The AMfe Documentation
AMfe is built in order to provide a simple and lean, yet powerful finite element framework for research purpose.
- Tutorials (Getting started Guide) are a good starting point to learn how to use AMfe. Start here if you are new to AMfe.
- Fundamentals (Topic guides) explains different parts and concepts of AMfe. It is the heart of documentation and mostly intended for users that are familiar with basics of AMfe or users that have already done the Tutorials
- Examples shows some examples for different tasks. This part of documentation can be used if is interested in how to solve specific problem. For many problems an example can be found here that can be altered for own needs.
- Reference is the API documentation of the whole package.
The idea of how AMfe is used is quite different to classic finite element programs. In classic finite element programs the process of analyses is strictly divided in three steps: Preprocessing, Solving, Postprocessing.
In AMfe you have to do these three tasks, too, but you are not limited to do them strictly in that order. While classic finite element programs need an input file which is passed to the solver, AMfe is interpreted. The advantage of this structure is that you are very flexible during the simulation. Assume you have done a transient analysis. But afterwards you realize that you need some further time steps to simulate. Then you only have to call the solver to solve the system for these timesteps. There is no new mesh generation or preallocation of global vectors needed. For linear systems even a new assembly is not needed. Classic programs do not allow this or have a very difficult API to achieve the same to use old simulation data for a new solution.