Chapter 4 Robust Design v9.0 2004 ANSYS, Inc.

Chapter 4 Robust Design v9.0  2004 ANSYS, Inc.

Chapter 4 Robust Design v9.0 2004 ANSYS, Inc. ANSYS, Inc. Proprietary Why Robust Design? Lockheed Martin used to spend an average of 200 workhours trying to get a part that covers the landing gear to fit. For years employees had brainstorming sessions, which resulted in seemingly logical solutions. None worked. The statistical discipline of Six Sigma discovered a part that deviated by one-thousandth of an inch. Now corrected, the company saves $14,000 a jet.1 1: Firms aim for Six Sigma efficiency; [FIRST Edition] Del Jones. USA TODAY. McLean, Va.: Jul 21, 1998. pg. 01.B ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-2 Why Robust Design? It will keep the company (Allied Signal) from having to build an $85 million plant to fill increasing demand for caprolactam used to make nylon, a total savings of $30 $40 million a year.1 Raytheon figures it spends 25% of each sales dollar fixing problems when it operates at four sigma, a lower level of efficiency. But if it raises its quality and efficiency to Six

Sigma, it would reduce spending on fixes to 1%.1 1: Firms aim for Six Sigma efficiency; [FIRST Edition] Del Jones. USA TODAY. McLean, Va.: Jul 21, 1998. pg. 01.B ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-3 Why Robust Design? The reason to do DFSS is ultimately financial. It generates shareholder value based on delivering customer value in the marketplace. Products developed under the discipline and rigor of a DFSS-enabled product development process will generate measurable value against quantitative business goals and customer requirements. DFSS helps fulfill the voice of the business by fulfilling the voice of the customer.2 2: Design for Six Sigma in Technology and Product Development, C.M. Creveling, J. L. Slutsky, and D. Antis, Jr. ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-4

Robust Design Overview Background of Robust Design What is Robust Design, DFSS, ? Design for Quality Robust Design in Engineering Analysis Illustration Example Sources of Uncertainty Effects of Uncertainty Compare Deterministic and Probabilistic Approach Enabling Technologies Demonstration Overview of Application Example Demo Questions ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004

Inventory #002156 4-5 Background of Robust Design What is Robust Design, DFSS, etc.? Robust Design is often synonymous to Design for Six Sigma or Reliability-based Optimization ANSYS, Inc. Proprietary 2004 ANSYS, Inc. Uncertainty Analysis Quantify the effect of uncertainties on the performance of a product (mean value, standard deviation, etc.) Reliability Analysis Quantify the reliability (failure probability, defects per million) Robust Design or Design For Six Sigma (DFSS) Optimize the design such that it is insensitive to unavoidable uncertainties (e.g. material,loads,) Reliability-based Optimization Optimize the design such that reliability is maximized or failure probability (defects per million) is minimized October 1, 2004

9.0 New Features Inventory #002156 4-6 Background of Robust Design Design for Quality Six Sigma Quality = Only 3.4 out of 1000000 parts fail Gaussian Distribution Product is ... Product is ... Bad Good Good Bad Area = Failure Probability USL = Upper Specification Limit LSL -6 -4 LSL = Lower

Specification Limit -2 USL 0 2 4 6 Sigma-Value Six Sigma Quality is inherently a probabilistic statement P.S.: Gaussian distribution is not realistic, but does convey the idea correctly ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-7 Background of Robust Design Design for Quality Six Sigma

Design For Six Sigma = Optimize manufacturing processes such that they automatically produce parts conforming with six sigma quality = Optimize the design such that the parts conform with six sigma quality, quality i.e. quality and reliability are explicit optimization goals Six Sigma 1000 100% 80% 100 60% 10 40% 1 20% 0.1 0% Research Design

Development Production PrototypeTests Product Development Phases ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features Degree of Fredom to affect the Product Lifetime Costs Rel. Cost of Design Change Design For Six Sigma Design for Six Sigma: Achieve Designed-In quality as opposed to letting customers find out about quality problems Make informed decision that are critical to quality early in the development process October 1, 2004 Inventory #002156 4-8 Background of Robust Design Design for Quality

Robust Design is a Paradigm Shift FROM: Reactive Quality Management TO: Predictive Quality Management Extensive Design Rework Assess Performance by build-test-build-test- Fix performance/quality problems after manufacturing Quality is Tested-In Controlled Design Parameters Estimate likelihood/rate of performance problems in design & development Address quality problems in design & development Designed for robust performance and quality Quality is Designed-In ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features

October 1, 2004 Inventory #002156 4-9 Background of Robust Design Robust Design in Engineering Analysis Purpose of Robust Design ANSYS ANSYS Input Input Material Properties Geometry Boundary Conditions Output Output Deformation Stresses / Strains Fatigue, Creep,... Its a reality that input parameters are subjected to scatter => automatically the output parameters are uncertain as well!! ANSYS, Inc. Proprietary 2004 ANSYS, Inc.

