#### aero

• Senior Member
• Posts: 2928
• 92129
• Liked: 783
• Likes Given: 272
##### Re: EM Drive Developments
« Reply #2560 on: 10/25/2014 01:00 AM »
Yes, I know that in the design none of the geometry variables are independent. That is, they are all interdependent. What I was showing is the effect of measurement error from the photographs. Those errors can be considered to be more or less independent of each other.

Measurement errors are mostly dependent on where we estimate the corners of the 2-d view of the cavity to be, then errors in the reference length be it from the image (Brady) or literature (Shawyer).
« Last Edit: 10/25/2014 01:06 AM by aero »
Retired, working interesting problems

#### Rodal

• Senior Member
• Posts: 5909
• USA
• Liked: 6115
• Likes Given: 5366
##### Re: EM Drive Developments
« Reply #2561 on: 10/25/2014 01:20 AM »
Yes, I know that in the design none of the geometry variables are independent. That is, they are all interdependent. What I was showing is the effect of measurement error from the photographs. Those errors can be considered to be more or less independent of each other.

Measurement errors are mostly dependent on where we estimate the corners of the 2-d view of the cavity to be, then errors in the reference length be it from the image (Brady) or literature (Shawyer).
Yes, I got that.  Actually for the Shawyer design factor I get that the most sensitive parameter is the small diameter, about 2 to 3 times more than the big diameter and the big diameter about 10 times more sensitive than the length which is the least sensitive parameter.  Too bad that we were not given the small diameter and we have to estimate, since that is the most sensitive parameter.

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2562 on: 10/25/2014 01:57 AM »
...I hope your records are better than mine because I don't have lengths recorded, (or the other dimensions, for that matter.)

Correction.  The Search function on this site sux. The googol sdearch engine is slightly and only slightly better.

And this is from ma guy who has just arrived home from a productive visit with his mixologist.  The english language and its capability for explanation has been derided on this thead by the math elite, to no pragmatic avail. If it can't be said in English, then the math ain't all that valuable.  But don't get me started on the value of meaning outside the mathematical paradigme.

I will take another stab at the dimesnsional analysis.  Soon, but no sooner than I feel like.
« Last Edit: 10/25/2014 02:05 AM by JohnFornaro »
Sometimes I just flat out don't get it.

#### ThinkerX

• Full Member
• Posts: 325
• Liked: 122
• Likes Given: 61
##### Re: EM Drive Developments
« Reply #2563 on: 10/25/2014 01:58 AM »
Seems to me we'd be further along here if the researchers had posted the dimensions of these devices in the first place.  As it is, we are taking a 'best guess' approach.

That said...been trying to think a wrong idea through here.

Ok, we have a truncated copper cone, with a microwave emitter at the small end.  Interior of the cone is polished.

Photons do not have mass, but do have inertia, which they can transfer to other objects (which is why Solar Sails work).

A photon must move at C (light-speed) - if it moves less than C, it stops being a photon.  (Is this right?)

So, to follow the initial path of a single photon in the device and its effects.

1) Photon leaves emitter.  Question: is there recoil?

2) Photon travels through the middle of the cone, hits the large flat (?) base of the cone.  As per a solar sail, the photons inertia (momentum) is transferred to the cone.  Then, instead of sticking, photon bounces.  If there is recoil in step 1, then said recoil should be cancelled or mostly cancelled at this point.

3) So photon flies back from the flat base of the cone, and hits the cone wall - which is both curved and angled.  The impact still transfers momentum, but NOT in the exact opposite direction of the impact in step 2.  Instead, the angle and curvature of the cone dictate that a good portion of the released force will be at an angle, maybe even sideways (along the cones curve).  (in open space, without constraints, would enough such impacts cause the cone to rotate?).  Anyhow, photon bounces again, towards the base.

4) photon hits base, which being flat, lets the photon transfer all or most of its momentum a second time, before bouncing back to the cone wall, repeating step 3.

Or...most/all photons hitting the cones base, regardless of direction, or number of times, contribute to forward force.
And most/all photons hitting the cone wall expend at least part of their energy in a 'sideways' direction.

Except this means a violation of the laws of thermodynamics and a cop named Maxwell is trying to write me a ticket.

Put a nozzle (opening in the cavity) in there somewhere (near the emitter), and you might have a turbocharged photon rocket.

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2564 on: 10/25/2014 02:05 AM »
If this is the only way air inside cavity can be heated then yes.

