Proceedings of the 3rd Pacific International Conference on Application of Lasers and Optics 2008

DIRECT DIODE LASER DEPOSITION OF FUNCTIONALLY GRADED TI-6AL-4V AND

INCONEL 718 COMPONENTS
Paper 504

Kamran Shah1, Andrew Pinkerton1, Richard Moat2, Lin Li1, Michael Preuss2
1
Laser Processing Research Centre, School of Mechanical, Aerospace and Civil Engineering, The
University of Manchester, PO Box 88, Manchester M60 1QD, United Kingdom
2
Manchester Materials Science Centre, School of Materials, The University of Manchester, Grosvenor
Street, Manchester M1 7HS, United Kingdom

Abstract

Laser direct metal deposition (LDMD) has developed

from a prototyping to a single metal manufacturing
tool. Its potential for creating multi-material and
functionally graded structures is now beginning to be
explored. In this work, three dimensional functionally
graded structures of Ti-6Al-4V and Inconel 718 are
fabricated by laser direct metal deposition (LDMD)
on Inconel 718 substrate. The multi-track parts are
built at a range of powder mass flow rates using a Figure 1 Schematic of laser direct metal deposition
side nozzle and 1.5 kW Diode laser operating in both
continuous and pulsed beam modes. Microstructure Laser direct metal deposition has traditionally been
characterization and phase identification are used to build small prototypes and has also found
performed by optical microscopy and X-ray various applications like component repair [1], and
diffraction. Only two samples are free of macro-scale short-run component fabrication [2]. The use of laser
cracks, one prepared with a continuous laser beam direct metal deposition in the aerospace industry is
and one prepared with the beam in pulsed mode. ever increasing; LDMD of titanium and nickel based
XRD results show the presence of the brittle Ti2Ni alloys is extensively applied to the manufacture of
phase in all samples. small components and to add details to large parts
[3].
Key words: Laser, direct metal deposition, pulsed
wave, continuous wave, Ti-6Al-4V, Inconel 718 The concept of functionally graded materials (FGMs)
was proposed in 1984 in Japan as a means of
Introduction preparing thermal barrier materials [4,5]. Since then
the FGM concept has been extended to a variety of
Laser direct metal deposition (LDMD) is a complex materials in various applications [5]. In a FGM,
process that is used to create metal parts by fusing composition and properties are customised to suit
metal powders or wires to a substrate. In LDMD, a specific engineering applications [6]. The
laser generates a melt pool on a substrate material development of functionally graded materials by
while a second material is delivered from either a laser direct metal deposition has received much
coaxial or side nozzle into the melt pool in the form attention in recent years [7, 8]. The freedom to
of powder and/or wire which melts and forms a selectively deposit different elemental powders at
metallurgical bond with the substrate. Multiple tracks discrete locations using multiple powder feeders
are deposited to make vertical walls, surfaces or makes LDMD well suited to the fabrication of FGMs.
combinations of the two Complex component shapes
can be realized directly from CAD files without using There have been a limited number of studies on the
any specialized tooling. Figure 1 illustrates the basic development of graded structures by LDMD of
components of the process. titanium and nickel alloys. In a detailed study of a
graded structure by deposition, Lin et al [9] W/mm2, to the pilot beam (20 mW) of the diode laser
investigated the solidification behaviour and phase and found to be 2.5 mm (slow axis) × 3.5 mm (fast
evolution of Ti-6Al-4V and Rene 88 DT. Although axis). During deposition the laser moved parallel to
crack free samples were obtained, all proportions of the fast axis. A side feed powder nozzle was
constituent alloys were not investigated and the positioned at a 45° angle to the vertical and in line
compositional gradient tested ranged from 100 pct with the direction of motion. Front feeding was
Ti-6Al-4V to Ti-6Al-4V with 38 pct Rene 88 DT. maintained throughout for all the samples.
Domack and Baughman [10] demonstrated
functionally graded deposition of Ti-6Al-4V and
Inconel 718 samples using a LENS system. The laser
was in continuous wave mode and powder was fed as
pre-mixed powder blends, with the composition
adjusted in steps of 10 percent by volume from 100
percent Ti-6Al-4V to 100 percent Inconel 718.
Macroscopic cracks formed before the full transition
from Ti-6Al-4V to Inconel 718 was achieved but the
cause of these was not investigated in detail in this
work.

