Quote from: SeeShells on 07/07/2015 05:24 pmNo the laser is secured to the center of the beam.Kinda like this... sorry for the quicky in paint.ShellYou can double your resolution by fixing the mirror to the beam pivot and bouncing the laser off of it. That's how the Nichols radiometer worked. If you are expecting a tiny displacement, then an optical lever http://badger.physics.wisc.edu/lab/manual/node33_mn.html is another option.Yes, I'm stalking.. but no new ideas today. Yet. ... And if you haven't read SevenEves yet, go grab it. We'll be here when you get back.
No the laser is secured to the center of the beam.Kinda like this... sorry for the quicky in paint.Shell
Quote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.OK, you have thoroughly covered the issues regarding different testing methods.What do you think of the conjecture that the EM Drive needs to be "motivated" by external vibrations (of unspecified magnitude and frequency) in order to show thrust? and how are you planning to motivate the EM Drive to exhibit such forces? As I understand it from TheTraveller, he is saying that unless you so motivate the EM Drive you will measure very small forces.TT may be right, I don't know, some love it some think it's poo. I'm indifferent. I'm after data. But Jose I plan on running it with out any external push or pull. If nothing happens in the vertical or 180 or even 90 degree rotations than I have a small weight to drop onto the beam in static conditions and than record it and then in activation of the frustum. Well take it from there.Shell
Quote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.OK, you have thoroughly covered the issues regarding different testing methods.What do you think of the conjecture that the EM Drive needs to be "motivated" by external vibrations (of unspecified magnitude and frequency) in order to show thrust? and how are you planning to motivate the EM Drive to exhibit such forces? As I understand it from TheTraveller, he is saying that unless you so motivate the EM Drive you will measure very small forces.
https://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.
..Unless you have taken efforts to eliminate vibration. your EMDrive should register some Force as long as it can move a little and compress your scale. But don't expect it to continuously generate Force as it has little distance to move.
Quote from: TheTraveller on 07/07/2015 06:16 pm..Unless you have taken efforts to eliminate vibration. your EMDrive should register some Force as long as it can move a little and compress your scale. But don't expect it to continuously generate Force as it has little distance to move.For a linear spring constant:stress = E strainforce/area = E (change in length)/originalSpringLengthforce = E (area/originalSpringLength) (change in length)force = K (change in length)where K = spring constant = E (area/originalSpringLength)is a constant.Hence the spring will continue to displace proportionally under increasing force, as long as it remains within its linear range, where K is a constant.The force will be proportional to the displacement.What limits the displacement in Shell's experiment when you state <<it has little distance to move.>>?For example, if she has a linear scale that can measure accurately up to 0.5 Newtons, why would the measurment be limited to a much smaller force?
Project update video:
..My comment was about using a scale to directly measure the 60 cycle pulsing Force being generated.Any scale compresses somewhat as the mass placed on it increases.In her case the EMDrive will generate Force for 1/2 a cycle and then power down for the following 1/2 cycle as the magnetron is powered by a simple 1/2 wave rectifier. During the NO power time, the Force is gone and the scale will restore itself to the pre Force state. As soon as the next 1/2 wave DC power pulse occur, the EMDrive will power up, enter MOTOR mode and generate Force for 1/120 a second and then power down. Only to repeat the process 60 times a second.
Quote from: SeeShells on 07/07/2015 06:01 pmQuote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.OK, you have thoroughly covered the issues regarding different testing methods.What do you think of the conjecture that the EM Drive needs to be "motivated" by external vibrations (of unspecified magnitude and frequency) in order to show thrust? and how are you planning to motivate the EM Drive to exhibit such forces? As I understand it from TheTraveller, he is saying that unless you so motivate the EM Drive you will measure very small forces.TT may be right, I don't know, some love it some think it's poo. I'm indifferent. I'm after data. But Jose I plan on running it with out any external push or pull. If nothing happens in the vertical or 180 or even 90 degree rotations than I have a small weight to drop onto the beam in static conditions and than record it and then in activation of the frustum. Well take it from there.Shell Unless you have taken efforts to eliminate vibration. your EMDrive should register some Force as long as it can move a little and compress your scale. But don't expect it to continuously generate Force as it has little distance to move.
