5.6.a Exp. & Exer

You might also like

Download as xlsx, pdf, or txt
Download as xlsx, pdf, or txt
You are on page 1of 33

EXAMPLE - FRACTIONAL FACTORIAL DESIGNS

1.-
Chelsea Footwear – Emily Walker's team is conducting an experiment to determine the optimal conditions and
levels for a shoe sole molding machine and obtain the least amount of defects possible. In order to do so, the
team decided to test 5 factors and two levels for each one.

A Molding Temperature (in °C) 180 220


B Molding Time (in min) 6 10
C Cooling Time (in min) 3 6
D Cooling Temperature (in °C) 100 120
E Type of Raw Material Crushed Not Crushed

The results from the experiments are as follows:

A B C D E Waste Percentage
Molding Temp.Molding Time Cooling Time Cooling Temp Raw Material R1 R2
180 6 3 100 Crushed 0.08 0.08
220 6 3 100 Crushed 0.14 0.15
180 10 3 100 Crushed 0.36 0.32
220 10 3 100 Crushed 0.54 0.56
180 6 6 100 Crushed 0.17 0.15
220 6 6 100 Crushed 0.2 0.2
180 10 6 100 Crushed 0.42 0.38
220 10 6 100 Crushed 0.59 0.61
180 6 3 120 Crushed 0.09 0.07
220 6 3 120 Crushed 0.15 0.15
180 10 3 120 Crushed 0.32 0.34
220 10 3 120 Crushed 0.55 0.54
180 6 6 120 Crushed 0.16 0.17
220 6 6 120 Crushed 0.21 0.21
180 10 6 120 Crushed 0.4 0.43
220 10 6 120 Crushed 0.6 0.62
180 6 3 100 Not Crushed 0.08 0.09
220 6 3 100 Not Crushed 0.16 0.15
180 10 3 100 Not Crushed 0.37 0.33
220 10 3 100 Not Crushed 0.53 0.55
180 6 6 100 Not Crushed 0.16 0.15
220 6 6 100 Not Crushed 0.21 0.23
180 10 6 100 Not Crushed 0.44 0.4
220 10 6 100 Not Crushed 0.65 0.62
180 6 3 120 Not Crushed 0.08 0.07
220 6 3 120 Not Crushed 0.15 0.14
180 10 3 120 Not Crushed 0.35 0.32
220 10 3 120 Not Crushed 0.53 0.54
180 6 6 120 Not Crushed 0.16 0.16
220 6 6 120 Not Crushed 0.22 0.18
180 10 6 120 Not Crushed 0.43 0.39
220 10 6 120 Not Crushed 0.63 0.63

a) Determine the significant factors for this operation.


b) For each factor, find the value that optimizes the process.
c) Create a 5-1 fractional factorial design to reduce the number of necessary
experiments by half.
Define before the experiment
Factors Numerical Non numerical
Levels 2 for each factor
Response Value you want to improve
Replicates 2
Blocking Noises

Expectations or answers after the experiment


Which factors are significant
Level of each factor that optimizes the response
Establish suggested further experiments

Rules for significant factors


3-way and above interactions are not considered
2-way interactions have priority over individual effects
Check for individual effects which are significant and their 2-way interactions are not

Significant factors
B, A, AB and C
AB
C
D Not significant
E Not significant

Levels that optimize the response


A 180
B 6
C 3
D The most economical one
E The most economical one

Suggested further experiments


A Decrease the molding temperature
B Decrease the molding time
C Decrease the cooling time
D Not significant
E Not significant
ANSWER TO EXAMPLE - FRACTIONAL FACTORIAL DESIGNS

1.- USE THE FULL DESIGN TO:


a) Determine the significant factors for this operation.

Pareto Chart of Standardized Effects


(response is Waste, alpha = 0.05, 30 defects tested)
Term
Factor Name
A Molding Temp.
B Molding Time
C Cooling Time
D Cooling Temp.
E Raw Material

Standardized Effect

Answer - Significant factors:


A Molding Temperature
B Molding Time
C Cooling Time
AB MoldingTemperature*MoldingTime
Double interaction AB and Arrangement (as an individual factor) are graphed.

b) For each factor, find the value that optimizes the process.

