For Integrated Realization of Engineered Materials and Products Pune, India. Session 8 Integration of Materials and Product Design

Size: px

Start display at page:

Download "For Integrated Realization of Engineered Materials and Products Pune, India. Session 8 Integration of Materials and Product Design"

Anabel Merry Neal
6 years ago
Views:

1 For Integrated Realization of Engineered Materials and Products Pune, India Session 8 Integration of Materials and Product Design Farrokh Mistree farrokh.mistree@ou.edu With contributions from Janet K. Allen and Jitesh Panchal janet.allen@ou.edu, panchal@purdue.edu December 20,

2 Way Forward The Integrated Realization of Engineered Materials and Products should be treated as a complex system accepting that the models are incomplete and perhaps inaccurate Platform embodies domain independent method that facilitates Managing uncertainty technical space Managing complexity process space Exploring the solution space 2

3 SESSION 8 Integration of Materials and Product Design Q1. What are the foundational principles that facilitate rapid and robust concept exploration when the models are incomplete and possibly inaccurate (particularly during the early stages of design)? Focus Q2. How do we invert the multiscale material models in a computationally efficient manner so that performance requirements can drive material and process design? Vedantam Q3. What is/are the level(s) of accuracy needed for reduced order models in various stages of simulation? What level of fidelity of simulation models is necessary from a design standpoint? Vedantam Q4. How should a designer allocate limited resources to modeling and design exploration activities while ensuring a good design outcome? Panchal Q5. Which couplings between design decisions are important? Which decisions/models can be decoupled without affecting the overall system performance? Panchal Q6. How can we leverage advanced manufacturing methods in design? Gautham Q7. How can we handle multi-material systems in design? Vedantam 3

4 Q1. What are the foundational principles that facilitate rapid and robust concept exploration when the models are incomplete and possibly inaccurate (particularly during the early stages of design)? 1. Foundational problem - characteristics a. Complex systems emergent properties. b. Models are incomplete and possibly inaccurate b. Proof of concept demonstrate design functionalities to industry 2. Foundational principles when models are incomplete and inaccurate a. Method i. Manage uncertainty. Manage complexity. ii. Robust design to facilitate the exploration of the design space and answer what if questions. iii. Model based design + templates+ Integration of templates. IDEM. v. Exploration of the solution space. b. Reduce cost of computation i. Process simplification. Q4 and Q5. Jitesh Panchal. ii. Surrogate models. Q3. Janet K. Allen. c. Platform i. Designers, engineers and scientists. BP. Gautham. ii. Modular, reusable, expandable. Janet K. Allen. 4

5 Integrated Materials & Products Design Integrated Realization of Engineered Materials and Products (IREMP) Objective: Prediction of properties from structure Focus: Accuracy in behavior prediction Challenge: Linking phenomena at different length and time scales Interface of Design and Engineering Sciences 2013 Indo-US Workshop Objective: Achievement of system-level objectives Focus: Efficiency in exploring design options Challenge: Joint exploration of the product, process, and materials design spaces Domain Specific Multiscale Modeling Materials Scientists Domain Independent Systems-based Multilevel Design System Designers 5 Panchal, J.H., H.-J. Choi, J.K. Allen, D.L. McDowell, and F. Mistree, 2005, Robust Multiscale Simulation-based Design of Multifunctional Materials, International Conference on Advanced Materials Design & Development (ICAMDD-05), Goa, India. School of Mechanical Engineering Purdue University 5

IREMP Foundational Problem Kumar P., Goyal S., Singh A. K., Allen J. K., Panchal J.

, 2013, PREMAP Exploring the Design Space for Continuous Casting of Steel,

Continuous cast steel Automotive Gear In progress at TCS / OU / Purdue

6 IREMP Foundational Problem Kumar P., Goyal S., Singh A. K., Allen J. K., Panchal J. H. and Mistree F., 2013, PREMAP Exploring the Design Space for Continuous Casting of Steel, ICORD 13, IIT Madras, Chennai, India. Continuous cast steel Automotive Gear In progress at TCS / OU / Purdue Kulkarni, N., Zagade, P.R., Gautham, B.P., Panchal, J.H., Allen, J.K. and Mistree, F., 2013, PREMΛP Exploring the Design and Materials Space for Gears, ICORD 13, IIT Madras, Chennai, India. 6

7 Component Manufacture Materials Design IREMP - Foundational Problem Mill Product From BP Gautham s presentation Rolling Slug Hot Forming Machining Shot Peening Heat Treatment Component Design TCS Public 7

8 Value Proposition Realization of engineering value across the chain of composition to design of component Design of the gear Design of the material Design of the manufacturing process for accelerated realization and cost reduction Addressing issues such as Segregation vs. forging distortion Inclusions vs. fatigue life Compactness vs. durability Shot peening vs. heat treatment Key Challenges Managing uncertainty Managing complexity 8

