Author Topic: EM Drive Developments - related to space flight applications - Thread 6  (Read 1493683 times)

Offline Space Ghost 1962

  • Senior Member
  • *****
  • Posts: 2780
  • Whatcha gonna do when the Ghost zaps you?
  • Liked: 2925
  • Likes Given: 2247
A few notes about Schlieren systems, near/far field video analysis, and other means to monitor heating.

These systems all have different benefits/issues. The above mentioned optical systems are actually looking at refraction due to air density causing different refraction of transmitted light. So what you can get is a pattern of heating (likely corresponding to nodes impedance heating), but getting actual temps and integratable flux maps will take considerable calibration. Half (or more) the battle with getting such results are proving the accuracy of your measurement, so its not a "code" thing quite so much as a details/apparatus thing.

If what you are going for is total flux, calorimetry is likely what you are after. If point flux is the question, some kind of absorber on a low/non interacting support can allow a means to probe it.

Schlieren systems are scaled by cross sectional area with the optical system on a stiff optical bench with no vibration. It takes effort to collimate the system, damp out vibrations, and to have an ambient temperature stable room before test. High humidity affects measurement and dewing of the optics scatters the beam reducing contrast.

If you google "microwave",  "heating", and "schlieren" you'll find some interesting papers.

Offline OnlyMe

  • Full Member
  • ***
  • Posts: 361
  • So. Calif.
  • Liked: 210
  • Likes Given: 195
Earlier there was a point made about buoyancy lift from the heated frustum. I did some calculations in an attempt to characterize buoyancy for Shell's EM drive configuration, (frustum, tuning cylinder, and two wave guides) since I do have the needed dimensional data in my meep model.

I calculate the volume to be less than 4.490367E-02 m3. Less than because the wave guides are cut at an angle by the frustum but I simply used the length of the long side. So there are two small wedges that are each counted twice.

So I found standard atmosphere data in the engineering toolbox and calculated the air density at constant volume and pressure using the Ideal gas law. I used the calculator here:
http://www.ajdesigner.com/idealgas/ideal_gas_law_density.php#ajscroll

I did this calculation at both standard temperature and 100 degrees C above standard. That is pretty hot, the temperature of boiling water. Oh, out of deference for Shells' high altitude in the mountains, I used 2000 meter data from the standard atmosphere.

With the above, I calculated buoyancy force = 0.1181471225 newtons, that's 118 mn and certainly not something that can be neglected. If someone is up to the task of checking my numbers, I'm sure it would be appreciated.

Now just need some actual heat data from Shell.

Rfmwguy, had the magnetron itself at around 150 degrees C, if I remember correctly, and much less the frustum itself, but that was a mesh frustum.

Once Shell has some actual data, it will start to get interesting for sure.

BTW Shell, since you moved inside, both room temperature and humidity, may be important.

Offline glennfish

  • Full Member
  • ****
  • Posts: 451
  • Liked: 351
  • Likes Given: 194

With the above, I calculated buoyancy force = 0.1181471225 newtons, that's 118 mn and certainly not something that can be neglected. If someone is up to the task of checking my numbers, I'm sure it would be appreciated.

That's in the range of comparable calcs done previously for similar designs.  Definitely a # to throw into the simulator I have.

While you're in a buoyant mood, can you look over HVAC calculators and see what the expected chimney airflow should look like for the same dimensions, depending on orientation, up, down, sideways?

The model I'm trying to work out assumes thermal effects in your order of magnitude that would differ slightly depending upon orientation.