Quote from: TheTraveller on 07/07/2015 06:16 pmQuote from: SeeShells on 07/07/2015 06:01 pmQuote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.OK, you have thoroughly covered the issues regarding different testing methods.What do you think of the conjecture that the EM Drive needs to be "motivated" by external vibrations (of unspecified magnitude and frequency) in order to show thrust? and how are you planning to motivate the EM Drive to exhibit such forces? As I understand it from TheTraveller, he is saying that unless you so motivate the EM Drive you will measure very small forces.TT may be right, I don't know, some love it some think it's poo. I'm indifferent. I'm after data. But Jose I plan on running it with out any external push or pull. If nothing happens in the vertical or 180 or even 90 degree rotations than I have a small weight to drop onto the beam in static conditions and than record it and then in activation of the frustum. Well take it from there.Shell Unless you have taken efforts to eliminate vibration. your EMDrive should register some Force as long as it can move a little and compress your scale. But don't expect it to continuously generate Force as it has little distance to move.So you're saying it will start to move and accelerate until the forces from the thrust balance out the weight, then according to you and RS drop back down at a slower rate? I've thought about this some and another reason I went with the fulcrum. So, if RS and you are right, I'll see an non-uniform sinusoidal movement in the pattern as the beam runs (ratchets) up and down differently compared to the natural harmonic? I'll see a change from Motor to Generator modes, up and down? The nonuniform sinusoidal action can show me a lot of data within the same plane operating continuously. Will the oscillations vary with time, will the amplitudes get larger, will the upper limit of movement get greater and mirror the downward movement or will it differ in speed? Ok, simplify. If I watch the beam after a little push I should see the beam move up to a point and then because as the beam reaches balance with the thrust the acceleration is no longer there and it will fall back down. If on the other hand I see the beam accelerate up and stay there suspended, then the ratchet theory is in jeopardy and out come the erasers for a new theory and we'll see what's behind door number 1 or 2 or 3!You see rotating around on a platform is ok, but you can't change what's needed to test in a dynamic environment. You can just rotate in one direction. But a fulcrum in it's active mode will measure much more in its dynamic moving environment. I understand what you're saying but this fulcrum can show, acceleration, resistance to movement, thrust forces, and it all will exist in the calculations derived from it attenuating the natural harmonic vibration of the beam and from its quiescent state.I'm jazzed to get data that you all can chew on.Shelllittle spellingg mstake
Roger Shawyer's IAC 2014 paper has been peer reviewed and accepted for publication in Acta Astronautica, the IAF journal.2014 Peer Review IAC 14 paper C4,8.5 final.pdf
Quote from: TheTraveller on 07/07/2015 12:18 pmRoger Shawyer's IAC 2014 paper has been peer reviewed and accepted for publication in Acta Astronautica, the IAF journal.2014 Peer Review IAC 14 paper C4,8.5 final.pdfI started making a page by page comparison between this newly uploaded paper and the earlier IAC 14 paper C4,8.5 final.pdf to determine what changes were introduced in the review process, but it turns out that the two pdf files are byte-identical.Traveller, can you please confirm that you uploaded the correct paper, and that there were, in fact, no changes introduced into the paper during the review process.~Kirk
Quote from: Ricvil on 07/07/2015 06:06 amQuote from: WarpTech on 07/07/2015 02:55 amQuote from: Ricvil on 07/06/2015 04:01 amhttps://en.m.wikipedia.org/wiki/Microwave_cavity#Cylindrical_cavityNow I see where the expressions comes from.The cut off frequency expressions for constant radius R cilindrical waveguide.It's wrong because, the expressions for constant radius waveguides are used like dispersions relations of a tappered waveguide , and the " constant radius R" is used as a function of spacial coordinates. Not only this, derivatives of this expressions are done acting on the"ad-hoc" spacial dependency introduced.Wrong!!!I don't think so. You, yourself posted,Quote from: Ricvil on 07/05/2015 03:23 amThis guys think another wayhttp://arxiv.org/abs/0708.3519In the first paragraph of section 4 of this paper, they decompose the wave vector into time-like and space-like, orthogonal components. This is a more precise derivation of exactly what I am doing. The space-like component must shift mass to match the boundary conditions as the time-like component travels down the waveguide. It seems you "believe" waveguides are linear and cannot deviate from this expectation. In the case of a tapered waveguide, it mimics gravity which is non-linear. So, I still do not see what is "wrong" with my equation. Although, now I do see a better way to derive it and reference it. Thank you!ToddThe field equations are linear ( any linear combination of field solutions is a solution too).Non linear are the dispersion