...That is not an option that I see in the h5topng manual. Maybe HDFview has that option but I think you're asking for some MatLab data processing. If so, then no, I can't, maybe someone else would like to accept the challenge. I know that Meep users commonly reduce data using MatLab programs so it's likely possible.
.... No end needs to be open, microwaves can escape with the whole resonator closed.
Quote from: Rodal on 06/23/2015 02:35 amQuote from: Rodal on 06/22/2015 10:32 pm....Let's try to work through this. The standard convention is to take z as the axis of symmetry (the longitudinal axis of the cone) but the way you labeled them, it looks like x is your axis of symmetry, is that correct?If x is the axis of symmetry, then y and z are perpendicular to it. On the trapezium-looking cross section with axis y perpendicular to it, the axis of the trapezium are x and z, is that correct?OK, if the answer is yes, on your y plot, what are you plotting:Ex ?Ez?Hx?Hz?Quote from: aero on 06/23/2015 12:56 amX is the axis of symmetry. The antenna is to the +Y edge of the cavity. Z completes the coordinate system.I am plotting a snap shot of the x, y and z corrdinate values of the Ez field. More than that you'd have to ask a physicist.Can you plot the Absolute Value of the E field:for example, for the cross-section with normal yinstead of Ez, can you have contour plot Sqrt[(Ex)^2 + (Ez)^2]That is not an option that I see in the h5topng manual. Maybe HDFview has that option but I think you're asking for some MatLab data processing. If so, then no, I can't, maybe someone else would like to accept the challenge. I know that Meep users commonly reduce data using MatLab programs so it's likely possible.
Quote from: Rodal on 06/22/2015 10:32 pm....Let's try to work through this. The standard convention is to take z as the axis of symmetry (the longitudinal axis of the cone) but the way you labeled them, it looks like x is your axis of symmetry, is that correct?If x is the axis of symmetry, then y and z are perpendicular to it. On the trapezium-looking cross section with axis y perpendicular to it, the axis of the trapezium are x and z, is that correct?OK, if the answer is yes, on your y plot, what are you plotting:Ex ?Ez?Hx?Hz?Quote from: aero on 06/23/2015 12:56 amX is the axis of symmetry. The antenna is to the +Y edge of the cavity. Z completes the coordinate system.I am plotting a snap shot of the x, y and z corrdinate values of the Ez field. More than that you'd have to ask a physicist.Can you plot the Absolute Value of the E field:for example, for the cross-section with normal yinstead of Ez, can you have contour plot Sqrt[(Ex)^2 + (Ez)^2]
....Let's try to work through this. The standard convention is to take z as the axis of symmetry (the longitudinal axis of the cone) but the way you labeled them, it looks like x is your axis of symmetry, is that correct?If x is the axis of symmetry, then y and z are perpendicular to it. On the trapezium-looking cross section with axis y perpendicular to it, the axis of the trapezium are x and z, is that correct?OK, if the answer is yes, on your y plot, what are you plotting:Ex ?Ez?Hx?Hz?
X is the axis of symmetry. The antenna is to the +Y edge of the cavity. Z completes the coordinate system.I am plotting a snap shot of the x, y and z corrdinate values of the Ez field. More than that you'd have to ask a physicist.
Quote from: aero on 06/23/2015 02:53 amQuote from: Rodal on 06/23/2015 02:35 amQuote from: Rodal on 06/22/2015 10:32 pm....Let's try to work through this. The standard convention is to take z as the axis of symmetry (the longitudinal axis of the cone) but the way you labeled them, it looks like x is your axis of symmetry, is that correct?If x is the axis of symmetry, then y and z are perpendicular to it. On the trapezium-looking cross section with axis y perpendicular to it, the axis of the trapezium are x and z, is that correct?OK, if the answer is yes, on your y plot, what are you plotting:Ex ?Ez?Hx?Hz?Quote from: aero on 06/23/2015 12:56 amX is the axis of symmetry. The antenna is to the +Y edge of the cavity. Z completes the coordinate system.I am plotting a snap shot of the x, y and z corrdinate values of the Ez field. More than that you'd have to ask a physicist.Can you plot the Absolute Value of the E field:for example, for the cross-section with normal yinstead of Ez, can you have contour plot Sqrt[(Ex)^2 + (Ez)^2]That is not an option that I see in the h5topng manual. Maybe HDFview has that option but I think you're asking for some MatLab data processing. If so, then no, I can't, maybe someone else would like to accept the challenge. I know that Meep users commonly reduce data using MatLab programs so it's likely possible.Your Ex component should be zero (I keep forgetting that you are using x for the longitudinal axis)There should only be a magnetic field Hx component in the longitudinal direction for x for TE modes.That's why it is called Transverse Electric: there should not be an electric field in the longitudinal directionIt is for a circular cross section with normal x you need to plot Sqrt[Ez^2+Ey^2]Can you please verify that your Ex is zero ?Can you give us a plot of the Hx field for the TE mode?
