[Meep-discuss] Bragg fibers, transmission
Hi all, I am using meep to calculate the wavelength/frequency dependent transmission or loss down a hollow core Bragg fiber (as described in the online book Photonic crystals: molding the flow of light). I have been able to replicate all figures in the book using cylindrical coordinates. Now, I want to calculate the transmission down a fiber - specifically one with low refractive index contrast (1.33 vs. 1.55) and low number of crystal repeat units ( 7). I calculate the transmission using two flux planes - one closer to the source normalized by one farther away down the fiber. I currently do not consider that the materials absorb because I want to first examine radiating loss from the photonic crystal structure. I am trying to get to a point where I can adjust the refractive indices and the photonic crystal dimensions to compare fiber transmission/loss for the different scenarios. We are interested in visible wavelengths which and the frequency range accordingly. The current parameters in the attached ctl file should guide approximately 600 nm light (f=1.67 meep units where a = 1 micron). As I understand it, I should place the first flux plane far enough away from the source to consider primarily guided frequencies. I see large differences in the transmission (and sometimes greater than 1) when I change things like the flux plane position from source and size as well as whether the source is a point or a plane. I have attempted to optimize the pml layer thickness and the resolution. My current questions are: 1. Am I modeling transmission down a bragg fiber correctly? 2. Should I use a point or plane source? 3. How far from the source should I put the first flux plane such that I primarily look at guided frequencies? I am currently looking at 2 microns from the source which is more than 2X the guided wavelength. 4. Should the flux planes be the size of the hollow core, the full fiber or larger? Thanks in advance for your advice, Amanda cwgtrans.ctl Description: Binary data ~~ Amanda Holt, Ph.D. | Postdoctoral Scholar University of Pennsylvania, Philadelphia Department of Physics and Astronomy alh...@physics.upenn.edu | 831.359.9822 ~~ ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
[Meep-discuss] cylindrical versus cartesian coordinates
Hi meep users,I am trying to compare results of transmission down a coaxial-Bragg fiber in cylindrical coordintates to results of transmission down the same fiber in cartesian coordinates (in 3D). I eventually want to explore aspects of the fiber which will require cartesian coordinates but first wanted to check my work by comparing results from the same fiber in the two coordinate systems. The results are not the same. I can mess with parameters in the cylindrical coordinate system and find that I get similar results if my flux plane is double in the cylindrical system. I have been assuming that when in cylindrical coordinates in meep, I specify the cylindrical flux plane by radius. If the size is (size 4.5 0 0) centered at (0 0 0), I have a circular flux plane with radius 4.5 and area pi*(4.5^2). In cartesian coordinates, I specify the flux plane as a square area: (size 9 9 0). Which has an area of 81 centered at (0 0 0). Is this correct? The results are only the same if the flux plane in cylindrical coordinates is (size 9 0 0).Since cylindrical coordinates allows me to use my laptop for simulations, I would really like to be sure that when I submit my code to a cluster, its going to give me an accurate result that I can compare with the simulations I run on my computer.I have attached two ctl files, one for a cylindrical waveguide in cylindrical coordinates (cwgtrans2.ctl) and one for an elliptical waveguide in cartesian coordinates (ewgtrans.ctl). I have verified that both codes generate the same dielectric structure when the axis of the ellipsoid are the same. Any ideas? Thanks,Amanda cwgtrans2.ctl Description: Binary data ewgtrans.ctl Description: Binary data ~~Amanda Holt, Ph.D. | Postdoctoral ScholarUniversity of Pennsylvania, PhiladelphiaDepartment of Physics and Astronomyalh...@physics.upenn.edu| 831.359.9822~~___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
[Meep-discuss] loading TEM images for 2D epsilon information
Hi meep users, I read a post from 2008-07-07 (general:1886) asking how to use a Transmission Electron Microscope image as input for modeling an arbitrary geometry (as an epsilon material function) but with no reply. Would anyone be able to explain to me the steps for using a TEM image to generate the epsilon function? I am not super familiar with Scheme so maybe that's where I need to look but I know its been done using meep as I've seen the results - so perhaps there are some functions already written for this purpose? I can process the image in other programs so that it can be binary if that is necessary. Thanks very much, Amanda ~~ Amanda Holt, Ph.D. | Postdoctoral Scholar University of Pennsylvania, Philadelphia Department of Physics and Astronomy alh...@physics.upenn.edu | 831.359.9822 ~~ ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
[Meep-discuss] epsilon-input-file
Hi meep users, If I input a 2D epsilon function in .h5 format - which is the cross section of a strangely shaped waveguide - and want to look at transmission down the waveguide in the 3rd dimension, will meep repeat the 2D cross section to use for epsilon in the 3rd dimension? In the reference section, the description of epsilon-input-file states that the epsilon file dataset will be scaled or linearly interpolated to map onto the computational grid. I am hoping this means the 3rd dimension will be gridded replicas of the 2D cross section. Any one able to verify? Thanks for your time, Amanda ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
[Meep-discuss] Meep and OS 10.9
Hi meep users, Anyone found a work around for the problem compiling Meep on Mac OS 10.9? We had to install Mavericks to use another program on our small in-house cluster and now want to get Meep back up and running. We want to check with the community to see if anyone has attempted this yet. Thanks for your input, Amanda ~~ Amanda Holt, Ph.D. | Postdoctoral Scholar University of Pennsylvania, Philadelphia Department of Physics and Astronomy alh...@physics.upenn.edu | 831.359.9822 ~~ ___ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
