I would like my simple question to be answered too:
What is causing the 1/2 wave to lengthen on one side and shorten on the other side, but a local variation of the wavelength?
That I can help with. The guide wavelength in a waveguide is basically due to the waves bouncing back and forth between the walls rather than straight down the middle. The superposition hides the non-axial portion, so you just see an apparently stretched axial wavelength, while the actual travelling waves maintain their original wavelength. In a smaller waveguide the angle of the bounces is steeper (for the same frequency of radiation), so the apparent wavelength gets longer. What Cullen originally showed was simply that the steeper the bounces, the less of the wave's momentum is in the axial direction, which is a very sensible result.
Something similar happens in an emDrive cavity as waves reflecting off the side walls cause the angles to get steeper towards the small end. You can't easily define a guide wavelength though because at any point the suprimposed waves are not all moving in the same direction, and it is not clear what modes (if any) the apparent standing wave has any meaningful relation to the angles the travelling waves are travelling at. The issue with this description and how Shawyer abuses Cullen's results is that by conservation of momentum, it is obvious that any change in axial momentum due to reflections off the sidewalls transfers that momentum to the sidewalls to end up with no net motion.
(Note:This post used simplified descriptions, to give an intuitive sense of what is going on. Boundary conditions and the wave equation make this more complicated than presented here.)
Isn't it the opposite (emphasis mine)? A "steeper" angle means a more
open angle, right? (English is not my primary language). Shawyer shows the sawtooth pattern gradually becomes sharper as the waveguide diameter becomes smaller, hence an apparent decreasing wavelength axially in the direction big to small in a frustum.
Even if the EM frequency stays the same, the wavelength undergoes an
apparent shrinking (in the axial direction) so the group velocity (which carries energy and information) decreases axially in an EM wave travelling big to small, as if the wave was undergoing an
apparent blueshift in direction of the small end plate.
[EDIT: wrong assumptions struck out, thanks to meberbs for his answer below. The bounce angles and the actual wave form are not the same, in fact quite the opposite, so the sharper the bounce angles, the longer the wave appears to be toward small end in the axial direction, and the shorter the wave appears toward big end.]
That won't work. All you're doing is moving the centre of mass about.
Many would say the same about emdrive...
edit: Moving the centre of mass would actually mean that the thing works.
Btw just to clear things up, I'm not going to be the one building it. I've masters in biochemical engineering(far from expertise required). I will be providing financial support as well as building equipment as part of signed agreement between my company and university.
A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. I may provide pictures of the build in 2017 as it progresses however testing data will be subject to NDA until paper is published/refused or experiment failed and no thrust was detected.
So there are universities involved in this research. Good. I thought Martin Tajmar at Dresden was the only (official) university researcher in this field.
You should also publish if you do not detect thrust (at least at ArXiv).
How do you plan to measure the force? There is a lot of attention for the microwave part at this forum. I have the impression the other necessary part of the game, the force/thrust measurement, is a little bit neglected.
My self, I'm a business man with deep passion for science and a dream to fly to space one day. However I do have lots of friends in university as they helped me immensely to start my business a decade ago.
This week I've signed a contract with university for a privately funded research project by my initiative. They like the idea of this kind of research but they wouldn't put their money into it.
Data will be published whether it's successful or not and the measurement methods will have to be worked out by people doing the science. I have masters in engineering and I know a thing or two but we will be discussing details this weekend/month. Official starting date is set in February when university returns to it's normal operation after holiday season, till than we have to finalize the build and start getting necessary equipment.
Dear ARW,
I noticed, your ideas about setting up the first measurements differ a lot from how I would do it/am going to do it. The people at the university who are going to work for you don't seem to have experience in this or a related field. I strongly recommend you to attract an advisory
committee group, maybe some the more experienced people active at this forum am willing to participate in it (not me, I'm not experienced enough, yet).
Best, Peter
That won't work. All you're doing is moving the centre of mass about.
Many would say the same about emdrive...
Indeed, but in the case of the EM drive the maths is a lot more complex making casual observations much harder.
I would like my simple question to be answered too:
What is causing the 1/2 wave to lengthen on one side and shorten on the other side, but a local variation of the wavelength?
That I can help with. The guide wavelength in a waveguide is basically due to the waves bouncing back and forth between the walls rather than straight down the middle. The superposition hides the non-axial portion, so you just see an apparently stretched axial wavelength, while the actual travelling waves maintain their original wavelength. In a smaller waveguide the angle of the bounces is steeper (for the same frequency of radiation), so the apparent wavelength gets longer. What Cullen originally showed was simply that the steeper the bounces, the less of the wave's momentum is in the axial direction, which is a very sensible result.
