>> (p.16)

QUESTIONS: 

A) Was the magnetic damper on at all times in the shown traces (for the calibration pulses as well as for the thrust pulses both to the left and right ?

B) What is the nature of the interaction << ....with magnetic damper system)>>  discussed in Fig. 20 (p.16) ?

C) Are you able to apply different levels of magnetic damping (and if so have you tested them, with what results) or all you are able to do is to have this level of magnetic damping either on or off ?

D) There is a range of thrust values that were measured for the same nominal conditions.  Were changes in the total supported mass and location of this mass made, and if so do they have any correlation with the range of thrust measured?

2) Do you know the cross-section of the 1.50" square aluminum pendulum arm ? (Is it solid, or if it is a channel, what are its internal dimensions?)

Jose' Rodal


Answers to Dr. Rodal's Questions:

1 A & B & C) The Eagleworks Lab's magnetic damper uses three or four permanent magnets mounted on the fixed portion of the torque pendulum with a partial steel flux return that generate a constant B-field that in turn interacts with a 1.5" wide by 0.25" thick aluminum angle that is mounted to the back side of the moving aluminum arm of the torque pendulum.  Thus the B-field generates damping eddy currents in the moving aluminum angle whenever it moves relative to the B-field.  So yes it was designed to be active at all times throughout the entire stroke of the torque pendulum.  This magnetic damper system provided less than critically damped performance requiring approx. 1.5 cycles to damp out any induced motion in the pendulum arm.  The natural oscillation period of the pendulum arm when loaded with the RF amplifier, its RF plumbing and the test article was around 4.5 seconds.

D) The mass on the torque pendulum and its orientation was held constant for a particular test series.  Of course slight changes to the placement of the test article occurred when we reversed the thrust vector of the test articles, but we tried to keep the balance approximately the same for all data runs.

 In regards to force calibration we used a set of NIST traceable, pre-calibrated meshed electrostatic fins that provided a constant attractive force between the fin pair for a given applied calibration voltage over a 25% to 75% meshed fins range.  This feature allows us to generate the same calibration force independent of the loading of the torque pendulum's C-flex torsional bearings or how much the fin set is meshed within the noted fin mesh range.  Thanks go to our previous Co-Op, Bill O'Neal, who is now at Purdue University for this design. 

2) The Fastek 6063-T6 aluminum extrusion cross section used for these tests is shown in the report or at its vendor and suppliers.

See:http://www.faztek.net/ and 
http://www.amazon.com/Faztek-15QE1515UL-Aluminum-T-Slotted-Extrusion/dp/B008MQA11C 
Star-Drive

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #434 on: 09/11/2014 04:15 PM »
3. Is the torsion pendulum the only way to test a device of this sort on Earth?

There are several types of small thrust measurement devices.   NASA has a rich tradition, with the Micropound Extended Range Thrust Stand (MERTS) at NASA' Goddard Spacefliight Center and RCA-New Jersey-Lab (http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19830026735.pdf ) (circa 1980 ?) and
Haag's plasma thrust stand at NASA's Lewis Research Center (http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19960008159.pdf ) in 1995. 


Basically there are three main types of devices (and subcategories):

1) The hanging pendulum -either in torsion or swinging-.  Torsion hanging pendulums have been used for sensitive, groundbreaking experiments in physics in several labs, for example, for gravity (inverse square law) and Casimir force measurements.

2) The inverted pendulum -either in torsion or swinging-.  An inverted torsion pendulum has been used by NASA Eagleworks under Dr. White.

3) The torsional style -not to be confused with torsion pendulums ! -, which uses a counterweight on one side to counteract the weight of gravity on the other side, to keep it in a neutral position.  In the neutral position there should be a horizontal line connecting the weights (that are located to the left and to the right of the central pivot point).  Its sensitivity depends on the stiffness of the pivot and the arm lengths.

The hanging pendulum and the inverted pendulum have advantages and disadvantages when compared to each other.

STABILITY: The hanging pendulum balance is inherently stable because gravity stabilizes any swinging oscillations it may experience. The inverted pendulum on the other hand is inherently unstable since the gravitational force is such as to swing the pendulum off-center.   Therefore, the inverted pendulum needs some form of upright stabilization: mainly, the bending stiffness of the vertical arm and the magnetic damping to eliminate parasitic frequencies.   (The critical ratio is the ratio of the bending stiffness divided by the supported mass as I have shown in a previous post). Magnetic dampers are also used for hanging torsion pendulums but the inverted torsion pendulum (due to its inherent instability) has more complicated nonlinear dynamics (capable of chaotic motions with attractors).