I can vouch for the phenomena that M/W air gets filled with warm water vapor, from hitting the 'potato' button the other day.  The air in the M/W was warm and moist, due to potato water that had been evaporated by the influx of M/W energy.  The moist air dispersed with the opening of the M/W door.
Sometimes I just flat out don't get it.

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2565 on: 10/25/2014 02:08 AM »
`
Seems to me we'd be further along here if the researchers had posted the dimensions of these devices in the first place.  As it is, we are taking a 'best guess' approach.

See my remark on the reflection upon the experimentors WRT the easy questions that have been asked.  The experimentors can only be said to mock the real efforts to understand that are mede on this thread.
Sometimes I just flat out don't get it.

#### Rodal

• Senior Member
• Posts: 5909
• USA
• Liked: 6115
• Likes Given: 5366
##### Re: EM Drive Developments
« Reply #2566 on: 10/25/2014 02:21 AM »
If this is the only way air inside cavity can be heated then yes.

I can vouch for the phenomena that M/W air gets filled with warm water vapor, from hitting the 'potato' button the other day.  The air in the M/W was warm and moist, due to potato water that had been evaporated by the influx of M/W energy.  The moist air dispersed with the opening of the M/W door.

Do you think that the NASA experimenters left some moist potatoes or something similar inside these cavities by accident?

You give me the material...I work with it ...
« Last Edit: 10/25/2014 02:47 AM by Rodal »

#### aero

• Senior Member
• Posts: 2928
• 92129
• Liked: 783
• Likes Given: 272
##### Re: EM Drive Developments
« Reply #2567 on: 10/25/2014 03:21 AM »
Probably just warming cold coffee.
-------------------
@Rodal - Is it possible to model Q analytically for an RF frequency and cavity geometry?

If so, then you could maximize the whole thing, F = PQs/c * ((1/w_big)-(1/w_small)) given an operating frequency. Of course as has been said, a good resonate cavity does not automatically mean a good thruster.
Retired, working interesting problems

#### ThinkerX

• Full Member
• Posts: 325
• Liked: 122
• Likes Given: 61
##### Re: EM Drive Developments
« Reply #2568 on: 10/25/2014 04:56 AM »
Hmmm...a low voltage DC 'lifter'?

could this be the culprit behind the EM Drive?

Maybe somebody here could build such a device...

#### frobnicat

• Full Member
• Posts: 518
• Liked: 500
• Likes Given: 151
##### Re: EM Drive Developments
« Reply #2569 on: 10/25/2014 12:59 PM »
Ha ha, what a good joke! Could tell where the thin invisible wire was attached though...
« Last Edit: 10/25/2014 12:59 PM by frobnicat »

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2570 on: 10/25/2014 01:38 PM »
If this is the only way air inside cavity can be heated then yes.

I ... [hit] the 'potato' button the other day.  ...

Do you think that the NASA experimenters left some moist potatoes or something similar inside these cavities by accident?

You give me the material...I work with it ...

I keep saying that the material you seek to work with is CHBW, which work best when dry.
Sometimes I just flat out don't get it.

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2571 on: 10/25/2014 01:39 PM »
...could this be the culprit behind the EM Drive?...

I'm not a Scot, so I cannae work on that druve,
Sometimes I just flat out don't get it.

#### frobnicat

• Full Member
• Posts: 518
• Liked: 500
• Likes Given: 151
##### Re: EM Drive Developments
« Reply #2572 on: 10/25/2014 02:24 PM »
First attempt to model thrust from cavity's air volumetric heating and jet effect through small aperture. Case studied : Brady a.

Model needs refining (please be patient) but first rough estimates put in the ballpark of 100µN effect during 30s with fast rise of 2s. For now I will give numerical values and derive feasibility from them, in reverse from the set of general equations from which those values derive because I'm still struggling to put some order with all parameters and dependences. Also I want a clean differential equation, takes some time.

Needed hypothesis : air inside cavity is volumetrically heated at Pow=4W (that is, air gets around 25% of microwave power input). There are hole(s) or crevice(s) between cavity and exterior of device, in the direction of thrust (air jetting...) for a total area of A=1.6 mm².

Mind you, volumetric air heating Pow=4W and hole area A=1.6mm² are unknowns, so these values were carefully chosen to get to some magnitude to explain... not the other way around.

Temperatures around 20°C  T = 293K
Cavity volume V = 0.027 m^3
Initial mass of air m = rho*V = 1.2 * .027 = .0324 kg (32 grams of air)
Assuming some air will be leaving but only a small part of that, so not significantly alter the heat capacity of the whole. Heat capacity supposed constant.

