Box-Behnken Design

Unlike the Central Composite Design, the Box-Behnken Design is quadratic and does not contain embedded factorial or fractional factorial design. As a result, Box-Behnken Design has a limited capability of orthogonal blocking, compared to Central Composite Design. The main difference of Box-Behnken Design from Central Composite Design is that Box-Behnken is a three-level quadratic design in which the explored space of factors is represented by . The "true" physical lower and upper limits corresponding to . In this design, however, the sample combinations are treated such that they are located at midpoints of edges formed by any two factors. The following is a geometry representation of Box-Behnken designs of three factors:

Example 4:  Box-Behnken Designs of Three Factors

The circumscribed CCD generally provides a high quality of response prediction. However, the circumscribed CCD requires factor level setting outside the physical range. Due to physical and economic constraints, some industrial studies prohibit the use of corner points, where all factors are at an extreme. In such cases, factor spacing/range can possibly be planned out in advance to ensure , so each coded factor falls within feasible levels.

The inscribed CCD uses points only within the specified factor levels. This means that it does not have the "corner point constraint" issue that the circumscribed CCD has. However, the improvement compromises the accuracy of the response prediction near the lower and upper limits of each factor. While the inscribed CCD provides a good response prediction over the central subset of factor space, the quality is not as high as the response prediction provided by the inscribed CCD.

Like the inscribed CCD, the face-centered CCD does not require points outside the original factor range. Compared to the inscribed CCD, the face-centered CCD provides a relatively high quality of response prediction over the entire explored factor space. The drawback of the face-centered CCD is that it gives a poor estimate of the pure quadratic coefficient, which is the coefficient of square term of a factor.

For relatively the same accuracy, Box-Behnken Design is more efficient than Central Composite Design in cases involving three or four factors because it requires fewer treatments of factor level combinations. However, like the inscribed CCD, the prediction at the extremes (corner points) is poor. The property of "missing corners" can be useful if these should be avoided due to physical or economic constraints, because potential of data loss in those cases can be prevented.


Release 18.2 - © ANSYS, Inc. All rights reserved.