Box Behnken RSM: Difference between revisions

From ReliaWiki
Jump to navigation Jump to search
No edit summary
No edit summary
 
Line 1: Line 1:
{{Reference Example|{{Banner DOE Reference Examples}}}}
{{Reference Example|{{Banner DOE Reference Examples}}}}
This example validates the calculation of the Box-Behnken RSM method in DOE++.
This example validates the calculation of the Box-Behnken RSM method in Weibull++.


{{Reference_Example_Heading1}}
{{Reference_Example_Heading1}}

Latest revision as of 21:08, 18 September 2023

DOE Reference Examples Banner.png


New format available! This reference is now available in a new format that offers faster page load, improved display for calculations and images and more targeted search.

As of January 2024, this Reliawiki page will not continue to be updated. Please update all links and bookmarks to the latest references at DOE examples and DOE reference examples.




Box Behnken RSM

This example validates the calculation of the Box-Behnken RSM method in Weibull++.

Reference Case

The data are from Example 7.5 on page 347 in the book Response Surface Methodology: Process and Product Optimization Using Designed Experiments, 2nd Edition, by Raymond H. Myers and Douglas C. Montgomery, John Wiley & Sons, Inc. 2002.

Data

Level (coded value) Temperature (A) Agitation (B) Rate (C)
High (+1) 200 10 25
Center (0) 175 7.5 20
Low (-1) 150 5 15

The viscosity of the resin was recorded as an indirect measure of molecular weight.

Standard order A B C Y
1 -1 -1 0 53
2 +1 -1 0 58
3 -1 +1 0 59
4 +1 +1 0 56
5 -1 0 -1 64
6 +1 0 -1 45
7 -1 0 +1 35
8 +1 0 +1 60
9 0 -1 -1 59
10 0 +1 -1 64
11 0 -1 +1 53
12 0 +1 +1 65
13 0 0 0 65
14 0 0 0 59
15 0 0 0 62


Result

From the book, the ANOVA table (partial sum of squares) is:

Source Partial SS DF Mean Square F value Prob > F
Model 882.48 9 98.05 9.57 0.0114
A 8 1 8.00 0.78 0.4174
B 55.12 1 55.12 5.38 0.0681
C 45.13 1 45.13 4.40 0.09
[math]\displaystyle{ A^2 }[/math] 200.83 1 200.83 19.59 0.0069
[math]\displaystyle{ B^2 }[/math] 12.98 1 12.98 1.27 0.3115
[math]\displaystyle{ C^2 }[/math] 48.52 1 48.52 4.73 0.0816
AB 16.00 1 16.00 1.56 0.2668
AC 484.00 1 484.00 47.22 0.001
BC 12.25 1 12.25 1.22 0.3241
Residual 51.25 5 10.25
Lack of Fit 33.25 3 11.08 1.23 0.4774
Pure error 18 2 9.00
Cor total 933.73 14

The final equation in terms of actual factors:

[math]\displaystyle{ \begin{align} Viscosity=-58.875+2.65 * Temp-0.65 * Agitation-11.125*Rate-0.0118*Temp^2+0.3*Agitation^2 \\ -0.145*Rate^2-0.032*Temp*Agitation+0.088*Temp*Rate+0.14*Agitation*Rate \end{align} }[/math]


Results in DOE++

The software results match the book results. The ANOVA table is:

Behnken anova.png

The final equation in terms of the actual factors is:

Behnken final.png

The maximum viscosity is shown in the following optimal solution plot.

Behnken plot.png