Like this? (run-until 200 (synchronized-magnetic output-poynting output-tot-pwr))
Quote from: Rodal on 07/03/2015 07:39 pmDue to the planar nature of the dipole antenna effect, the Poynting vector field in the xy plane has significantly smaller amplitude than the one in the xz plane. However, it in increasing with time, the amplitude at Time Step 13 clearly being much larger than at TS03 and TS08 (which are similarly located in the time cycle).3D animation rendering, please stand by
Due to the planar nature of the dipole antenna effect, the Poynting vector field in the xy plane has significantly smaller amplitude than the one in the xz plane. However, it in increasing with time, the amplitude at Time Step 13 clearly being much larger than at TS03 and TS08 (which are similarly located in the time cycle).
Quote from: lmbfan on 07/03/2015 08:04 pmConsidering that Dr. Rodal's analysis shows only 3 slices of the frustum, what are the odds that there are some vectors in the field that are not show actually point the other way? Is there some way to sum the entire Poynting field, including those not shown in the cross sectional views? I notice that in Meep there is the option to output the Poynting vectors here: http://ab-initio.mit.edu/wiki/index.php/Meep_Reference#Output_functionsNotably this section:Quoteoutput-Xfield-x, output-Xfield-y, output-Xfield-z, output-Xfield-r, output-Xfield-p Output the x, y, z, r, or φ component respectively, of the field X, where X is either h, b, e, d, or s for the magnetic, electric, displacement, or Poynting field, respectively. If the field is complex, outputs two datasets, e.g. ex.r and ex.i, within the same HDF5 file for the real and imaginary parts, respectively. Note that for outputting the Poynting field, you might want to wrap the step function in synchronized-magnetic to compute it more accurately; see Synchronizing the magnetic and electric fields. I wonder if there is a HDF5 reduction tool that can sum up fields in the file and reduce the entire field to one vector, or how complicated it would be to write such a tool. Seems to me to be just adding up a bunch of numbers... which computers are rumored to be good at.Nothing left to chance here. In Cartesian coordinates three views should suffice, plus common knowledge of transverse electromagnetic variation with the azimuthal angle. There are only two planes shown at azimuthal angles of 0 and 90 degrees, but the variation with azimuthal angle is shown in the base view and it is a conventional harmonic m=1, n=1 variation.It is for higher modes, m>1 that what you are discussing would apply, not for m=1, n=1.We know what the antenna looks like: it is a dipole antenna, the plane views are consistent with it.To double check this all that is needed is to provide other circular cross-sections: I would favor one at the antenna location, another one close to it, within the same longitudinal wave-pattern, and another one in the next longitudinal wave pattern away from it towards the big base.Concerning <<seems to me to be just adding up a bunch of numbers... which computers are rumored to be good at.>>, the csv files are available to the public, so anybody can perform their own postprocessing calculations based on the csv, all you have to do is to calculate this can be done by anybody without using HDF5
Considering that Dr. Rodal's analysis shows only 3 slices of the frustum, what are the odds that there are some vectors in the field that are not show actually point the other way? Is there some way to sum the entire Poynting field, including those not shown in the cross sectional views? I notice that in Meep there is the option to output the Poynting vectors here: http://ab-initio.mit.edu/wiki/index.php/Meep_Reference#Output_functionsNotably this section:Quoteoutput-Xfield-x, output-Xfield-y, output-Xfield-z, output-Xfield-r, output-Xfield-p Output the x, y, z, r, or φ component respectively, of the field X, where X is either h, b, e, d, or s for the magnetic, electric, displacement, or Poynting field, respectively. If the field is complex, outputs two datasets, e.g. ex.r and ex.i, within the same HDF5 file for the real and imaginary parts, respectively. Note that for outputting the Poynting field, you might want to wrap the step function in synchronized-magnetic to compute it more accurately; see Synchronizing the magnetic and electric fields. I wonder if there is a HDF5 reduction tool that can sum up fields in the file and reduce the entire field to one vector, or how complicated it would be to write such a tool. Seems to me to be just adding up a bunch of numbers... which computers are rumored to be good at.
output-Xfield-x, output-Xfield-y, output-Xfield-z, output-Xfield-r, output-Xfield-p Output the x, y, z, r, or φ component respectively, of the field X, where X is either h, b, e, d, or s for the magnetic, electric, displacement, or Poynting field, respectively. If the field is complex, outputs two datasets, e.g. ex.r and ex.i, within the same HDF5 file for the real and imaginary parts, respectively. Note that for outputting the Poynting field, you might want to wrap the step function in synchronized-magnetic to compute it more accurately; see Synchronizing the magnetic and electric fields.