relations between frequency (omega) and the wave number(kappa), and this relationship is ditacted solving the maxwell /helmoltz equations for each geometry, boundary conditions and medium of the problem under analysis, and both omega and kappa never has a espatial or temporal dependence just because they are eigenvalues of the differential operators envolved, and the modes are the eigenvectors.When you put a ad-hoc coordinate dependence you are automaticaly, not solving the original equation.The waveguide or cavity can be any shape, and linearity is not related to it.When you take the wikipedia expression for cut off frequency of e regular cylindrical waveguide, and change the constant radius R of the expression for a function of z coordinate in your expression, automatically your expression is not more consistent with the equations, basically because if c a constant, f is a function and d is a differential operator, them d(cf)=cd(f) , but if c is now a function then d(cf)=d(c)f+cd(f), and this new term d(c)f will make the equations not be satisfied, neither the boundary conditions.The article has a decomposition like any orthogonal decomposition, and if you had noted, all wave numbers kappa are constants.You have done a transformation inofensive for algebric equation solutions, but for differential equations is wrong.And if are trying describe fotons in a gravitational field using that same "procedures" you are in error again, because fotons in general relativity also must satisfy differential equations called null geodesic equations.Please show us the "correct" dispersion relationship for a tapered waveguide then. It is simple to solve the boundary condition for a straight non-tapered waveguide. Show us how to do it for a tapered waveguide, without parameterizing the radius with respect to z. Show us how to derive and solve the differential equations for a tapered waveguide. I defer to your expertise!Thank you.Just FYI: My degree is in Quantum Optics, Lasers and Holography, and I have 35 years of experience in power electronics. So I understand resonance and I understand what you are saying. I know how to solve differential equations. If you are expecting a linear solution in a tapered waveguide, you will not find one.Todd
Quote from: WarpTech on 07/07/2015 02:55 amQuote from: Ricvil on 07/06/2015 04:01 amhttps://en.m.wikipedia.org/wiki/Microwave_cavity#Cylindrical_cavityNow I see where the expressions comes from.The cut off frequency expressions for constant radius R cilindrical waveguide.It's wrong because, the expressions for constant radius waveguides are used like dispersions relations of a tappered waveguide , and the " constant radius R" is used as a function of spacial coordinates. Not only this, derivatives of this expressions are done acting on the"ad-hoc" spacial dependency introduced.Wrong!!!I don't think so. You, yourself posted,Quote from: Ricvil on 07/05/2015 03:23 amThis guys think another wayhttp://arxiv.org/abs/0708.3519In the first paragraph of section 4 of this paper, they decompose the wave vector into time-like and space-like, orthogonal components. This is a more precise derivation of exactly what I am doing. The space-like component must shift mass to match the boundary conditions as the time-like component travels down the waveguide. It seems you "believe" waveguides are linear and cannot deviate from this expectation. In the case of a tapered waveguide, it mimics gravity which is non-linear. So, I still do not see what is "wrong" with my equation. Although, now I do see a better way to derive it and reference it. Thank you!ToddThe field equations are linear ( any linear combination of field solutions is a solution too).Non linear are the dispersion relations between frequency (omega) and the wave number(kappa), and this relationship is ditacted solving the maxwell /helmoltz equations for each geometry, boundary conditions and medium of the problem under analysis, and both omega and kappa never has a espatial or temporal dependence just because they are eigenvalues of the differential operators envolved, and the modes are the eigenvectors.When you put a ad-hoc coordinate dependence you are automaticaly, not solving the original equation.The waveguide or cavity can be any shape, and linearity is not related to it.When you take the wikipedia expression for cut off frequency of e regular cylindrical waveguide, and change the constant radius R of the expression for a function of z coordinate in your expression, automatically your expression is not more consistent with the equations, basically because if c a constant, f is a function and d is a differential operator, them d(cf)=cd(f) , but if c is now a function then d(cf)=d(c)f+cd(f), and this new term d(c)f will make the equations not be satisfied, neither the boundary conditions.The article has a decomposition like any orthogonal decomposition, and if you had noted, all wave numbers kappa are constants.You have done a transformation inofensive for algebric equation solutions, but for differential equations is wrong.And if are trying describe fotons in a gravitational field using that same "procedures" you are in error again, because fotons in general relativity also must satisfy differential equations called null geodesic equations.
