P. Paillard, L. Couturier
During welding and metal additive manufacturing operations, deformation of assemblies and parts can occur. In the case of welding, the parts are generally clamped, which reduces deformations but generates stresses withinin the materials. These two phenomena are due to the thermal cycles that the assemblies and parts undergo. In order to reduce the strains and stresses as much as possible, it is advisable to quantify them and, to know the predominant parameters affecting them in order toso strategies set up strategies for reducing these phenomena may be set up. To this aimFor this we adopt bothhave experimental and numerical approaches.
Keywords: Welding, Metallic Additive Manufacturing, Residual stresses, Strains
During the metal additive manufacturing of parts, for example using WAAM (Wire and Arc Additive Manufacturing) for the building of the parallelepiped displayed in Figure 1, the latter are likely to deform and deform the substrate. We seek to develop numerical models to predict deformations and residual stresses in Metal Additive Manufacturing parts (Wire-Arc and Wire-Laser) as well as in welded assemblies. In order for our models to be as predictive as possible, we take into account the influences of thermics, metallurgy and mechanics in our models.
|Figure 1 : WAAM deposit of a solid piece of austenitic 304L stainless steel on a 4 mm thick 304L substrate|
Figure 2 shows a measurement of deformation of a single bead wall and its substrate realised with a 3D laser scanner. We have shown on a number of examples, on the deformation the influence of the deposition parameters such as clamping, welding energy, deposition strategy, ...
|Figure 2 : Laser scanner measurement of the deformation of a WAAM (single-bead wall) piece in 304L stainless steel|
In parallel with these experiences; we have developed digital models in collaboration with colleagues from IRDL in Lorient. After selecting and testing the type of energy distribution on the substrate, we succeeded in reproducing the geometry of the deposit of a single bead (Figure 3) and then a serie of vertically stacked beads thus forming a wall (Figure 4). The thermal part of the numerical modeling being correct (Figure 5), we simulated the mechanical part in order to determine the deformations of both the substrate and the part (Figure 6). In this first approach, we have neglected the effect of metallurgy. Indeed, the case study using austenitic stainless steel (304L), we were able to neglect the metallurgical transformations in the solid state of the material. On the other hand, as can be seen in Figure 6, the choice of the type of energy source is not without consequences.
|Figure 3 : finite element thermal modeling of a first WAAM deposit bead|
|Figure 4 : modeling of a multi-bead WAAM part and comparison with experience
|Figure 5 : comparison between simulated and measured part temperatures
|Figure 6 : comparison between the measured deformations of the austenitic stainless steel substrate and the simulated deformations according to 2 models: Goldak, Mobile Melted Zone
Afin de complexifier nos études, nous sommes passés sur l’étude d’un matériau qui possède des transformations métallurgiques à l’état solide (austénitisation au chauffage et transformation martensitique au refroidissement) : l’acier inoxydable martensitique 415. De ce fait, nous allons pouvoir faire des modélisations numériques en prenant en compte la thermique, la métallurgie et la mécanique. Les premières expérimentations que nous avons réalisées sont présentées sur la Figure 7. On voit sur cette figure la grande influence de la stratégie de dépôt (les cordons faits les uns derrière les autre sans pause (a), une pause de 150 secondes entre chaque cordon mais sans atteindre la température de début de transformation martensitique lors du dépôt de chacun des cordons (b), un refroidissement complet entre chaque cordon et donc autant de transformations martensitiques que de cordon (c)) sur la forme des pièces obtenues (d, e, f) et au final sur la déformation de la pièce et du substrat (g, h, i).
|a b c|
|d e f|
|g h i|
|Figure 7 : influence of the deposition strategy on the deformations of martensitic stainless steel parts: (a), (d) and (g) the layers are deposited without pause between the layers; (b), (e) and (h) a 150 seconds pause is made between each of the layers; (c), (f) and (i) a pause enabling the complete cool-down of the part between each layer is used. (a), (b) and (c): recordings of thermal cycles; (d), (e) and (f): macrographs; (g), (h) and (i): strain measurements|
The above work is taken from the thesis of Lauriane GUILMOIS (PERFORM project of the IRT Jules Verne, collaboration with the IRDL) and will be continued as part of the thesis of Clément LE FAHLER (PERFORM project of the IRT Jules Verne, collaboration with the IRDL).
As part of Juliette THEODORE's thesis (PERFORM project of the IRT Jules Verne, collaboration with the GeM lab), we will address the phenomena of deformation and residual stresses in the context of assemblies of thick plates. In this case, the approach will be more metallurgical with the use of one or more input metals in order to reduce these phenomena without altering the mechanical strength of the assembly.
IRT Jules Verne, IRDL, GeM, NAVAL GROUP