Something similar happens in an emDrive cavity as waves reflecting off the side walls cause the angles to get steeper towards the small end. You can't easily define a guide wavelength though because at any point the suprimposed waves are not all moving in the same direction, and it is not clear what modes (if any) the apparent standing wave has any meaningful relation to the angles the travelling waves are travelling at. The issue with this description and how Shawyer abuses Cullen's results is that by conservation of momentum, it is obvious that any change in axial momentum due to reflections off the sidewalls transfers that momentum to the sidewalls to end up with no net motion.
(Note:This post used simplified descriptions, to give an intuitive sense of what is going on. Boundary conditions and the wave equation make this more complicated than presented here.)
Isn't it the opposite (emphasis mine)? A "steeper" angle means a more open angle, right? (English is not my primary language). Shawyer shows the sawtooth pattern gradually becomes sharper as the waveguide diameter becomes smaller, hence an apparent decreasing wavelength axially in the direction big to small in a frustum.
Even if the EM frequency stays the same, the wavelength undergoes an apparent shrinking (in the axial direction) so the group velocity (which carries energy and information) decreases axially in an EM wave travelling big to small, as if the wave was undergoing an apparent blueshift in direction of the small end plate.
No it's the opposite. i.e. it depends on the component of ψ you are looking at. I think you are looking at the wavelength/better shell of constant phase right? Then the pic below shows whats happens.
To quote TT (and I do this really not often
) the "guide wavelength" will be longer, the smaller the diameter is.
The wave is more stretched the smaller the diameter is.
The point is this description is true regarding/whatever the terms he is using
ωnmp=c*√(j(’)mn/a)²+(pπ/l)²decrease "a" you will see it using the equation for a cylindrical cavity. (frequency will be as higher as smaller "a" is)
"
I recall this issue being raised before, but it was a while ago and the situation may have changed. Is it reasonably possible to 3D print an emdrive, or a major portion of it? I mean direct fab and not making a wax model, a mold, and then casting it.
I am not asking if the typical $800 hobbyist unit can do this, though it would be interesting if such a device could. I am talking about the big expensive machines from places like Stratasys...can they do this? If so, does it economically make sense?
The reason I ask is because I have tax reasons to quickly spend money (six figures) on industrial equipment (before the end of the year). If buying a high-end 3d printer and using it to make some money by producing emdrive prototypes for people has a chance of being economically feasible, I'd like to know and get started ASAP.
I think the casting places typically have their own 3d printer. All you do is provide the CAD file for the part. They will update it to work with their type of casting, and the type of finish required, and give you a chance to approve it before the part is cast. It's not that expensive, nowhere near six figures! 
But for a high Q you need a post-printing/casting step for polishing. 3D printing can give you a near-net-shape, but the surface finish is nowhere near good enough. A 3D metal printer big enough to build an aluminum EMDrive in the 2.4 GHz range is upwards of $1M. We just priced them at work 
Or you could do old fashioned CNC/casting + polishing for a fraction of your suggested cost.
This is all quite discouraging. I want to invest in equipment which will make it easier for designers to take the particular geometry they want to try, and get a quality finished cavity for not a lot of money.
Cost of 3D printer: As the build volume increases, so does the price. So the tiny, benchtop devices just don't have a good enough build area. I am looking at a Stratasys Fortus 400 with the optional larger 16" x 14" x 16" build size. This seems sufficient for at least some emdrive designs. What portion of designs will it be insufficient for?
Metal: I thought the above Fortus had an optional DMLS (metal) hot end but I may have misread.
In either case, it looks like I'd be getting into plating which involves dangerous chemicals, which I am not qualified to handle.
So, perhaps back to square one, and to ask the broader question: In what way can I quickly invest some capital in equipment and provide a useful service to emdrive developers?
Perhaps a used vacuum chamber large enough to handle the design needs of a majority of the DIY builders. Maybe Paul March could provide some feedback on size and vacuum requirements, since he worked with the chamber at EW.
I am sure that a chamber and the vacuum equipment would meet or even exceed your $$ needs, time deadlines for investment may be a more significant problem...
Yes, vacuum chamber and torsion balance. along with miscellaneous other equipment and supplies. Thermal imager. voltage calibrator, VNA, etc. Any help with the list of equipment, and its required specifications, would be helpful. I'd have to build the torsion balance, yes, so specs and plans could be useful.