Quote from: Ricvil on 07/06/2015 04:01 amhttps://en.m.wikipedia.org/wiki/Microwave_cavity#Cylindrical_cavityNow I see where the expressions comes from.The cut off frequency expressions for constant radius R cilindrical waveguide.It's wrong because, the expressions for constant radius waveguides are used like dispersions relations of a tappered waveguide , and the " constant radius R" is used as a function of spacial coordinates. Not only this, derivatives of this expressions are done acting on the"ad-hoc" spacial dependency introduced.Wrong!!!I don't think so. You, yourself posted,Quote from: Ricvil on 07/05/2015 03:23 amThis guys think another wayhttp://arxiv.org/abs/0708.3519In the first paragraph of section 4 of this paper, they decompose the wave vector into time-like and space-like, orthogonal components. This is a more precise derivation of exactly what I am doing. The space-like component must shift mass to match the boundary conditions as the time-like component travels down the waveguide. It seems you "believe" waveguides are linear and cannot deviate from this expectation. In the case of a tapered waveguide, it mimics gravity which is non-linear. So, I still do not see what is "wrong" with my equation. Although, now I do see a better way to derive it and reference it. Thank you!Todd
https://en.m.wikipedia.org/wiki/Microwave_cavity#Cylindrical_cavityNow I see where the expressions comes from.The cut off frequency expressions for constant radius R cilindrical waveguide.It's wrong because, the expressions for constant radius waveguides are used like dispersions relations of a tappered waveguide , and the " constant radius R" is used as a function of spacial coordinates. Not only this, derivatives of this expressions are done acting on the"ad-hoc" spacial dependency introduced.Wrong!!!
This guys think another wayhttp://arxiv.org/abs/0708.3519
Quote from: SeeShells on 07/07/2015 07:01 pmQuote from: TheTraveller on 07/07/2015 06:16 pmQuote from: SeeShells on 07/07/2015 06:01 pmQuote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems. Shelllittle spellingg mstake60 times a second your cavity will have no Rf energy inside it for 1/2 cycle. So you will get 60 x 1/2 cycle long periods of powered down EMDrive per second and 60 x 1/2 cycle per second where the cavity will be full of energy.During the powered down 1/2 cycles I expect your system will react to the lack of Force and try to restore itself back to a non Force state.So yes there may be some oscillatory element imposed on your measurement system by the 60 1/2 cycle pulses per second of Force generation and the 60 1/2 cycle periods per second of a powered down EMDrive.Added to that you need a cavity bandwidth wide enough to swallow your magnetron energy envelope or any energy outside the cavity bandwidth will not be accepted and bounced back. This is why the Chinese had the reflected load built into their tests setup. Reflected power will of course reduce the Ws of power actually inside the cavity versus that outputted by the magnetron.
Quote from: TheTraveller on 07/07/2015 06:16 pmQuote from: SeeShells on 07/07/2015 06:01 pmQuote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems. Shelllittle spellingg mstake
Quote from: SeeShells on 07/07/2015 06:01 pmQuote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.
