Reliability HotWire

Reliability HotWire

Issue 111, May 2010

Tool Tips

*What is the difference between Taguchi orthogonal array designs and Taguchi robust designs in DOE++?

Both Taguchi orthogonal array designs and Taguchi robust designs are based on the work of Dr. Genichi Taguchi. The design types serve different purposes, however, and are created in different ways.

Taguchi orthogonal arrays are a type of general fractional factorial design. These highly fractional orthogonal designs, based on a design matrix proposed by Dr. Genichi Taguchi, allow you to consider a selected subset of combinations of multiple factors at multiple levels. Orthogonal arrays are balanced to ensure that all levels of all factors are considered equally. For this reason, the factors can be evaluated independently of each other despite the fractionality of the design.

To create a Taguchi orthogonal array (OA) design in ReliaSoft’s DOE++ software , select Factorial Design in Step 1 of the Design Wizard and then select Taguchi OA Factorial Design in Step 2.

Taguchi OA designs deal exclusively with "control factors," or factors of interest. If you need to deal with "noise factors," or factors beyond the control of the operator, you should use a Taguchi robust design.

The purpose of Taguchi robust designs is to minimize the variability of the response in spite of noise factors. This is done by combining an inner array of control factors with an outer array of noise factors. All control factor settings combinations specified in the inner array are tested at each noise factor settings combination specified in the outer array. In DOE++, the inner array can use a Taguchi OA, two level full factorial, two level fractional factorial, Plackett-Burman or general full factorial design. The outer array can use a two level full factorial, two level fractional factorial, Plackett-Burman or general full factorial design.

To create a Taguchi robust design in DOE++, select Taguchi Robust Design in Step 1 of the Design Wizard.

Once you have created the Taguchi robust design and entered the data, you can specify the equation to be used in calculating the signal-to-noise ratio. Your selection for this option is determined by the purpose of your analysis:

  • Nominal (or "nominal-the-best") should be used if you have a specific target value for the response.
  • Larger (or "larger-the-better") should be used if you want to maximize the value of the response.
  • Smaller (or "smaller-the-better") should be used if you want to minimize the value of the response.

*How can I find out when ReliaSoft has issued a free Service Release?

ReliaSoft uses minor version "service releases" to resolve any critical issues with the software. These updates are free to registered users. We recommend that you check occasionally to ensure that you always have the latest fixes and enhancements. There are two easy ways to stay informed about the latest updates:

  1. Check the ReliaSoft.com Web site. New service releases are announced on the site home page and the latest build number (e.g. 7.5.9) is identified in the upper right corner of each product home page. If the build date displayed in the About window for the software on your computer does not match the latest build, click the Update link on the Web page to download the latest Service Release installation.

  1. Sign up for a free RSS feed that provides a notification for each software release.