Automated computer derivation of behavioral constraints from a generalized mechanical design representation
Abstract
State of the art mechanical design computer systems promote the creation of an intelligent design that can respond to designer inputs. In effect, such a design representation is a declarative computer program that executes when variables or values are entered. The next logical step is to begin to make procedural languages available to the designer so that design exploration can be planned by the designer and executed by the computer. Of the many design tools available, there exists no general-purpose tool to bridge the gap between synthesis and analysis, making it difficult to program an intelligent refinement loop. This thesis presents a tool to perform automatic behavioral constraint derivation for desired performance parameters from a design representation that can be specialized to represent any design through class extension, but that also has common characteristics that allow a derivation procedure to navigate it. The behavioral constraint laws are represented separately from the derivation control and are defined in a specialized law language. The derivation procedure is a recursive application of behavioral relationships to unconstrained variables, ending in full constraint or failure. A working computer system demonstrates the methods with two examples.
Degree
Ph.D.
Advisors
Anderson, Purdue University.
Subject Area
Mechanical engineering|Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.