Customization

Implementing new features

Although vanilla CRATE provides a fairly reasonable set of essential features (see Available features), it is expected that the interested user may want to implement specific features seeking a particular application or implement new features in the context of research and development.

Given that CRATE is implemented in Python and following an object-oriented programming (OOP) paradigm, several interfaces emerged naturally during the computational design process and are now available to easily accommodate new implementations. An interface is intrinsically associated with abstraction and polymorphism in OOP, i.e., it is fundamentally a blueprint to implement new classes that respect a given contract and that can be dynamically treated in a unified sense. This contract establishes that certain attributes and/or methods must be defined by new classes according to their specific type and uniqueness.

Of course that the implementation of new features is not limited to the available CRATE interfaces, and other minor improvements and/or major developments are also possible according with the envisaged research/application goals.

Note

Implementing new features in CRATE requires a medium level of expertise in Python programming under an object-oriented programming approach. There is a vast body of literature on such topic (e.g., Phillips (2018), Lott (2019)) as well as open-source web resources (e.g., Real Python, Refactoring Guru).

Available interfaces

Below is a list of some CRATE available interfaces alongside a brief description and a hyperlink to the corresponding documentation page:

  • Interface: Direct Numerical Simulation solver - Implement a direct numerical simulation (DNS) multi-scale method solver to perform the required offline-stage simulations (interface documentation);

  • Interface: Clustering feature - Implement a clustering feature required to perform the RVE cluster analysis (interface documentation);

  • Interface: Clustering algorithm - Implement a clustering algorithm to perform the RVE cluster analysis (interface documentation);

  • Interface: Constitutive model - Implement a constitutive model to describe the physical behavior of a given material phase (interface documentation).

Other implementations

Below is a description of possible implementations that can be regarded as a minor enrichment of existent CRATE toolkits: