Composition of Multi-disciplinary, Multi-scale Models for the Design of Fluid-Power Systems

Advisor

Chris Paredis

Other students working on the project

Tommy Johnson, Jonathan Jobe, Alex Kerzhner

Desired research interests

Modeling and simulation,  design, knowledge modeling

Desired (but not absolutely required) skills

Information and knowledge modeling, Abstract logical thinking, Computing (Java), UML or SysML.

Project description

Modeling and simulation have become essential components of engineering design.  In comparison to physical prototyping, they allow designers to predict the outcomes of their design decisions quickly and inexpensively.  However, developing a model that is well-suited for supporting a particular design decision is still a cumbersome task, especially when the design decision requires multi-disciplinary analyses at different levels of abstraction. Rather than developing such complex models from scratch every time, substantial gains in efficiency and applicability are expected from re-using models that have been stored in model repositories.  The key to such reuse, is a formal representation of the models and the relationships between models such that one can quickly retrieve and compose simple, single-discipline, single-formalism, single-component models into complex, multi-discipline, multi-formalism, system-level models.  The research challenges are:

• to identify the relationships between models of different system components, models in different disciplines, and models at different levels of abstraction; 
• to formally represent the models and their relationships; and
• to develop algorithms for composing such models into multi-disciplinary, multi-scale, system-level models.

In solving these research challenges, we will build on past work in the area of object-oriented systems modeling, and on design-analysis integration for finite-element modeling.  Specifically, we will further develop our current work on model representation and integration in the Systems Modeling Language (OMG SysML^TM: http://www.omgsysml.org).

The modeling and design framework resulting from this research will be demonstrated in the context of compact and efficient fluid-power systems, specifically, the redesign of the hydraulic system of a excavator or backhoe.  This hydraulic system will integrate several new fluid-power technologies that are under development within the NSF engineering research center on compact and efficient fluid power.