Model Answers to 13 Sequential DOE in Six Sigma Statistics using Minitab 17, Green Belt Edition.

Set-Up 1

Exercise 13.7.1 Starship Engine Power Sequential DOE

You are helping Jock, the Chief Engineer, increase the engine power of the Starship Emoji. Jock has identified 8 factors and level that he wants investigated. Your mission is to boldly identify significant factors and maximise the response, Power, using a Screening DOE. Fold the design to gain better resolution in the Effect that can be detected.

Once significant factors have been identified run a Modeling DOE with the intention of maximising the response using existing levels. If curvature is detected conduct additional runs to characterise the curvature. Clearly identify all significant factors to maximise the response and the levels they should be set. All data for this chapter is in ‘13 DOE.xlsx’.

The screening DOE factors are listed below using the following format  Factor Full name (Column Heading, Lo setting, Hi Setting).

Engine Temperature (Temp, 9k, 12k)

Magnetic Field Density (MFD, 180, 230)

Magnetic Field Flux (MFF 26, 49)

Scones, Cream or Jam (Scones 0,1) Categorical

Reaction Initiation, Crank or Fusion  (Initiation 0,1) Categorical

Core Wavelength (Wave 13k, 21k)

Plasma Input Angle (PIA 14, 26)

Plasma Input Flux (PIF 80, 120)

Analysis 1

  1. Start with a new project and then click Assistant<<DOE<<Plan and Create.
  2. Click on the Create Screening Design  box.

  1. Enter ‘Power’ as the response variable.
  2. Jock has identified 8 possible factors for an experiment so we will enter 8 as the number of factors.

  1. Enter all the factor details. We need to enter the short factor name, the data type and then the high and low levels to be used for the factors in the experiment. Note that Scone and Initiation are categorical factors.
  2. Change the total runs in the design to 24.
  3. Click OK to produce the worksheet and the create Summary Report and Report Card.

On the Summary Report the power bar tells us that to achieve a power of 80% the minimum effect that we will be to detect is 1.06 StDevs. In the guidance given below we are told that we will be able to detect down to small to moderate effects of the factors.

After setup of the worksheet the next step is for the trials to be conducted so that the response data can be gathered. Luckily we have a sheet already prepared for you.

Set-Up 2

The data for this section is  13 DOE.xlsx worksheet ‘Power Screen’.

  1. Go to the Minitab worksheet where our modeling DOE worksheet was created. Delete everything but the title bar and then transfer the data from the Excel worksheet into Minitab.
  2. Click Assistant<<DOE<<Analyze and Interpret.
  3. Click on the Fit Screening Model box.

4.  Click ‘Yes’ on the confirmation menu box to produce the Screening Analysis Report.

On the top left of the Summary Report the Pareto of Standardized Effects shows us which Main Effects are significant and would be worth exploring further.

Analysis 2

We are told that changes in these five factors could explain 88.08% of changes in the response.

The Report Card does not show any warnings that indicate problems with the model thus far.

Set-Up 3

The next step is to take the significant factors into the Modeling design and to check for significant interactions and to check for curvature.

  1. Click Assistant<<DOE<<Analyze and Interpret.
  2. Click on the Create Modeling Design box.

3.  A menu box will appear and ask you if you wish to enter new factor levels. Click on No and you are taken to the Modeling Worksheet Creation Menu.

4.  Most of the required information should be automatically completed by Minitab, Just check the response variable is ‘Power’ and the goal is to ‘Maximize the response’.

5.  Also check the levels and types for the 5 factors we are taking forward to the Modeling Design. Also, we are going to use 2 replicates making 40 runs in total.

6. Click OK to produce the worksheet and the create Summary Report and Report Card. When asked, we don’t actually want to print the worksheet, so click No

Next Page