Factorial Interaction Plot for Waste


Data Fitted Means
Molding Temp. * Molding Time Molding Time
Mean (of Waste)
Mean (of Wast

Molding Temp.

Main Effects Plot for Waste


Data Fitted Means
Mean (of Waste)

Cooling Time

OPTIMAL
A Molding Temperature 180 °C
B Molding Time 6 min
C Cooling Time 3 min
D Cooling Temperature The most economical
E Type of Raw Material The most economical

It is recommended to conduct test with a lower Molding Temperature, lower Molding Time, and lower Cooling Time

c) Create a 5-1 fractional factorial design to reduce the number of necessary


experiments by half.

Pareto Chart of Standardized Effects


(response is Waste, alpha = 0.05)
Term
Factor Name
A Molding Temp.
B Molding Time
C Cooling Time
D Cooling Temp.
E Raw Material
Factor Name
A Molding Temp.
B Molding Time
C Cooling Time
D Cooling Temp.
E Raw Material

Standardized Effect

Answer - Significant factors:


A Molding Temperature
B Molding Time
C Cooling Time
AB MoldingTemperature*MoldingTime
Double interaction AB and Arrangement (as an individual factor) are graphed.
Significant factors: AB and C - Same answer/result as Full Factorial.

Factorial Interaction Plot for Waste


Data Fitted Means
Molding Temp. * Molding Time Molding Time
Mean (of Waste)

Molding Temp.

Main Effects Plot for Waste


Data Fitted Means
Mean (of Waste)

Cooling Time

OPTIMAL
A Molding Temperature 180 °C
B Molding Time 6 min
C Cooling Time 3 min
D Cooling Temperature The most economical
E Type of Raw Material The most economical

It is recommended to conduct tests with a lower Molding Temperature, lower Molding Time, and lower Cooling Tim

Same results/conclusion as Full Factorial Design.


Coded Coefficients
Term
Constant
Molding Temp.
Molding Time
Cooling Time
Cooling Temp.
Raw Material
Molding Temp. * Molding Time
Molding Temp. * Cooling Time
Molding Temp. * Cooling Temp.
Molding Temp. * Raw Material
Molding Time * Cooling Time
Molding Time * Cooling Temp.
Molding Time * Raw Material
Cooling Time * Cooling Temp.
Cooling Time * Raw Material
Cooling Temp. * Raw Material
Molding Temp. * Molding Time * Cooling Time
Molding Temp. * Molding Time * Cooling Temp.
Molding Temp. * Molding Time * Raw Material
Molding Temp. * Cooling Time * Cooling Temp.
Molding Temp. * Cooling Time * Raw Material
Molding Temp. Cooling Temp. * Raw Material
Molding Time * Cooling Time * Cooling Temp.
Molding Time * Cooling Time * Raw Material
Molding Time * Cooling Temp. * Raw Material
Cooling Time * Cooling Temp. * Raw Material
Molding Temp. * Molding Time * Cooling Time * Cooling Temp.
Molding Temp. * Molding Time * Cooling Time * Raw Material
Molding Temp. * Molding Time * Cooling Temp. * Raw Material
Molding Temp. * Cooling Time * Cooling Temp. * Raw Material
Molding Time * Cooling Time * Cooling Temp. * Raw Material
Molding Temp. * Molding Time * Cooling Time * Cooling Temp. * Raw Material
Term
Constant
Molding Temp.
Molding Time
Cooling Time
Cooling Temp.
Raw Material
Molding Temp. * Molding Time
Molding Temp. * Cooling Time
Molding Temp. * Cooling Temp.
Molding Temp. * Raw Material
Molding Time * Cooling Time
Molding Time * Cooling Temp.
Molding Time * Raw Material
Cooling Time * Cooling Temp.
Cooling Time * Raw Material
Cooling Temp. * Raw Material
Molding Temp. * Molding Time * Cooling Time
Molding Temp. * Molding Time * Cooling Temp.
Molding Temp. * Molding Time * Raw Material
Molding Temp. * Cooling Time * Cooling Temp.
Molding Temp. * Cooling Time * Raw Material
Molding Temp. Cooling Temp. * Raw Material
Molding Time * Cooling Time * Cooling Temp.
Molding Time * Cooling Time * Raw Material
Molding Time * Cooling Temp. * Raw Material
Cooling Time * Cooling Temp. * Raw Material
Molding Temp. * Molding Time * Cooling Time * Cooling Temp.
Molding Temp. * Molding Time * Cooling Time * Raw Material
Molding Temp. * Molding Time * Cooling Temp. * Raw Material
Molding Temp. * Cooling Time * Cooling Temp. * Raw Material
Molding Time * Cooling Time * Cooling Temp. * Raw Material
Molding Temp. * Molding Time * Cooling Time * Cooling Temp. * Raw Material

e, and lower Cooling Time.


