Lecture 9 - Design of Experiments (DoE) Flashcards
What does DoE stand for?
Design of Experiments
DoE is defined as…
A series of purposeful changes to the process inputs (factors)
in order to observe the corresponding changes in the outputs (responses)
Key Processes +
______ Processes +
______ Processes = Output
Key processes +
Controlled processes +
Uncontrolled Processes = Output
The objectives of DoE is to…
- Determines which factors (inputs) have an effect on response average, variability, and those which do not
- Gain maximum amount of information with minimum resources
- Validate Results
- Optimise the process
In DoE, when we “Optimise the process” means…
Improve performance characteristics, reduce costs and time associated with product development and production
Give three negatives of using a OFAT (One Factor at at Time) approach
- Expensive and Time Consuming
- Unable to evaluate interaction effects
- Can miss optimum settings
What are the main advantages to DoE
- Efficiency (Reduced time and resources to conduct experiments)
- Evaluates interaction effects between the factors
DoE should ONLY be used after the application of…
IPO / PF / CE / SOP / FMEA
if we had 20 factors, each of which had 4 different possible values, how many combinations of factors would we need to evaluate if we were to use the OFAT approach?
Using “One Factor at a Time” we would need to evaluate
4^20 = 1,099,511,627,776 samples.
Therefore OFAT is a very inefficient process!
Strategies for Experimentation include
- Screening to separate contributing and non contributing factors (Factorial Designs)
- Optimisation to be able to predict the response for a given combination of factors (Response Surface Designs)
Factorial Designs are…
• Good preliminary experiments which is separated into two types of factorials:
Full and Fractional
“Full” factorials involve…
testing every combination of factor levels
“Fractional” factorials are…
used to find the “vital view” significant factors out of a large group of potential factors
DoE Terminology:
Controlled Experiment means…
A study where treatments are imposed on experimental units, in order to observe a response
DoE Terminology:
Factor means…
A variable that potentially affects the response
DoE Terminology:
Treatment means…
A combination of one or more factors
DoE Terminology:
Levels means…
The values a factor can take on
DoE Terminology:
Effect means…
How much a main factor or interaction between factors influences the mean response
DoE Terminology:
Replicates means…
When you repeat your experiment design a 2nd, 3rd, 4th etc time.
It is NOT the same as repeated measurements!
What are the 6 steps to the DoE Process?
- Identifying the opportunity / problem variable to understand (Improve)
- Defining you experiment
- Planning your experiment for implementation
- Conducting your experiment
- Analysing your experiment
- Implementing Action Plan
True / False?
The general scenario for performing a DoE for non-industrial processes is the
same for industrial processes
True
You still need to understand your customer, products and processes.
What is the fundamental difference between DoE for industrial purposes vs non-industrial purposes?
In an industrial process you are measuring things
In a non-industrial process you are measuring people