A statistical approach for tolerancing from design stage to measurements analysis

Ambre Diet (1) , Nicolas Couellan (2, 3) , Xavier Gendre (3, 4) , Julien Martin (1) , Jean-Philippe Navarro (1)

1 Airbus Operation S.A.S.
2 ENAC - Ecole Nationale de l'Aviation Civile
3 IMT - Institut de Mathématiques de Toulouse UMR5219
4 ISAE-SUPAERO - Institut Supérieur de l'Aéronautique et de l'Espace

Starting from the development phase, tolerance design must be accurate enough to not only hedge against various uncertainties in order to ensure assembly feasibility but also minimize production cost and avoid expensive over-quality. Once tolerances are agreed, the production allows tolerance features observations and we propose a verification and correction on initial model based on the knowledge of measurement data. The feedback consideration also enables risk evaluation of each tolerance and a more accurate limit definition knowing measures of other assembly contributors is proposed. In addition, we propose an algorithm to optimize the tolerance sharing within a stack chain based on various relevant cost criteria. Finally, an example of tolerancing industrial applications on aerostructures use-cases is detailed to illustrate the methodology from tolerance design to feedback measurement analysis.

Mots clés: Tolerance design; manufacturing; tolerance analys; Assembly; Design method; Family identification

Domaines: Statistiques [stat]

Ambre Diet, Nicolas Couellan, Xavier Gendre, Julien Martin, Jean-Philippe Navarro. A statistical approach for tolerancing from design stage to measurements analysis. CIRP CAT 2020 16th CIRP Conference on Computer Aided Tolerancing, Jun 2020, Virtual event, United States. pp.33-38, ⟨10.1016/j.procir.2020.05.171⟩. ⟨hal-02976044⟩