I was also thinking of getting a COMSOL multiphysics license and then rent out use of the computer upon which it was installed (networked license, or not. With the non-network license, just transfer the data by USB stick from the non-networked computer to the networked one which clients would transfer data to/from).
The industrial space of have made inquiries about are variously located in Great Barrington, Pittsfield, and North Adams Massachusetts. These are approximately an hour's drive from the Albany airport.
I would like my simple question to be answered too:
What is causing the 1/2 wave to lengthen on one side and shorten on the other side, but a local variation of the wavelength?
That I can help with. The guide wavelength in a waveguide is basically due to the waves bouncing back and forth between the walls rather than straight down the middle. The superposition hides the non-axial portion, so you just see an apparently stretched axial wavelength, while the actual travelling waves maintain their original wavelength. In a smaller waveguide the angle of the bounces is steeper (for the same frequency of radiation), so the apparent wavelength gets longer. What Cullen originally showed was simply that the steeper the bounces, the less of the wave's momentum is in the axial direction, which is a very sensible result.
Something similar happens in an emDrive cavity as waves reflecting off the side walls cause the angles to get steeper towards the small end. You can't easily define a guide wavelength though because at any point the suprimposed waves are not all moving in the same direction, and it is not clear what modes (if any) the apparent standing wave has any meaningful relation to the angles the travelling waves are travelling at. The issue with this description and how Shawyer abuses Cullen's results is that by conservation of momentum, it is obvious that any change in axial momentum due to reflections off the sidewalls transfers that momentum to the sidewalls to end up with no net motion.
(Note:This post used simplified descriptions, to give an intuitive sense of what is going on. Boundary conditions and the wave equation make this more complicated than presented here.)
Isn't it the opposite (emphasis mine)? A "steeper" angle means a more open angle, right? (English is not my primary language). Shawyer shows the sawtooth pattern gradually becomes sharper as the waveguide diameter becomes smaller, hence an apparent decreasing wavelength axially in the direction big to small in a frustum.
Even if the EM frequency stays the same, the wavelength undergoes an apparent shrinking (in the axial direction) so the group velocity (which carries energy and information) decreases axially in an EM wave travelling big to small, as if the wave was undergoing an apparent blueshift in direction of the small end plate.
No its the opposite. i.e. it depends on the component of ψ you are looking at. I think you are looking at the wavelength/better shell of constant phase right? Then the pic below shows whats happens.
No, flux_capacitor's drawing was correct, yours clearly does not have equal incident and reflected angles.
Flux capacitor is wrong about the effect on wavelength his drawing implies though. (This is confusing, so these mistakes are understandable.)
While the bounces are getting more frequent, the wavelength actually gets longer, because projected distance between 2 constant phase surfaces gets larger. see attached for a picture. (Note that in reality there are constraints of the ratio of the wavelength and the angle of the bounces (depending on mode) I ignored these to make it easier to draw quickly)
The green lines are the direction of travel, the red lines are the wavefronts. If I drew it right, the distance between the red lines along the green lines should be the same for both (this is the free space wavelength). Due to superposition in a waveguide, what you see as the guide wavelength is the distance between where the red lines intersect the black lines (marked in grey), which gets larger as the direction of travel gets steeper.
I would like my simple question to be answered too:
What is causing the 1/2 wave to lengthen on one side and shorten on the other side, but a local variation of the wavelength?
That I can help with. The guide wavelength in a waveguide is basically due to the waves bouncing back and forth between the walls rather than straight down the middle. The superposition hides the non-axial portion, so you just see an apparently stretched axial wavelength, while the actual travelling waves maintain their original wavelength. In a smaller waveguide the angle of the bounces is steeper (for the same frequency of radiation), so the apparent wavelength gets longer. What Cullen originally showed was simply that the steeper the bounces, the less of the wave's momentum is in the axial direction, which is a very sensible result.
Something similar happens in an emDrive cavity as waves reflecting off the side walls cause the angles to get steeper towards the small end. You can't easily define a guide wavelength though because at any point the suprimposed waves are not all moving in the same direction, and it is not clear what modes (if any) the apparent standing wave has any meaningful relation to the angles the travelling waves are travelling at. The issue with this description and how Shawyer abuses Cullen's results is that by conservation of momentum, it is obvious that any change in axial momentum due to reflections off the sidewalls transfers that momentum to the sidewalls to end up with no net motion.