SENSITIVITY:  The sensitivity of the hanging pendulum is hindered by the length of the arm -an important issue when trying to make measurements inside a vacuum chamber, since the balance has to fit inside it-.   The inverted pendulum can be made sensitive by matching the pivot's bending stiffness with the gravitational force, but this depends on: A) the supported mass, B) the bending stiffness of the pendulum arm (for swinging motion) and C) the length of the pendulum arm, as I have shown in a previous post.
« Last Edit: 09/11/2014 05:44 PM by Rodal »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #435 on: 09/11/2014 04:20 PM »
The quoted article by Brady, White, March, et.al., has an outstanding ending, that reads:

<<The current plan is to support an IV&V test campaign at the Glenn Research Center (GRC) using their low thrust torsion pendulum followed by a repeat campaign at the Jet Propulsion Laboratory (JPL) using their low thrust torsion pendulum. The Johns Hopkins University Applied Physics Laboratory has also expressed an interest in performing a Cavendish Balance style test with the IV&V shipset.>>

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5900
  • Likes Given: 5248
Re: EM Drive Developments
« Reply #436 on: 09/11/2014 04:37 PM »
Under Professor Manuel Martinez-Sanchez (an authority in electric propulsion) students at MIT have developed different thrust balances (usually developed for their Master Thesis).

One of these balances, built a few years ago under Professor Manuel Martinez-Sanchez, had these advantages:
 
A) the ability to produce a translation, as opposed to a rotation under thrust. (Recall that the thrust balances used by Dr. White and the ones planning to be used at Glenn, JPL and John Hopkins, actually measure a rotation rather than a translation). This was accomplished by including  flexible points of known stiffness at both the top and the bottom of the stand. The engine (whose thrust is being measured) remains horizontal as the stand tilts, keeping the thrust in a known angular direction and reducing the uncertainty in the stand's measurements.

B) The MIT inverted pendulum thrust balance includes a counterweight.  It has the effect of removing (or diminishing) the sensitivity to base vibrations by creating a statically balanced design. The counterweight also removes the gravity term from the denominator of the equation for the displacement. This is very important for nonlinear dynamics stability reasons (attractors).

However my understanding is that the sensitivity of the MIT thrust measurement balance may be insufficient for Dr. White's purposes, my recollection is that the MIT Martinez-Sanchez balances range up to dozens of milliNewtons (?) with an uncertainty of a few hundred microNewtons which is sufficient for conventional electric thrusters but it is way insufficient for Dr. White's purposes.  But anyone interested should check with MIT's Aero & Astro department as they may have newer and/or more sensitive thrust balances and my understanding of their sensitivity may be incorrect !.
« Last Edit: 09/11/2014 07:00 PM by Rodal »

Offline aceshigh

  • Full Member
  • ****
  • Posts: 606
  • Liked: 171
  • Likes Given: 16
Re: EM Drive Developments
« Reply #437 on: 09/11/2014 05:53 PM »
at NextBigFuture, GoatGuy (a known Skeptic and quite good at demolishing a lot of nonsensical stuff with good science and calculations, but also sometimes he turns his cannons even to science done well just because he doesn´t believe it), posted some stuff

Quote
feel like I'm watching a dumbshow.

So… anyone … and I mean anyone at all … do you think that there would be a electromechanical interaction of a big honking antenna inside a big conductive metal container? Ya think? No? Why not!?

I know this is somewhat the "wrong analogy" but consider, just for a second: you have a magnet that you've attached to a string. you want to measure its torsion. You place it in a copper vacuum chamber a diameter or two larger than the magnet. You start it swinging. What's the first thing that happens?

The magnet induces substantial eddy currents in the surrounding copper enclosure. This in turn generate counteracting magnetic fields. These in turn rapidly quench the oscillation of the magnet. There's real force there. REAL. Indeed, this arrangement (inverted) is used to quench the oscillations of a beam-balance's beam. Force, without stiction.

What I see is a nice big copper horn, inside a nice big metal cylinder (the vacuum cylinder). It is supposedly a high Q device, so that the microwave field will be approximately Q • P watts in energy density. Since those watts have to "go somewhere", guess what … they fill the cylinder, which has the geometry of a particularly nice Faraday shield. The chamber is not anechoic, so, they just bounce around like mad, making nodes and antinodes.

Now you think those nodes and antinodes aren't going to interact mechanically with all that metal stuff which is the apparatus?

I bet when the thing is lofted (at GREAT NASA expense) to space, it'll fail to deliver the micronewtons that it supposedly delivers. Or, to put it differently, it'll be one of the greatest days of experimental physics if it does develop the nanonewtons, when free-floating in space.

And it will be dâmned easy to measure, too. Send it off at a few meters per second "away" before turning on power. Get a good fix on its ultra-precise transmitter (laser). Turn on power. Watch it accelerate away by observing the doppler shift in the laser (sensitive to micrometers per second per second). Turn it off. Watch the change. Turn it on, watch the change. Leave it on until the power supply quits. Measure the ΔV again. It should mathematically prove, or disprove the effect.


Quote
And PS: Whenever I see in an article the stretch of imagination being used to build missions to Deimos in 50 days, I just want to puke. Its like … I need to write a book … “Seriously Bogus Science” or something.