Quote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.
Quote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems.
Quote from: TheTraveller on 07/07/2015 07:16 pmQuote from: SeeShells on 07/07/2015 07:01 pmQuote from: TheTraveller on 07/07/2015 06:16 pmQuote from: SeeShells on 07/07/2015 06:01 pmQuote from: Rodal on 07/07/2015 05:26 pmQuote from: SeeShells on 07/07/2015 05:16 pmhttps://en.wikipedia.org/wiki/Weighing_scaleA good article summation on simple ways to measure stuff... and the problems. Shelllittle spellingg mstake60 times a second your cavity will have no Rf energy inside it for 1/2 cycle. So you will get 60 x 1/2 cycle long periods of powered down EMDrive per second and 60 x 1/2 cycle per second where the cavity will be full of energy.During the powered down 1/2 cycles I expect your system will react to the lack of Force and try to restore itself back to a non Force state.So yes there may be some oscillatory element imposed on your measurement system by the 60 1/2 cycle pulses per second of Force generation and the 60 1/2 cycle periods per second of a powered down EMDrive.Added to that you need a cavity bandwidth wide enough to swallow your magnetron energy envelope or any energy outside the cavity bandwidth will not be accepted and bounced back. This is why the Chinese had the reflected load built into their tests setup. Reflected power will of course reduce the Ws of power actually inside the cavity versus that outputted by the magnetron.That's a good thing if i measure any thrust, it will be some factor around half of what can be generated, as I only have a 50% duty cycle. A homer Simpson "whoohooo"!So I'll modify my magnetron (which can be done) get 100% duty and double the thrust? I'm not really jazzed about that as WarpTech (and I) think that the DC component from the magnetron assists in the thrust. I'll take this first step with a 50% acceleration component if that's what it is.
Quote from: CW on 07/07/2015 07:23 am.......I also tried to point out many a post ago in (I think) the last EM drive thread, that a standing wave might be interpretable as an 'exotic' type of dynamically created massive particle. This whole tapered frustum actually looks to me as if this geometry squeezed one side of this dynamically created exotic particle equivalent. Assuming this point of view, this squeezed dynamic particle should then react and try to escape the squeezing towards the wider end of the frustum, hence being sort of accelerated, while the equivalent but opposite impulse is being imparted towards the smaller frustum end. If this were the case, the type of squeezing reaction might even depend on the type of exotic particle equivalent that is being dynamically generated in terms of frequency, energy density and field distribution/modes. Just the same as static types of particles have different properties that particle physics knows.Yes, yes yes! Perfect. What wonderfully weird actions within 1/3 of a wavelength are occurring by the antenna? A action the can impart spin and momentum and mass to that virtual particle that is being created within the small endcap? It is a evanescent wave action generating first order forces! Evanescent waves can move MIE particles, gold spheres and even charge your cell phone! This isn't a small force we're imparting to the virtual massive particles, this is an evanescent wave action can move particles (MIE spheres with billions of atoms). The poynting vectors which give the momentum and spin of the massive virtual a direction to "push" and that's to the large end.They head out to the large (*edit) end imparting force but being virtual particles that cannot exist without a corresponding real field from the small end cap they decay and disappear back into the Quantum vacuum. This action doesn't violate anything by making a virtual massive particle at the antenna adding energy and direction and thrust towards the large cavity end and then decaying back into the Quantum Vacuum.I read somewhere and it went like this... short-lived high-mass force-carrier particles seem to violate the laws of conservation of energy and mass -- their mass just can't come out of nowhere!" They can and they do, they are a result of the Heisenberg Uncertainty principle. These high-mass particles come into being and seem to escape the standard model's notice somehow. These massive virtual particles don't violate COE, in the end of the small cavity within 1/3 wavelength. Assisted by the evanescent waves they come into existence, The poynting vectors give them direction and upon decaying their kinetic energy plus their mass is imparted to the Frustum giving it acceleration and thrust.......