(Note:This post used simplified descriptions, to give an intuitive sense of what is going on. Boundary conditions and the wave equation make this more complicated than presented here.)
Isn't it the opposite (emphasis mine)? A "steeper" angle means a more open angle, right? (English is not my primary language). Shawyer shows the sawtooth pattern gradually becomes sharper as the waveguide diameter becomes smaller, hence an apparent decreasing wavelength axially in the direction big to small in a frustum.
Even if the EM frequency stays the same, the wavelength undergoes an apparent shrinking (in the axial direction) so the group velocity (which carries energy and information) decreases axially in an EM wave travelling big to small, as if the wave was undergoing an apparent blueshift in direction of the small end plate.
No its the opposite. i.e. it depends on the component of ψ you are looking at. I think you are looking at the wavelength/better shell of constant phase right? Then the pic below shows whats happens.
No, flux_capacitor's drawing was correct, yours clearly does not have equal incident and reflected angles.
Flux capacitor is wrong about the effect on wavelength his drawing implies though. (This is confusing, so these mistakes are understandable.)
While the bounces are getting more frequent, the wavelength actually gets longer, because projected distance between 2 constant phase surfaces gets larger. see attached for a picture. (Note that in reality there are constraints of the ratio of the wavelength and the angle of the bounces (depending on mode) I ignored these to make it easier to draw quickly)
The green lines are the direction of travel, the red lines are the wavefronts. If I drew it right, the distance between the red lines along the green lines should be the same for both (this is the free space wavelength). Due to superposition in a waveguide, what you see as the guide wavelength is the distance between where the red lines intersect the black lines (marked in grey), which gets larger as the direction of travel gets steeper.
Please try to understand that the ray trajectory theory is not applicable 1 by 1 since the wavelength is of the same order as the structure of the cavity it self. We are talking about near field conditions.
The freespace wavelength is always smaller compared to the one in a waveguide with ε=1, μ=1 and having conductive walls, if the diameter is in the order of the wavelength.
Yes, vacuum chamber and torsion balance. along with miscellaneous other equipment and supplies. Thermal imager. voltage calibrator, VNA, etc. Any help with the list of equipment, and its required specifications, would be helpful. I'd have to build the torsion balance, yes, so specs and plans could be useful.
I was also thinking of getting a COMSOL multiphysics license and then rent out use of the computer upon which it was installed (networked license, or not. With the non-network license, just transfer the data by USB stick from the non-networked computer to the networked one which clients would transfer data to/from).
The industrial space of have made inquiries about are variously located in Great Barrington, Pittsfield, and North Adams Massachusetts. These are approximately an hour's drive from the Albany airport.
The vacuum chamber should really be large enough to hold a rotational table/test rig. Tiny vacuum chambers are the absolute bane of EM Drive research. Rotational table/test rig is the recommended configuration for testing ion drives and the like. The Cannae folks have this set up.
Another consideration is vacuum hardened electronics, particularly the RF drive section. I recall Paul detailing the serious problems that EW had in this area.
Many folks have been getting good traction with Fekko. Perhaps that should be given consideration over COMSOL?
It might be oversimplified, but here is a rough draft on how modifying the mass of light in a cavity with regards to its wavelength might give it some push.
https://www.researchgate.net/profile/Dustin_Macdermott "Is the frustum EM Drive4 decelerating light for propellantless propulsion?" It may still be flawed. Will revise as I go.
That won't work. All you're doing is moving the centre of mass about.
Many would say the same about emdrive...
edit: Moving the centre of mass would actually mean that the thing works.
Damn, yes. I edited my first post, should have just stuck to what I said initially. Anyway the thing won't work, it will vibrate a bit but not move anywhere. Back to lurking and waiting for more data.
I would like my simple question to be answered too:
What is causing the 1/2 wave to lengthen on one side and shorten on the other side, but a local variation of the wavelength?
That I can help with. The guide wavelength in a waveguide is basically due to the waves bouncing back and forth between the walls rather than straight down the middle. The superposition hides the non-axial portion, so you just see an apparently stretched axial wavelength, while the actual travelling waves maintain their original wavelength. In a smaller waveguide the angle of the bounces is steeper (for the same frequency of radiation), so the apparent wavelength gets longer. What Cullen originally showed was simply that the steeper the bounces, the less of the wave's momentum is in the axial direction, which is a very sensible result.