Those dimensions BIG DIAMETER = 0.27246 m SMALL DIAMETER = 0.068115 m LENGTH = 0.4890240258390259 mhave lots of natural frequencies around that range. Here are just a few, for flat ends:Mode frequency (GHz)TE011 1.73146TE012 2.0553TE013 2.3431TM211 1.9874TM212 2.40296TM213 2.72512TE111 0.965122TE112 1.24641TE113 1.50459TM111 1.51277TM112 1.89759TM113 2.20088
...Good points. I worried about the pendulum effect that I've seen on other tests and the slow movement of other outside forces is a concern. I remember setting up anti-vibration tables in a lab and watching cars and trucks drive by a 100 foot away. Even when we would set up our semiconductor equipment in a lab with a concrete floor we could detect the bending movement of the floor and into our machines by someone walking next to it. Very small movements, but a issue when they were expecting submicron accuracies. ....
Quote from: rfmwguy on 06/23/2015 03:10 amQuote from: SeeShells on 06/23/2015 01:35 amThe critical issue I've seen id getting power cables to the Drive without having to worry about the power cable causing issues. I have the power box under the fulcrum in a shielded cage. The power line from it goes up to connect to a point above the center of the beam. *see drawing and never touches the beam which could cause deflection.Thoughts? Questions?Shellthere will be 4 wires needed for magnetron, 2 bias and 2 filament. Make a liquid "wire splice" for the 4 wires. Basically 4 test tubes of conductive mercury. An electrode from power supply wire goes in at bottom of tube. Another electrode from fulcrum drops down into mercury. The mercury conducts the voltage regardless of the depth of the fulcrum side electrode, it simply swims in it. Depth of mercury and length of electrode should allow Moment arm displacement without breaking contact (fulcrum side electrode rising out of mercury).A non scientific description of a near frictionless wire splice...hope I explained it well enough.Very well indeed! I'll look into it. I think I have some old home thermostats that used mercury. I need to be very careful of it.Shell
Quote from: SeeShells on 06/23/2015 01:35 amThe critical issue I've seen id getting power cables to the Drive without having to worry about the power cable causing issues. I have the power box under the fulcrum in a shielded cage. The power line from it goes up to connect to a point above the center of the beam. *see drawing and never touches the beam which could cause deflection.Thoughts? Questions?Shellthere will be 4 wires needed for magnetron, 2 bias and 2 filament. Make a liquid "wire splice" for the 4 wires. Basically 4 test tubes of conductive mercury. An electrode from power supply wire goes in at bottom of tube. Another electrode from fulcrum drops down into mercury. The mercury conducts the voltage regardless of the depth of the fulcrum side electrode, it simply swims in it. Depth of mercury and length of electrode should allow Moment arm displacement without breaking contact (fulcrum side electrode rising out of mercury).A non scientific description of a near frictionless wire splice...hope I explained it well enough.
