Re: [Meep-discuss] E and H for the same point
On Jun 3, 2008, at 2:27 AM, Steven G. Johnson wrote: To get the fields for the same *time*, you will need to do some averaging yourself. In particular, each time step first steps H and then steps E (and D), so the E (and D) field is always half a time-step ahead of the H field. So, for example, you could output HDF5 files for E and H from two consecutive time steps, and use the *first* E field with the *average* of the two H fields. Update: this is now easy to do in Meep version 0.20: just wrap your step function in (synchronized-magnetic ...) and Meep will synchronize H (and B) with E (and D) in time for you, and then you can do your field computations normally (e.g. to compute the stress tensor as for the original poster in this thread). See also: http://ab-initio.mit.edu/wiki/index.php/Synchronizing_the_magnetic_and_electric_fields Steven ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
Re: [Meep-discuss] E and H for the same point
In two dimensions, the fields (by symmetry since the xy plane is a mirror plane) can be divided into two polarizations, TE (Hz, Ex, Ey) and TM (Ez, Hx, Hy). When you use an Ey source, it only produces TE- polarized fields. You get a magnetic field (and a nonzero stress tensor), but it is in the z direction only. Conversely, an Hx source produces only TM fields, by symmetry. (See also chapter 3 of our new textbook, available online at http://ab-initio.mit.edu/book) Obviously, the optical forces (the stress tensor) will depend on what source you use (and consequently what field patterns you excite). I have no idea what physical situation you are interested in. However, there is nothing physically wrong with using just a TE source, computing the stress tensor from the resulting (Ex,Ey,Hz) fields, and hence obtaining the radiation pressure on some object. 1. A quick question: how to implement (dielectric constant = -24.9 + 1.57i) 2.Back to the previous topic I consider a model based on nearly touching gold spheres illuminated with a plane wave incident perpendicular to an axis through the particles' centers and with the electrical field parallel to the axis. In my case, wavelengh of source is 550nm, the radius of gold sphere is 25nm and the gap between two spheres is 1nm. The electrical field is implemented as Ey and the propagation is along X+. As I am running my simulation on the Sun Grid with parallel meep, the resolution is set high (400) in order to get good visualization result. My target is to calculate the stress tensor in the X-Y plane. == ;define the simulation cell ;basic unit is 100nm ;planewave source ;frequency 1/(550nm), fwidth 0.2 (346nm~1127nm) ; Drude model of Au (set! eps-averaging? false) (define-param wp_au 0.67) ; resonant plasma frequency (define-param gama_au 6.85e-3) ; gamma of Gold (define-param e_au 9.84) ; epsilon of Gold9.84 (define-param r 0.25) (define-param d 0.01) (define-param th 0.5) (define Au (make dielectric (epsilon e_au) (polarizations (make polarizability (omega 1e-20) (gamma gama_au) (delta-epsilon (* wp_au wp_au 1e+20 1e+20)) (set! geometry-lattice (make lattice (size (+ th 15) (+ th 15) no-size))) (set! geometry (list (make block (center 0 0) (size 3 3 infinity) (material (make dielectric (epsilon 1.77 (make sphere (center 0 0) (radius r) (material Au)) (make sphere (center 0 (+ (* 2 r) d)) (radius r) (material Au)) )) (set! sources (list (make source (src (make continuous-src (wavelength 5.5))) (component Ey) (center -7 0) (size 0 14 (set! pml-layers (list (make pml (thickness th (set! resolution 400) (use-output-directory) (run-until 200 (at-beginning (in-volume( volume (center 0 0) (size 1 1)) output-epsilon)) (to-appended ey (at-every 1 (in-volume (volume (center 0 0) (size 1 1)) output-efield-y))) (to-appended ex (at-every 1 (in-volume (volume (center 0 0) (size 1 1)) output-efield-x))) (to-appended hy (at-every 1 (in-volume (volume (center 0 0) (size 1 1)) output-hfield-y))) (to-appended hx (at-every 1 (in-volume (volume (center 0 0) (size 1 1)) output-hfield-x))) ) ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
Re: [Meep-discuss] E and H for the same point
