3

pdf

School

Arizona State University *

*We aren’t endorsed by this school

Course

572

Subject

Mathematics

Date

Feb 20, 2024

Type

pdf

Pages

10

Uploaded by BailiffLlama2777

Report
Characterization of the Response Surface * Find out where our stationary point is » Find what type of surface we have Graphical Analysis Canonical Analysis « Determine the sensitivity of the response variable to the optimum value Canonical Analysis Chapter 11 Design & Analysis of Experiments 10E 2020 21 Montgomery
Finding the Stationary Point » After fitting a second order model take the partial derivatives with respect to the x;'s and set to zero —Qy/dx;=...= 0y/dx,=0 Stationary point represents... Maximum Point Minimum Point Saddle Point Chapter 11 Design & Analysis of Experiments 10E 2020 Montgomery
Stationary Point x,= —B7'b X Bl Bll'BL!’z X > x=|"]1 b= B.' and B = Xi B sym. Chapter 11 Design & Analysis of Experiments 10E 2020 Montgomery B\\.- . v vee s Bul2 B/2 B kk 23
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Canonical Analysis + Used for sensitivity analysis and stationary point identification » Based on the analysis of a transformed model called: canonical form of the model « Canonical Model form: Y = Ys AW+ AW + L+ A w2 Chapter 11 Design & Analysis of Experiments 10E 2020 Montgomery 24
x, u FIGURE 11.9 Canonical form of \ the second-order model *1 Chapter 11 Design & Analysis of Experiments 10E 2020 25 Montgomery
Eigenvalues » The nature of the response can be determined by the signs and magnitudes of the eigenvalues {e} all positive: a minimum is found {e} all negative: a maximum is found {e} mixed: a saddle point is found » Eigenvalues can be used to determine the sensitivity of the response with respect to the design factors » The response surface is steepest in the direction (canonical) corresponding to the largest absolute eigenvalue Chapter 11 Design & Analysis of Experiments 10E 2020 26 Montgomery
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Chapter 11 exampLE 11.2 We will continue the analysis of the chemical process in Example 11.1. A second-order model in the vanables x, and v; cannot be fit using the design in Table 11.4. The experi- menter decides to augment this design with enough points to fit a second-order model.' She obtains four observations at (x, =0,5 = *1414) and (x, = £ 1414, x; = 0). The complete experiment is shown in Table 11.6, and the design is dis- played in Figure 11.10. This design is called a central com- posite design (or a CCD) and will be discussed in more detail in Section 11.4.2. In this second phase of the study, two = TABLE 11.6 Central Composite Design for Example 11.2 Natural Vartables Coded Varfables Responses & & x, X ¥y (yhedd) ¥y A viscosity ) ¥y (mwbecular weighty %0 170 1 I 76.5 62 2040 0 150 1 ! 7.0 0 470 170 I -1 50 o w0 180 I 1 0.5 » 85 175 0 0 99 ~ 175 0 ) 0.3 0 85 175 0 0 0.0 o8 ' 175 0 0 9.7 70 290 5 175 0 0 9% 1 1500 920 175 1414 0 54 o8 1360 7798 175 1414 0 756 T 20 85 182.0 0 1414 8.5 8 1630 85 167.93 0 ~1.414 77.0 57 1150 Design & Analysis of Experiments 10E 2020 Montgomery 27