Something similar happens in an emDrive cavity as waves reflecting off the side walls cause the angles to get steeper towards the small end. You can't easily define a guide wavelength though because at any point the suprimposed waves are not all moving in the same direction, and it is not clear what modes (if any) the apparent standing wave has any meaningful relation to the angles the travelling waves are travelling at. The issue with this description and how Shawyer abuses Cullen's results is that by conservation of momentum, it is obvious that any change in axial momentum due to reflections off the sidewalls transfers that momentum to the sidewalls to end up with no net motion.
(Note:This post used simplified descriptions, to give an intuitive sense of what is going on. Boundary conditions and the wave equation make this more complicated than presented here.)
Isn't it the opposite (emphasis mine)? A "steeper" angle means a more open angle, right? (English is not my primary language). Shawyer shows the sawtooth pattern gradually becomes sharper as the waveguide diameter becomes smaller, hence an apparent decreasing wavelength axially in the direction big to small in a frustum.
Even if the EM frequency stays the same, the wavelength undergoes an apparent shrinking (in the axial direction) so the group velocity (which carries energy and information) decreases axially in an EM wave travelling big to small, as if the wave was undergoing an apparent blueshift in direction of the small end plate.
No its the opposite. i.e. it depends on the component of ψ you are looking at. I think you are looking at the wavelength/better shell of constant phase right? Then the pic below shows whats happens.
No, flux_capacitor's drawing was correct, yours clearly does not have equal incident and reflected angles.
Flux capacitor is wrong about the effect on wavelength his drawing implies though. (This is confusing, so these mistakes are understandable.)
While the bounces are getting more frequent, the wavelength actually gets longer, because projected distance between 2 constant phase surfaces gets larger. see attached for a picture. (Note that in reality there are constraints of the ratio of the wavelength and the angle of the bounces (depending on mode) I ignored these to make it easier to draw quickly)
The green lines are the direction of travel, the red lines are the wavefronts. If I drew it right, the distance between the red lines along the green lines should be the same for both (this is the free space wavelength). Due to superposition in a waveguide, what you see as the guide wavelength is the distance between where the red lines intersect the black lines (marked in grey), which gets larger as the direction of travel gets steeper.
Please try to understand that the ray trajectory theory is not applicable 1 by 1 since the wavelength is of the same order as the structure of the cavity it self. We are talking about near field conditions.
The freespace wavelength is always smaller compared to the one in a waveguide with ε=1, μ=1 and having conductive walls, if the diameter is in the order of the wavelength.
Agreed, that is why I made a comment about ignoring the constraints of the specific angles it must propagate at, which is due to the near field nature of the propagation. This does correctly give the results for the relations between guide wavelength and propagation velocity. I want to say that you can produce the full near field result by summing superimposed plane waves, but it has been a while since I went through the full details, and I could be mis-remembering.
Yes, vacuum chamber and torsion balance. along with miscellaneous other equipment and supplies. Thermal imager. voltage calibrator, VNA, etc. Any help with the list of equipment, and its required specifications, would be helpful. I'd have to build the torsion balance, yes, so specs and plans could be useful.
I was also thinking of getting a COMSOL multiphysics license and then rent out use of the computer upon which it was installed (networked license, or not. With the non-network license, just transfer the data by USB stick from the non-networked computer to the networked one which clients would transfer data to/from).
The industrial space of have made inquiries about are variously located in Great Barrington, Pittsfield, and North Adams Massachusetts. These are approximately an hour's drive from the Albany airport.
The vacuum chamber should really be large enough to hold a rotational table/test rig. Tiny vacuum chambers are the absolute bane of EM Drive research. Rotational table/test rig is the recommended configuration for testing ion drives and the like. The Cannae folks have this set up.
Another consideration is vacuum hardened electronics, particularly the RF drive section. I recall Paul detailing the serious problems that EW had in this area.
Many folks have been getting good traction with Fekko. Perhaps that should be given consideration over COMSOL?
FEKO is setup to do exactly that.
"FEKO is licensed under Altair’s HyperWorks Unit (HWU) Licensing system which is a token style licensing systems that draws a certain number of unit for the different software options available.
The minimum number of units required to run FEKO is 25 HWU and allows processing on up to 4 cores."
As I understand it, once you have the HWU, they can be shared amongst individuals in a group, but only one seat at a time can be active. The cost was $15k USD to me as a hobbiest. No way I can do it alone but if 10 or 20 people wanted to share the units, that would be almost affordable. Note: Training is EXTRA!
