Reduce the cost of your DOE’s while still obtaining satisfactory results -

Reduce the cost of your DOE’s while still obtaining satisfactory results

A Remarkable Resolution V DOE with six factors in 22 runs

It is always desirable to have fractional factorial DOE’s at Resolution V upwards. Main effects are estimated and two factor interactions are only aliased with three factor or higher order interactions. To achieve this desirable condition with six factors in standard designs requires 32 runs (a Resolution VI design). See the alias structure in Table 1.

Using Design Expert DOE software it is possible to achieve a resolution V design with just 22 runs. See the alias structure in Table 2. This represents a saving of 10 runs while still avoiding aliasing among the two factor interactions!!! This is a remarkable result.

Learn more about the most efficient designs available by attending the SQT Advanced Design of Experiments training course with DOE expert Albert Plant.

Table 1
Alias structure for Six Factors in 32 runs – a Resolution VI design – two factor interactions are not aliased with other two factor interactions

Table 2

Alias structure for Six Factors in 22 runs – a Resolution V design – most importantly, two factor interactions are not aliased with other two factor interactions.

The table includes the alias structure for just two of the fifteen two factor interactions. However, a display of the full alias structure will show that all of the fifteen two factor interactions are free of aliasing with other two factor interactions.