These thermal effects are not bad.  They are to be assumed present.  What I'm trying to work out is a model that would show what kind of test protocol is required to statistically demonstrate thrust given that thrust may be a small percentage of thermal effects.  i.e. it is possible to create a test protocol where the thrust is 10 mn and the boyancy is 118 mn, and you can at some point declare thrust is real.  See attached simulator.  If you can help produce initial estimated values and propose other error sources, that's double plus good.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Just FYI, there exists OpenFOAM, a computational fluid dynamics (CFD) suite, which appears to be fairly mature.  I don't know enough about CFD to comment on whether or not OpenFOAM has the requisite solvers for modeling whichever thermal effects are present in/around a frustum.

http://openfoam.org/
https://en.wikipedia.org/wiki/OpenFOAM

Existing solvers:

http://openfoam.org/features/standard-solvers.php

I would be willing to create IGES, STEP, or STL (or a limited number of other, less common formats) files of frustums and/or experimental setups.  I the software I use has the capability to "develop" solids for finite element analysis (FEA) and export it as ANSYS PREP7, and so therefore I may be able to help with that (if that is even necessary for CFD).  Caveat - I have never "developed" a model for FEA analysis.

Thank you, but CFD (computer fluid dynamics) analysis would, by comparison, make the Meep analysis pale by comparison.  CFD analysis would entail running millions of cycles (hence much longer computational time), iteration sub-cycles (for nonlinear Navier Stokes equation solution), and dealing with numerical relaxation and numerical convergence settings (in addition to dealing with convergence of the mesh).  Couple to that the analyst's time to analyze the input and output of these numerical solutions.
« Last Edit: 12/17/2015 06:52 pm by Rodal »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564

With the above, I calculated buoyancy force = 0.1181471225 newtons, that's 118 mn and certainly not something that can be neglected. If someone is up to the task of checking my numbers, I'm sure it would be appreciated.

That's in the range of comparable calcs done previously for similar designs.  Definitely a # to throw into the simulator I have.

While you're in a buoyant mood, can you look over HVAC calculators and see what the expected chimney airflow should look like for the same dimensions, depending on orientation, up, down, sideways?

The model I'm trying to work out assumes thermal effects in your order of magnitude that would differ slightly depending upon orientation.

These thermal effects are not bad.  They are to be assumed present.  What I'm trying to work out is a model that would show what kind of test protocol is required to statistically demonstrate thrust given that thrust may be a small percentage of thermal effects.  i.e. it is possible to create a test protocol where the thrust is 10 mn and the boyancy is 118 mn, and you can at some point declare thrust is real.  See attached simulator.  If you can help produce initial estimated values and propose other error sources, that's double plus good.

The bigger problem is the transient vortex shedding due to fluid dynamics, interacting with the ON/OFF timing of the magnetron (http://forum.nasaspaceflight.com/index.php?topic=39004.msg1459128#msg1459128 ).

The lift force due to natural convection (with the magnetron constantly OFF) is time-dependent and not a steady constant.

 The constant bouyancy calculations do not address the timing of the transient natural convection (not yet taken into account), and yet the statistical analysis is predicated on statistical analysis of the ON/OFF timing of the magnetron.
« Last Edit: 12/17/2015 06:56 pm by Rodal »

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360

With the above, I calculated buoyancy force = 0.1181471225 newtons, that's 118 mn and certainly not something that can be neglected. If someone is up to the task of checking my numbers, I'm sure it would be appreciated.

That's in the range of comparable calcs done previously for similar designs.  Definitely a # to throw into the simulator I have.

While you're in a buoyant mood, can you look over HVAC calculators and see what the expected chimney airflow should look like for the same dimensions, depending on orientation, up, down, sideways?

The model I'm trying to work out assumes thermal effects in your order of magnitude that would differ slightly depending upon orientation.

These thermal effects are not bad.  They are to be assumed present.  What I'm trying to work out is a model that would show what kind of test protocol is required to statistically demonstrate thrust given that thrust may be a small percentage of thermal effects.  i.e. it is possible to create a test protocol where the thrust is 10 mn and the boyancy is 118 mn, and you can at some point declare thrust is real.  See attached simulator.  If you can help produce initial estimated values and propose other error sources, that's double plus good.