9 What is a complex system? A complex system is composed of interconnected parts that as a whole exhibit one or more properties (behavior among the possible properties) not obvious from the properties of the individual parts. A system s complexity may be of one of two forms: disorganized complexity and organized complexity. In essence, disorganized complexity is a matter of a very large number of parts, and organized complexity is a matter of the subject system (quite possibly with only a limited number of parts) exhibiting emergent properties. Emergent properties add complexity that necessitates a change in design philosophy. 9

10 10 10

11 Managing Uncertainty and Complexity Managing Uncertainty Imprecise control of process parameters Incomplete knowledge of phenomena Incomplete models & information aggregation Need to explore alternatives Robust Design via Inductive Design Exploration Method (IDEM) 1. Solution (multi-domain) Space Exploration 2. Uncertainty Management Managing Complexity Precision vs. cost of computation Degree of interdependencies Distribution of resource allocation Value of Information Ramifications McDowell, D. L., Panchal, J. H., Choi, H.-J., Seepersad, C. C., Allen, J. K., and Mistree, F., 2009, Integrated Design of Multiscale, Multifunctional Materials and Products, Elsevier. ISBN: Choi, H-J, and Allen, J.K., 2009, A Metamodeling Approach for Uncertainty Analysis of Nondeterministic Systems, Journal of Mechanical Design, vol. 131,

12 Types of Robust Design In Type I robust design, design variable values are identified to satisfy a set of performance requirement targets regardless of noise factors. Noise factors are not under a designer s control. In Type II robust design, design variable values are determined that satisfy a set of performance requirement targets regardless of anticipated variations in those design variables. (Includes topology design) In Type III robust design, design variable values are determined which satisfy a set of performance requirements regardless of variations in the mathematical models used to describe that performance. In Type IV (Combination of Types I, II, and III) robust design, design variable values are determined which satisfy a set of performance requirements in spite of variability introduced by a hierarchical, multiscale or multidisciplinary formulation of the product, 12

13 Modeling Uncertainty in the Compromise Decision Support Problem (what can be uncertain?) Given Find Input Parameters Noise Knowledge of Physics, Standards, etc. System Variables Deviation Variables Satisfy Constraints Capability Demand Goals Performance Target Bounds Minimize Archimedian Formulation Z = W 1 (d 1+, d 1- ) W n (d n+, d n- )

14 Q5 The Robust Concept Exploration Method (RCEM) Infrastructure Compromise DSP Overall Design Requirements A. Factors and Ranges Noise z Control Factors Factors Product/ Response x Process y B. Point Generator Design of Experiments Plackett-Burman Full Factorial Design Taguchi Orthogonal Array Central Composite Design Etc. F. The Compromise DSP Find Control Variables Satisfy Constraints Goals Mean on Target Minimize Deviation Bounds Minimize Deviation Function C. Simulation Programs (Rigorous Analysis Tools) D. Experiments Analyzer Eliminate unimportant factors Reduce the design space to the region of interest Plan additional experiments E. Response Surface Model y Robust, Top-Level Design Specifications x z 2 x 1 y f ( x, z) f( x, ) y 2 2 k l 2 f 2 f 2 y z i xi i1 zi i1 xi Input and Output Processor Simulation Program z 14

Design and Material Space Exploration for Gear Reliability: Important from user s safety S1: Highly reliable S4: Reliable and Compact S6: Reliable and Low cost S7: Reliable, compact and

15 Design and Material Space Exploration for Gear Reliability: Important from user s safety S1: Highly reliable S4: Reliable and Compact S6: Reliable and Low cost S7: Reliable, compact and low cost S2: Compact S3: Low cost Compact: Occupies less volume, less weight, low dynamic forces and vibrations S5: Compact and low cost Cost: Low cost important from market competition 15

Design and Material Space Exploration for Gear Maximize compactness (minimize center distance) Minimize cost Maximize reliability Feasible

16 Design and Material Space Exploration for Gear Maximize compactness (minimize center distance) Minimize cost Maximize reliability Feasible designs collated and ternary plots created Center region shows area of compromise These plots can be used to make initial decision on the weights 16

17 An Overview of a Robust Concept/Topology Exploration Method A. The Compromise DSP Overall Design Requirements Product/Process Information and Knowledge from Other Design Phases Find Control Variables Satisfy Constraints Goals Mean on Target Minimize Deviation Bounds Minimize Deviation Function Robust Design Specifications Product/Process Information and Knowledge to Other Design Phases C. Topology Representation and Modification D. Variability Assessment E. Search Techniques/ Algorithms Simulation Infrastructure B. Characterize Factors & Ranges Noise Factors Control Factors x z Product/ Process Responses y F. Analysis Programs Input and Output Processor Analysis Program Today: Managing Uncertainly by J.K. Allen 17

