Design Expert offers you the following choices for adding or augmenting new runs to an existing design:

1. Generate a Foldover, a Semifold-over, or a Factorial D-optimal of your initial design to de-alias 2-factor interactions (2FIs) and resolve any ambiguities.

2. Improve the predictive power of your model for optimization by either adding axial (star) points to transform a screening design to a Central composite response surface method (RSM) design, or using the RSM D-optimal option to select the additional runs needed to update your model.

We need to consider augmenting our designs to de-alias the effects that are of interest. Traditionally, the foldover technique has been used to augment screening designs. However, in the cases of yellow designs and repairing designs & alias structures after failed experiments, alternative strategies such as semi-fold over or D-optimal augmentation offer more efficient ways to add sufficient runs to de-alias key effects.

To handle the augmenting of designs for optimization within the Design Expert DX7 software tool, we have further tipsheets to help you:

   Augmenting for De-aliasing: Foldover a Red Design
   Augmenting for De-aliasing: Semi-fold over a Yellow Design
   Augmenting for De-aliasing: Using Factorial D-optimal 

These tipsheets are linked to this case study

• Move
• Close
  

Top Spray Granulation Process

This Case Study is from pharmaceutical development and covers a series of designs that, when viewed together, illustrate the sequential nature of Design of Experiments (DoE). If you follow these designs through our tipsheets, you will also see how DoE can be successfully integrated into the activities of a Quality by Design (QbD) work flow.

The process developers identified a fluid bed top spray granulation process as the most appropriate commercial process type to manufacture a modified release tablet. A decision was then made about which parameters to Control; treat as Noisy, or eXperiment with (referred to as CNX). It was decided to investigate the effects of six input parameters:

• Batch Size
• Liquid Volume
• Spray Rate
• Screen Size
• Inlet Air Temperature
• Fluid Air Volume

on three output parameters:

• Compressibility
• Dissolution
• Hardness

in order to establish operating ranges (Goals) for the parameters capable of routinely meeting these acceptance criteria:

• Compressibility: minimise < 18%
• Dissolution: target 240min; in range 220-260min/li>
• Hardness: max >11kp

Fold Over, Semi-fold Over and Factorial D-optimal Augmentation to De-alias Effects: At the early stages of an investigation, red and yellow designs (resolution III & IV) are recommended for screening. For this study they help identify main effects and, particularly in the case of yellow designs, let us know if any interactions exist. However, due to the lower resolution of these designs, it is necessary to de-alias 2-Factor Interactions in order to update and improve your process model and process knowledge.

Once you have augmented a design and BEFORE you run the experiments we strongly recommend you go to the Design Evaluation & double-check the new alias structure to make sure you will be able to estimate the effects you want. It is also recommended that you discuss the aliasing with a Statistician or Lead User. 

Augmenting an existing design creates a second block of runs, unless you stipulated otherwise to Design Expert. With the blocking retained, the analysis removes any block-to-block differences due to equipment, environmental or operational changes.

When you augment a design using the three options above, it is unlikely that all the effects will be completely de-aliased or independent. You will see a message that This design is not orthogonal when you click on the Effects button. 

You will see a message that This design is not orthogonal when you click on the Effects button. The fewer additional runs you perform, the more complicated the aliasing will be. After you select your effects on the Normal plots or from the Effects List, do as design expert requests and click Recalculate to see how the effects change in light of what effects you selected. This change will be more dramatic if you chose to run a few additional experiments using Factorial D-optimal compared to the Semifold or complete Foldover.