This preview shows pages 1–2. Sign up to view the full content.
16.881 – Robust System Design
Solution to Quiz #3
Steps in Robust Design (Data Analysis)
The factor effect plots below represent the results of an orthogonal array based matrix
experiment.
1)
How many total experiments were required to gather the data represented in the
figures assuming one noise factor with two levels in the outer array and no replicates?
The data represents 4 control factors with 3 levels each. The required degrees of freedom
are therefore
1+(31)*4=9DOF
This experiment happens to fit the L
9
(3
4
) exactly. A noise strategy including one noise
factor in an outer array with no replicates will require just two eperiments per row of the
L
9
for
a total of 18 experiments
.
2)
Which control factor should be used as the scaling factor assuming the additive model
holds?
Factor C is the best choice for scaling factor.
A good scaling factor should have
minimal effect on the S/N ratio. So you’re looking for a flat line on the control factor plot
for S/N ratio. A good scaling factor should also have a strong effect on the mean so that it
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
This is the end of the preview. Sign up
to
access the rest of the document.
This note was uploaded on 11/08/2011 for the course AERO 16.851 taught by Professor Ldavidmiller during the Fall '03 term at MIT.
 Fall '03
 lDavidMiller

Click to edit the document details