The critical issue I've seen id getting power cables to the Drive without having to worry about the power cable causing issues. I have the power box under the fulcrum in a shielded cage. The power line from it goes up to connect to a point above the center of the beam. *see drawing and never touches the beam which could cause deflection.Thoughts? Questions?Shell
Quote from: demofsky on 06/22/2015 11:21 pmQuote from: kml on 06/22/2015 09:10 pmMost likely there is an RFI problem with the scale. I'm going to attempt to shield the scale with the unit suspended above it to prove that it is RFI.I have to say that these are very intriguing results! While we are all waiting to see if you can eliminate RFI interference on your scale as the cause for these results, I would like to point out a couple of notable things:1). The performance with and without dielectrics reflects Shawyer's and Yang's experience.2). Much more interesting from my perspective are the declining lows in your first experiment with the dielectrics. This is a very similar pattern to the Eagleworks tests With a dielectric insert. It was speculated at the time that this might have been due to out gassing but I can't see that happening with a ceramic dielectric!!So what the heck is happening with the dielectrics?There is a more subtle pattern of declining lows without the dielectrics, but it is more pronounced with them present. So either there is some type of ongoing electrical (electrostatic?) interference with the digital scale once the power is off, significant out gassing from somewhere, or we are looking some very interesting theoretical issues here folks - and evanescent waves are just not going to cut it...Edit: After checking the scale, it is apparent that the unit is getting lighter and so we have to add ballooning to the possible effects, with the dielectrics thermal mass sustaining the effect.Don't read too much into the overall declining slope. It takes a while for the scale to fully stabilize after placing the unit on it. The faster I start the test the steeper the slope is, usually. I have learned to wait longer before starting as you can see from the later test #'s.Keep in mind the "suspended" test is without any contact between the test unit and the scale. There should be absolutely no force registered in this configuration. The weight on the scale is from several ceramic flooring tiles used to simulate the weight of the unit. I did test with very little weight on the scale with the unit suspended above it and there was no change in weight indicated:20150621-test-42-nod-up-suspended-paper.pngThis indicates that the RFI induced error may be related to the dynamic correction system used to offset heavy loads.:http://www.scalenet.com/and/gx/sensor.htmlAlso, the tests done in the "down" orientation show much less force, though still in the "weighs less" direction. This may be due to the better RF sealing on the fixed end which is down in the "down" orientation.
Quote from: kml on 06/22/2015 09:10 pmMost likely there is an RFI problem with the scale. I'm going to attempt to shield the scale with the unit suspended above it to prove that it is RFI.I have to say that these are very intriguing results! While we are all waiting to see if you can eliminate RFI interference on your scale as the cause for these results, I would like to point out a couple of notable things:1). The performance with and without dielectrics reflects Shawyer's and Yang's experience.2). Much more interesting from my perspective are the declining lows in your first experiment with the dielectrics. This is a very similar pattern to the Eagleworks tests With a dielectric insert. It was speculated at the time that this might have been due to out gassing but I can't see that happening with a ceramic dielectric!!So what the heck is happening with the dielectrics?There is a more subtle pattern of declining lows without the dielectrics, but it is more pronounced with them present. So either there is some type of ongoing electrical (electrostatic?) interference with the digital scale once the power is off, significant out gassing from somewhere, or we are looking some very interesting theoretical issues here folks - and evanescent waves are just not going to cut it...Edit: After checking the scale, it is apparent that the unit is getting lighter and so we have to add ballooning to the possible effects, with the dielectrics thermal mass sustaining the effect.
Most likely there is an RFI problem with the scale. I'm going to attempt to shield the scale with the unit suspended above it to prove that it is RFI.
MOMENTUM ENHANCEMENTI think I just solved the momentum enhancement problem, and learned something new in the process. I'm working on my DC analysis, since now I'm 99% convinced microwaves are not causing thrust. In the course of my day, I was trying to figure out how much momentum is carried away by a quantum of magnetic flux. Very interesting answer, probably best described quantum mechanically, though that is not how I came to this conclusion.In a superconductor, the momentum of the cooper pair is given by; p = h/λA quantum of magnetic flux is given by: Φ=h/2eTherefore, momentum/volt-sec = p/Φ = 2e/λIn macroscopic terms, the momentum carried away by the magnetic flux, depends on the recoil momentum of the charge per unit length of the electrons flowing in the copper. This value is an intrinsic property of the copper based on the free electron density, which is only slightly altered by collisions, heat and relativistic effects (velocity). So far, I have shown that due to the inductance gradient of the cone geometry, there is a force acting on the current and magnetic flux, pushing it toward the big end. I've also shown that the drift velocity at the small end is much larger than at the big end, so as the current is pushed toward the big end, it is losing momentum in the form of magnetic flux. The amount of momentum gained by the frustum will depend on the difference in drift velocity, the momentum stored as magnetic flux AND, on how much of it can escape.Todd
Quote from: SeeShells on 06/23/2015 03:21 amQuote from: rfmwguy on 06/23/2015 03:10 amQuote from: SeeShells on 06/23/2015 01:35 amYou might check with Mulleton. As I recall he bought some liquid metal not involving mercury specially sold for completing electrical contacts on his DYI device. I don't know how far his project has progressed but I'm sure he could give you a name of the material and maybe a source contact.I had a thought while taking out the trash that I'd like to share. How large is your garage? If you need the extra sensitive with your laser detecter system could you use a couple (or more pairs) of mirrors to reflect the beam across your garage so that any movement will be amplified by the extra distance the light beam travels. You might even be able to see vibration noise this way.This also applies to rfmwguy's setup.