We put 4W of power into that, with a specific heat of  C=1000J/kg/K  Tdot = Pow / ( C * m ) = 0.123 K/s

Consider first a short transient period with pressure buildup (like the hole is closed), m constant, p*V=m*Rs*T (Rs around 287 J/kg/K for air)
pdot = m*Rs/V *Tdot = 42 Pa/s  (Pascals per second, please remember there is 10^5 Pa in one atm. pressure). So in 2s we would reach 84 Pa more pressure inside vessel than outside. Note that Tdot uses an isobar value of C (1000) while strictly in this phase we would be in isochore conditions (C=720) : that would only lower the time. So in less than 2 seconds 84 Pa differential. Pd = 84 Pa.

This is the transient. At this level of pressure differential the rate of air escaping through the hole becomes important enough that this pressure will be kept constant while the temperature continues to rise (why I choose isobar heat capacity overall) :
The mass flow through aperture mdot = Cf A sqrt(2 rho Pd) where Cf is a flow coefficient depending on geometry and I understand is around 0.6 for rough holes.
This is from orifice plate article on wikipedia (repress any second thought here), I took the most simplified forms assuming incompressible flow (pressure differential / absolute pressure < 1/1000 so I guess this is a good first approximation). On this chapter, equation (2).

=> mdot = 1.36 10^-5 kg/s  of air expelled through the 1.6 mm² hole(s).
This is a volume flow of mdot/rho = 1.13 10^-5 m^3/s
Speed of ejection is volume flow divided by hole section : v = 7.1 m/s
Clearly not sonic or supersonic.

The reaction force imparted (thrust) = mdot * v = 9.65 10^-5
Thrust = 96 µN

Reported by Brady a : 91.2 µN

After 30 seconds of this "steady state" of constant rate of heating and expelling air :
T would be 3.7 ° above initial conditions
Mass of air expelled 0.41 g (a bit more than 1% of initial air mass of 32g)

Sanity check on total momentum :
From momentum expelled by jet : 0.41e-3 kg * 7.1 m/s = 2.9e-3 kg m/s
From required thrust during 30s : 96e-6 N * 30 s = 2.9e-3 kg m/s

When power off, Tdot falls to 0, sharp fall to 0 thrust ( differential pressure is quickly released, pressure equilibrium restored with same time constant as rise time < 2s). Conduction slowly release the heat of gas through the (colder) copper walls with a long time constant. No visible effect of contracting gas sucking air through the hole(s) when restoring T equilibrium.

« Last Edit: 10/25/2014 02:26 PM by frobnicat »

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2573 on: 10/25/2014 02:40 PM »
Needed hypothesis : air inside cavity is volumetrically heated at Pow=4W (that is, air gets around 25% of microwave power input). There are hole(s) or crevice(s) between cavity and exterior of device, in the direction of thrust (air jetting...) for a total area of A=1.6 mm².

I would say that the perimeter of the big end leaks.  Not sure how you'd model that leaky connection of the PCB to the copper frustrum flange.

At any rate (rate?  get it?) the are coming out radially at right angles to the thrust axis of the thingy.

[Hint from the inappropriate humor department:  Listener maximizes humor by familiarity with Brian Eno's album cover art.  Linked artwork is not true color, but that's immaterial.]
Sometimes I just flat out don't get it.

#### Rodal

• Senior Member
• Posts: 5909
• USA
• Liked: 6115
• Likes Given: 5366
##### Re: EM Drive Developments
« Reply #2574 on: 10/25/2014 03:01 PM »
Needed hypothesis : air inside cavity is volumetrically heated at Pow=4W (that is, air gets around 25% of microwave power input). There are hole(s) or crevice(s) between cavity and exterior of device, in the direction of thrust (air jetting...) for a total area of A=1.6 mm².

I would say that the perimeter of the big end leaks.  Not sure how you'd model that leaky connection of the PCB to the copper frustrum flange.

At any rate (rate?  get it?) the are coming out radially at right angles to the thrust axis of the thingy.

[Hint from the inappropriate humor department:  Listener maximizes humor by familiarity with Brian Eno's album cover art.  Linked artwork is not true color, but that's immaterial.]

