Weighted Inertial Tolerancing

Maurice Pillet (1, 2) , Daniel Duret (1) , Alain Sergent (1)

1 SYMME - Laboratoire SYstèmes et Matériaux pour la MEcatronique
2 LISTIC - Laboratoire d'Informatique, Systèmes, Traitement de l'Information et de la Connaissance

The objective of tolerancing methods is to limit the variations of a characteristic while trying to minimize the cost of realization. Traditionally expressed in the form of an interval (min, max), it can also be expressed in a different form as in the case of inertial tolerancing. The principle of inertial tolerancing consists of tolerancing the mean square deviation in relationship to the target. This new tolerancing method has many properties that the property of additivity of the mean square deviation offers. However, when several characteristics are added to give a resulting characteristic, it leads to a significant tightening of the variations around the target in the case of a process with a small dispersion. Our proposal consists of defining a new alternative of inertial tolerancing: weighted inertial tolerancing. Its goal is to obtain the best possible compromise between statistical tolerancing and worst-case tolerancing method. One will be able to use it when it is not useful to guarantee an inertia on the resulting characteristic, but simply to limit the variations compared with the target.

Mots clés: Statistical tolerancing; Worst-case tolerancing; Inertial tolerancing; Capability; Taguchi loss function

Domaines: Matériaux et structures en mécanique [physics.class-ph]; Matériaux et structures en mécanique [physics.class-ph]

Maurice Pillet, Daniel Duret, Alain Sergent. Weighted Inertial Tolerancing. Quality Engineering, 2005, 17 (4), pp.687-693. ⟨10.1080/08982110500251279⟩. ⟨hal-00258059⟩