me, and lower Cooling Time.
Effect Coeff SE Coeff T-Value
0.31344 0.00200 156.64
0.13187 0.06594 0.00200 32.95
0.32875 0.16437 0.00200 82.15
0.07187 0.03594 0.00200 17.96
-0.00250 -0.00125 0.00200 -0.62
-0.00438 -0.00219 0.00200 -1.09
0.07375 0.03688 0.00200 18.43
-0.00437 -0.00219 0.00200 -1.09
0.00000 0.00000 0.00200 0.00
-0.00188 -0.00094 0.00200 -0.47
0.00250 0.00125 0.00200 0.62
-0.00062 -0.00031 0.00200 -0.16
-0.00375 -0.00188 0.00200 -0.94
0.00375 0.00187 0.00200 0.94
-0.00375 -0.00219 0.00200 -1.09
0.00625 0.00313 0.00200 1.56
0.00625 0.00312 0.00200 1.56
0.00187 0.00094 0.00200 0.47
0.00125 0.00062 0.00200 0.31
-0.0025 -0.00125 0.00200 -0.62
-0.00563 -0.00281 0.00200 -1.41
0.00125 0.00062 0.00200 0.31
0.00188 0.00094 0.00200 0.47
-0.00500 -0.0025 0.00200 -1.25
-0.00063 -0.00031 0.00200 -0.16
0.00250 0.00125 0.00200 0.62
0.00062 0.00031 0.00200 0.16
-0.00375 -0.00188 0.00200 -0.94
-0.00313 -0.00156 0.00200 -0.78
0.00125 0.00062 0.00200 0.31
0.00188 0.00094 0.00200 0.47
-0.00188 -0.00094 0.00200 -0.47
P-Value VIF
0.000
0.000 1.00
0.000 1.00
0.000 1.00
0.537 1.00
0.282 1.00
0.000 1.00
0.282 1.00
1.000 1.00
0.643 1.00
0.537 1.00
0.877 1.00
0.356 1.00
0.356 1.00
0.282 1.00
0.128 1.00
0.128 1.00
0.643 1.00
0.757 1.00
0.537 1.00
0.169 1.00
0.757 1.00
0.643 1.00
0.221 1.00
0.877 1.00
0.537 1.00
0.877 1.00
0.356 1.00
0.441 1.00
0.757 1.00
0.643 1.00
0.643 1.00

Minitab provided the 32 tests that must be conducted, instead of the original 64:
(16 combinations, 2 replicates for each)

Molding Temp Molding Time Cooling Time Cooling Temp Raw Mat.
180 6 6 120 Not Crushed
180 10 6 100 Not Crushed
220 6 3 120 Not Crushed
220 10 3 100 Not Crushed
220 10 6 120 Not Crushed
180 6 3 120 Crushed
180 6 6 120 Not Crushed
180 10 6 100 Not Crushed
220 6 6 100 Not Crushed
220 6 6 100 Not Crushed
220 6 6 120 Crushed
220 10 6 100 Crushed
180 10 3 100 Crushed
220 6 3 100 Crushed
220 10 3 100 Not Crushed
220 10 3 120 Crushed
220 10 6 100 Crushed
180 6 3 120 Crushed
180 10 3 120 Not Crushed
180 10 3 120 Not Crushed
220 10 3 120 Crushed
220 10 6 120 Not Crushed
220 6 6 120 Crushed
180 10 6 120 Crushed
180 10 3 100 Crushed
180 6 3 100 Not Crushed
180 6 6 100 Crushed
220 6 3 100 Crushed
180 10 6 120 Crushed
180 6 6 100 Crushed
180 6 3 100 Not Crushed
220 6 3 120 Not Crushed

Coded Coefficients
Term Effect Coeff SE Coeff T-Value
Constant 0.31438 0.00269 116.95
Molding Temp. 0.13000 0.06500 0.00269 24.18
Molding Time 0.32750 0.16375 0.00269 60.91
Cooling Time 0.07500 0.03750 0.00269 13.95
Cooling Temp. 0.00125 0.00063 0.00269 0.23
Raw Material 0.00500 0.00250 0.00269 0.93
Molding Temp. * Molding Time 0.07125 0.03562 0.00269 13.25
Molding Temp. * Cooling Time -0.00375 -0.00188 0.00269 -0.7
Molding Temp. * Cooling Temp. 0.00500 0.00250 0.00269 0.93
Molding Temp. Raw Material 0.00375 0.00188 0.00269 0.7
Molding Time * Cooling Time 0.00125 0.00062 0.00269 0.23
Molding Time * Cooling Temp. 0.00500 0.00250 0.00269 0.93
Molding Time * Raw Material 0.00125 0.00063 0.00269 0.23
Cooling Time * Cooling Temp. 0.00250 0.00125 0.00269 0.46
Cooling Time * Raw Material 0.00625 0.00313 0.00269 1.16
Cooling Temp. * Raw Material 0 0 0.00269 0
Waste
0.16
0.44
0.15
0.53
0.63
0.09
0.16
0.4
0.21
0.23
0.21
0.59
0.36
0.14
0.55
0.55
0.61
0.07
0.35
0.32
0.54
0.63
0.21
0.4
0.32
0.08
0.17
0.15
0.43
0.15
0.09
0.14

P-Value VIF
0.000
0.000 1.00
0.000 1.00
0.000 1.00
0.819 1.00
0.366 1.00
0.000 1.00
0.496 1.00
0.366 1.00
0.496 1.00
0.819 1.00
0.366 1.00
0.819 1.00
0.648 1.00
0.262 1.00
1.000 1.00
EXERCISE - FRACTIONAL FACTORIAL DESIGNS

1.- A test/experiment is being conducted at Logistics Company in order to minimize the scrap percentage per damaged
product. The factors that were analyzed and their respective levels are the following:
A Number of pickers 4 6
B Picking method Cubes Units
C Warehouse arrangement Vertical Horizontal
D Picking route A B
E Number of arrangements per pallet 5 8

The results from the experiments are as follows:

A B C D E Scrap Percentage
Pickers Method Warehouse A. Route No. of Arrangements R1 R2
4 Cubes Vertical A 5 0.97 0.96
6 Cubes Vertical A 5 0.55 0.50
4 Units Vertical A 5 0.26 0.26
6 Units Vertical A 5 0.22 0.20
4 Cubes Horizontal A 5 0.71 0.74
6 Cubes Horizontal A 5 0.48 0.42
4 Units Horizontal A 5 0.18 0.20
6 Units Horizontal A 5 0.11 0.11
4 Cubes Vertical B 5 0.79 0.82
6 Cubes Vertical B 5 0.53 0.57
4 Units Vertical B 5 0.28 0.28
6 Units Vertical B 5 0.21 0.22
4 Cubes Horizontal B 5 0.73 0.71
6 Cubes Horizontal B 5 0.42 0.45
4 Units Horizontal B 5 0.20 0.20
6 Units Horizontal B 5 0.12 0.09
4 Cubes Vertical A 8 0.86 0.82
6 Cubes Vertical A 8 0.58 0.53
4 Units Vertical A 8 0.28 0.30
6 Units Vertical A 8 0.21 0.20
4 Cubes Horizontal A 8 0.70 0.73
6 Cubes Horizontal A 8 0.49 0.44
4 Units Horizontal A 8 0.21 0.20
6 Units Horizontal A 8 0.11 0.12
4 Cubes Vertical B 8 0.83 0.83
6 Cubes Vertical B 8 0.57 0.51
4 Units Vertical B 8 0.29 0.24
6 Units Vertical B 8 0.21 0.21
4 Cubes Horizontal B 8 0.73 0.71
6 Cubes Horizontal B 8 0.46 0.42
4 Units Horizontal B 8 0.20 0.18
6 Units Horizontal B 8 0.11 0.09
a) Determine the significant factors for this operation.
b) For each factor, find the value that optimizes the process.
c) Create a 5-1 fractional factorial design to reduce the number of necessary
experiments by half.
r damaged
ANSWER TO EXERCISE - FRACTIONAL FACTORIAL DESIGNS

1.- USE THE FULL DESIGN TO:


a) Determine the significant factors for this operation.