In the present study, two different modes of laser

beam, continuous and pulsed, were used to
investigate the effects on the development of
cracking. Constant mean power was used in all of the Figure 2 Wave definition
experiments to remove this variable from the study
and highlight the effects of duty cycle and powder The build materials were Ti-6Al-4V (5.5-6.75% Al,
mass flow rate. 3.5-4.5% V, 0.4 % Fe, 0.2% O, and balance Ti) and
Inconel718 (50-55% Ni, 17.2-21.0% Cr, 18.5% Fe,
Experimental Procedure 3% Mo, 5.1% Nb and 0.20-0.80%Al). The powders
were both gas atomised, giving particles with
A matrix of experiments was performed to
approximately spherical morphology, and had a
investigate properties of a transition from Inconel 718
powder size distribution of 53 – 150 µm. A disc-type
to Ti-6Al-4V introduced during the building of thin
powder feeder was used to deliver them to the nozzle.
walls by LDMD. Walls of approximately 20 mm
The deposition area was also shrouded by an Argon
height were built at various parameter combinations,
gas flow from a side nozzle directed at the melt pool.
with Inconel 718 used for the lower half and Ti-6Al-
The gas pressure for all experiments was set to 5
4V for the upper half in each case.
bars.
The starting point for deposition was Inconel 718
A range of samples were deposited. Two parameters
plates of approximately of dimensions 50 ×50 × 2
were varied: powder flow rate and the duty cycle, A
mm. These were first sand blasted in a Guyson sand
mean power of 600 W had been determined from
blaster and then degreased using ethanol. The
previous experiments as giving a specific energy
absorptivity of diode laser wavelengths (808 and
appropriate for the deposition of both Inconel 718
940nm), measured using a SD2000 fiber optic
and Ti-6Al-4V at the chosen traverse speed and beam
spectrometer was 65 ~ 67 %. The substrate was
area and was kept constant throughout. The
tightly secured on a clamp, which was then mounted
experimental parameters used are shown in Table 1.
onto a 3 axis position and motion control system. A
velocity of 8.75 mm/s was maintained during the
deposition of all the samples.

A 1.5 kW Laserline LDL 160-1500 diode laser,

operated in both continuous and pulsed mode during
the experiment was used to carry out the laser
deposition. In pulsed mode the output was
approximately square wave and defined by the period
and pulse length as illustrated in Figure 2. The beam
size was experimentally measured by exposing an
infrared detector card with a sensitivity of 1.75 × 10-9
 Table 1: Experimental Parameter used during the was mainly in a horizontal direction (as viewed in
experiment transverse cross-section) and, as shown in figure 4,
tended to be located above the most obvious

Powder mass flow

Peak Power (W) A
transition from Inconel 718 to Ti-6Al-4V

Pulse length (ms)

B
Sample No.

Period (ms)

Duty Cycle

rate (g/s)
1 600 - - 1 0.358
2 1500 20 50 0.4 0.358
3 857 35 50 0.7 0.358
4 600 - - 1 0.516
5 1500 20 50 0.4 0.516
6 857 35 50 0.7 0.516
7 600 - - 1 0.674
8 1500 20 50 0.4 0.674
9 857 35 50 0.7 0.674 Figure 3 Micrograph of crack free sample 2.
A
Mean power = 600 W in all cases
B
Duty cycle = pulse length / period

After the experiment, deposited walls were

transversely sectioned and mounted in Struers epoxy
resin. All samples were polished to Ra 1µm using
standard metallographic techniques. The sample was
firstly etched using HF solution to reveal Ti-6Al-4V
microstructure. Microstructures were evaluated using
optical microscopy. Samples were then repolished
and areas containing Inconel 718 electrolytically
etched in 10% oxalic acid solution using 6V DC and
re-evaluate using microscopy. Figure 4 Micrograph of cracked sample 9
X-ray diffraction phase analyses were carried out
using a Philips X'Pert MPD X-ray diffractometer. Both the Inconel 718 and Ti-6Al-4V side of the
with a Cu anode to provide radiation of 1.5405 Å deposited walls showed a dendritic structure, as
wavelength. The samples were scanned between shown in Figure 5.
°2Th angles of 6 and 95 degrees with a 30 second
scan step time.

