Robust Design and Tolerance Analysis

Use DOE to create products and processes robust to varying conditions, and tolerance analysis to assure your specifications are met. This workshop is a must for Design for Six Sigma (DFSS). Anyone with prior DOE experience and the desire to create an ideal product or process should attend.

This class is only available as a private on-site workshop.

Robust Design and Tolerance Analysis

Details

Robust Design and Tolerance Analysis*, 1 day, (0.8 PDH, In-House Only)

The Best of Both Worlds

Use DOE to create products and processes robust to varying conditions, and tolerance analysis to assure your specifications are met. This workshop is a must for Design for Six Sigma (DFSS). Anyone with prior DOE experience and the desire to create an ideal product or process should attend.

In just one day you will discover how to take your DOE knowledge to the next level, making improved design decisions that lead to defect-free products. Apply the techniques you learn to minimize expensive design rework and accelerate product scale-up and commercialization.

Practice building and running designed experiments that pinpoint settings which are robust to sources of noise and create a viable transfer function. Then use that transfer function to determine the best tolerances for your product design. Through “knowledgeable design” discover which product or process design parameters are critical to performance. Achieve optimum values with minimal variation.

Together, DOE and tolerance analysis lead to robust products and processes:
1. DOE helps to identify factor levels that maximize the performance of a response and minimize the effect of uncontrolled variables. It also identifies factors that have no significant effect on performance, allowing tolerances to be relaxed—leading to reduced costs.
2. Propagation of error (transmitted variation) is a math-based tool that, when added to DOE, helps identify factor settings that are robust to variation from the input variables.
3. Tolerance analysis pinpoints the factors that are critical to process capability. It also reveals the impact of reducing the variation transmitted from factors to responses.

Concepts You’ll Master

  • How to apply robust design in the Design for Six Sigma (DFSS) setting
  • Which design of experiment is best for verification studies
  • Putting propagation of error (POE) to use for optimizing factor levels and minimizing sensitivity to noise
  • Comparing the pros and cons of worst-case tolerancing, statistical tolerancing, and process tolerancing
  • Improving the process capability indexes (Cpk’s) for your process via tolerance analysis

Course Outline

 

Day 1

Section 1—Tolerance Analysis

 
  • Concepts of tolerance analysis (TA)
    • Case Study - Assembly 
  • Process Capability Review
 

Section 2—Robust Design

 
  • Robust Design (Injection Molding)
  • Robust Design Concepts
  • Four Approaches to Robust Design (Lemon Bars)

 

Section 3—Factorial Designs and POE

 
  • Intro to Propagation of Error (POE)
  • Using Interactions to Achieve Robustness
    • Case Study - Leaf spring
  • Empirical Tolerance Analysis
 Lunch

Section 4—Response Surface Methods with POE and TA

 
  • RSM Analysis with Tolerance Analysis
    • Case study - Lathe machined parts
    • Case study - HDTV signal
  • Transformation and POE
    • Case study - Adhesive RSM
 

Section 5—Dual Response Approach

 
  • Analyze Mean and Standard Deviation
    • Case Study - Printing
      • Design of Experiments (DOE)
      • Tolerance Analysis (TA)
  Section 6—Repeated Measures
 
  • Repeated Measures
    • Case study - Stent Coating
      • DOE
      • TA
  • RDTA Summary
   Section 7—Optional Case Studies (as time allows)
 
  • Multiple Linear Constraints (Stop on a Shaft)
  • Log-Linear model with POE (Well drilling)
  • Categoric Factors (Packaging heat seal)
  Section 8—Appendix & Papers
 
  • Robust Design – Reducing Transmitted Variation
  • POE math: general, for factors, transformed, for mixtures
  • Comparing Three Approaches to Robust Design

Prerequisites

You need not be a statistician, but you must have a working knowledge of factorial design and response surface methods. (If you are working for a company with a Six Sigma program, you should be at the black-belt level of proficiency in statistics). We recommend you take our Experiment Design Made Easy and Response Surface Methods for Process Optimization workshops as prerequisites.

For more information, contact the Workshop Coordinator by e-mail or by calling 1.612.746.2038

Additional Information

PDHs 8
Additional Information No