On Jun 5, 2008, at 7:47 AM, Juntao Xi wrote: -- From: Steven G. Johnson [EMAIL PROTECTED] Sent: Thursday, June 05, 2008 9:16 PM To: meep-discuss@ab-initio.mit.edu Subject: Re: [Meep-discuss] E and H for the same point Both E and H are necessary in order to calculate stress tensor. Do I need to run the simulation twice with different source component? No. A single source produces both E and H. (From Maxwell's equations, an oscillating E field creates H fields and vice versa.) However, all Hx and Hy field are zero when src-component is set as Ey. all Ex and Ey field are zero when scr-component is set as Hxuss In two dimensions, the fields (by symmetry since the xy plane is a mirror plane) can be divided into two polarizations, TE (Hz, Ex, Ey) and TM (Ez, Hx, Hy). When you use an Ey source, it only produces TE- polarized fields. You get a magnetic field (and a nonzero stress tensor), but it is in the z direction only. Conversely, an Hx source produces only TM fields, by symmetry. (See also chapter 3 of our new textbook, available online at http://ab-initio.mit.edu/book) Obviously, the optical forces (the stress tensor) will depend on what source you use (and consequently what field patterns you excite). I have no idea what physical situation you are interested in. However, there is nothing physically wrong with using just a TE source, computing the stress tensor from the resulting (Ex,Ey,Hz) fields, and hence obtaining the radiation pressure on some object. Steven ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
Re: [Meep-discuss] E and H for the same point
To get the fields for the same point in space, you can just use one of the supplied interpolation routines as defined in the manual, such as get-field-pt. The HDF5 output routines also interpolate all the fields onto the same spatial grid. Also, the built-in field- integration routines (see the reference manual) also interpolate the fields onto the same spatial grid for you --these routines may be useful to you if you want to integrate the stress tensor. To get the fields for the same *time*, you will need to do some averaging yourself. In particular, each time step first steps H and then steps E (and D), so the E (and D) field is always half a time- step ahead of the H field. So, for example, you could output HDF5 files for E and H from two consecutive time steps, and use the *first* E field with the *average* of the two H fields. To do this averaging within Meep is possible, but is a bit tricky right now from the Scheme interface. To see how it is done in C++, see the implementation of fields::flux_in_box in src/ energy_and_flux.cpp. Or you could just ignore the time difference and use E and H at slightly different times; this will still converge as you increase resolution, but is only first-order-accurate in the resolution. Steven On Jun 3, 2008, at 1:13 AM, matt wrote: Do these texts answer your question? http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#Finite-difference_time-domain_methods http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#The_illusion_of_continuity_in_Meep On Tue, 3 Jun 2008, Juntao Xi wrote: Hi, Steven and Meepers, Now, I am working on the optical force simulation project. And mostly, the optical force is presented by the Maxwell stress tensor. (http://en.wikipedia.org/wiki/Maxwell_stress_tensor). As you know, Meep is using FDTD algorithms and the E output and H output have half-step difference. My question is that how I can get the E and H for the same point Regards Jun ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
[Meep-discuss] E and H for the same point
Hi, Steven and Meepers, Now, I am working on the optical force simulation project. And mostly, the optical force is presented by the Maxwell stress tensor. (http://en.wikipedia.org/wiki/Maxwell_stress_tensor). As you know, Meep is using FDTD algorithms and the E output and H output have half-step difference. My question is that how I can get the E and H for the same point Regards Jun ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
Re: [Meep-discuss] E and H for the same point
Do these texts answer your question? http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#Finite-difference_time-domain_methods http://ab-initio.mit.edu/wiki/index.php/Meep_Introduction#The_illusion_of_continuity_in_Meep Best, Matt On Tue, 3 Jun 2008, Juntao Xi wrote: Hi, Steven and Meepers, Now, I am working on the optical force simulation project. And mostly, the optical force is presented by the Maxwell stress tensor. (http://en.wikipedia.org/wiki/Maxwell_stress_tensor). As you know, Meep is using FDTD algorithms and the E output and H output have half-step difference. My question is that how I can get the E and H for the same point Regards Jun ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
