Designed experiments are a powerful approach to improving a process. To use this approach,
it is necessary that everyone involved in the experiment have a clear idea in advance of the
objective of the experiment, exactly what factors are to be studied, how the experiment is to
be conducted, and at least a qualitative understanding of how the data will be analyzed.
Montgomery (2005) gives an outline of the recommended procedure, reproduced in Fig. 13.4.
We now briefly amplify each point in this checklist.
1. Recognition of and statement of the problem. In practice, it is often difficult to
realize that a problem requiring formal designed experiments exists, so it may not be easy to
develop a clear and generally accepted statement of the problem. However, it is absolutely
essential to fully develop all ideas about the problem and about the specific objectives of the
experiment. Usually, it is important to solicit input from all concerned parties—engineering,
quality, marketing, the customer, management, and the operators (who usually have much
insight that is all too often ignored). A clear statement of the problem and the objectives of
the experiment often contributes substantially to better process understanding and eventual
solution of the problem.
power supply E. These tolerance bands can be used to define
appropriate factor levels, and an experiment can be performed
to determine which circuit components have the
most critical tolerances and how much they must be tightened
to produce adequate gauge capability. The information
from this experiment will result in a design specification that
tightens only the most critical tolerances the minimum
amount possible consistent with desired measurement capability.
Consequently, a lower cost design that is easier to
manufacture will be possible.
Notice that in this experiment it is unnecessary actually to
build hardware, since the response from the circuit can be calculated
via equation (13.1). The actual response variable for
the experiment should be the standard deviation of Y. However,
an equation for the transmitted variation in Y from the circuit
can be found using the methods of Section 8.6.2. Therefore,
the entire experiment can be performed using a computer
model of the Wheatstone bridge.
■ F IGURE 1 3 . 3 A
Wheatstone bridge.
Y A B
X
D C
F
E
Pre-experimental 1. Recognition of and statement of the problem
planning 2. Choice of factors and levels
3. Selection of the response variable
4. Choice of experimental design
5. Performing the experiment
6. Data analysis
7. Conclusions and recommendations
đang được dịch, vui lòng đợi..