Research Papers: Design Automation

An Inductive Design Exploration Method for Robust Multiscale Materials Design

[+] Author and Article Information
Haejin Choi

The School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore, 629798 Singapore

David L. McDowell

The G. W. Woodruff School of Mechanical Engineering, Georgia Tech, Atlanta, GA 30332

Janet K. Allen, Farrokh Mistree

Systems Realization Laboratory, The G. W. Woodruff School of Mechanical Engineering, Georgia Tech, Savannah, GA 31407

David Rosen

Systems Realization Laboratory, The G. W. Woodruff School of Mechanical Engineering, Georgia Tech, Atlanta, GA 30332

Design variables that may be introduced in the middle of a model chain (e.g., design variables that are inputs to model “g” but outputs from neither f1 nor f2 in Fig. 3)

The parameter that is an input to a model as well as an output from another model.

J. Mech. Des 130(3), 031402 (Feb 05, 2008) (13 pages) doi:10.1115/1.2829860 History: Received June 24, 2006; Revised July 27, 2007; Published February 05, 2008

Synthesis of hierarchical materials and products is an emerging systems design paradigm, which includes multiscale (quantum to continuum level) material simulation and product analysis models, uncertainty in the models, and the propagation of this uncertainty through the model chain. In order to support integrated multiscale materials and product design under uncertainty, we propose an inductive design exploration method (IDEM) in this paper. In IDEM, feasible ranged sets of specifications are found in a step-by-step, top-down (inductive) manner. In this method, a designer identifies feasible ranges for the interconnecting variables between the final two models in a model chain. Once feasible ranges of interconnecting variables between these two models are found, then, using this information, feasible ranges of interconnecting variables between the next to the last model and the model immediately preceding it are found. This process is continued until feasible ranged values for the input variables for the first model in the model chain are found. In IDEM, ranged sets of design specifications, instead of an optimal point solution, are identified for each segment of a multilevel design process. Hence, computer intensive calculations can be easily parallelized since the process of uncertainty analysis is decoupled from the design exploration process in IDEM. The method is demonstrated with the example of designing multifunctional energetic structural materials based on a chain of microscale and continuum level simulation models.

Copyright © 2008 by American Society of Mechanical Engineers
Topics: Design , Uncertainty
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 10

The feasible range obtained in Tignit and x3 space using DCE

Grahic Jump Location
Figure 11

The feasible range obtained in design space with HḎEMIτignit>=1 at x2=0.2E−3mm

Grahic Jump Location
Figure 12

Reducing the feasible region by increasing the required HḎEMI for Tignit: (a) 0.9, (b) 1.0, (c) 1.1, and (d) 1.2

Grahic Jump Location
Figure 13

Probability plot of Monte Carlo simulation results at the best IDEM and robust optimum

Grahic Jump Location
Figure 1

An example of information flow in a model chain

Grahic Jump Location
Figure 2

All-in-one versus multilevel system boundary

Grahic Jump Location
Figure 3

Solution search procedure for multilevel robust design

Grahic Jump Location
Figure 4

An example of mapping between input and output spaces

Grahic Jump Location
Figure 5

An example of function evaluation with multiple inputs and outputs with shared variables

Grahic Jump Location
Figure 6

HḎEMI calculation in a direction

Grahic Jump Location
Figure 7

Generation of exact constraint boundary points

Grahic Jump Location
Figure 8

Multiscale MESM model chain

Grahic Jump Location
Figure 9

Multiscale robust MESM design based on IDEM



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In