.......I also tried to point out many a post ago in (I think) the last EM drive thread, that a standing wave might be interpretable as an 'exotic' type of dynamically created massive particle. This whole tapered frustum actually looks to me as if this geometry squeezed one side of this dynamically created exotic particle equivalent. Assuming this point of view, this squeezed dynamic particle should then react and try to escape the squeezing towards the wider end of the frustum, hence being sort of accelerated, while the equivalent but opposite impulse is being imparted towards the smaller frustum end. If this were the case, the type of squeezing reaction might even depend on the type of exotic particle equivalent that is being dynamically generated in terms of frequency, energy density and field distribution/modes. Just the same as static types of particles have different properties that particle physics knows.
@Rodal-Meep can output the time slices as frequently as desired, up to once per time step, or 6527 h5 file time indices (data sets) per run. So instead 10 h5 data sets/cycle, there would result just over 200 h5 data sets/cycle. Neither I nor my computer can deal with that much data over a full run. It would be much smaller impact to increase the run time beyond 32 cycles. As for increasing the csv data frequency, better to install the latest Meep from source, then explore the use of the newer features of the software to reduce the data at run time, prior to output. That means we would need to drop back 5 and punt, for now, but later we might have a stronger meep to support our efforts.
My latest run has completed. This is the output at the final step.; run time set to 13.052188647619047 meep time This is printed output from control file set.;Meep progress: 13.054/13.052188647619047 = 100.0% done in 6479.3s, -0.9s to go;run 0 finished at t = 13.054 (6527 timesteps)
* There is a net, non-zero Poynting vector over an integer number of periods. This negates what Greg Egan showed, where he concluded that the Poynting vector is zero for an EM Drive.
* The Poynting vector is increasing with time. This is required in more general situations (for radiation pressure all that is required is a non-zero average). This is very significant. It remains to be explored whether it is just a transient.
Quote from: SeeShells on 07/07/2015 07:32 pm.......So I'll modify my magnetron (which can be done) get 100% duty and double the thrust? I'm not really jazzed about that as WarpTech (and I) think that the DC component from the magnetron assists in the thrust. I'll take this first step with a 50% acceleration component if that's what it is.If using a scale for Force generation measurement I see value in using a pulsed power supply.Please note both Shawyer/UK and the Chinese run on 50Hz, so they had longer on periods and longer off periods. Depending to the time the measurement system need to reset from the Force pulse, the pulse rate per second may favour 50Hz over 60Hz.Something to consider.I have now added a duty cycle test to my test schedule, using a scale to see if I can fine tune the duty cycle, using a 20W Rf amp that I can turn on and off via USB, to get optimal Force measurement. This could be exciting as I may be able to use scales in my independent Force measurement program and avoid the necessity to replicate the rotary test setup.
.......So I'll modify my magnetron (which can be done) get 100% duty and double the thrust? I'm not really jazzed about that as WarpTech (and I) think that the DC component from the magnetron assists in the thrust. I'll take this first step with a 50% acceleration component if that's what it is.
I will repeat my questions here, as it seems they disapeared in the flood of other issues:Quote from: Rodal on 07/07/2015 02:11 pm* There is a net, non-zero Poynting vector over an integer number of periods. This negates what Greg Egan showed, where he concluded that the Poynting vector is zero for an EM Drive. Has this been demonstrated/accurately calculated? It has been shown that some slices of the data, when represented graphically appear to have large non-zeo Poynting vector. I guess noone will object to this. This does not mean that a) all slices (... infinitesimally small) exhibit the same behavior and b) that the many small (again, infinitesimally small) Poynting vectors which may or may not be graphically represented do not cancel out the large ones which dwarf them graphically. Or am I missing something ?Quote from: Rodal on 07/07/2015 02:11 pm* The Poynting vector is increasing with time. This is required in more general situations (for radiation pressure all that is required is a non-zero average). This is very significant. It remains to be explored whether it is just a transient.Roughly the same comment on this, but added to the potential need to consider much longer and finer time steps. Has this been taken into account?