2. Strategy

As a third party library, the Library of Advanced Materials for Engineering (LAMÉ) is renewing a commitment to accurate, robust, and efficient constitutive modeling for solid mechanics. There are three distinct groups that require services from LAMÉ: constitutive model developers, application code developers, and analysts. Each group has different, but related, requirements on LAMÉ.

Constitutive modelers who develop the mathematical models describing the response of interest and the associated numerical methodologies require a framework for developing and implementing these models. This framework must be well documented so that these developers can easily develop and implement a model that can, when the model is sufficiently robust, be used reliably in production calculations.

Application code developers, on the other hand, require that the library of constitutive models share a common, simple interface. This requires that the conceptual division between a constitutive model and the application code be well understood. The application codes also require verified behavior of the models along with certain performance requirements as these models can be called billions of times in an analysis.

Analysts require constitutive models that are both verified and well documented. The responses of interest can very greatly depending on the material utilized and the conditions under which it is loaded. This accounts for the wide range and sheer volume of constitutive models in the literature. Furthermore, there may be nothing that affects the results of an analysis more then the constitutive model. Therefore, the analyst needs a thorough knowledge of the behavior of the models along with how to use it in an application code. Such an understanding is also essential for the accurate determination and calibration of different material and model parameters. In some cases, even subtle changes in a material specimen’s history can lead to large variation in properties and responses. As such, a clear description and understanding of this input data is essential for appropriate utilization of different models.

Due to the varied requirements on LAMÉ, a strategy for supporting these various user groups has been developed. This strategy is described here.