FDTD code for fibre probe in matlab

EDAboard.com Forum Index -> Electromagnetic Design and Simulation -> FDTD code for fibre probe in matlab

Author

Message

ravi8331

Joined: 04 Jul 2008
Posts: 337
Helped: 95
Location: India

26 Aug 2009 11:35 fdtd staircase

hii
anyone can tell me how can i define cylindrical structure in FDTD.

skysearcher

Joined: 20 Jul 2009
Posts: 10

26 Aug 2009 22:36 fdtd cylinder

You need to define it wherever you're setting your material parameters. Your code probably has some coefficients that take care of that - you need to change the ones that represent the nodes where your cylinder is physically located.

ali110

Joined: 02 May 2005
Posts: 229
Helped: 68

27 Aug 2009 9:51 what is fdtd and how to use it?

Hi ravi8331
You can use two methods. In first method you can approximate you cylinder boundaries with fine rectangulars and use FDTD equation in Cartesian coordinate. In second method you should explain Max well's equation in cylindrical coordinates.

I hope it'll be helpful

ravi8331

Joined: 04 Jul 2008
Posts: 337
Helped: 95
Location: India

27 Aug 2009 10:57 fdtd head equation

hello sir
first of all thanks for reply .........
Sir you have any document or any material which provide me some help in this case just for example ....so on th ebasis of that i can make program for that......

skysearcher

Joined: 20 Jul 2009
Posts: 10

27 Aug 2009 16:20 how to make a circle in fdtd

Check out the example programs in Sullivan's book Electromagnetic Simulation Using the FDTD Method. There's an example using a 2D circle and a 3D sphere - you can extend that out pretty easily to a cylinder example. He also talks a little about a basic averaging method for handling the edges. The FDTD method has some problems with round edges, and there are various methods to handle that.

Google
AdSense
Google Adsense

27 Aug 2009 16:20 Ads

ravi8331

Joined: 04 Jul 2008
Posts: 337
Helped: 95
Location: India

28 Aug 2009 9:46 Re: cylinder in FDTD

thanks for suggestion ....i will try this.................

iyami

Joined: 12 Apr 2008
Posts: 120
Helped: 16

28 Aug 2009 14:59 Re: cylinder in FDTD

Depending on your excitation you might want to check BOR-FDTD as well
(see e.g. Taflove, BOR=bodies of revolution)
If it fits, you can reduce simulation time considerably.

ravi8331

Joined: 04 Jul 2008
Posts: 337
Helped: 95
Location: India

29 Aug 2009 8:14 Re: cylinder in FDTD

here i am using 50 ohm coaxial probe for feeding the antenna ...........
so for designing that probe i want to define cylindrical object in 3D FDTD...
so help me .............

medaziz

Joined: 13 Dec 2006
Posts: 69
Helped: 2

29 Aug 2009 13:18 Re: cylinder in FDTD

Hi

the description of the feed cylinder with the classical FDTD ( staircase ) can make errors, your 50 ohms will be never respected. Try to using small FDTD cells in edges. you can also using CFDTD ( see Mittra and Yu).

good luk

magbach

Joined: 13 Oct 2008
Posts: 14

30 Aug 2009 2:12 Re: cylinder in FDTD

ravi if my program are rigth plz send me a visible men head model i asked you this in this topic
http://www.edaboard.com/viewtopic.php?t=348760&highlight=sar
the program in Matlab you can to conver it in C++ if you use it
the speace for simulation are borned by : ia,ib;ja,jb;ka,kb (sulivan notation) dont forget PML
ic,jc,kc are the center cylinder

h=height_cylinder;
radius= radius_cylinder;

program :

clc
t=1000;
epsi0=8.8419e-12;
N=124;
ia=9;
ja=9;
ka=9;
ic=N/2;
jc=N/2;
kc=N/2;

ib=N-ia+1;
jb=N-ja+1;
kb=N-ka+1;

ddx=0.0025;
dt=ddx/6e8;
epsilon=44;
sigma=0.9;

for j=1:1:N

for k=1:1:N

for i=1:1:N

ga(i,j,k)=1;
gb(i,j,k)=0;

end
end
end
h=40;
radius= 10;
for i=ia:1:ib
for j=ja:1:jb
for k=ka:1:kb
xdist=(ic-i);
ydist=(jc-j);
kdist=abs(kc-k);
dist=sqrt(xdist^2+ydist^2);
if (dist<=radius)&&(kdist<=h/2)
ga(i,j,k)=1/(epsilon+(sigma*dt/epsi0));
gb(i,j,k)=sigma*dt/epsi0;
end
end
end
end

i test it its work
it is not necessery simulate coaxial. you simulate juste your antenna and you fixe the E filed or D filed directly in the cell wher the coaxial connected that is wat the people do

this document will help you
bat remenber to send me the visible man head please
email : tassoki(at)yahoo.fr
or post it in this topic plz
you use matlab or c++ or other
if you have any difficulte to convert the program on c++ i will help you
excuse my english i dont speak it

Sorry, but you need login in to view this attachment

ravi8331

Joined: 04 Jul 2008
Posts: 337
Helped: 95
Location: India

31 Aug 2009 9:58 Re: cylinder in FDTD

hello
i have done SAR Analysis in three layer Bio-media in CST and HFSS simulation software . if you wanty these program then i will give you .for head i used the structure given in varous research papers just by using dielectric properties and size of the human head...if u desired this then i will post it..............
ok.............