9.0 New Features October 1, 2004 Inventory #002156 4-10 Background of Robust Design Robust Design in Engineering Analysis Purpose of Robust Design ANSYS ANSYS DesignXplorer DesignXplorer Questions answered with Robust Design: How large is the scatter of the output parameters? What is the probability that output parameters do not fulfill design criteria (failure probability defects per million)? How much does the scatter of the input parameters contribute to the scatter of the output (sensitivities critical-to-quality)? ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-11

Background of Robust Design Sources of Uncertainty Property SD/Mean % Metallic materiales, yield 15 Carbon fiber composites, rupture 17 Metallic shells, buckling strength 14 Junction by screws, rivet, welding 8 Bond insert, axial load 12 Honeycomb, tension 16 Honeycomb, shear, compression

10 Honeycomb, face wrinkling 8 Launch vehicle , thrust 5 Transient loads 50 Thermal loads 7.5 Deployment shock 10 Acoustic loads 40 Source: Klein, Schueller et.al. Probabilistic Approach to Structural Factors of Safety in Aerospace. Vibration loads Proc. CNES Spacecraft Structures and Mechanical Testing Conf., Paris 1994 ANSYS, Inc. Proprietary

2004 ANSYS, Inc. 9.0 New Features 20 October 1, 2004 Inventory #002156 4-12 Background of Robust Design Effects of Uncertainty 5-100% Thermal Analysis CAD Geometry FEM CFD 0.1-10% Materials, Bound.Cond., ... 5-50% ANSYS, Inc. Proprietary 2004 ANSYS, Inc.

Materials, Bound.Cond., Loads, ... LCF Structural Analysis ??% FEM Materials, Bound.Cond., 5-100% Loads, ... 9.0 New Features Materials 30-60% October 1, 2004 Inventory #002156 4-13 Background of Robust Design Effects of Uncertainty Influence of Youngs Modulus and Thermal Expansion Coefficient on thermal stresses: thermal = E T Deterministic Approach: Mean = EMean Mean T Mean = typically used results Probabilistic Approach:

Probability that ( thermal >= 105% Mean) ( thermal >= 110% Mean) E scatters 5% 16% (~1 out of 6) 2.3% (~1 out of 40) E and scatter 5% 25% (~1 out of 4) 8% (~1 out of 12) E, & T scatter 5% 28% (~1 out of 4) 13% (~1 out of 8) ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-14 Background of Robust Design

Compare Deterministic/Probabilistic Turbine What-If Analysis Series ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-15 Background of Robust Design Enabling Technology: Parameterization Robust Design for all parameters including: APDL Parameters CAD Parameters (Workbench) Paramesh db APDL Parameters Initial mesh Parameter value /syp,parabatch.exe,'testpb.rsx','testpb.cdb','location',%value%,'testpb_mod.cdb' /inp,testpb_mod,cdb ! Input the modified geometry Parameter name

Output mesh Import Output mesh ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 NewParameters Features ParaMesh October 1, 2004 Inventory #002156 4-16 Background of Robust Design Enabling Technology: DesignXplorer DesignXplorer manages the parameters and the uncertainties ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-17 Robust Design Demonstration Overview of Application Example

CAD Geometry ANSYS, Inc. Proprietary 2004 ANSYS, Inc. FEM Mesh 9.0 New Features FEM Boundary Conditions October 1, 2004 Inventory #002156 4-18 Robust Design Demonstration Overview of Application Example Results for Maximum Principal Stress Pressure Side Suction Side Peak Value s Peak Value p Design Variables and Uncertainties Axial Leaning Tang. Leaning

Material Density (Gaussian) Fillet Radius (Lognormal) Dove Tail Width Mass Imbalance: (p s)2 ANSYS, Inc. Proprietary 2004 ANSYS, Inc. Avg.Stress: 0.5(p + s) 9.0 New Features October 1, 2004 Inventory #002156 4-19 Robust Design Demonstration Demonstration Robust Design Demonstration Turbine Blade Workshop ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features