Quote from: rfmwguy on 06/23/2015 03:10 amQuote from: SeeShells on 06/23/2015 01:35 am
Quote from: SeeShells on 06/23/2015 01:35 am
Quote from: WarpTech on 06/23/2015 05:07 amMOMENTUM ENHANCEMENTI think I just solved the momentum enhancement problem, and learned something new in the process. I'm working on my DC analysis, since now I'm 99% convinced microwaves are not causing thrust. In the course of my day, I was trying to figure out how much momentum is carried away by a quantum of magnetic flux. Very interesting answer, probably best described quantum mechanically, though that is not how I came to this conclusion.In a superconductor, the momentum of the cooper pair is given by; p = h/λA quantum of magnetic flux is given by: Φ=h/2eTherefore, momentum/volt-sec = p/Φ = 2e/λIn macroscopic terms, the momentum carried away by the magnetic flux, depends on the recoil momentum of the charge per unit length of the electrons flowing in the copper. This value is an intrinsic property of the copper based on the free electron density, which is only slightly altered by collisions, heat and relativistic effects (velocity). So far, I have shown that due to the inductance gradient of the cone geometry, there is a force acting on the current and magnetic flux, pushing it toward the big end. I've also shown that the drift velocity at the small end is much larger than at the big end, so as the current is pushed toward the big end, it is losing momentum in the form of magnetic flux. The amount of momentum gained by the frustum will depend on the difference in drift velocity, the momentum stored as magnetic flux AND, on how much of it can escape.ToddWOW!!! This is VERY interesting!! When you say "AND, on how much of it can escape." are you referring to the magnetic flux that is storing the momentum? If so what is the mechanism that it uses to escape in your view? Thanks!
Quote from: aero on 06/23/2015 05:08 amQuote from: SeeShells on 06/23/2015 03:21 amQuote from: rfmwguy on 06/23/2015 03:10 amQuote from: SeeShells on 06/23/2015 01:35 amYou might check with Mulleton. As I recall he bought some liquid metal not involving mercury specially sold for completing electrical contacts on his DYI device. I don't know how far his project has progressed but I'm sure he could give you a name of the material and maybe a source contact.I had a thought while taking out the trash that I'd like to share. How large is your garage? If you need the extra sensitive with your laser detecter system could you use a couple (or more pairs) of mirrors to reflect the beam across your garage so that any movement will be amplified by the extra distance the light beam travels. You might even be able to see vibration noise this way.This also applies to rfmwguy's setup.It is a good idea. more angles and lengths to calculate, I hate crunching numbers but if need be I will hack away with a smile. ShellPS: Nite all. I've just too many things swirling between my ears.