X2 +2 0, 1.4149) =11 (LN | ] -2 (-1.414,0) (0,0 (1.414,0) +2 =1,-1) (1,-1 0.-1.413) 2 s FIGURE 11.10 Central composite design for Example 11.2 additional responses were of interest: the viscosity and the molecular weight of the product. The responses are also shown in Table 11.6. We will focus on fitting a quadratic model to the yield response v, (the other responses will be discussed in Section 11.3.4). We generally use computer software to fit a response surface and to construct the contour plots. Table 11.7 contains the output from Design-Expert. From exam- ining this table, we notice that this software package first computes the “sequential or extra sums of squares™ for the Chapter 11 Design & Analysis of Experiments 10E 2020 Montgomery 28
TABLE 11.7 Y ST Ty S S T P I 7 E T S T S e —— Computer Output from Design-Expert for Fitting a Model to the Yield Response in Exam, Response: yield ***WARNING: The Cubic Model is Aliased!*** Sequential Model Sum of Squares Source Sum of Squares DF Mean Square FValue Prob > F Mean 80062.16 1 80062.16 Linear 10.04 2 5.02 2.69 0.1166 2FI 0.25 1 0.25 0.12 0.7350 Quadratic 17.95 2 8.98 126.88 <0.001 Suggested Cubic 2.042E-003 2 1.021E-003 0.010 0.9897 Aliased Residual 0.49 5 0.099 Total 80090.90 13 6160.84 “Sequential Model Sum of Squares”: Select the highest order polynomial where the additional terms are significant. Lack-of-Fit Tests Source Sum of Squares DF Mean Square FValue Prob > F Linear 18.49 6 3.08 58.14 0.0008 2FI 18.24 5 3.65 68.82 0.0006 Quadratic 0.28 3 0.094 1.78 0.2897 Suggested Cubic 0.28 1 0.28 5.31 0.0826 Aliased Pure Error 0.21 4 0.053 “Lack-of-Fit Tests”: Want the selected model to have insignificant lack-of-fit. Model summary Statistics Source Std. Dev. R-Squared Adjusted R-Squared Predicted R-Squared PRESS Linear 137 0.3494 0.2193 -0.0435 29.99 2FI 1.43 0.3581 0.1441 —-0.2730 36.59 Quadratic 0.27 0.9828 0.9705 0.9184 2.35 Suggested Cubic 0.31 0.9828 0.9588 0.3622 18.33 Aliased “Model Summary Statistics”: Focus on the model minimizing the “PRESS,” or equivalently maximizing the “PRED R-SQR.” Response: yield ANOVA for Response Surface Quadratic Model Analysis of variance table [Partial sum of squares] Source Sum of Squares DF Mean Square FValue Prob > F Model 28.25 5 5.65 79.85 <0.0001 A 7.92 1 7.92 111.93 <0.0001 B 2.12 1 2.12 30.01 0.0009 A2 13.18 1 13.18 186.22 <0.0001 B? 6.97 1 6.97 98.56 <0.0001 AB 0.25 1 0.25 3.53 0.1022
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Response: yiel& ] S ANOVA for Response Surface Quadratic Model ] Analysis of variance table [Partial sum of squares] Source Sum of Squares ) DF M;an Square F Vaiue Prob > F Model S 28.25 5 o 5.65 79.85 <0.0001 A 7.92 7.92 111.93 <0.0001 B 2.12 1 2.12 30.01 0.0009 A2 13.18 1 13.18 186.22 <0.0001 B? 6.97 1 6.97 98.56 <0.0001 AB 0.25 1 0.25 3.53 0.1022 Residual 0.50 7 0.071 Lack of Fit 0.28 3 0.094 1.78 0.2897 Pure Error 0.21 4 0.053 Cor Total 28.74 12 Std. Dev. 0.27 R-Squared 0.9828 Mean 78.48 Adj R-Squared 0.9705 C.V. 0.34 Pred R-Squared 0.9184 PRESS 2.35 Adeq Precision 23.018 F;lctor DF Standard Error 95% CI Low 95% CI High I;xtercep‘t 79.94 N o 0.12 79.66 80.22 A-time 0.99 1 0.094 0.77 1.22 li;ten;l.). 0.52 1 0.094 0.29 0.74 “A{ -1.38 1 0.10 -1.61 -1.14 F -1.00 1 0.10 -1.24 -0.76 AB 0.25 1 0.13 —0.064 0.56 Final Equation in Terms of Coded i‘:actors: [ yield = +79.94 +0.99 +0.52 —1.38 —1.00 +0.25 e VWM _at s A AT