One can see from Equation 7 that the interactions effects are aliased with the main effects. Expected interactions are:
- It is likely that a given thermal factor (\(a_{1-4}\)) shows interactions with the other thermal factors. Therefore are relevant the thermal cross-factors: \(a_{12}\), \(a_{13}\), \(a_{14}\), \(a_{23}\), \(a_{24}\), \(a_{34}\).
- It is likely that a given mechanical factor (\(a_{5-7}\)) shows interactions with the other mechanical factors, hence for the mechanical counterpart: \(a_{56}\), \(a_{57}\), \(a_{67}\).
- The relation between the thermal and the mechanical parameters is not obvious. Imagining that it may depend on combined effect of a thermomechanical fatigue-related phenomenon, it would concern the rest of the 2x2 coefficents that appear in Equation 7.
- It could be possible that the contrasts \(l_{1-4}\) only comprise the thermal 2x2 cross-factors, while \(l_{5-7}\) only are aliased by the mechanical 2x2 cross-factors.
Since the interactions effects are pure alias, a Pareto chart can also be used to represent the main effects' importance. By considering only the magnitude (absolute value) of the half effects (since a Pareto chart by definition can only have positive values), one gets (Figure 10).