<< are coming out radially at right angles to the thrust axis of the thingy.>>
A=1.6 mm² leaking could be axial due to the gap between the bolts and the boltholes

1.6 mm² is only 0.0025 square inches or a square having 0.05 inches per side

Look deeply, look at all those bolts.  I count 24 bolts, so this would be a gap of only 0.0001 square inches per bolt

24 nozzles of big end escaping gas
« Last Edit: 10/25/2014 03:10 PM by Rodal »

#### frobnicat

• Full Member
• Posts: 518
• Liked: 500
• Likes Given: 151
##### Re: EM Drive Developments
« Reply #2575 on: 10/25/2014 03:21 PM »
Needed hypothesis : air inside cavity is volumetrically heated at Pow=4W (that is, air gets around 25% of microwave power input). There are hole(s) or crevice(s) between cavity and exterior of device, in the direction of thrust (air jetting...) for a total area of A=1.6 mm².

I would say that the perimeter of the big end leaks.  Not sure how you'd model that leaky connection of the PCB to the copper frustrum flange.

At any rate (rate?  get it?) the are coming out radially at right angles to the thrust axis of the thingy.

[Hint from the inappropriate humor department:  Listener maximizes humor by familiarity with Brian Eno's album cover art.  Linked artwork is not true color, but that's immaterial.]

<< are coming out radially at right angles to the thrust axis of the thingy.>>
A=1.6 mm² leaking could be axial due to the gap between the bolts and the boltholes

1.6 mm² is only 0.0025 square inches or a square having 0.05 inches per side

Look deeply, look at all those bolts.  I count 24 bolts, so this would be a gap of only 0.0001 square inches per bolt

Or the seam that appear on the side of cone (assuming it's not welded). If leaks are not perfectly axial there would still be a net thrust but with a lower efficiency (cosine of the angle relative to axis). The magnitude jet effects can get are pretty close what is to be explained, would be hard to account for less than ideal jet directions... yet it's tantalizingly close. Equations together (will try to summarize that this week-end) can give higher thrust with lower leak area but at the price of higher time constant to reach delta pressure equilibrium (more than 2s).

Also a jet that would come out a seam between two planes perpendicular to axis would have a significant axial component when the flange is asymmetric (for instance the copper part ends when the PCB extends a little bit further...)

In what direction (average) the jets would have to go ? I'm still all confused with the thrust directions.

The big caveat : 4W volumetric heating of (presumably not perfectly dry) air out of 16W microwave power in "empty oven".
« Last Edit: 10/25/2014 03:23 PM by frobnicat »

#### Mulletron

• Full Member
• Posts: 1140
• Liked: 820
• Likes Given: 1062
##### Re: EM Drive Developments
« Reply #2576 on: 10/25/2014 03:22 PM »
It is incorrect to state that only DC fields can produced ionic winds.  I don't know whether such misunderstanding comes from getting information from Wikipedia.

AC fields can also produce ionic wind in a variety of ways.  For example, the point electrode and ring electrode system is capable of generating electric winds (with velocities of few m/s) for both DC and AC applied voltages.  In the AC regime, ions generated within the corona move in the field and migrate a distance before recombining; the net flow of ions away from the corona creates a time-averaged force that drives the steady flows.  AC coronas can sustain wind velocities of over 1m/s independent of electrode separation in marked contrast to DC coronas.

Another arrangement in which AC fields can produce ionic wind is dielectric barrier discharge actuators.  AC applied across the electrodes through the dielectric produces a variety of electric breakdown phenomena (e.g., corona, streamers, and plasma).  Spark breakdown is prevented by the dielectric barrier.  The dielectric material needs to be in contact with electrodes such that the electrodes contact each surface of the dielectric.

Transient migration of charged species within AC fields also gives rise to steady electric winds.

In contrast to winds driven by DC fields, AC fields (as in the point electrode and ring electrode system ) generate wind velocities comparable with (or better than) the strongest DC winds for any value of the electrode separation.  In the high-frequency  AC regime (>1 KHz), the electric force is localized within a region near the tip of the point electrode.

From a fundamental perspective, any type of electric wind (DC or AC) derives from the same basic mechanism whereby a steady flux of ions transfers momentum to the surrounding fluid to drive steady gas flows.

No it isn't because I'm some (as you're eluding to) unsophisticated Wikipedia scholar. The subject matter in which I discuss is a matter of my career, waveguides and electronics, informed by my practical experience from being on the job. You know how much flack I gave Ron about posting links to info on Wikipedia. My reference to high voltage DC is strictly to mention that you need high voltage in order to create the corona effect you enjoy from ionizing air. I've worked with a lot of high voltage power supplies up to =15Kvdc and with a simultaneous -6Kvdc.....and none of them were ever AC high voltage power supplies. Not even 1 AC high voltage power supply, because the utility of a high voltage AC power supply isn't useful for biasing tubes or anything else. You can literally feel and smell the ionized air in the vicinity of a charged HVPS. High voltage is the sole reason those lifters work. This isn't applicable to emdrive, hence why I said we can rule out ion wind with emdrive.

