FAQ

The below FAQ was written for the original public release of the software in 2004. I have updated it in parts.

Q1: What are the CUDOS MOF Utilities for?

The CUDOS MOF Utilities are a software package to compute and analyze the modes of microstructured optical fibres, also called photonic crystal fibres. The CUDOS MOF Utilities can compute MOF modes, their dispersion curves, their field distribution, their structural losses, their effective area, and perform other tasks useful for the studies of MOF modes, such as computing a mode's Bloch transform or Wijngaard test. The CUDOS MOF Utilities use the multipole method to compute the modes. You can find further details on the multipole method in the references.

Q2: How easy are the CUDOS MOF Utilities to use?

Simple answer: Understanding how to write a parameter file to compute the fundamental mode of a simple MOF structure and how to visualize the modes with WinField shouldn't take you more than a few minutes.

Extended answer: The macro language used for parameter files is very simple and easy to read. For simple tasks, such as finding the fundamental mode of solid core MOFs, a lot of settings are automatic, so that you don't need to fiddle around with countless obscure parameters before you get accurate results. If you intend to use the software to characterize higher order modes, modes of large structures (say more than 6 rings of holes), modes of hollow core fibres or modes of MOFs with unconventional geometries (eg highly birefringent fibres or asymetric structures) things will get more complicated. You will find advice on how to tackle some situations in the User's Guide, but at that stage you will probably also want to know more about the multipole method. You will find appropriate references below. WinField, the windows based modal field explorer, is a dialog based application and is intuitive to use. It can provide field plots of publishable quality, and if you know how to use a mouse you basically know how to use it. To understand what the more advanced options (appearing after clicking on the "more options" button) do, you will however need to know a little bit about the multipole method. See the appropriate references.

Q3: How fast are the CUDOS MOF Utilities?

Simple answer: Finding the fundamental mode of a six-hole solid core MOF from scratch (letting the software select all appropriate parameters) takes less than one second on a Pentium 4 (at 2GHz). For a larger MOF, with 4 rings of holes, the same simulation will take about 24 seconds. Computing the dispersion curve for a usual MOF with three rings of holes between 1 and 2 micrometers takes about 10 minutes. Computing the dispersion curves for very large structures (8 rings and more) will take at least several hours.

Extended answer: The time consuming step in the multipole method is to compute the determinant and/or compute the eigenvalues of the system's matrix. The size of the matrix scales as the number of inclusions times the number of multipoles, and the determinant/eigenvalue algorithm has a complexity scaling as the cube of the matrix size. Using symmetry properties of the structure reduces the size of the matrix. The multipole method is hence very efficient for symmetric MOFs with a few rings of holes, but can become slow as the number of inclusions is increased; other methods may then be more suitable. Another factor influencing speed of computation is the number of times the determinant has to be computed before the root-finding algorithm converges to a solution. The most time consuming step here is to compute the initial determinant map. It helps a lot to have a relatively good idea of the value of the mode's effective index to narrow down the region where the mode has to be sought. The CUDOS MOF Utilities can take advantage of asymptotic expansions and heuristic formulas to "guess" the effective index of the fundamental mode for common structures.

Q4: Do I need a supercomputer?

No. Any reasonably recent PC has enough power to run the CUDOS MOF Utilities. See also hardware/software requirements.

Q5: I only have access to Linux/Unix/Macintosh based computers. Can I run the CUDOS MOF Utilities?

No, unfortunately, unless you can emulate Windows on your system. Currently, only executable files for Microsoft Windows are available for download.

Q6: Where can I find the source code for the CUDOS MOF Utilities?

Unfortunately, because of IP issues, at the moment you can't. However some specific cases I might be able to provide it, in particular in the case of collaborations.

==Q7: Can I use CUDOS MOF Utilities for commercial research/development?

No. The CUDOS MOF Utilities have been developed by academic researchers for public research and educational purposes. Commercial simulation tools for MOFs are available and should be used instead. See also the licence agreement. However, private companies can use the software freely for research (not development) that will be made public through publications. If in doubt contact me.

Q8: What kind of geometries can the CUDOS MOF Utilities deal with?

This first public release of the CUDOS MOF Utilities can deal with inclusions with circular cross-sections in a homogeneous, isotropic matrix surrounded by an optional cladding and jacket. Elliptic holes, hexagonal cores, or other non-circular inclusions can not be simulated. Another requirement is that holes do not intersect or touch each other. This may sound a bit restrictive knowing that state of the art hollow core photonic crystal fibres often have honeycomb structures and cores which are far from being circular. There is, however, still a lot of research to be done exploring all possible configurations of MOFs with circular inclusions.

More recent versions can simulate elliptic inclusions, as well as inclusions made out of an arbitrary number of concentric layers of materials (eg coated inclusions, or graded index inclusions).

Q9: Can the CUDOS MOF Utilities take into account material dispersion?

Yes, but the public version only does so for silica. You can define any other (real or complex) refractive index for the inclusions, the matrix, the cladding and the jacket, but you can not define them as a function of wavelength.

In more recent versions materials can be described using a large number of expansions, including Sellmeier and Cauchy expansions and polynomials or simple resonator model based expansions for metals.

Q10: Can I publish results obtained using the CUDOS MOF Utilities?

Yes, please! Our aim in making the Utilities freely downloadable is to further MOF research and to avoid that the task of implementing the multipole method be repeated more than is necessary. For the sake of scientific accuracy, and out of fairness with regard to the authors, we require however that you acknowledge the fact that you have used the CUDOS MOF Utilities. See the page on citing the software for details (as well as the license agreement for the public version).

Q11: Who wrote the CUDOS MOF Utilities?

Thanks for asking. The software as it is was written by Boris Kuhlmey. The initial draft of software which has evolved into the CUDOS MOF Utilities was written during his PhD studies under the supervision of Gilles Renversez, Ross McPhedran, Daniel Maystre, and de facto also C. Martijn de Sterke. Tom White wrote a similar piece of software almost simultaneously as part of his honours project, and although the current version of the software doesn't include code from Tom White, it has benefited from his pioneering work. After his Phd Boris Kuhlmey joined CUDOS, and rewrote a substantial part of the code to make it more stable and more user-friendly with the aim of making it publicly available. Prof. Ross McPhedran and Prof. C. Martijn de Sterke are also with CUDOS, and Dr. Gilles Renversez and Dr. Daniel Maystre are with the Institut Fresnel, UMR 6133, Marseille, France. Tom White was at the timefinishing his PhD with CUDOS/the University of Sydney, and is now with the Austrlian National University, Canberra. Boris Kuhlmey would also like to acknowledge the support of the system administrators of both the School of Physics/University of Sydney (Dr. Sebastian Juraszek, Dr. Tony Monger and George Shan) and the Institut Fresnel (Frederic Forestier).