Quote from: demofsky on 06/23/2015 05:37 amQuote from: WarpTech on 06/23/2015 05:07 amMOMENTUM ENHANCEMENTI think I just solved the momentum enhancement problem, and learned something new in the process. I'm working on my DC analysis, since now I'm 99% convinced microwaves are not causing thrust. In the course of my day, I was trying to figure out how much momentum is carried away by a quantum of magnetic flux. Very interesting answer, probably best described quantum mechanically, though that is not how I came to this conclusion.In a superconductor, the momentum of the cooper pair is given by; p = h/λA quantum of magnetic flux is given by: Φ=h/2eTherefore, momentum/volt-sec = p/Φ = 2e/λIn macroscopic terms, the momentum carried away by the magnetic flux, depends on the recoil momentum of the charge per unit length of the electrons flowing in the copper. This value is an intrinsic property of the copper based on the free electron density, which is only slightly altered by collisions, heat and relativistic effects (velocity). So far, I have shown that due to the inductance gradient of the cone geometry, there is a force acting on the current and magnetic flux, pushing it toward the big end. I've also shown that the drift velocity at the small end is much larger than at the big end, so as the current is pushed toward the big end, it is losing momentum in the form of magnetic flux. The amount of momentum gained by the frustum will depend on the difference in drift velocity, the momentum stored as magnetic flux AND, on how much of it can escape.ToddWOW!!! This is VERY interesting!! When you say "AND, on how much of it can escape." are you referring to the magnetic flux that is storing the momentum? If so what is the mechanism that it uses to escape in your view? Thanks!Yes, but I'm still working on it. It's partially due to resistance of the copper. When there is a voltage drop, it means flux is escaping the loop. For DC it's no problem, but for microwaves, the only thing that might explain it would be excessive heat causing the skin effect to degrade. IMO, thrust is more likely due to the DC offset from the half-wave 60Hz rectification driving the magnetron, than it is from the microwaves.Todd
Quote from: WarpTech on 06/23/2015 06:05 amQuote from: demofsky on 06/23/2015 05:37 amQuote from: WarpTech on 06/23/2015 05:07 amMOMENTUM ENHANCEMENTI think I just solved the momentum enhancement problem, and learned something new in the process. I'm working on my DC analysis, since now I'm 99% convinced microwaves are not causing thrust. In the course of my day, I was trying to figure out how much momentum is carried away by a quantum of magnetic flux. Very interesting answer, probably best described quantum mechanically, though that is not how I came to this conclusion.In a superconductor, the momentum of the cooper pair is given by; p = h/λA quantum of magnetic flux is given by: Φ=h/2eTherefore, momentum/volt-sec = p/Φ = 2e/λIn macroscopic terms, the momentum carried away by the magnetic flux, depends on the recoil momentum of the charge per unit length of the electrons flowing in the copper. This value is an intrinsic property of the copper based on the free electron density, which is only slightly altered by collisions, heat and relativistic effects (velocity). So far, I have shown that due to the inductance gradient of the cone geometry, there is a force acting on the current and magnetic flux, pushing it toward the big end. I've also shown that the drift velocity at the small end is much larger than at the big end, so as the current is pushed toward the big end, it is losing momentum in the form of magnetic flux. The amount of momentum gained by the frustum will depend on the difference in drift velocity, the momentum stored as magnetic flux AND, on how much of it can escape.ToddWOW!!! This is VERY interesting!! When you say "AND, on how much of it can escape." are you referring to the magnetic flux that is storing the momentum? If so what is the mechanism that it uses to escape in your view? Thanks!Yes, but I'm still working on it. It's partially due to resistance of the copper. When there is a voltage drop, it means flux is escaping the loop. For DC it's no problem, but for microwaves, the only thing that might explain it would be excessive heat causing the skin effect to degrade. IMO, thrust is more likely due to the DC offset from the half-wave 60Hz rectification driving the magnetron, than it is from the microwaves.ToddThat says thrust is proportional to frequency. Laser cavities here we come. Or maybe just very high frequency driven magnatrons. The reason they switch at 60 Hz is because 60 Hz is at the wall socket but there isn't any good reason not to use a 400 Hz generator or go even much higher. Might need to re-design the magnatron but that's no biggie compared to the payoff, if the thruster works that way.
Would a simple switched RF source work as well? Don't know.
Hi Dr. White,I'm an active member of the NSF EMDrive discussion forum and of the Reddit EMDrive group and an engineer by training.On both forums your Appendix A titled"Analysis of Conservation of Energy for Interplanetary Space Missions using Electric Propulsion"http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20140013174.pdfhas been claimed to be in error, being1) The calculations for the start-final KE analysis is negative but is shown as positive.2) The analysis fails to use the Oberth effect on the KE of the ships fuel.Your comments are most welcome as possible CofE violation for long term thrusting spacecraft needs to be understood, especially if applicable to non propellantless drive technology.I would also like to further understand this statement of yours:"When this situation occurs, in order to ensure that the input energy is equal to the change in kinetic energy, the thrust to power performance will have to decrease over time."How will this happen? How will the ship know when to stop obeying A = F/M and start reducing force generation to not break CofE? What will happen if the ship continually accelerates?I actively support your work at Eagleworks and believe those that say your Q Thruster / EMDrive can't work as claimed because then CofE will be violated are incorrect.Hopefully your reply will help others to understand apparent CofE violation with propellantless thrusters is not the show stopper they think it is.Best regards