B l9 array considers four control factors each at

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: parameter (the signal factor) once the variance has been reduced. Parameter Design (cont’d) •  What is the trade ­off? –  Min (# of tests) vs. Loss (Detailed informaOon of interacOons). •  Two commonly used orthogonal arrays. (a)  L4 deals with three control factors at two levels. A total of 4 runs of DOE is needed. (b) L9 array considers four control factors each at three levels. A total of 9 runs of DOE is needed. Taguchi Robust Design Method What are the steps? •  Step 1: Problem definiOon. –  Select the parameter to be opOmized and the objecOve funcOon. •  Step 2: SelecOon of design parameters; –  The control parameters (controlled under the designer) and the noise parameters (contribute to the variaOon caused by the environment. •  Step 3: Experiment design –  Select the appropriate fracOonal factorial array, the number of levels to be used, and the range of the parameters that correspond to these levels. •  Step 4: Do the experiments –  Follow DOE; •  Step 5: Results analysis; –  Calculate the S/N. •  Step 5: Repeat steps 1 to 4 if no clear opOmum value. OR: •  Step 5: Validate the results; –  Perform a confirming experiment when the method gives a set of opOmal parameter values. Robust Design Example •  Problem: –  A new prototype of new game box has indicator light failure •  Causes: –  Poor solder joints. •  Root cause: –  Use of improper solder paste (i.e., solder balls and flux). •  ObjecOve: –  Design the best condiOons for making strong solder joints...
View Full Document

This note was uploaded on 02/26/2012 for the course 650 486 taught by Professor Zou during the Fall '11 term at Rutgers.

Ask a homework question - tutors are online