Modélisation numérique du billage ultrasonore
Areva, ESI France et l’Insa de Lyon nous démontrent dans cet article comment la modélisation numérique du billage ultrasonore permet à la fois de simuler le procédé au travers du mouvement des billes dans l’enceinte de l’outil et de prédire le niveau des contraintes résiduelles dans la pièce traitée.
© Areva
Figure 1: Experimental USP device [6].
Figure 2: USP principle.
Figure 3: Effect on surface.
Figure 4: Fixation system of the specimen by bolts and expected residual stress profile due to USP treatment.
Figure 5: Formation of arc height due to USP treatment.
Figure 6: Specimens treated with different exposure times (the specimens are initially painted before the USP treatment).
Figure 7: Example of contact validation.
Figure 8: Evolution of restitution coefficient on the part.
Figure 9: Evolution of restitution coefficient on the bead.
Figure 10: Evolution of contact radius with the speed.
Figure 12: Numerical distribution of impacts.
Figure 13: Evolution of the average speed.
Figure 14: Distribution of normal contact speed.
Figure 15: Evolution of coverage.
Figure 16: Evolution of the average speed.
Figure 17: Distribution of normal speed at contact.
Figure 18: Deformation of a representative zone after Finite Element simulation of the impact history.
Figure 19: Maximum strain rate [s-1] during a normal impact (initial velocity = 4 m/s, ball diameter = 4 mm).
Figure 20: Local representative area and sequence of impacts (the pattern presents a 100% peening coverage).
Figure 21: Benchmark results: residual surface displacement as a function of the radial coordinate x, normalized by the maximum contact radius a*.
Figure 22: Benchmark results: accumulated plastic strain at the centre of the impact as a function of the depth z, normalized by the maximal contact radius a*.
Figure 23: Transfer of the averaged inelastic strain tensor from the SA model to the FE model
Figure 24: Almen-type experimental results for initial velocity = 4 m/s. Comparison of simulated and XRD determined residual stresses for T = 100% (a), T = 150% (b) and T = 200% (c).