Include Top

Taguchi L9 (3 Level) Four Factor - Paper Airplane Example

This Paper Airplane Experiment is adapted from the article "Teaching Taguchi's Approach to Parameter Design", by Sanjiv Sarin, Quality Progress, May 1997.

  1. Click SigmaXL > Help > Template Examples > Taguchi Examples > L9 Four Factor - Paper Airplane to open the example file. If you wish to start with a blank template and populate the values, click SigmaXL > Design of Experiments > Basic Taguchi DOE Templates > Taguchi L9 (3 Level) > Four Factor.
  2. The response is Distance of Flight in Inches. The goal is to maximize Distance.

  3. We will maximize the Signal-to-Noise Ratio SN: Larger is Better, selected from the list as shown:

  4.  image1

  5. The Inner Array Controllable Input factors used in this study are: Paper Weight (1 Sheet, 2 Sheets, 3 Sheets), Design (Design 1, Design 2, Design 3), Paper Width (4 inches, 6 inches, 8 inches) and Paper Length (6 inches, 8 inches, 10 inches).

  6.  image3

  7. The Uncontrollable Factors are Launch Height, Launch Angle and Ground Surface. The Outer Array was created using an L4 Three Factor design. Scroll right to view the screen capture:
  8. image4
  9. The distance values are entered into the Outer Array of the L9 as Yrep1 to Yrep4:


    Scroll to the right to view the calculated statistics:

  10. Since there are no interactions, we will focus our analysis on the Main Effects Plots for Signal-To-Noise Ratio and Mean Distance as done in the paper. Scroll right and down to view the Main Effects Plot for SN Ratios:


    The optimum settings to maximize SN Ratio are: Paper Weight = 3, Design = 2, Paper Width = 6 and Paper Length = 10.
  11. Scroll left to look at the Main Effects Plot for Means:

    The optimum settings to maximize the Mean are: Paper Weight = 3, Design = 1, Paper Width = 6 and Paper Length = 10. (Note, The Y-Axis minimum has been adjusted to 40 to improve the scale).
  12. Commenting on this difference, Sarin states:

  13.  image10

  14. We will compare the two settings using the Predicted Output Calculator. Scroll up and enter (or select from the drop-down list) as shown:


    Now change the Design to 1 as shown:


  15. As noted by the author, depending on the priority given to Mean or Signal-To-Noise Ratio, one would select Design 1 or 2. In either case, confirmation runs need to be performed to validate the experimental results.

Define, Measure, Analyze, Improve, Control

Lean Six Sigma Software Excel Add-in

Simulate, Optimize,

Lean Six Sigma Software Excel Add-in

Web Demos

Our CTO and Co-Founder, John Noguera, regularly hosts free Web Demos featuring SigmaXL and DiscoverSim
Click here to view some now!

Contact Us

Phone: 1.888.SigmaXL (744.6295)