I need more details on what boundary conditions you are imposing and exactly how are you imposing them at the nodes to understand the model. I am NOT discussing the copper modeling. I am interested in the boundary conditions you implement such that the transverse electric fields Ez, Ez, Ey are zero at a boundary and that the normal magnetizing field Hx, Hy and Hz is zero at a boundary as appropriate to each boundary. In particular for the side conical walls.Could you give us an example of the actual code showing the imposition of boundary conditions at a finite difference gridpoint on the conical wall?
The illusion of continuity in Meep Although FDTD inherently uses discretized space and time, as much as possible Meep attempts to maintain the illusion that you are using a continuous system. At the beginning of the simulation, you specify the spatial resolution, but from that point onwards you generally work in continuous coordinates in your chosen units (see units in Meep, above). For example, you specify the dielectric function as a function ε(x) of continuous x, or as a set of solid objects like spheres, cylinders, etcetera, and Meep is responsible for figuring out how they are to be represented on a discrete grid. Or if you want to specify a point source, you simply specify the point x where you want the source to reside—Meep will figure out the closest grid points to x and add currents to those points, weighted according to their distance from x. If you change x continously, the current in Meep will also change continuously (by changing the weights). If you ask for the flux through a certain rectangle, then Meep will linearly interpolate the field values from the grid onto that rectangle. In general, the philosophy of the Meep interface is pervasive interpolation, so that if you change any input continously then the response of the Meep simulation will change continuously as well, and so that it will converge as rapidly and as smoothly as possible to the continuous solution as you increase the spatial resolution. For example, the ε function used internally by Meep is not simply a discretely sampled version of the ε(x) specified by the user. Rather, each grid point is a kind of average of the ε in the surrounding pixel. Our subpixel average is specially designed in order to minimize the "staircasing" and other errors caused by sharp interfaces, and we believe it is a substantial improvement over past methods used for FDTD. See the paper by Farjadpour et al. in Citing Meep.
Quote from: VAXHeadroom on 07/03/2015 08:32 pmQuote from: Rodal on 07/03/2015 07:39 pmDue to the planar nature of the dipole antenna effect, the Poynting vector field in the xy plane has significantly smaller amplitude than the one in the xz plane. However, it in increasing with time, the amplitude at Time Step 13 clearly being much larger than at TS03 and TS08 (which are similarly located in the time cycle).3D animation rendering, please stand by I'm thinking I need to maybe switch the 'flat' and 'upright' images in the animation... I swear the Poynting vectors look like they're rowing
Quote from: VAXHeadroom on 07/03/2015 08:48 pmQuote from: VAXHeadroom on 07/03/2015 08:32 pmQuote from: Rodal on 07/03/2015 07:39 pmDue to the planar nature of the dipole antenna effect, the Poynting vector field in the xy plane has significantly smaller amplitude than the one in the xz plane. However, it in increasing with time, the amplitude at Time Step 13 clearly being much larger than at TS03 and TS08 (which are similarly located in the time cycle).3D animation rendering, please stand by I'm thinking I need to maybe switch the 'flat' and 'upright' images in the animation... I swear the Poynting vectors look like they're rowing This is great !!!Could you 1) look at this from an angle so that one has a better view of the horizontal plane?2) Can you rotate it more so that one gets to see the circular cross-section better?Or would it be too much to ask to get 3 separate movies for each plane separately?
Could you give us an example of the actual code showing the imposition of boundary conditions at a finite difference gridpoint on the conical wall?Sure, here - https://github.com/stevengj/meep/blob/master/src/meep.hpp#L169-L172Dr. Rodal, you are asking about the internals of Meep. I have no idea how to answer that. In my understanding, wall thickness does not matter with perfect metal. With copper, wall thickness and a PML box encompassing the drive would be needed.deuteragenie - I think the first sentence is right. As for the second sentence, well, any fields that escape the cavity will bounce around in the computational lattice. But look at some of the .csv files. You can see that the energy outside the cavity is down by nearly 30 orders of magnitude compared to the energy inside the cavity. To me, that seems negligible.
Quote from: aero on 07/03/2015 08:55 pmCould you give us an example of the actual code showing the imposition of boundary conditions at a finite difference gridpoint on the conical wall?Sure, here - https://github.com/stevengj/meep/blob/master/src/meep.hpp#L169-L172Dr. Rodal, you are asking about the internals of Meep. I have no idea how to answer that. In my understanding, wall thickness does not matter with perfect metal. With copper, wall thickness and a PML box encompassing the drive would be needed.deuteragenie - I think the first sentence is right. As for the second sentence, well, any fields that escape the cavity will bounce around in the computational lattice. But look at some of the .csv files. You can see that the energy outside the cavity is down by nearly 30 orders of magnitude compared to the energy inside the cavity. To me, that seems negligible.I am asking about your model.Let me put the question this different way:You input a square grid for the cross-section. The truncated cone circular cross-section is inside this square grid.How did you let Meep know where the circular boundary gridpoints of the metal are located ? How did you let Meep know where the copper is located ?
