Central Composite Design Vs Box-Behnken

Some reaction surface approaches include the Central Composite Design and Box-Behnken. Response surface approaches are a collection of computational and mathematical techniques for developing analytical models. Response surface methods aid in the careful design of experiments with the aim of maximizing an output variable that is affected by a variety of independent variables. Both the Central Composite Design and the Box-Behnken approaches to response surface design are commonly used in refining models after the establishment of critical factors through factorial or screening designs. Central Composite Design is a response surface design method that comprises of either fractional or factorial design embedded with center points which are denoted with star points in groups which enables curvature estimation. In the event that “the distance from the center of the design space to a factorial point is ±1 unit for each factor, the distance from the center of the design space to a star point is |α| > 1” (Bajpai, et al. 436). However, under the Central Composite Design, the precise α value relies on particular characteristics desired for the design as well as the number of incorporated factors. In addition, the number of center points runs the central composite design is to have also relied on particular properties needed for the design.

Basically, the central composite design methodology is the often utilized response surface designed experiment. This method can be applied inefficiently estimate both the first and second order terms. It is also used in modeling a response or output variable with curvature through adding a center as well as axial points to an earlier performed factorial design. In addition, the central composite design approaches are essentially very critical in sequential experiments. This is because one can at most of the times develop on past factorial experiments through the addition of both the axial and center points.

Usually, central composite designs method fits in an entire quadratic model. This response surface design is usually used in the event that design plan requires sequential experimentation. This is due to the fact that central composite designs may constitute of data from a well-planned factorial experiment. Basically, central composite design at all time comprises of ”twice as many star points as there are factors in the design” (Bajpai, Shailendra, et al. 436). In addition, central composite design comprises of an adequate capability of orthogonal blocking.

Box-Behnken Design

On the other hand, the Box-Behnken design response surface design method is an independent quadratic design. This is because this response surface design does not comprise of an embedded fractional or factorial design (Ferreira, et al. 179). Under Box-Behnken design, the treatment combinations are basically at the midpoints of edges of the process space as well as at the center. In addition, Box-Behnken designs are rotatable and need 3 levels of every factor. However, Box-Behnken designs comprise of inadequate capability for orthogonal blocking.

Basically, Box-Behnken designs are quite useful in the same environment as the central composite designs. Box Behnken major advantage consists of addressing the challenge of where the response surface experimental boundaries should be as well as to prevent treatment combinations which are excess or extreme. The term extreme is used in Box-Behnken design to refer to the corner points as well as the star points that are extreme points based on the region where the experiment is being done. Usually, Box-Behnken design runs away from all the corner points as well as the star points

Central Composite Design Vs Box-Behnken Design

A comparison of Central composite and Box-Behnken designs reveals that Box-Behnken design has fewer design points as compared to central composite designs. Hence, as compared to central composite design, the Box-Behnken design is less expensive to operate with the same number of factors. The other difference between the two common response surface designs is that at all times Box-Behnken designs have 3 levels per factor, this is contrary to central composite designs that can have up to 5 levels per factor.

Furthermore, Box-Behnken design never consists of runs where all elements are at the extreme condition, this is contrary to central composite designs that include runs at low settings. Moreover, as compared to central composite design, Box-Behnken design has limited capability for orthogonal blocking. The only similarity between the two response surface designs is that both of them can efficiently estimate the first- and second-order coefficients.

Benefits and Drawbacks

Even though, the central composite design is expensive, it is more suitable for response surface design. This is because unlike Box-Behnken design that is cheap, central composite design can be run sequentially into first and second subset estimates and it is very efficient in offering adequate information on experiment variable effects as well as overall experimental error in a few required runs (Verseput, 33). In addition, the central composite design is quite flexible due to the availability of a number of central composite methods which promotes their usage under the variant experimental setting of interest as well as operability.

Work Cited

Bajpai, Shailendra, et al. ”Application of Central Composite Design approach for removal of chromium (VI) from aqueous solution using weakly anionic resin: modeling, optimization, and study of interactive variables.“ Journal of hazardous materials 227 (2012): 436-444.

Ferreira, SL Costa, et al. ”Box-Behnken design: an alternative for the optimization of analytical methods.“ Analytica chimica acta 597.2 (2007): 179-186.

Verseput, Richard. ”Digging into DOE Selecting the right central composite design for response surface methodology applications.“ Quality Digest 21.6 (2001): 33-36.