Results

Microstructure and Visible Cracking

Examination of the final Ti-6Al-4V and Inconel 718

structures showed that only two out of nine samples
were found crack-free. Sample 1 (continuous wave,
low powder mass flow rate) and sample 5 (0.4 duty
cycle, medium powder mass flow rate) showed no
signs of cracking under optical microscopy. Figure 3
shows the crack free interface between Inconel 718
and Ti-6Al-4V of sample 2.

Other samples developed severe macroscopic cracks Figure 5 Dendritic structures in the Inconel 718 and
during the deposition process. In these the cracking Ti-6Al-4V sides of the wall in Sample 9
Dendrites were generally elongated in the vertically
direction and normal to any clear Ni-Ti alloy
interface. Porosity was observed in all of the samples.
The pores were approximately circular and their size
varied between 10-30 µm. Generally more porosity
was observed in the Ti-6Al-4V than in the Inconel
718.

Incompletely melted Ti-6Al-4V powder particles

were observed near the sides of the walls as shown in
Figure 6.

Figure 6 Residual Ti-6Al-4V particles found in

Sample 6 near the side of the wall approximately
3mm above the interface

Phase Analysis using X-ray Diffraction

Despite the small size of the samples, clear spectra

with well defined diffraction peaks were obtained.
Figure 7 shows X-ray diffraction spectra from a crack
free sample (sample 2, a) and a cracked sample
(samples 6, b).
Figure 7 Reference XRD diffraction angles and
The two initially appear very different but reveal the measured spectra showing the presence of various
presence of the same phases: Titanium Chromium, phases in deposited samples (a) crack free sample 2
Ti0.8Cr0.2; Titanium Aluminum, Ti0.85Al0.15; (b) cracked sample 6.
Titanium Nickel, Ti2Ni and Ti0.14Ni0.86 and γ`-
Ti0.14 Ni0.86 in both of the samples. Sample 2 had several theories have been presented, identifying
strong peaks at approximately 40o indicating a numerous factors for crack initiation. These factors
significant presence of Titanium Chromium, can be divided into the metallurgical and the
Ti0.8Cr0.2 and also Titanium Aluminum, mechanical. Metallurgical factors mainly concern
Ti0.85Al0.15. These were not so strong in sample 6, phase relationships, while mechanical factors focus
which showed a strong presence of γ`-Ti0.14 Ni0.86. on stress and strain behaviour.
Quantification of the proportions of the phases Considering metallurgical factors; XRD results show
present was not possible. the presence of Ti2Ni in both cracked and un-cracked
samples. Ti2Ni is recognised as one of the brittle
Discussion phases responsible for cracking in welding and
appropriate selection of laser parameters may not be
The majority of Ti-6Al-4V - Inconel 718 walls sufficient to avoid the production of such
developed cracks that are similar to the solidification intermetallics. Chatterjee et al [14] reported that
cracks reported by several researchers during laser during laser welding of nickel and titanium it grows
and gas tungsten arc (GTA) welding [11-13]. toward the Ti base metal, along the direction of heat
Solidification cracking is a complex phenomenon and
removal. These authors’ results demonstrated the both cracked and un-cracked walls. Thermal
importance of the interaction of phase diagram expansion coefficient differences during cooling are
considerations with the transport phenomena in the one probably cause of the stresses that are causing the
weld pool. Transport phenomena in a deposition or cracks.
weld pool are changed significantly by the laser pulse
parameters and to a lesser extent by powder mass Acknowledgements
injection rate [15]. In this case these may have
affected phase composition sufficiently to affect The authors would like to thank Ms Judith
susceptibility to cracking and this in re-enforced by Shackleton of Manchester Materials Science Centre
the slightly different proportion of phases that can be for help with obtaining and interpreting the XRD
inferred from the XRD results. All samples indicated results.
good intermixing within the melt pool, but this was
particularly true of samples at low duty cycle. References