inline direction stop_at_direction(ndim dim) { return (direction) (dim + 1 + 2 * (dim == D1));}I could bet dim never equals to D1 in practice; if that is the case the whole expression can be simplified and moved upstream. But we should discuss this in a Meep forum, not here.int num_direction(direction d) const { return num[((int) d) % 3]; };Maybe a candidate for optimization: is it using ivec ? Can the direction be made of a special form which guarantee that the modulo is always 0 or 1 ? or that mod can be optimized because of the special form of d (ie d is a Mersenne prime etc.)See http://stackoverflow.com/questions/1697358/fast-modulo-3-or-division-algorithm?lq=1idiv cycles: approx. 43shift/add approach cycles: approx 30, so at first glance there is something to be won to optimize %3 Does it beat -O3 ?Also, what is the value of direction for one run? Does it change or is it set once the .ctl file is loaded? If it is constant troughout a run, this can be simplified a lot of course.
Three dimensional runs are by far the slowest so they could use some help.
Quick notes:We have shown that there is a definite Poynting vector, meaning an energy flux in the longitudinal direction, that persists over integer periods of time.This has very important consequences regarding the Stress-Energy tensor, regarding momentum. See: https://en.wikipedia.org/wiki/Stress%E2%80%93energy_tensorThe Poynting vector is the momentum density part of the Stress-Energy tensor.We still have to show that the flux goes into momentum of the copper and does not get all dissipated into the walls as heat.Even if there would be momentum we would have to show that the amount is anywhere close to what experimenters are claiming.As to how momentum could be produced, notice that Aero has a FD grid outside the cavity.Aero: do you have any gaps through which evanescent waves can escape? when you had an evanescent wave field in the past, was it achieved through intentional gaps you had placed on purpose?A possible gap in the EM Drive of experimenters is between the bases and the truncated cone walls: not a perfect conductive seal, as I recall. NASA uses fiber reinforced epoxy bases coated with a very thin layer of copper on the inside. Other experimenters seem to have a gasket between the base and the truncated cone walls.
Quote from: Rodal on 07/03/2015 09:08 pmQuote from: VAXHeadroom on 07/03/2015 08:48 pmQuote from: VAXHeadroom on 07/03/2015 08:32 pmQuote from: Rodal on 07/03/2015 07:39 pmDue to the planar nature of the dipole antenna effect, the Poynting vector field in the xy plane has significantly smaller amplitude than the one in the xz plane. However, it in increasing with time, the amplitude at Time Step 13 clearly being much larger than at TS03 and TS08 (which are similarly located in the time cycle).3D animation rendering, please stand by I'm thinking I need to maybe switch the 'flat' and 'upright' images in the animation... I swear the Poynting vectors look like they're rowing This is great !!!Could you 1) look at this from an angle so that one has a better view of the horizontal plane?2) Can you rotate it more so that one gets to see the circular cross-section better?Or would it be too much to ask to get 3 separate movies for each plane separately?Your wish is my command. *vanishes in a greasy puff of smoke*
Aero: do you have any gaps through which evanescent waves can escape? when you had an evanescent wave field in the past, was it achieved through intentional gaps you had placed on purpose?No, there are no gaps, as is evident in the code snippet posted above.Yes, in the past my frustum was modelled as perfect metal, hence intentional gaps were the only method to allow evanescent waves to escape. I did model the Copper Kettle with perfect metal and the gasket they had used at one point. See image attached. I don't remember anything more about that gasket, but it did make a pretty picture
Take a look at the image above and tell me if they are evanescent waves or propagating waves outside of the cavity. They must squeeze through the gap as evanescent waves but maybe they self-organize into propagating waves on escaping. What ever they they are, they certainly go farther than 1/3 wavelength from the cavity. Cavity is 9 - 10 inches high and the frequency is like 1.95 GHz.
Quote from: aero on 07/03/2015 04:52 pmDr. Rodal- Time slices 3 thru 7 have been added to the csv folder on Google Drive. Same place as the other ones, I just changed the name to reflect 3 thru 13.@aero:two important questions to investigate this further:1) How do you impose boundary conditions? What are your boundary conditions and how do you actually implement them in Meep (I read that boundary conditions can be imposed such that the problem becomes nonlinear, which would also partly explain the results)2) TS013 : does this mean that you only marched the FD solution through 13 time steps total? If so, this is way insufficient to make sure that this is not just a transient, if so we would need to investigate marching forwards many more time steps to investigate the time evolution
Dr. Rodal- Time slices 3 thru 7 have been added to the csv folder on Google Drive. Same place as the other ones, I just changed the name to reflect 3 thru 13.