CloudCEM

Electromagnetic Modelling (EM) is found in many industrial and scientific organisation and is used in various fields such as aeroplane, satellite, ship and car design. Due to the complexity of the structure analysed, EM software requires enormous computer resources, both in terms of memory and computation, resulting in long lead times for developing simulations, lead times that are not compatible with the time available in the industry for design and analysis.

The Coud Fasting Scalable EM Modelling (CloudCEM) project intended to take advantage of the benefits resulting from the adoption of a Cloud Computing model for solving EM problems so as to reduce the time necessary for processing data, limiting the cost and optimising computation resoureces.

The project intended to create a Cloud Computing infrastructure that permits the rapid and cost-effective creation of simulations of electromgnetic effects on very complex structure. The model used is a kind of hybrid in order to take advantage of the potential of both public and private Cloud Computing.

Cloud infrastructure monitoring and metering tools have been examined and implemented, tools that allowed us to know exactly what proportion of computational resources are used. The most interesting aspect concered the monitoring of application and computing resources that work on different (private - public) Clouds.

An algorithm that is both innovative and efficient from a computational point of view was implemented for the solution of Marxwell equations, a Computer Aided Engineering (CAE) platform based on innovative algorithms for pre-and post-processing (meshing, visualitation) was created to prepare the geometrical structures to be analysed, and finally a Web Graphic User Interface (GUI), both intuitive and easy to use, was developed that allows even inexprienced users to process their models in Cloud Computing easily and automatically, without the need fot IT or electromagnetic expertise.

Advanced Computing and Electromagnetics