October 1, 2004 Inventory #002156 4-20 DesignXplorer Family v9.0 2004 ANSYS, Inc. ANSYS, Inc. Proprietary DX Family 9.0 New Features DesignXplorer & DesignXplorer VT Robust Design GUI Structure Parameterize DFSS Results Optimize DFSS results New Trade-Off Study Genetic Algorithm for Sample Generation Variational Technology (DesignXplorer VT) ANSYS, Inc. Proprietary 2004 ANSYS, Inc.

Support of Discrete Variables in Workbench RSX File Viewing Additional Contact Support Frequency dependent material properties Support inertial load parameters 2D Analysis 9.0 New Features October 1, 2004 Inventory #002156 4-22 DX Family 9.0 New Features Robust Design - GUI Structure Measures DFSS of robustness Charts and Table Pages CDF Plot Y-Axis can be scaled as Gaussian, Log-Normal, Weibull, Exponential ANSYS, Inc. Proprietary 2004 ANSYS, Inc. Customize Tables 9.0 New Features (add, delete)

Statistics Available Sigma-Level in Tables October 1, 2004 Inventory #002156 4-23 DX Family 9.0 New Features Parameterize DFSS Results New Robust Design View New DFSS Parameters showing up in Parameter View ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-24 DX Family 9.0 New Features Optimize DFSS results

1) 2) 1) Random Variables: Uncontrollable used to obtain DFSS results 2) Design Variables: Useable for Robust Design Optimization ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-25 DX Family 9.0 New Features Optimize DFSS results Same optimization functionality as for GDS ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-26

DX Family 9.0 New Features Trade-Off Studies Conflicting goals lie along Mouse-Over the Pareto Front, Tradeoffs Results Studies occur here ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features Both 2D and 3D Tradeoff Plots are available October 1, 2004 Inventory #002156 4-27 DX Family 9.0 New Features Trade-Off Studies Feasible Infeasible MouseOver Results Pareto Front

for Conflicting Goals ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features Infeasible Points in Pareto Front October 1, 2004 Inventory #002156 4-28 DX Family 9.0 New Features Genetic Algorithm Sample Generation ANSYS, Inc. Proprietary 2004 ANSYS, Inc. Choice of Basic (Screening), which is pseudo-random sampling method typically done first, or Advanced (Genetic) Algorithm which is typically October 1, 2004 Inventory #002156 done 9.0 second New Features 4-29 DX Family 9.0 New Features

Genetic Algorithm Sample Generation Advanced sample options Advanced Samples Screening Samples ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-30 DesignXplorer VT 9.0 Support of Discrete Variables in Workbench Time for a 30 minute base solve 1E+15 1E+13 1E+11 DOE 1E+09 DXVT 1 Day

1E+07 T im e(D a y s ) Variational Technology is much faster than DOE for discrete parameters Supports: 1 Year 100000 1 Decade 1000 1 Century 10 1 Mil ennium 0.1 0.001 0

5 10 15 20 25 30 35 40 45 50 Number of Boolean Parameters ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features

Spot Welds Solid Bodies Sheet Bodies Line Bodies Parts October 1, 2004 Inventory #002156 4-31 DesignXplorer VT 9.0 Support of Discrete Variables in Workbench Efficient Postprocessing Multi-objective Boolean optimizer based on Bayesian sampling faster optimization of discrete parameters Boolean scatter chart representing the solution points of all the parameter combinations of the selected parameters. ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features

October 1, 2004 Inventory #002156 4-32 DesignXplorer VT 9.0 Support of Discrete Variables in Workbench 6 Supports 6 Supports 4 Suppressed None Suppressed ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-33 DesignXplorer VT 9.0 RSX File Viewing Allows user to view Variational Technology results from a variety of sources including

DesignXplorer VT (of course) ANSYS using VT Hoover model solved with VT within the ANSYS Environment ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-34 DesignXplorer VT 9.0 Support for Face-to-Edge Contact ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-35 DesignXplorer VT 9.0 Support for Edge-to-Edge Contact ANSYS, Inc. Proprietary 2004 ANSYS, Inc.

9.0 New Features October 1, 2004 Inventory #002156 4-36 DesignXplorer VT 9.0 Support of Frequency Dependent Properties For harmonic analyses, VT supports frequency dependant modulus and damping Modulus of Elasticity vs. Frequency Frequency Response 3.0E+07 2.8E+07 Modulus of Elasticity 2.6E+07 10 2.4E+07 2.2E+07 2.0E+07 1.8E+07 Variable E 1

1.6E+07 E for 0 Hz 1.2E+07 1.0E+07 0 100 200 300 400 500 600 700 800 Frequency (Hz) Imaginary Modulus Damping Ratio v s. Frequency Imanginary Modulus Damping Ratio 3.5%

Y Displacement 1.4E+07 E for 800 Hz 0.1 0.01 3.0% 0.001 2.5% 2.0% 1.5% 0.0001 0 1.0% 100 200 300 400 500

600 700 800 Frequency (Hz) 0.5% 0.0% 0 100 200 ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 300 400 500 Frequency (Hz) 600 700

800 9.0 New Features October 1, 2004 Inventory #002156 4-37 DesignXplorer VT 9.0 Inertial Load Paramenters & 2D Analysis Support Inertial Load Parameters Acceleration (Magnitude and Components) Rotational Velocity (Magnitude and Components) Rotational Acceleration (Magnitude and Components) Allows Variational Technology analysis of 2D Simulation studies to include: Axisymmetric Plane Strain Plane Stress ANSYS, Inc. Proprietary 2004 ANSYS, Inc. 9.0 New Features October 1, 2004 Inventory #002156 4-38

Recently Viewed Presentations

  • Designing and Implementing Classes Topic 6 Don't know

    Designing and Implementing Classes Topic 6 Don't know

    Designing and Implementing Classes Topic 6 Don't know much geography Don't know much trigonometry Don't know much about algebra Don't know what a slide rule is for
  • EARTHS INTERNAL STRUCTURE And processes What Was Early

    EARTHS INTERNAL STRUCTURE And processes What Was Early

    So far we have broken the layers down by composition and density. (Crust, Mantle, Core) We can also divide them up by how they "behave" The upper most layer is called the lithosphere. It is rigid, meaning it can bend...
  • Painting M. Ryan Academic Decathlon 2005-06 Painting Paint

    Painting M. Ryan Academic Decathlon 2005-06 Painting Paint

    Painting Watercolor: most common water-based paint, transparent. White paint is seldom used to lighten paint (water is used ) Light colors are applied first, then dark, working from background to foreground. Painting Gouache: water-based opaque paint, it has more body...
  • Professional Issues - CASE

    Professional Issues - CASE

    The students critically reflect on their own practice-related knowledge and enhance skills for advancing/extending professional practice. Students need to consider the issues in relation to their specific clinical situation and profession specific content required by professional bodies (CASE or CoR).
  • Mentoring for Success in Academic Libraries

    Mentoring for Success in Academic Libraries

    Mentoring for Success in Academic Libraries OLA Super Conference February 4, 2005 What is Mentoring ? Developmental, caring, sharing and helping relationship Investment of time, know-how and effort For growth, knowledge, skills To meet critical needs To prepare for the...
  • Atenea o Minerva - Junta de Andalucía

    Atenea o Minerva - Junta de Andalucía

    Paris dio la manzana a Afrodita, granjeándose así la fiel protección de la diosa para él y los suyos para siempre, y la enemistad de las otras dos diosas lo que quedará reflejado en la Guerra de Troya. Rubens, Peter...
  • MHSW Committee update Grace McKenzie, Peel MHSW Chair

    MHSW Committee update Grace McKenzie, Peel MHSW Chair

    MHSW Committee update Grace McKenzie, Peel MHSW Chair Co-Chairs: Dennis Lam/David Metcalfe Vice Chair - Vacant MWA Spring Workshop May 2019 MWA Spring Workshop May 2019 Meetings held: 2018 May 18, 2018 - Peel 2019 March 28, 2019 - Teleconference...
  • Vowel variables in mid-term ass.

    Vowel variables in mid-term ass.

    Vowel variables in mid-term ass Vowel variables revised list BATH-TRAP Split FOOT-STRUT Split FOOT-GOOSE Merger Long Mid Diphthonging: FACE and GOAT become diphthongs NURSE Merger Diphthong Shift HappY Tensing Full HappY Tensing (NappIES Tensing) Vowel variables revised list BATH-TRAP Split...