I found this:
http://www.engineeringtoolbox.com/natural-draught-ventilation-d_122.html

but I don't see how to apply it to this problem. We don't have a chimney, rather just a heat source with air flow entering from both the bottom and the 4 sides. More like the air flow around a suspended light bulb. Maybe I'll look further.
Retired, working interesting problems

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2205
  • Liked: 2713
  • Likes Given: 1134
Great discussions...yes, my mag temp idled between 150 and 170 degrees C depending on whether it was 30% or 50% power cycle respectively. 100% power for long duration was not measured on the original mag, but another pullout mag I tested got to 200 deg C quickly at 100% power. Thus confirmed my concern that getting a "matched" set of mags (one on each side of the balance beam) would be difficult...they're just not built precise enough.

Here's a simple test somebody can plug into some software...imagine a 4 inch square metallic box with 2 opposite sides open (vertical sides).

Heat the thing to 170 degrees C from a CENTRAL point within the box, conducting heat to the remaining walls via air convection and direct mechanical attachment.

Ambient air was 28 degrees C and humidity was about 56% in the tests I ran within a few days of each other.

The metal was galvanized steel. It is safe to assume 14 gauge steel or 0.0677 inches thick on 4 sides (5 & 6 are open). The weight was 750 mg (magnetron alone).

For those wishing to dig deeper, the non-hermetic box on top is about 3 inches tall and 3.75 inches square...made of the same material.

Below is a cropped image someone sent me of my mag, suggesting a heatsink idea.

These simple details should be enough to quantify the amount of vertical lift component (perhaps not turbulence).

Are we talking millinewtons? micronewtons? How similar is this to the 40 or so micronewtons of horizontal (torsional) Lorentz force? Remember vertical Lorentz force was not measured nor estimated in Mr Li's paper.


Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360

With the above, I calculated buoyancy force = 0.1181471225 newtons, that's 118 mn and certainly not something that can be neglected. If someone is up to the task of checking my numbers, I'm sure it would be appreciated.

That's in the range of comparable calcs done previously for similar designs.  Definitely a # to throw into the simulator I have.

While you're in a buoyant mood, can you look over HVAC calculators and see what the expected chimney airflow should look like for the same dimensions, depending on orientation, up, down, sideways?

The model I'm trying to work out assumes thermal effects in your order of magnitude that would differ slightly depending upon orientation.

These thermal effects are not bad.  They are to be assumed present.  What I'm trying to work out is a model that would show what kind of test protocol is required to statistically demonstrate thrust given that thrust may be a small percentage of thermal effects.  i.e. it is possible to create a test protocol where the thrust is 10 mn and the boyancy is 118 mn, and you can at some point declare thrust is real.  See attached simulator.  If you can help produce initial estimated values and propose other error sources, that's double plus good.

The bigger problem is the transient vortex shedding due to fluid dynamics, interacting with the ON/OFF timing of the magnetron (http://forum.nasaspaceflight.com/index.php?topic=39004.msg1459128#msg1459128 ).

The lift force due to natural convection is time-dependent and not a steady constant.

 The constant bouyancy calculations do not address the timing of the transient natural convection (not yet taken into account), and yet the statistical analysis is predicated on statistical analysis of the ON/OFF timing of the magnetron.

The problem of buoyancy lift and natural convection lift (drag) forces are separable. Buoyancy is an effect from inside the EM drive while convection occurs outside the drive.

Under the conditions that I assume operation, buoyancy is a function of internal temperature only. Yes, temperature will change as the magnetron (heat source) is turned on and off. Characterizing the internal temperature of the drive as a function of time seems difficult to me, perhaps you have some ideas?

Natural convection is also a function of temperature, in this case, the temperature of the copper. Characterizing this temperature would be easier given the internal temperature as a function of time. The problem then arises, given the temperature of the copper as a function of time (power on/off) what do we do with it? I am not ready to solve the Navier-Stokes equations, I had enough of them while in college.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564

With the above, I calculated buoyancy force = 0.1181471225 newtons, that's 118 mn and certainly not something that can be neglected. If someone is up to the task of checking my numbers, I'm sure it would be appreciated.

That's in the range of comparable calcs done previously for similar designs.  Definitely a # to throw into the simulator I have.

While you're in a buoyant mood, can you look over HVAC calculators and see what the expected chimney airflow should look like for the same dimensions, depending on orientation, up, down, sideways?

The model I'm trying to work out assumes thermal effects in your order of magnitude that would differ slightly depending upon orientation.

These thermal effects are not bad.  They are to be assumed present.  What I'm trying to work out is a model that would show what kind of test protocol is required to statistically demonstrate thrust given that thrust may be a small percentage of thermal effects.  i.e. it is possible to create a test protocol where the thrust is 10 mn and the boyancy is 118 mn, and you can at some point declare thrust is real.  See attached simulator.  If you can help produce initial estimated values and propose other error sources, that's double plus good.

The bigger problem is the transient vortex shedding due to fluid dynamics, interacting with the ON/OFF timing of the magnetron (http://forum.nasaspaceflight.com/index.php?topic=39004.msg1459128#msg1459128 ).

The lift force due to natural convection is time-dependent and not a steady constant.

 The constant bouyancy calculations do not address the timing of the transient natural convection (not yet taken into account), and yet the statistical analysis is predicated on statistical analysis of the ON/OFF timing of the magnetron.

The problem of buoyancy lift and natural convection lift (drag) forces are separable. Buoyancy is an effect from inside the EM drive while convection occurs outside the drive.

Under the conditions that I assume operation, buoyancy is a function of internal temperature only. Yes, temperature will change as the magnetron (heat source) is turned on and off. Characterizing the internal temperature of the drive as a function of time seems difficult to me, perhaps you have some ideas?

Natural convection is also a function of temperature, in this case, the temperature of the copper. Characterizing this temperature would be easier given the internal temperature as a function of time. The problem then arises, given the temperature of the copper as a function of time (power on/off) what do we do with it? I am not ready to solve the Navier-Stokes equations, I had enough of them while in college.

If I understand you correctly you have been addressing only the buoyancy of the EM Drive cavity per se, as if it would be a hot air balloon.

But even a piece of burning paper will experience buoyancy due to the natural convection lift forces on it.

And in essence natural thermal convection is due to fluid dynamics where hotter air molecules experience buoyancy as compared to colder air molecules.  (No need for a cavity to discuss buoyancy).

I am calling buoyancy the lift force due to natural thermal convection.


The magnetron on RFMWGUY's EM Drive was sitting on top of the EM Drive cavity.  The EM Drive cavity had perforated mesh walls.  The thermal camera showed that what got really hot was the magnetron sitting on top of the EM Drive and not the EM Drive cavity.

So I was referring to bouyancy of the magnetron.   The magnetron is also a partially closed cavity, that experiences buoyancy effects.


Also the top plate of the EM Drive, under the magnetron, experiences a lift effect due to the natural convection flow.  So I was also discussing buoyancy of that plate under the magnetron, as well.

These buoyancy forces (on the magnetron, and on the plate under the magnetron) are not constant (even with the magnetron constantly ON, or thereafter constantly OFF), but are a function of time, transient, due to vortex shedding and fluid dynamics transport of the natural convection of air molecules with different temperature.(http://forum.nasaspaceflight.com/index.php?topic=39004.msg1459128#msg1459128)
« Last Edit: 12/17/2015 07:47 pm by Rodal »

Offline glennfish

  • Full Member
  • ****
  • Posts: 451
  • Liked: 351
  • Likes Given: 194


Are we talking millinewtons? micronewtons? How similar is this to the 40 or so micronewtons of horizontal (torsional) Lorentz force? Remember vertical Lorentz force was not measured nor estimated in Mr Li's paper.

Hundred(s) of milli newtons in the baseline frustums so far.  Your mesh meshed things up since it was porous.

Here's where my brain is going and hopefully Aero or someone can plug in some more calcs for modeling.

Aero has provided a 1st approximation of the hot-air balloon effect for Shell's design.  If her mag is rock solid and she flips orientation frequently, a 5-10mn thrust signal will emerge quickly.  If it's in the micro N range, it will emerge very slowly.  If the mag output is wobbly (new technical term), the wobblies could drown out any signal.

Statistically, a physics type likes a 6 sigma, separation between noise and signal.

For the DIY domain, I'd be ecstatic with a 3 sigma.

The issue is separating out thrust signals from thermal and other error sources.  If the thrust is small, the number of required samples required becomes large.  If the thrust is large, the number of samples required becomes small. 

A rule of thumb for a 3 sigma finding. If the thrust is 10% of the other error sources, the number of samples required would be about 50.  It's not a linear relationship, but a good mental model is, for every factor of 10 reduction in thrust as a percentage of other error source, the number of samples required to claim detection increases by a factor of 10.

If errors can be characterized, and their Standard Deviation minimized there's a reduction in the number of samples required.

I dream of a nice quiet stable magnetron, miminal airflow, stable ambient air temperature, and even if Lorenz is in the room hiding behind Elvis, he shouldn't matter much because there isn't going to be any noticeable shift in where the magnetic poles are during the runs.

Imagine it this way:

Assume thermal is 800 mn
Assume Lorenz is 5 mn
Assume other stuff is 10 mn

You have net junk thrust of 815 mn.

If your actual thrust is 5 mn and you can flip the frustum up or down and the only change is the direction of the thrust, then

If up you would see total thrust of 820 mn and down you would see 810 mn.  If the thermal, Lorenz and stuff are rock solid, you'll see that 5 mn clearly and soon. 

The key in my mind is very clean very stable unvarying sources of error.  If their SDs are small, they simply go away from a statistics point of view.
« Last Edit: 12/17/2015 07:50 pm by glennfish »

Offline Space Ghost 1962

  • Senior Member
  • *****
  • Posts: 2780
  • Whatcha gonna do when the Ghost zaps you?
  • Liked: 2925
  • Likes Given: 2247
Take a dead magnetron, core it out, stuff it with resistors or a electric dryer's heating element, and replace for the active one. If you dissipate the same heat, the same way, you've got a adequate "heat dummy" to do a difference with. 

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Take a dead magnetron, core it out, stuff it with resistors or a electric dryer's heating element, and replace for the active one. If you dissipate the same heat, the same way, you've got a adequate "heat dummy" to do a difference with.
Or how about conducting two tests?

PURPOSE: Test whether the "effect" is purely thermal.

Vary the magnetron power vs. time for both cases so that the temperature of the magnetron is similar in both tests.

1) Magnetron RF exciting the EM Drive cavity (as already done by RFMWGUY)

2)Close the RF entrance into the EM Drive.  Magnetron RF NOT exciting the EM Drive cavity. 

_________

After that, another test;

3) Put one magnetron at one end of the balanced beam and the other magnetron at the other end of the balanced beam.  One magnetron's RF is exciting the EM Drive cavity and the other one is not.
« Last Edit: 12/17/2015 07:55 pm by Rodal »

Offline Space Ghost 1962

  • Senior Member
  • *****
  • Posts: 2780
  • Whatcha gonna do when the Ghost zaps you?
  • Liked: 2925
  • Likes Given: 2247
Much more practical, but then what happens is that the magnetics of the magnetron are still in play, and you don't want to start "red herring" theories about the influence of them and the cavity.

By having a non microwave source, it's just about heat.

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360
Quote
But even a piece of burning paper will experience buoyancy due to the natural convection lift forces on it.

No, that is not buoyancy. It is a matter of terminology. It is a lift force due to the natural convection currents created by the heat of burning paper. As I wrote above, buoyancy and convection are separable problems.

The magnetron itself will experience natural convection lift forces, but not buoyancy, because it is a sealed unit. It does not out-gas hence it's mass is constant. Perhaps there is a very small buoyancy due to the expansion of the metal of the magnetron with temperature but the expansion is small and the mass of the air displaced by that expansion is very, very small, I think negligible, when compared to the natural convection lift forces it experiences.

In my post, I was considering SeaShells' test rig and cavity. She has placed the magnetron at the pivot so that lift forces off the hot magnetron itself will be negated, be they buoyancy or natural convection lift forces.

I suggest we focus uniquely on a test set-up for our posts because the error sources are different for different test set-ups. Heat from the magnetron was definitely a factor with rfmwguy's test rig, but should not be a factor with Shells' rig. Similarly, buoyancy of the cavity will be a definite factor with Shells' rig but should not have been as much of a factor, if any, with rfmwguy's mesh cavity design.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Much more practical, but then what happens is that the magnetics of the magnetron are still in play, and you don't want to start "red herring" theories about the influence of them and the cavity.

By having a non microwave source, it's just about heat.

Good point. That is sort of what Yang reported in her last series of tests (she just heated her EM Drive). 
She appears to not have reported anything else after that... (and it has been a long time)

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Quote
But even a piece of burning paper will experience buoyancy due to the natural convection lift forces on it.

No, that is not buoyancy. It is a matter of terminology. It is a lift force due to the natural convection currents created by the heat of burning paper. As I wrote above, buoyancy and convection are separable problems.

The magnetron itself will experience natural convection lift forces, but not buoyancy, because it is a sealed unit. It does not out-gas hence it's mass is constant. Perhaps there is a very small buoyancy due to the expansion of the metal of the magnetron with temperature but the expansion is small and the mass of the air displaced by that expansion is very, very small, I think negligible, when compared to the natural convection lift forces it experiences.

In my post, I was considering SeaShells' test rig and cavity. She has placed the magnetron at the pivot so that lift forces off the hot magnetron itself will be negated, be they buoyancy or natural convection lift forces.

I suggest we focus uniquely on a test set-up for our posts because the error sources are different for different test set-ups. Heat from the magnetron was definitely a factor with rfmwguy's test rig, but should not be a factor with Shells' rig. Similarly, buoyancy of the cavity will be a definite factor with Shells' rig but should not have been as much of a factor, if any, with rfmwguy's mesh cavity design.

That may not be buoyancy according to you, but it is according to what I learnt,

Buoyancy is a force resulting from differences in fluid density


 and also according to:

http://www.merriam-webster.com/dictionary/buoyancy

Quote
a :  the tendency of a body to float or to rise when submerged in a fluid
b :  the power of a fluid to exert an upward force on a body placed in it; also :  the upward force exerted

https://en.wikipedia.org/wiki/Buoyancy

Quote
In science, buoyancy also known as upthrust) is an upward force exerted by a fluid that opposes the weight of an immersed object. In a column of fluid, pressure increases with depth as a result of the weight of the overlying fluid. Thus the pressure at the bottom of a column of fluid is greater than at the top of the column. Similarly, the pressure at the bottom of an object submerged in a fluid is greater than at the top of the object. This pressure difference results in a net upwards force on the object. The magnitude of that force exerted is proportional to that pressure difference, and (as explained by Archimedes' principle) is equivalent to the weight of the fluid that would otherwise occupy the volume of the object, i.e. the displaced fluid.

For this reason, an object whose density is greater than that of the fluid in which it is submerged tends to sink. If the object is either less dense than the liquid or is shaped appropriately (as in a boat), the force can keep the object afloat. This can occur only in a reference frame which either has a gravitational field or is accelerating due to a force other than gravity defining a "downward" direction (that is, a non-inertial reference frame). In a situation of fluid statics, the net upward buoyancy force is equal to the magnitude of the weight of fluid displaced by the body

https://en.wikipedia.org/wiki/Natural_convection

Quote
Natural convection is a mechanism, or type of heat transport, in which the fluid motion is not generated by any external source (like a pump, fan, suction device, etc.) but only by density differences in the fluid occurring due to temperature gradients. In natural convection, fluid surrounding a heat source receives heat, becomes less dense and rises. The surrounding, cooler fluid then moves to replace it. This cooler fluid is then heated and the process continues, forming a convection current; this process transfers heat energy from the bottom of the convection cell to top. The driving force for natural convection is buoyancy, a result of differences in fluid density.

////////////

The "body" experiencing buoyancy does not need to have metal walls or be a balloon, to experience buoyancy.  It can be a hot gas molecule, or group of molecules.  It can be a differential volume element in Continuum Mechanics.  That is why the driving force for natural convection is buoyancy, a result of differences in fluid density.
« Last Edit: 12/17/2015 08:17 pm by Rodal »

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2205
  • Liked: 2713
  • Likes Given: 1134
Whatever we work out, I'm at a point where DIY experimentation may not be worthwhile unless we can figure a way to characterize thermal dynamics.

From my perspective, I knew there was lift, what I was looking for was an interruption of it during mag on conditions. I believe I saw this clearly as either attenuation, reversal or a hold of the lift progression.

Remember, the thermal mass of the magnetron assembly does not permit instantaneous heating or cooling. If you look at the thermal videos, you see it is quite gradual...much slower than the instantaneous changes to the thermal lift profile I observed when mag switched between on and off, so my posit is that quick changes to a lift profile CANNOT be due to instantaneous heating or cooling at mag transition.

Where am I off base here?

Offline aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1146
  • Likes Given: 360
Quote
But even a piece of burning paper will experience buoyancy due to the natural convection lift forces on it.

No, that is not buoyancy. It is a matter of terminology. It is a lift force due to the natural convection currents created by the heat of burning paper. As I wrote above, buoyancy and convection are separable problems.

The magnetron itself will experience natural convection lift forces, but not buoyancy, because it is a sealed unit. It does not out-gas hence it's mass is constant. Perhaps there is a very small buoyancy due to the expansion of the metal of the magnetron with temperature but the expansion is small and the mass of the air displaced by that expansion is very, very small, I think negligible, when compared to the natural convection lift forces it experiences.

In my post, I was considering SeaShells' test rig and cavity. She has placed the magnetron at the pivot so that lift forces off the hot magnetron itself will be negated, be they buoyancy or natural convection lift forces.

I suggest we focus uniquely on a test set-up for our posts because the error sources are different for different test set-ups. Heat from the magnetron was definitely a factor with rfmwguy's test rig, but should not be a factor with Shells' rig. Similarly, buoyancy of the cavity will be a definite factor with Shells' rig but should not have been as much of a factor, if any, with rfmwguy's mesh cavity design.

That may not be buoyancy according to you, but it is according to what I learnt, and also according to:

https://en.wikipedia.org/wiki/Buoyancy

From your reference, the applicable definition is "Buoyancy = weight of displaced fluid." In Shells' cavity, air is displaced by the added heat causing lower density as the pressure is equalized by displaced air flow out the relief tube.
Quote

Quote
In science, buoyancy also known as upthrust) is an upward force exerted by a fluid that opposes the weight of an immersed object. In a column of fluid, pressure increases with depth as a result of the weight of the overlying fluid. Thus the pressure at the bottom of a column of fluid is greater than at the top of the column. Similarly, the pressure at the bottom of an object submerged in a fluid is greater than at the top of the object. This pressure difference results in a net upwards force on the object. The magnitude of that force exerted is proportional to that pressure difference, and (as explained by Archimedes' principle) is equivalent to the weight of the fluid that would otherwise occupy the volume of the object, i.e. the displaced fluid.

