dc.description.abstract |
At present the commercially available rapid prototyping (RP) machines can mainly
produce parts that can be used either as models for visualization or for rapid tooling.
The emphasis of the on going research in this field is to produce parts that can
physically imitate and work like a component produced by a conventional
manufacturing technique. Thus the idea is to produce “form-fit-functional” parts
rather than prototypes for visualization. Parts made by metals are of specific interest
and welding based RP has good prospects in this regard; with the specific possibility
to produce fully dense metallic parts and tools. However, the big draw back of using
welding as the deposition process is the large heat input to the substrate or to the
previously deposited layers, thus causing high temperature gradients and resulting in
deformations, warpage, residual stresses, delamination and poor surface quality. In
addition the layer by layer additive manufacturing nature results in non-homogeneous
structures, porosity and anisotropic material properties. Parts thus produced are of
near net shape and out of tolerance. In order to predict and minimize these problems,
knowledge of thermal gradients and temperature history during manufacture is
important. Moreover, to overcome the problem of surface quality and out of tolerance
parts a hybrid welding/CNC milling based RP system can be a good option. These
problems associated with the use of welding as RP tool needs to be minimized by the
proper investigation of the different deposition parameters and process conditions e.g.
intermediate machining, deposition patterns, heat sink size, interpass cooling time,
preheating and constant control temperatures on the material properties and
mechanical behaviors of the finally produced parts.
This dissertation presents an analysis based on a numerical and experimental
approach for the effects of different deposition and process parameters on welding
based rapid prototyping process. The entire work is divided into two main parts. The
first part is an experimental comparison of microstructure and material properties of
the simple GMAW based layered manufacturing (LM) with the hybrid
vwelding/milling based LM process. Material properties were investigated both on a
macro and microscopic level. The microstructure for the two deposition procedures
were studied and compared. The hardness test results for the two procedures were
investigated and the results were studied in the light of the respective microstructures.
Tensile test samples were developed and testing was performed to investigate the
directional properties in the deposited materials. Reaustenitised and un-reaustenitised
regions were found in the entire body of deposition without machining (DWM) while
these are confined to the top layer of deposition with intermediate machining (DWIM)
changing alternatively across the weld direction with intervals equal to the inter-bead
spacing. The central layers of the DWIM deposit comprise only of reaustenitised
region varying sequentially in grain size in both longitudinal and perpendicular
direction. This sequential variation is in accordance with the inter-bead spacing in the
across direction, and with the layer thickness in the perpendicular direction. The
hardness results are in good agreement with the variation of the microstructure both
for DWIM and DWM. The hardness values are higher at the top and interface layer
while it is comparatively less in the central layers of DWIM samples. However, in
DWM samples the hardness values are relatively higher in the top layer only. The
correlation for hardness values as related to the tensile strength also holds within
normal expectations. The tensile test results show no variation in the yield strengths of
samples produced longitudinal and perpendicular to the deposition direction; however
there is a slight difference in elongation. Moreover a sharp yield point was observed
in the DWIM samples in contrast to the DWM samples.
The second part presents a finite element (FE) based 3D analysis to study the thermal
and structural effects of different deposition parameters and deposition patterns in
welding based LM. A commercial finite element software ANSYS is coupled with a
user programmed subroutine to implement the welding parameters like Goldak double
ellipsoidal heat source, material addition, temperature dependent material properties.
The effects of interpass cooling duration were studied and it was found that an
intermediate value of interpass time is suitable for a nominal level of deformations
and stresses. A similar finding was made from the studies about different weld bead
starting temperatures. The studies regarding different boundary conditions revealed
that the deformations are least for adiabatic case while isothermal case produced the
maximum deformations. Simulations carried out with various deposition sequences
virevealed that the thermal and structural effects, on the work piece, are different for
different deposition patterns. The sequence starting from outside and ending at the
center is identified as the one which produces minimum warpage. The results
presented are for deposition by gas metal arc welding but can be applied to other
deposition process employing moving heat source. The parametric results suggest that
in order to minimize the harmful effect of residual stresses, proper combination of
deposition parameters is essential. Proper selection of deposition patterns, substrate
thermal insulation, and nominal interpass cooling / control temperature can reduce the
part warpage due to residual stresses. |
en_US |