Btw just to clear things up, I'm not going to be the one building it. I've masters in biochemical engineering(far from expertise required). I will be providing financial support as well as building equipment as part of signed agreement between my company and university.
A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. I may provide pictures of the build in 2017 as it progresses however testing data will be subject to NDA until paper is published/refused or experiment failed and no thrust was detected.
So there are universities involved in this research. Good. I thought Martin Tajmar at Dresden was the only (official) university researcher in this field.
You should also publish if you do not detect thrust (at least at ArXiv).
How do you plan to measure the force? There is a lot of attention for the microwave part at this forum. I have the impression the other necessary part of the game, the force/thrust measurement, is a little bit neglected.
My self, I'm a business man with deep passion for science and a dream to fly to space one day. However I do have lots of friends in university as they helped me immensely to start my business a decade ago.
This week I've signed a contract with university for a privately funded research project by my initiative. They like the idea of this kind of research but they wouldn't put their money into it.
Data will be published whether it's successful or not and the measurement methods will have to be worked out by people doing the science. I have masters in engineering and I know a thing or two but we will be discussing details this weekend/month. Official starting date is set in February when university returns to it's normal operation after holiday season, till than we have to finalize the build and start getting necessary equipment.
Dear ARW,
I noticed, your ideas about setting up the first measurements differ a lot from how I would do it/am going to do it. The people at the university who are going to work for you don't seem to have experience in this or a related field. I strongly recommend you to attract an advisory committee, maybe some the more experienced people active at this forum am willing to participate in it (not me, I'm not experienced enough, yet).
Best, Peter
... A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. ...Even with the understanding that this is to be a project based at or associated witha University Physics Department, perhaps one of our builders with experience could repost the cautions and dangers involved.
Most of what has been posted in these discussions has been working with DC voltages of 2 to 3 kv and raw microwave energies up to perhaps 1000watts. Both can be deadly if not handled carefully by experienced and qualified individuals. Once you start talking about 10kw of MW energies the danger increases more than 10 fold.
Again maybe one of the DIY builders in these discussions with the experience I do not have can properly emphasize the need for caution and careful experienced handling of the DC and high frequency currents involved, even for a project being formulated at a university lab.
It might be oversimplified, but here is a rough draft on how modifying the mass of light in a cavity with regards to its wavelength might give it some push. https://www.researchgate.net/profile/Dustin_Macdermott "Is the frustum EM Drive4 decelerating light for propellantless propulsion?" It may still be flawed. Will revise as I go.
Equations (5) and (6) seem to be contradictory. On the left they are equal, on the right they are not. I see what you're doing, but it's poorly expressed.
Equations (19) and (20) where you derive K. I tried this. Did you notice that your K≤1, where K in the PV Model is always K≥1? In the PV Model, K≤1 is FTL, whereas in a waveguide v≤c. The logic doesn't fit the facts, at least not in a straight forward 1:1 comparison.
Btw just to clear things up, I'm not going to be the one building it. I've masters in biochemical engineering(far from expertise required). I will be providing financial support as well as building equipment as part of signed agreement between my company and university.
A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. I may provide pictures of the build in 2017 as it progresses however testing data will be subject to NDA until paper is published/refused or experiment failed and no thrust was detected.
So there are universities involved in this research. Good. I thought Martin Tajmar at Dresden was the only (official) university researcher in this field.
You should also publish if you do not detect thrust (at least at ArXiv).
How do you plan to measure the force? There is a lot of attention for the microwave part at this forum. I have the impression the other necessary part of the game, the force/thrust measurement, is a little bit neglected.
My self, I'm a business man with deep passion for science and a dream to fly to space one day. However I do have lots of friends in university as they helped me immensely to start my business a decade ago.
This week I've signed a contract with university for a privately funded research project by my initiative. They like the idea of this kind of research but they wouldn't put their money into it.
Data will be published whether it's successful or not and the measurement methods will have to be worked out by people doing the science. I have masters in engineering and I know a thing or two but we will be discussing details this weekend/month. Official starting date is set in February when university returns to it's normal operation after holiday season, till than we have to finalize the build and start getting necessary equipment.
Dear ARW,
I noticed, your ideas about setting up the first measurements differ a lot from how I would do it/am going to do it. The people at the university who are going to work for you don't seem to have experience in this or a related field. I strongly recommend you to attract an advisory committee, maybe some the more experienced people active at this forum am willing to participate in it (not me, I'm not experienced enough, yet).
Best, Peter
... A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. ...