The science which is serious enough to with straight-face, entertain all these creative things, without being critical enough of the experimental "conditions" to detect the systematic error of reasoning contained therein. The kind of science which is serious enough to pander Mills' endless succession of tripe-papers building on prior tripe-papers, purporting to have come up with a form of hydrogen in a degenerate state, that if it could exist at all would be the most common form of the stuff, and the whole Universe would have collapsed, before it was even the size of a watermelon. Oh, darn.

It is as if simply talking, and talking, and talking some more about “Q-thrusters” is somehow making them plausible, and is arguing away the systematic errors.

Folks, it is not. You don't measure micro-earthquakes during a major one. In electrical engineering as well as quantitative statistics of measurements, its called "signal to noise ratio" and "systemic errors".

Quote
To buy a $61,000,000 per lob launch-window with SpaceX, and to lob a nice space-worthy version of the device out there, to see how it works. $61,000,000 may sound like a lot to us groundhogs, but in space sciences, this is almost a rounding error. Maybe 3 rounding errors.

Furthermore, the expense is so minor, that one might reasonably get the trip "for free" from SpaceX themselves, as they clearly need to have "live fire" tests of their Falcon Heavy rocket, upcoming.

Let's say the testbed costs about a million to make competently. With 2 parts, with nice lasers, with big batteries, and all that. The thing on the test-bed didn't look like it would have cost more than $25,000 to make. I mean, under 100 watt transmitting tubes, at low gigahertz frequencies, and a bunch of commonplace copper to hold it all together. So, the test thing will be cheap 'n' dirty.

IF the Falcon Heavy launch is successful, then hey … the science cost $1,000,000. If it fails, then build another one at 10% the price (now that the kinks are worked out), and try again. SpaceX won't be put out, they can tolerate the extra (good for them) mass of a the test-bed.

If they want to get really edgy, then incorporate a bunch of micro-satellite projects from school-kids across the land. You know, growing beans in space, and whether milk will make graham crackers turn to mush in zero G.

LOL


Offline Hanelyp

  • Full Member
  • ***
  • Posts: 369
  • Liked: 64
  • Likes Given: 252
Re: EM Drive Developments
« Reply #438 on: 09/11/2014 07:24 PM »
I'm not quite following GoatGuy's reasoning, but there are issues that can arise from a high-Q RF system.  My own focus of concern is unbalanced forces from RF in the feed line.  Without knowing the routing of the feed line I can't evaluate these potential errors properly, but plausible case can easily produce the force levels I've seen cited.

Offline Star-Drive

  • Member
  • Full Member
  • ****
  • Posts: 823
  • TX/USA
  • Liked: 856
  • Likes Given: 9
Re: EM Drive Developments
« Reply #439 on: 09/11/2014 08:02 PM »
If the measured thrust pulse were due to displacement of a mass m relative to the fixation point of the device on the arm, a constant thrust (during the pulse) implies a constant acceleration of this mass (again, relative to the mobile arm) a=F/m, and an integrated displacement at the end of the pulse x = 1/2at˛ = 1/2 F/m t˛ = 0.064/m  (approx. with F=80µN and t=40s, and mass m given in kg). After this phase at constant acceleration, the displacement of mass would have to continue at constant velocity for some more time to mimic a sharp fall in thrust but not an opposite thrust pulse (has would be the case if displacing mass suddenly stopped from its integrated vt velocity). So 0.064/m is a minimum displacement, and it's 6.4 cm assuming something 1kg is moving, or 6.4mm for 10kg. Even if a thermally expanding part of the device were driving a more massive part (not needing heating the whole mass to move it), the required displacement seems much too huge for a thermal expansion, given the scales and max temperatures overall.

Nice to have some first hand feedback on this thread.

Maybe I missed that but anyone inquired or commented on the apparently systematic slope changes (downward) after the relatively clean square thrust periods ? Is there a higher period (>200s) harmonic mode in the system that gives those overall slopes on the order of 1µN/s, or is this just long period "random drifts" due to sensitivity of system ? The charts show the signals measured relative to this slowly drifting baseline (drawn as dotted yellow curve, like piecewise linear best fits). The statistically small sample makes hypothesis risky, but visually there appear a systematic downward change of slope after thruster's pulses, and (also not quite clearly) no slope changes after the (arguably smaller magnitude) 30 µN calibration periods. What is the relevance or irrelevance of this drifting baseline ?


We found that this slope change after the test article and RF amplifer were turned on for 10-to-20 seconds was apprently due to IR radiation from the amplifier's heatsink that is mounted on the back side of the torque penlulum on an 8" square platform was affecting the top C-flex bearing more than the lower one.  We tried aluminum shielding the top bearing assembly from the heatsink IR source and managed to reverse the metioned thermal slope in the thrust plots, but after shielding the bottom one we could reduce it but still coundn't completely get rid of this thremal drift artifact.  Currently we are just living with it.
Star-Drive

Tags: