Reduced Basis Components

Akselos technology is unique: it merges a component-based approach (substructuring/superelements) with the Reduced Basis (RB) method. This component-based approach allows users to efficiently deal with industry workflows by simplifying the design, meshing and solving steps. It provides these features in a modular way, and with extraordinary computational speedup. This means that you no longer have to simulate your models piece-by-piece; you can now quickly and easily simulate complete models in full detail. The component-based Reduced Basis technology that makes this all possible was recently developed at MIT. Akselos has an exclusive license for this technology.

Efficient port representation

The connection between components is achieved using efficient basis of functions that provide both fast and accurate descriptions of the port functional variability.

FEA as the underlying technology

You retain the tried and true benefits of the FEA technology that has been used in simulation software for decades. RB technology adds an extensive layer of mathematics and computational algorithms on top of FEA. The Reduced Basis solutions are then obtained by reusing pre-computed FEA data for each component.

Acceleration & accuracy

How can we accelerate computations so much and yet stay very accurate? Because we have a big data footprint in the background. Our algorithms start from pre-computed components with a large data footprint due to the many precomputed FEA solutions. We've done the heavy-duty work upfront and stored it all on the cloud so you don't have to worry about speed or accuracy.

Built on top of state-of-the-art open source supercomputing software

PETSc, libMesh. PETSc provides some of the best-in-class parallel linear algebra solvers. libMesh provides a very efficient and scalable C++ framework for FEA that interfaces with PETSc, and is tried and tested for extremely complex and large-scale simulation challenges.