Even with the understanding that this is to be a project based at or associated witha University Physics Department, perhaps one of our builders with experience could repost the cautions and dangers involved.
Most of what has been posted in these discussions has been working with DC voltages of 2 to 3 kv and raw microwave energies up to perhaps 1000watts. Both can be deadly if not handled carefully by experienced and qualified individuals. Once you start talking about 10kw of MW energies the danger increases more than 10 fold.
Again maybe one of the DIY builders in these discussions with the experience I do not have can properly emphasize the need for caution and careful experienced handling of the DC and high frequency currents involved, even for a project being formulated at a university lab.
Can't emphasize the safety aspects enough. If you go the magnetron route there needs to be someone with experience in high voltage electronics and high energy microwaves or someone could be killed.
Folks are converging on the idea that this is a highly narrowband phenomena and high energies may work against you unless you can keep the signal very narrow within the fustrum.
A recent post by TheTraveller sums up for me a current state of the art approach:
http://forum.nasaspaceflight.com/index.php?topic=41732.msg1615509#msg1615509Note that TT and others are planning to use 100W solid state rf sources that are highly tuneable (and reasonably inexpensive).
The narrowband aspect of this phenomena may mean that there is a lot of benefit to actively tuning the frequency as it is expected to change as the frustum accelerates.
A very interesting question is whether performance could be improved by anticipating where the frequency should be for a desired impulse and tuning accordingly. This would be a bit like feed forward signal conditioning.
I think the volume losses could only model the solid dielectric, but could not model the losses of the skin effect in the copper, due to the planar nature of these losses. The surface losses look more accurate, but I think the total losses would be the combination of the two.
If we wanted to optimize cavity for thrust, we would want the highest surface losses at one end and the highest energy stored without losses, at the other end. The magnetic flux stored in a cavity exerts pressure on the walls of the cavity. a voltage drop, resulting in losses in the walls of the cavity, is like poking a hole in it and letting the pressure out.
So,.... What? An inside shiny metal surface for one half of the cavity, and a "thin" dielectric coating for the other half? Would that work? A "thin" layered dielectric still reflecting the MW waves like interference based mirrors do... That wouldn't be thin
That would be kind of like a reflection not based on induced E and B from free electron rich metal surface but on return capacitance of the surface... Am I making any sense?
Not really. Where is the voltage drop, in the dielectric? Personally, I think it will do better without a dielectric. It might do better with nickel at one end, copper at the other end, but it will perform best with the largest amount of stored energy, exerting the maximum amount of pressure on the cavity.
? A voltage drop is how you lose energy; voltage drop - eddy current - etc. This is what you want at one end (metallic) of the cavity. We do have a voltage drop across the dielectric coatingt; metal surface behind on one side and MW on the other side, causing minimal charge movement ... essentially a variable polarization of the dielectric,
i.e. much much less loss of energy.... Better than nickel ?
You cannot use a dielectric insert without a trade-off. In this case, an insert will lower the Q of the cavity, effectively widening the 3dB BW of the return loss response. More testing will be required to determine if that tradeoff is worth it, or...someone will have to develop a bullet-proof theory that "inserts provide X so thrust can become Y".
Do you know of experiments done with a cylindrical cavity and dielectric inserts? I plan do these with shaped dielectrics (cones, to start with).
It might be oversimplified, but here is a rough draft on how modifying the mass of light in a cavity with regards to its wavelength might give it some push. https://www.researchgate.net/profile/Dustin_Macdermott "Is the frustum EM Drive4 decelerating light for propellantless propulsion?" It may still be flawed. Will revise as I go.
Equations (5) and (6) seem to be contradictory. On the left they are equal, on the right they are not. I see what you're doing, but it's poorly expressed.
Equations (19) and (20) where you derive K. I tried this. Did you notice that your K≤1, where K in the PV Model is always K≥1? In the PV Model, K≤1 is FTL, whereas in a waveguide v≤c. The logic doesn't fit the facts, at least not in a straight forward 1:1 comparison.