If one were to use an AC high voltage power supply, there wouldn't be an ion wind, just an ion wind oscillation.
And I can feel the change in the wind right now - Rod Stewart

#### Mulletron

• Full Member
• Posts: 1140
• Liked: 820
• Likes Given: 1062
##### Re: EM Drive Developments
« Reply #2577 on: 10/25/2014 03:31 PM »
I recall that Paul March wrote that the EMDrives tested at NASA Eagleworks had a temperature that never rose more than 1 deg (F ? or C?) above room temperature.

Anybody recall that statement?  Is the temperature measurement in the NASA Eagleworks report? Using search I cannot find it in the text.  Is it in the pictures?

I also recall AcesHigh reporting on information elsewhere reporting March's statement he made on this thread regarding temperature.  Was that at nextbigfuture? Does anybody still have a link for that?

Yes I remember March saying the temperature didn't rise more than 1 degree F. It was F, not C.
And I can feel the change in the wind right now - Rod Stewart

#### Rodal

• Senior Member
• Posts: 5909
• USA
• Liked: 6115
• Likes Given: 5366
##### Re: EM Drive Developments
« Reply #2578 on: 10/25/2014 03:32 PM »
It is incorrect to state that only DC fields can produced ionic winds.  I don't know whether such misunderstanding comes from getting information from Wikipedia.

AC fields can also produce ionic wind in a variety of ways.  For example, the point electrode and ring electrode system is capable of generating electric winds (with velocities of few m/s) for both DC and AC applied voltages.  In the AC regime, ions generated within the corona move in the field and migrate a distance before recombining; the net flow of ions away from the corona creates a time-averaged force that drives the steady flows.  AC coronas can sustain wind velocities of over 1m/s independent of electrode separation in marked contrast to DC coronas.

Another arrangement in which AC fields can produce ionic wind is dielectric barrier discharge actuators.  AC applied across the electrodes through the dielectric produces a variety of electric breakdown phenomena (e.g., corona, streamers, and plasma).  Spark breakdown is prevented by the dielectric barrier.  The dielectric material needs to be in contact with electrodes such that the electrodes contact each surface of the dielectric.

Transient migration of charged species within AC fields also gives rise to steady electric winds.

In contrast to winds driven by DC fields, AC fields (as in the point electrode and ring electrode system ) generate wind velocities comparable with (or better than) the strongest DC winds for any value of the electrode separation.  In the high-frequency  AC regime (>1 KHz), the electric force is localized within a region near the tip of the point electrode.

From a fundamental perspective, any type of electric wind (DC or AC) derives from the same basic mechanism whereby a steady flux of ions transfers momentum to the surrounding fluid to drive steady gas flows.
.....
If one were to use an AC high voltage power supply, there wouldn't be an ion wind, just an ion wind oscillation.
Anybody in this thread that wants correct information on ionic wind produced by AC, please read for example:

"Electric winds driven by time oscillating corona discharges"
JOURNAL OF APPLIED PHYSICS 114, 143302

Aaron M. Drews,1 Ludovico Cademartiri,2,3,4,5 George M. Whitesides,2,6
and Kyle J. M. Bishop1,a)
1Department of Chemical Engineering, Pennsylvania State University, University Park, Pennsylvania 16802,
USA
2Department of Chemistry & Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, USA
3Department of Materials Science & Engineering, Iowa State University, Ames, Iowa 50011, USA
4Department of Chemical & Biological Engineering, Iowa State University, Ames, Iowa 50011, USA
5Ames Laboratory, US Department of Energy, Ames, Iowa 50011, USA
6Kavli Institute, Harvard University, Cambridge, Massachusetts 02138, USA

http://gmwgroup.harvard.edu/pubs/pdf/1194.pdf

<<Transient migration of charged species within AC fields also gives rise to steady electric winds>>
« Last Edit: 10/25/2014 03:45 PM by Rodal »

#### JohnFornaro

• Not an expert
• Senior Member
• Posts: 9161
• Delta-t is the salient metric.
• Planet Eaarth
• Liked: 617
• Likes Given: 320
##### Re: EM Drive Developments
« Reply #2579 on: 10/25/2014 03:36 PM »
You.  Play nice.

...I don't know whether such misunderstanding comes from getting information from Wikipedia. ...

You.  Get a dictionary.

No it isn't because I'm some (as you're [alluding] to) unsophisticated Wikipedia scholar.
Sometimes I just flat out don't get it.

Tags: