Scribd
Upload
134 views

Design of Experiment

The document describes an experiment to measure vibration using a factorial design of experiments. It tests how two factors - spring stiffness and the distance of a fiber optic from a reflective surface - affect the response of vibration. The experiment found that the distance of the fiber optic had a significant effect on vibration according to the ANOVA table, while the spring stiffness did not. Plots of the main effects and interactions supported this, showing that changes in the fiber optic distance produced greater changes in the measured vibration response.

Uploaded by

Fairus Affini
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
134 views

Design of Experiment

The document describes an experiment to measure vibration using a factorial design of experiments. It tests how two factors - spring stiffness and the distance of a fiber optic from a reflective surface - affect the response of vibration. The experiment found that the distance of the fiber optic had a significant effect on vibration according to the ANOVA table, while the spring stiffness did not. Plots of the main effects and interactions supported this, showing that changes in the fiber optic distance produced greater changes in the measured vibration response.

Uploaded by

Fairus Affini
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 21

DESIGN OF EXPERIMENT

MEASURING
VIBRATION
PRESENTED BY;
MOHD AZIZI BIN MAT SAH
EXPERIMENTS??
Used to study the performance of process
and systems.
(Latin: ex-+-periri, "of (or from) trying"),
Set of actions and observations, solving a
particular problem or question
To support or falsify a hypothesis or
research concerning phenomena.

STARTEGIES EXPERIMENTS?
1. Best Guess approach
Based on knowledge & experience
2. One Factor at a time
Each factor at a time observe the response
3. Factorial experiment
Most popular nowadays
Each factors observed at the same time
Safe time & cost
1. Planning
Objective
Factors, range of experiment, level
Type of experiment
2. Execution
How the experiment conducted?
3. Analysis
Conclusion of experiment
Guidelines designing experiment..
PLANNING..
1. OBJECTIVE:
To determine the best method that will affects
and improve the response of vibration.
To observe the output response when input
variables is change.
To determine which variable are most influential
on the response.

FACTORS:
Spring stiffness
Distance fiber optic from reflective
surface

LEVEL: 2 for each factors
REPLICATION: 3 repeating experiment
EXPERIMENTAL CHOICE: Factorial
design
VIBRATION DEVICE:

EXECUTION..
A (Spring) B (Displacement) Treatment
combination
Order
Replicate 1 Replicate 2 Replicate 3
-
(TWO SPRING)
-
(LOWEST)
A low, B low 1 9 5
-
(TWO SPRING)
+
(HIGHEST)

A low, B high 7 3 11
+
(ONE SPRING)

-
(LOWEST)
A high, B low 6 2 10
+
(ONE SPRING)

+
(HIGHEST)

A high,B high 4 8 12
RANDOMIZE ORDER:
Treatment
combination
Response (N) Total Data
Means
Replicate 1 Replicate 2 Replicate 3
(1) A low, B low 0.019 0.03 0.05 0.099 0.033
b A low, B high 183.82 245.52 127.26 565.6 188.533
a A high, B low 0.09 0.03 0.03 0.15 0.05
ab A high,B high 197.96 134.33 148.47 480.76 160.253
y = 1046.509
ANALYSIS DATA..
ANOVA TABLE
1. Find the significant factor that effect the
response
NORMAL PROBABILITY PLOT
1. Checking the distribution of data
MAIN EFFECT PLOT
1. Determine which factors effects the response
INTERACTION PLOT
1. The impact that changing the settings of one
factor has on another factor

ANOVA TABLE:
It is clearly shown that only factor B is
significant because it has P-value of
0.000.
Conclusion, only factor B will effect the
response of this system
ANOVA calculations and equation.

Source
Degree of Freedom Sum of squares Mean square Fo
A (a-1) = 2-1
= 1
SSA = 599 MSA = SSA
a-1
= 599
1
= 599
Fo = MSA
MSError
= 599
1295.4
= 0.16
B (b-1) = 2-1
= 1
SSB = 91,196 MSB = SSB
b-1
= 91916
1
= 91916
Fo = MSB
MSError
= 91916
1295.4
= 70.40
AB (Interaction) (a-1)(b-1) = 1 SSAB = 601 MSAB = SSAB
(a-1)(b-1)
= 601
1
= 601
Fo = MSAB
MSError
= 601
1295.4
= 0.46
Error ab(n-1) = 2(2)(3-1)
= 8
SSError = 10,364 MSError = SSError
ab(n-1)
= 10364
8
= 1295.4
Total abn -1 =
2(2*3)-1 = 11
SSTotal = 102,459
2
(599 91196 601)
0.8991
102759
R


NORMAL PROBABILITY PLOT:
Residual
P
e
r
c
e
n
t
100 50 0 -50
99
95
90
80
70
60
50
40
30
20
10
5
1
Normal Probability Plot of the Residuals
(response is Response)
Error distribution is approximately not
really normal where many of the data
focus on 0 residual.
MAIN EFFECT PLOT:
M
e
a
n

o
f

R
e
s
p
o
n
s
e
1 -1
200
150
100
50
0
1 -1
A B
Main Effects Plot (data means) for Response
Effect of factor B is greater than factor A
INTERACTION PLOT
B
M
e
a
n
1 -1
200
150
100
50
0
A
-1
1
Interaction Plot (data means) for Response
B+ has higher yield than B-
A- has higher yield than A+
No response difference at lower level

Contour plot

0
1
100
200
Response
1
0
A 0
B
-1
-1
B
A
1.0 0.5 0.0 -0.5 -1.0
1.0
0.5
0.0
-0.5
-1.0
Response
70 - 120
120 - 170
170 - 220
> 220
< 20
20 - 70
Surface Plot of Response vs A, B
Contour Plot of Response vs A, B
Residual versus Fitted Values

Fitted Value
R
e
s
i
d
u
a
l
200 150 100 50 0
75
50
25
0
-25
-50
Residuals Versus the Fitted Values
(response is Response)
Residual versus factors
B
R
e
s
i
d
u
a
l
1.0 0.5 0.0 -0.5 -1.0
75
50
25
0
-25
-50
Residuals Versus B
(response is Response)
A
R
e
s
i
d
u
a
l
1.0 0.5 0.0 -0.5 -1.0
75
50
25
0
-25
-50
Residuals Versus A
(response is Response)
Interaction equation of A and B
A
R
e
s
p
o
n
s
e
1.0 0.5 0.0 -0.5 -1.0
240
180
120
60
0
S 101.074
R-Sq 0.6%
R-Sq(adj) 0.0%
B
R
e
s
p
o
n
s
e
1.0 0.5 0.0 -0.5 -1.0
240
180
120
60
0
S 34.0047
R-Sq 88.7%
R-Sq(adj) 87.6%
Fitted Line Plot
Response = 87.22 - 7.07 A
Fitted Line Plot
Response = 87.22 + 87.18 B
SUMMARY..
Factor B, fiber optic
displacement has greater impact
on the response
DOE important to select the
most significant factor

END SESION

You might also like