Javier Garcia - Verdugo Sanchez Six Sigma Training. Week 3. Catapult DOE (Practical exercise).

1. DOE
Practical assignment
Rubber band -
fix point
Rubber band -
fix point
Rubber band -Rubber band -
tension adjustment
4 4
Ball typeBall type
Rubber bandRubber band5
6
Rubber band -
fix point
Rubber band -
fix point
Rubber band -Rubber band -
tension adjustment
4 4
Ball typeBall type
Rubber bandRubber band5
6
Rubber band -
fix point
Rubber band -
fix point
Rubber band -Rubber band -
tension adjustment
4 44
Ball typeBall type
Rubber bandRubber band5
6
Stop -PositionStop -Position
tension adjustmenttension adjustment
2
345
3
4
1
2
3
4
Tension angleTension angle
1
2
Stop -PositionStop -Position
tension adjustmenttension adjustment
2
345
3
4
1
2
3
4
Tension angleTension angle
1
2
Stop -PositionStop -Position
tension adjustmenttension adjustment
2
345
3
4
1
2
3
4
1
2
3
4
Tension angleTension angle
1
2
1
2
35
6 1
2
35
6 1
2
35
6
The catapultp
optimization Week 3
Knorr-Bremse Group
The Catapult Optimization
Ball type
R bb b d
yp
6Rubber band -
fix point Rubber band -
tension adjustment
3
4 4
Rubber band5
Stop -Position3
2
3
1
2
1
2
345
1Tension angle
6
1
1
Further Factors: Operator Measurement System + + + ?
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 2/8
Further Factors: Operator, Measurement System... + + + ?

2. Assignment
Your assignment is to hit a target in a distance of a range of
100 to 300 cm with the most accuracy The target has a size100 to 300 cm with the most accuracy. The target has a size
of 20 x 20 cm. This is a compromise which was worked out
with the “marketing” department of King Kunibert. However, if
smaller targets could be hit accurate as well this would be ansmaller targets could be hit accurate as well this would be an
strong argument for King Kunibert to buy more than currently
ordered. If targets with a size of 10 x 10 cm could be hit
accurate the King would double their current orders It wasaccurate the King would double their current orders. It was
decided to order catapult models and to perform studies with
them. The time is short for the King and also the budget.
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 3/8
Assignment
Use your knowledge about experimentation (DOE) and
include all factors in your thoughtsinclude all factors in your thoughts.
Start with a fractional factorial design, which can be extendedg ,
accordingly
Use techniques like fold over and blocks
Make sure that nothing occurs just by chance what could beMake sure that nothing occurs just by chance what could be
predicted in front, consider all factors
Your budget allows 60 shots maximum. You know that 30
shots are needed to verify the accuracy of your catapult iny y y p
distance and variation.
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 4/8

3. Assignment
This exercise has several steps:
1. Fractional factorial design which contains blocks or fold
over
2. Extension of the design to a full factorial design, which
can include center points or additional levels
3 Extension to a RSM design3. Extension to a RSM design
4 Performance of a short term process capability study4. Performance of a short term process capability study
Please, before your start, think about the numbers of trials you
have available! Your budget is limited to 60 max.
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 5/8
g
Design types
Screening designs:
Derivates the less, significant, factors from the many
unimportant ones. Investigates a relative large number of
factors with a low number of runs (fractional factorial design)factors with a low number of runs (fractional factorial design)
Characterizing designs:g g
Identified the significant variables with an effect on the
F i ti ti f l it ti ith lresponse. For investigation of complex situation with a low
number of factors (2-6). General full factorial designs with a
higher resolution as a fractional factorial designs.g g
Optimization designs:
To define the optimal operation window. Full factorial with
RSM
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 6/8
RSM

4. Steps of a DOE planning
Define the problem
Establish the goalEstablish the goal
Select the output variables (response)
Select the input variables (factors)
Define the levels of the factors for the experiment
Choice an experimental design
Collect the dataCollect the data
Analyze the data
Draw your statistical conclusions
Replicate your results
Develop practical solutions
Realize your solution/improvements
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 7/8
Realize your solution/improvements
Final report
The report should contain the following:
• Summary
• Problem statement and backgroundg
• Goals
• Output variables
• Input variablesInput variables
• Experimental design
• Procedure
• Results and data analysisResults and data analysis
• Conclusions
Knorr-Bremse Group BB W3 catapult DOE 08, D. Szemkus/H. Winkler Page 8/8