I'll double check equations 5 and 6. Thanks. All equations 5 and 6 are stating is that both change in frames are equal after the impact. I'll make sure it is correct, but it should be. The 2nd solution for dv1 and dv2 is that the two colliding objects pass right through each other, so only the first solution makes sense. Regarding equations 19 and 20, K is supposed to be less than 1 at the narrow end. Supposedly the speed of light at that end becomes greater. This seems to go along with TT stating that they have measured a weaker impulse from photons at the narrow end but I don't know where he is getting this from. It also goes along with the wavelength growing larger and in the PV model the mass of the photon decreasing. Also with K dropping the energy should increase and generally where the wavelength appears to increase, the energy density of the radiation increases. That is consider
http://gregegan.customer.netspace.net.au/SCIENCE/Cavity/Cavity.html where he shows under, "Energy, pressure and forces" he shows the energy density and above that title, is how it effects the EM modes wavelength.
You will also notice at this url I gave in the paper, "
https://www.microwaves101.com/encyclopedias/waveguide-mathematics" they state,
"The guide wavelength in waveguide is longer than wavelength in free space. This isn't intuitive, it seems like the dielectric constant in waveguide must be less than unity for this to happen... don't think about this too hard you will get a headache".
I always thought the phase velocity<c not the group velocity so I'll have to look a bit more into this. I could almost swear there were ways to make the group velocity faster than c. But here is another quote from:
http://www.mathpages.com/home/kmath210/kmath210.htm Hence, not only is the phase velocity generally greater than c, it approaches infinity as ω approaches the cutoff frequency ω0.
This idea I am working with predicts, with the wavelength a force toward the big end unless a dielectric is introduced which changes the wavelength distribution. Similar to what TT has been saying.
I appreciate your comments and I'll use it to help me refine and look for changes I should make.
Both solutions for 5 and 6 from the energy equation appear to give the correct solution when substituted into the momentum equation 1 so it should be ok.
Btw just to clear things up, I'm not going to be the one building it. I've masters in biochemical engineering(far from expertise required). I will be providing financial support as well as building equipment as part of signed agreement between my company and university.
A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. I may provide pictures of the build in 2017 as it progresses however testing data will be subject to NDA until paper is published/refused or experiment failed and no thrust was detected.
So there are universities involved in this research. Good. I thought Martin Tajmar at Dresden was the only (official) university researcher in this field.
You should also publish if you do not detect thrust (at least at ArXiv).
How do you plan to measure the force? There is a lot of attention for the microwave part at this forum. I have the impression the other necessary part of the game, the force/thrust measurement, is a little bit neglected.
My self, I'm a business man with deep passion for science and a dream to fly to space one day. However I do have lots of friends in university as they helped me immensely to start my business a decade ago.
This week I've signed a contract with university for a privately funded research project by my initiative. They like the idea of this kind of research but they wouldn't put their money into it.
Data will be published whether it's successful or not and the measurement methods will have to be worked out by people doing the science. I have masters in engineering and I know a thing or two but we will be discussing details this weekend/month. Official starting date is set in February when university returns to it's normal operation after holiday season, till than we have to finalize the build and start getting necessary equipment.
Dear ARW,
I noticed, your ideas about setting up the first measurements differ a lot from how I would do it/am going to do it. The people at the university who are going to work for you don't seem to have experience in this or a related field. I strongly recommend you to attract an advisory committee, maybe some the more experienced people active at this forum am willing to participate in it (not me, I'm not experienced enough, yet).
Best, Peter
... A good friend of mine is head of physics department in a university. He and few phd students will be building and testing it. ...
Even with the understanding that this is to be a project based at or associated witha University Physics Department, perhaps one of our builders with experience could repost the cautions and dangers involved.
Most of what has been posted in these discussions has been working with DC voltages of 2 to 3 kv and raw microwave energies up to perhaps 1000watts. Both can be deadly if not handled carefully by experienced and qualified individuals. Once you start talking about 10kw of MW energies the danger increases more than 10 fold.
Again maybe one of the DIY builders in these discussions with the experience I do not have can properly emphasize the need for caution and careful experienced handling of the DC and high frequency currents involved, even for a project being formulated at a university lab.
I offer my advice for this. They can PM me here. I will soon be very busy with more project work however.
It might be oversimplified, but here is a rough draft on how modifying the mass of light in a cavity with regards to its wavelength might give it some push. https://www.researchgate.net/profile/Dustin_Macdermott "Is the frustum EM Drive4 decelerating light for propellantless propulsion?" It may still be flawed. Will revise as I go.
... Was actually drawn by your Barnett type experiment thesis which I downloaded before realizing I already had your .pptx presentation of same. I researched this a bit for my own experiments.. What was Barnett`s conjecture again about the magnetic field ... ? That it was some sort of emission? What are others conjectures about its dynamic nature?
) the "guide wavelength" will be longer, the smaller the diameter is. 
