Core only optical Fiber power transmission


Hi everyone

I am new in using Lumerical.

I am trying to simulate the behavior of an optical fiber with refractive index of 1.45. This fiber is submerged in a body fluid with refractive index of 1.33 which serves as the cladding layer. This fiber dissolves in the body fluid within 3-4 days. I know the dissolution rate for this. I want to simulate the following properties:

Fiber diameter is initially 100 micron and length is 75 mm. Within 3-4 days it completely dissolves. I use a 632 nm polarized laser source for excitation. I have the experimental data but i would like to simulate the response and wish to get the following result.

  1. The effect of diameter change in transmission of light
  2. absorption loss, scattering loss ( due to irregular surface at the core clad surface after the fiber started dissolution) and other loss parameter

I couldn’t find any example. Can anyone please give me some idea which solver to use, how to define the mesh, input/output port or sources and monitors.

Thanks in advance for all of your help.



The coupling efficiency of the laser beam into the fiber for different diameters of the fiber can be simulated by using the FDE solver in MODE Solutions. You would first calculate the supported mode of the fiber. Then you can use the built-in overlap analysis tool to calculate the power coupling between a beam with user-specified properties and the selected fiber mode:

The beam properties which can be set in the “Beam” tab of the overlap analysis tool are the same as the ones defined under the “Beam options tab” section of this page:

It could be possible to simulate the loss due to roughness of the fiber using the EME solver by directly simulating a distance of the 3D fiber structure to obtain the loss per distance. However, I’ve never tried such a simulation myself so I’m not sure how many modes or cells would be required in the EME solver region in order to get accurate results. You would need to perform some convergence testing:

I would recommend going through the EME Spot size converter tutorial here if you haven’t done so already for an introduction to the workflow since it contains step-by-step instructions on the “Modeling instructions” page:

Another issue is that you would need to model the geometry of the rough surface of the fiber. You could do this by adding a rough wire like in this example:

However, you would need to know the characteristics of the surface roughness like the height of the roughness and correlation length. You might want to see if you can find any publications about similar fibers for more information.

Since the structure has dielectric materials only with refractive indices of 1.33 and 1.45 I wouldn’t expect any absorption loss, so the coupling loss between the beam and fiber and scattering loss should be the only sources of loss.

Hopefully this helps!



Thanks for your help. So using the FDE solver…i can simulate the coupling efficiency. But what would be the source power…i expect that higher diameter fiber will have greater coupling efficiency.

Again i want to know the input power and output power of this structure. I thought i can use port in two end using EME solver and figure out the S21 or transmission coeffecient and S11 or reflection coefficient. But i can’t find any good tutorial for this. If you can share me any then it would be best.

I am going through the links right now…surface roughness can wait …at this moment…my priority is to simulate the structure with fixed cladding radius but sweeping the core radius from 50 micron to 0 micron and see the effect on the S11 and S21 response.

Also can i use a variable name in the property dialog box when designing a fiber so that i can sweep that variable in the variable diaglog box. I read the sweeping parameter tutorial but not quite sure that i understood.

Thanks and Regards



The power coupling result from the FDE solver gives the fraction of the power which can couple from the beam to the fiber mode, so in order to get the power into the fiber you would multiply the power coupling by the power of the laser source that you would use in experiment.

You could also use the EME solver with two ports to calculate the S-parameters as you mentioned. To do this, you could set up 3 cell group regions with the fiber in the center cell and air in the cells on either side. You can then specify the Gaussian beam profile as the input and output port modes. The method for loading a custom field profile into a port can be found here:

If you wanted to use the same Gaussian beam profile generated using the FDE solver’s overlap analysis tool, it’s possible to save the generated beam profile data into the necessary .mat file format using code like the following:

# get data from d-card
x = getdata("gaussian1","x");
y = getdata("gaussian1","y");
z = getdata("gaussian1","z");
Ex = getdata("gaussian1","Ex");
Ey = getdata("gaussian1","Ey");
Ez = getdata("gaussian1","Ez");
Hx = getdata("gaussian1","Hx");
Hy = getdata("gaussian1","Hy");
Hz = getdata("gaussian1","Hz");

# create dataset
EM = rectilineardataset("EM fields",x,y,z);

# save dataset to .mat file which can be imported

After running the simulation and propagating the fields, you can get the S-parameters from the user s-matrix result of the EME solver.

Let me know if this sounds like what you’d like to simulate, and if you have further questions about the setup. It might look somewhat similar to the following example:

For the parameter sweep, you might look into how the parameter sweep is set up in this example:

In the example, the geometry is varied and the user s-matrix is collected. Then from the script the desired S-parameter elements are extracted from the user s-matrix.