18 Hierarchical Mapping of Design Space Exploration 18

19 Uncertainty Management for Integrated Design 19

20 Multiscale System for AUV Sinha A, Panchal, J.H., Allen J.K. and Mistree F., 2013, Microstructure Mediated Materials Design and Product, International Conference on Integrated Computational Materials Engineering, Salt Lake City. 20

21 IDEM Inductive Design Exploration Method Ov er verall Design equirements 1.Framing the Multiscale Problem Define Design Space Hierarchical Models Independent Parameters Dependent Parameters Goals Characterize Reducible Uncertainty Irreducible Uncertainty 3.Inductive Design Exploration Method Step 1: Discretize Design Spaces in Hierarchical Domains Step 2: Conduct Parallel Discrete Function Evaluation Step 3: Inductive Discrete Constraints Evaluation (IDCE) 2.HIERACHIAL MATERIAL MODELING AND ANALYSIS 4.Compromise Decision Support Problem Given Targets Weights Find System Variables Weights Satisfy Goals Constraints Bounds Minimize: Deviation Function Robust Process Specifications Design of Experiments Identify Assumptions Range of Operation Input Factors Output Factors Simulation Design Central Composite Latin Hypercube Orthogonal Arrays Simulation Develop Simulation Infrastructure Perform Simulation Extract Data Analyze Data Integrated Metamodel and Prediction Interval Estimate Metamodel Transformation Eliminate unimportant factors Response Surface Modeling Interval Estimates Confidence Intervals Taylor Series 21

22 Results using IDEM Module 3 Module 2 Module 1 Precipitate Size Module 1 Grain Size 22

23 Way Forward Question 1 1. Foundational problem - characteristics a. Complex systems emergent properties. b. Models are incomplete and possibly inaccurate b. Proof of concept demonstrate design functionalities to industry 2. Foundational principles when models are incomplete and inaccurate a. Method i. Manage uncertainty. Manage complexity. ii. Robust design to facilitate the exploration of the design space and answer what if questions. iii. Model based design + templates+ Integration of templates. v. Exploration of the solution space. b. Reduce cost of computation i. Process simplification. Q4 and Q5. Jitesh Panchal ii. Surrogate models. Q3. Janet K. Allen c. Platform i. Designers, engineers and scientists. BP Gautham ii. Modular, reusable, expandable. Janet K. Allen 23

24 Way Forward The Integrated Realization of Engineered Materials and Products should be treated as a complex system accepting that the models are incomplete and perhaps inaccurate Platform embodies domain independent method that facilitates Managing uncertainty technical space Managing complexity process space Exploring the solution space 24

25 Acknowledgments We gratefully acknowledge Colleagues at TCS especially BP Gautham, Pradip, Amarendra Singh, Nagesh Kulkarni, Pramod Zagade and Sharad Goyal. Our former students especially Carolyn Seepersad, Haejin Choi, Marco Fernandez, Bert Bras and Warren Smith. NSF EAGER Grants and

26 Key Publications McDowell D.L., Panchal, J. H., Choi, H.-J., Seepersad C. C., Allen J. K. and Mistree F., Integrated Design of Multiscale Materials and Products, Elsevier, New York, (2010). ISBN-13: Choi, H-J., McDowell, D.L., Allen, J.K., and Mistree, F., 2008, An Inductive Design Exploration Method for Hierarchical Systems Design Under Uncertainty, Engineering Optimization, vol. 40., no. 4, pp Messer, M., Panchal, J.H., Krishnamurthy, Klein, B., Yoder, P.D., Allen, J.K. and Mistree, F., 2010, Evaluating and Selecting Embodiment Design Processes Using a Process Performance Indicator for Applications in Integrated Product and Materials Design, ASME Journal of Mechanical Design, vol. 132, Choi H-J., McDowell D. L., Rosen D., Allen J. K. and Mistree F., 2008, An Inductive Design Exploration Method for Robust Multiscale Materials Design, ASME Journal of Mechanical Design, vol. 130, no. 3, pp /13. Panchal, J. H., Choi, H.-J., Allen, J. K., McDowell, D. L. and Mistree, F., 2007, A Systems Based Approach for Integrated Design of Materials, Products & Design Process Chains, Journal of Computer-Aided Materials Design, vol. 14, n supplement 1., pp Chen, W., Allen, J.K, Tsui, K-L and Mistree, F., "A Procedure for Robust Design: Minimizing Variations Caused by Noise Factors and Control Factors," ASME Journal of Mechanical Design, vol. 118, no. 4, pp ,

Reduction Of Cost Of Quality By Using Robust Design: A Research Methodology

Reduction Of Cost Of Quality By Using Robust Design: A Research Methodology 1 Jalal Khan, 2 S.N. Teli, 3 B.P. Hada Assistant Professor, SCOE, Kharghar, India Abstract: CoQ and Taguchi Design are to be