Pareto Chart of Standardized Effects


(response is Scrap, alpha = 0.05, 30 defects tested)
Term
Factor Name
A Pickers
B Method
C Warehouse Arr.
D Route
E No. of Arrang.

Standardized Effect

Answer - Significant factors:


A Pickers
B Method
C Warehouse Arrangement
D Route
AB Pickers*Method
BC Method*WarehouseArrangement
BD Method*Route
The three double interactions are graphed.

b) For each factor, find the value that optimizes the process.

Factorial Interaction Plot for Scrap


Data Fitted Means
Pickers * Method Method
Cubes
Units
ean (of Scrap)
Mean (of Scrap)

Pickers

Factorial Interaction Plot for Scrap


Data Fitted Means
Method * Warehouse A. Warehouse A.
Mean (of Scrap)

Cubes Units
Method

Factorial Interaction Plot for Scrap


Data Fitted Means
Method * Route Route
Mean (of Scrap)
M
Cubes Units
Method

OPTIMAL
A Pickers 6
B Method Units
C Warehouse Arrangement Horizontal
D Route The most economical Provides practically the same result
E Number of Arrangements The most economical Not Significant

c) Create a 5-1 fractional factorial design to reduce the number of necessary


experiments by half.

Pareto Chart of Standardized Effects


(response is Scrap, alpha = 0.05)
Term
Factor Name
A Pickers
B Method
C Warehouse Arr.
D Route
E No. of Arrang.

Standardized Effect

Answer - Significant factors:


A Pickers
B Method
C Warehouse Arrangement
AB Pickers*Method
Double interaction AB and Warehouse Arrangement (as an individual factor) will be graphed.

Factorial Interaction Plot for Scrap


Data Fitted Means
Pickers * Method Method
Cubes
Units
rap)
Mean (of Scrap)

Pickers

Factorial
FactorialInteraction
InteractionPlot
Plotfor
forScrap
Scrap
Data Fitted Means
Data Fitted Means
Mean (of Scrap)

Warehouse Arrangement

OPTIMAL
A Pickers 6
B Method Units
C Warehouse Arrangement Horizontal
D Route The most economical Not Significant
E Number of Arrangements The most economical Not Significant

Same results/conclusion as Full Factorial Design.


Coded Coefficients
Term Effect Coef
Constant 0.41969
Pickers -0.18562 -0.09281
Method -0.44562 -0.22281
Warehouse Arrangement -0.10375 -0.05187
Route -0.01375 -0.00688
No. of Arrangements -0.00375 -0.00188
Pickers * Method 0.10938 0.05469
Pickers * Warehouse A. 0.00500 0.00250
Pickers * Route 0.00875 0.00437
Pickers * No. of Arrang. 0.00750 0.00375
Method * Warehouse A. 0.01375 0.00688
Method * Route 0.01125 0.00563
Method * No. of Arrang. 0.00500 0.00250
Warehouse A. * Route 0.00562 0.00281
Warehouse A. * No. of Arrang. 0.00562 0.00281
Route * No. of Arrang. 0.00188 0.00094
Pickers * Method * Warehouse A. -0.0175 -0.00875
Pickers * Method * Route -0.00875 -0.00438
Pickers * Method * No. of Arrang. -0.01125 -0.00781
Pickers * Warehouse A. * Route -0.01562 -0.00781
Pickers * Warehouse A. * No. of Arrang. -0.00438 -0.00219
Pickers * Route * No. of Arrang. -0.00938 -0.00469
Method * Warehouse A. * Route -0.00938 -0.00469
Method * Warehouse A. * No. of Arrang. -0.00563 -0.00281
Method * Route * No. of Arrang. -0.01188 -0.00594
Warehouse A. * Route * No. of Arrang. -0.00625 -0.00313
Pickers * Method * Warehouse A. * Route 0.01187 0.00594
Pickers * Method * Warehouse A. * No. of Arrang. 0.00687 0.00344
Pickers * Method * Route * No. of Arrang. 0.01687 0.00844
Pickers * Warehouse A. * Route * No. of Arrang. 0.00875 0.00437
Method * Warehouse A. * Route * No. of Arrang. 0.00750 0.00375
Pickers * Method * Warehouse A. * Route * No. of Arrang. -0.01250 -0.00625
tically the same result

Minitab provided the 32 tests that must be cond


(16 combinations, 2 replicates for each)

Pickers Method
4 Cubes
6 Units
4 Cubes
4 Units
4 Units
6 Units
4 Units
4 Cubes
6 Units
4 Cubes
6 Cubes
6 Units
6 Cubes
6 Cubes
4 Cubes
4 Units
4 Units
6 Cubes
6 Cubes
6 Units
6 Cubes
4 Units
4 Cubes
6 Units
4 Units
6 Units
6 Cubes
Method
4 Cubes
Cubes 6 Units
Units
4 Cubes
4 Units
6 Cubes

Coded Coefficients
Term Effect Coef
Constant 0.41344
Pickers -0.17812 -0.08906
Method -0.43687 -0.21844
Warehouse Arrangement -0.08688 -0.04344
Route -0.00688 -0.00344
No. of Arrangements 0.00812 0.00406
Pickers * Method 0.10312 0.05156
Pickers * Warehouse A. -0.00687 -0.00344
Pickers * Route 0.00313 0.00156
Pickers * No. of Arrang. -0.00188 -0.00094
Method * Warehouse A. 0.00437 0.00219
Method * Route 0.00687 0.00344
Method * No. of Arrang. -0.01062 -0.00531
Warehouse A. * Route -0.00563 -0.00281
Warehouse A. * No. of Arrang. -0.00312 -0.00156
Route * No. of Arrang. -0.01562 -0.00781
SE Coef T-Value P-Value VIF
0.00270 155.6 0.000
0.00270 -34.41 0.000 1.00
0.00270 -82.61 0.000 1.00
0.00270 -19.23 0.000 1.00
0.00270 -2.55 0.016 1.00
0.00270 -0.70 0.492 1.00
0.00270 20.27 0.000 1.00
0.00270 0.93 0.361 1.00
0.00270 1.62 0.115 1.00
0.00270 1.39 0.174 1.00
0.00270 2.55 0.016 1.00
0.00270 2.09 0.045 1.00
0.00270 0.93 0.361 1.00
0.00270 1.04 0.305 1.00
0.00270 1.04 0.305 1.00
0.00270 0.35 0.73 1.00
0.00270 -3.24 0.003 1.00
0.00270 -1.62 0.115 1.00
0.00270 -2.09 0.045 1.00
0.00270 -2.9 0.007 1.00
0.00270 -0.81 0.423 1.00
0.00270 -1.74 0.092 1.00
0.00270 -1.74 0.092 1.00
0.00270 -1.04 0.305 1.00
0.00270 -2.20 0.035 1.00
0.00270 -1.16 0.255 1.00
0.00270 2.20 0.035 1.00
0.00270 1.27 0.212 1.00
0.00270 3.13 0.004 1.00
0.00270 1.62 0.115 1.00
0.00270 1.39 0.174 1.00
0.00270 -2.32 0.027 1.00
ed the 32 tests that must be conducted, instead of the original 64:
ns, 2 replicates for each)

Warehouse A. Route No. of Arrangements Scrap


Horizontal B 8 0.73
Horizontal A 5 0.11
Horizontal A 5 0.71
Vertical A 5 0.26
Vertical B 8 0.29
Vertical A 8 0.21
Vertical B 8 0.24
Vertical B 5 0.79
Vertical A 8 0.20
Horizontal B 8 0.71
Horizontal A 8 0.49
Horizontal B 8 0.11
Vertical B 8 0.57
Horizontal B 5 0.42
Vertical A 8 0.86
Horizontal B 5 0.2
Horizontal B 5 0.2
Horizontal B 5 0.45
Vertical B 8 0.51
Vertical B 5 0.21
Horizontal A 8 0.44
Vertical A 5 0.26
Vertical B 5 0.82
Horizontal B 8 0.09
Horizontal A 8 0.21
Horizontal A 5 0.11
Vertical A 5 0.55
Vertical A 8 0.82
Vertical B 5 0.22
Horizontal A 5 0.74
Horizontal A 8 0.20
Vertical A 5 0.50

SE Coef T-Value P-Value VIF


0.00401 103.00 0.000
0.00401 -22.19 0.000 1.00
0.00401 -54.42 0.000 1.00
0.00401 -10.82 0.000 1.00
0.00401 -0.86 0.000 1.00
0.00401 1.01 0.327 1.00
0.00401 12.85 0.000 1.00
0.00401 -0.86 0.404 1.00
0.00401 0.39 0.702 1.00
0.00401 -0.23 0.818 1.00
0.00401 0.54 0.593 1.00
0.00401 0.86 0.404 1.00
0.00401 -1.32 0.204 1.00
0.00401 -0.70 0.494 1.00
0.00401 -0.39 0.702 1.00
0.00401 -1.95 0.069 1.00
HOMEWORK - FRACTIONAL FACTORIAL DESIGNS

1.- In order to maximize filtration flow, a company conducts an experiment to learn


more about the optimal operating conditions when filtering a chemical product.
The four factors that were analyzed and their respective levels are the following:
Temperature 40°C 60°C
Pressure 100 psi 200 psi
Solvent Concentration 500 ml/lt 750 ml/lt
Stirring Speed 3 rev/min 6 rev/min

The results from the experiments are as follows:

Filtration Flow
Temp Pres Solv Speed R1 R2
40 100 500 3 46 48
60 100 500 3 70 72
40 200 500 3 48 48
60 200 500 3 70 68
40 100 750 3 78 77
60 100 750 3 58 62
40 200 750 3 80 80
60 200 750 3 60 60
40 100 500 6 42 44
60 100 500 6 102 104
40 200 500 6 43 45
60 200 500 6 103 105
40 100 750 6 75 75
60 100 750 6 92 96
40 200 750 6 73 75
60 200 750 6 94 98

a) Determine the significant factors for this operation.


b) For each factor, find the value that optimizes the process.
c) Create a 4-1 fractional factorial design to reduce the number of necessary
experiments by half.
HOMEWORK - FRACTIONAL FACTORIAL DESIGNS

2.- A semiconductor company runs an experiment in order to improve process performance.


The factors that were analyzed and their respective factors are:
A Opening (in cms) 30 60
B Exposure time (in min) 12 28
C Revealing time (in min) 3 9
D Photomask dimension (in cms) 100 200
E Engraving time (in min) 4 10

The results from the experiments are as follows:

A B C D E Performance
Opening T. Exposure T. Revealing Photomask T. Engraving R1 R2
30 12 3 100 4 37.5 30.0
60 12 3 100 4 65.0 66.7
30 28 3 100 4 44.4 44.4
60 28 3 100 4 59.8 62.5
30 12 9 100 4 72.2 71.3
60 12 9 100 4 53.7 57.4
30 28 9 100 4 74.1 74.1
60 28 9 100 4 55.6 55.6
30 12 3 200 4 38.9 40.7
60 12 3 200 4 94.4 96.3
30 28 3 200 4 39.8 41.7
60 28 3 200 4 95.4 97.2
30 12 9 200 4 69.4 69.4
60 12 9 200 4 85.2 88.9
30 28 9 200 4 67.6 69.4
60 28 9 200 4 87.0 90.7
30 12 3 100 10 41.8 43.6
60 12 3 100 10 63.6 65.5
30 28 3 100 10 43.6 43.6
60 28 3 100 10 63.6 61.8
30 12 9 100 10 70.9 70.0
60 12 9 100 10 52.7 56.4
30 28 9 100 10 72.7 72.7
60 28 9 100 10 54.5 54.5
30 12 3 200 10 38.2 40.0
60 12 3 200 10 92.7 94.5
30 28 3 200 10 39.1 40.9
60 28 3 200 10 93.6 95.5
30 12 9 200 10 68.2 68.2
60 12 9 200 10 83.6 87.3
30 28 9 200 10 66.4 68.2
60 28 9 200 10 85.5 89.1
a) Determine the significant factors for this operation.
b) For each factor, find the value that optimizes the process.
c) Create a 5-1 fractional factorial design to reduce the number of necessary
experiments by half.

You might also like