For this reason, an object whose density is greater than that of the fluid in which it is submerged tends to sink. If the object is either less dense than the liquid or is shaped appropriately (as in a boat), the force can keep the object afloat. This can occur only in a reference frame which either has a gravitational field or is accelerating due to a force other than gravity defining a "downward" direction (that is, a non-inertial reference frame). In a situation of fluid statics, the net upward buoyancy force is equal to the magnitude of the weight of fluid displaced by the body

https://en.wikipedia.org/wiki/Natural_convection

Quote
Natural convection is a mechanism, or type of heat transport, in which the fluid motion is not generated by any external source (like a pump, fan, suction device, etc.) but only by density differences in the fluid occurring due to temperature gradients. In natural convection, fluid surrounding a heat source receives heat, becomes less dense and rises. The surrounding, cooler fluid then moves to replace it. This cooler fluid is then heated and the process continues, forming a convection current; this process transfers heat energy from the bottom of the convection cell to top. The driving force for natural convection is buoyancy, a result of differences in fluid density.

The driving force is buoyancy of the warm air, due to its lower density. That is the cause of natural convection. From your reference quoted above.  "In natural convection, fluid surrounding a heat source receives heat, becomes less dense and rises."

As I wrote above, it is a misunderstanding of the terminology. Attempting to model buoyancy forces using the equations for natural convection, or alternatively, attempting to model natural convection using equations for buoyancy forces is a lost cause. Neither is applicable to the other. They are separable issues.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Whatever we work out, I'm at a point where DIY experimentation may not be worthwhile unless we can figure a way to characterize thermal dynamics.

From my perspective, I knew there was lift, what I was looking for was an interruption of it during mag on conditions. I believe I saw this clearly as either attenuation, reversal or a hold of the lift progression.

Remember, the thermal mass of the magnetron assembly does not permit instantaneous heating or cooling. If you look at the thermal videos, you see it is quite gradual...much slower than the instantaneous changes to the thermal lift profile I observed when mag switched between on and off, so my posit is that quick changes to a lift profile CANNOT be due to instantaneous heating or cooling at mag transition.

Where am I off base here?

Your experiments do NOT show that effect consistently with every ON and OFF turning of the magnetron: the thermal lift profile observed when magnetron was switched on and off was not the same through time  You had to appeal to statistics to arrive at a conclusion.  The statistics are based on a small sample population hence questionable from a statistical viewpoint.

Physically, the effect you observed may be due to transient thermal natural convection.  There was no analysis of the transient thermal natural convection effect in your experiment (due to the hot magnetron sitting on top of the upper plate of the EM Drive). Transient thermal natural convection effects will result in statistical-looking effects like the one you measured.  If the sample population was representative of the true statistical population (which was NOT shown), the statistical test may be only showing the effect of transient thermal natural convection effects.
« Last Edit: 12/17/2015 08:54 pm by Rodal »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Whatever we work out, I'm at a point where DIY experimentation may not be worthwhile unless we can figure a way to characterize thermal dynamics.

From my perspective, I knew there was lift, what I was looking for was an interruption of it during mag on conditions. I believe I saw this clearly as either attenuation, reversal or a hold of the lift progression.

Remember, the thermal mass of the magnetron assembly does not permit instantaneous heating or cooling. If you look at the thermal videos, you see it is quite gradual...much slower than the instantaneous changes to the thermal lift profile I observed when mag switched between on and off, so my posit is that quick changes to a lift profile CANNOT be due to instantaneous heating or cooling at mag transition.

Where am I off base here?

To be "on base" you would have to follow with these tests:

http://forum.nasaspaceflight.com/index.php?topic=39004.msg1459176#msg1459176

http://forum.nasaspaceflight.com/index.php?topic=39004.msg1459173#msg1459173

to show that there is an anomalous force in your tests that cannot be explained by transient thermal convection effects.
« Last Edit: 12/17/2015 08:48 pm by Rodal »

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0