Abstract:
Wire+arc additive manufacturing is a technique suitable for the deposition of large
components; a variety of materials can be processed, including titanium.
For the alloy Ti{6Al{4V, an experimental model based on design of experiment
and linear regression was developed to control layer geometry during deposition.
The modelled variables were wall width and layer height; the former was dependent
on the heat input, and the latter on the heat input as well as on the wire feed
speed to travel speed ratio. Equations enabled the automatic selection of process
parameters based on geometric requirements speci c to the part being built. This
could ensure minimisation of production time and material waste.
Additively manufactured parts are a ected by distortion and residual stress; the
e ect of high pressure rolling on these two, as well as on geometry, microstructure
and mechanical properties was studied. Due to plastic deformation, rolled linear
deposits were characterised by a larger width and smaller height. The variability of
the layer height was reduced, a bene cial e ect from a production implementation
viewpoint.
Distortion was less than half in rolled components, a change associated with the
reduction in residual stress which were still tensile in the bottom of the parts and
compressive in their top; however their overall magnitude was smaller than in the
unrolled samples. The contour method showed relatively good agreement with the
neutron di raction measurements, and although destructive it proved to be a fast
way to characterise residual stress in additively manufactured components.
Microstructurally, the columnar prior grains con guration observed in all unrolled
deposits, also a ected by a strong texture developed in the building direction,
was changed to equiaxed grains due to the recrystallisation triggered by both the
strain introduced by rolling and the repeated thermal cycles induced by each layer
deposition. The microstructure was overall considerably ner and the texture randomised.
A fundamental study was performed to discern the extent of the deformed
zone from the one a ected thermally. While the deformed zone could not be identi
ed precisely, the thermally in uenced zone showed a relationship between rolling
load and depth of the recrystallised volume.
Finally, testing of hardness and tensile strength showed superior properties of
rolled specimens than in the unrolled specimens. The mechanical performance of
rolled samples was fully isotropic too.
This project was entirely sponsored by Airbus Group Innovations (formerly
EADS Innovation Works).
