GAFFE A toolbox for solving evolutionary nonlinear PDEs

This toolbox implements the well known split-step Fourier technique for solving nonlinear PDEs.

Description

The Generalised Adaptive Fast-Fourier Evolver (GAFFE) toolbox is a framework that greatly simplifies the solution of complex partial differential equations (PDEs) in an adaptive manner.

By default both the step-size and the mesh adapt to the problem at hand to optimise the speed of execution for a given nominal target accuracy.

The technique is N dimensional and can therefore be used to model diverse problems such as temporal solitons, spatial self-focusing or exotic space-time effects.

Currently the technique is limited to scalar fields.

• MATLAB release: MATLAB 7.9 (2009b)

[Alexandro Ruiz]

This is a very efficient toolbox. I have been using it to propagate a laser inside a dielectric material using the nonlinear schrodinger equation. It usually takes just a few tens of seconds to complete the propagation.

[Alexander]

Simply the best toolbox for PDEs

[Mohsin Shah]

Hi you have done a great job. i wana know that is there anything for birefringence or polarization in this toolbox? can we simulate pulse propagation in a birefringent fiber by this toolbox?

[Edward Grace]

Mohsin. "..is there anything for birefringence or polarization in this toolbox?". No, not explicitly.

"can we simulate pulse propagation in a birefringent fiber by this toolbox?"

Like many things the answer is "it depends".

As long as you can approximate whatever it is that you want to do as being scalar then the answer is "yes" -- just use an effective refractive index for each (decomposed) polarisation. And model it as two separate (decoupled) problems.

If on the other hand the important effect that you want to capture is due to off-diagonal terms in the nonlinear permettivity tensor that strongly couple between polarisation states then the answer is a resounding "No!"; it's tough enough generalising to an N dimensional scalar nonlinear problems, let alone vectorial / tensorial ones!

[Dong-sig Shin]

It's exactly what I find toolbox. But I don't know where I should input parameters like pulse energy and duration.

[Edward Grace]

As defined the example model for pulse propagation will work in normalised units. I strongly recommend reading the example alongside the cited section of Agrawal's "Nonlinear Fiber Optics".

[Amdad]

GREAT