Considering mechanical factors; stresses can be

generated by constrained elastic expansion or [1] Yamada, K., Morisita, S., Kutsuna, M. & Ikeda,
contraction due to transient temperature gradients, T. (2003) Direct diode laser cladding of Co based
constrained elastic expansion or contraction due to alloy to dual phase stainless steel for repairing the
thermal expansion coefficient mismatch and by machinery parts, in First International Symposium on
changes in specific density due to solid phase High-Power Laser Macroprocessing, Osaka, Japan,
transformations. In this case, cooling rates are fast so 65-70.
the first factor will be significant and there is a large
difference in the thermal expansion coefficients of [2] Krantz, D., Nasla, S., Byrne, J. & Rosenberger,
Ti-6Al-4V (8 × 10-6 K-1) and Inconel 718 (1.3 × 10-5 B. (2001) On-demand spares fabrication during space
K-1) so the second will also contribute. Henderson et missions using laser direct metal deposition,
al [16] reported that solidification cracks are initiated Albuquerque, NM, USA, 170-175.
if the thermo-mechanical strains caused by the time-
dependent temperature distribution during welding [3] Gao, S.Y., Zhang, Y.Z., Shi, L.K., Du, B.L., Xi,
and the material-specific shrinkage during liquid/ M.Z. et al (2007) Research on Laser Direct
solid phase transition cannot be compensated by the Deposition Process of Ti-6Al-4V Alloy, Acta
ductility of the metal or by reflow of melt. It was Metallurgica Sinica (English Letters) 20(3), 171-180.
probably the combination of the different stresses
from these sources and the brittle phases generated [4] Koizumi, M. (1997) FGM activities in Japan,
that led to the cracks seen. Composites Part B:Engineering 28(1-2), 1-4.

As reported by Moat et al [17], in this parameter [5] Wholer, T. (2003) Wohlers Report "Rapid
range there is no significant difference in the levels of Prototyping and Tooling State of the Industry,
residual stresses generated during deposition of Annual Worldwide Progress Report.
nickel alloys by pulsed or continuous beam.
Although that study was with a single material, it [6] Liu, W. & DuPont, J.N. (2003) Fabrication of
infers that any effect the pulse parameters had on the functionally graded TiC/Ti composites by Laser
stresses generated in this case was not an important Engineered Net Shaping, Scripta Materialia 48(9),
factor. Finite element modeling or neutron stress 1337.
analysis of uncracked samples would help to confirm
this inference. [7] Pei, Y.T. & De Hosson, J.T.M. (2000)
Functionally graded materials produced by laser
Conclusion cladding, Acta Materialia 48(10), 2617.
The effect of pulsed and continuous wave deposition [8] Yarrapareddy, E., Zekovic, S., Hamid, S. &
of functionally graded Ti-6Al-4V and Inconel 718 Kovacevic, R. (2006) The development of nickel-
wall has been investigated. Formations of cracks tungsten carbide functionally graded materials by a
were observed in most of the walls. Only two laser-based direct metal deposition process for
samples were crack free, one prepared with a industrial slurry erosion applications, Proceedings of
continuous wave beam and low powder mass flow the Institution of Mechanical Engineers, Part B
rate and the other with a pulsed beam with 0.4 duty (Journal of Engineering Manufacture) 220(B12),
cycle and moderate powder mass flow rate. XRD 1923-1936.
results showed the presence of brittle intermetallics in
[9] Lin, X., Yue, T.M., Yang, H.O. & Huang, W.D.
(2007) Solidification behavior and the evolution of
phase in laser rapid forming of graded Ti6Al4V-
Rene88DT alloy, Metallurgical and Materials
Transactions A: Physical Metallurgy and Materials Meet the Authors
Science 38(1), 127-137.
Mr Kamran Shah obtained an MSc in advanced
[10] Domack, M.S. & Baughman, J.M. (2005) Manufacturing technology and Systems Management
Development of nickel-titanium graded composition from The University of Manchester in 2005. He is
components, Rapid Prototyping Journal 11(1), 41-51. currently a research student, investigating direct
manufacture of functionally graded materials and
[11] Nelson, T.W., Lippold, J.C. & Mills, M.J. combined wire-powder laser deposition.
(2000) Nature and evolution of the fusion boundary
in ferritic-austenitic dissimilar metal welds - Part 2: Dr Andrew Pinkerton is a University of Manchester
on-cooling transformations, Welding Journal (Miami, Research Fellow who has studied laser materials
Fla) 79(10), 267-277. processing for six years. He specializes in laser direct
metal deposition and analytical modelling methods.
[12] Sun, Z. & Moisio, T. (1994) Weld metal/ferritic
Mr. Richard Moat graduated in 2004 and completed
steel interface in laser welded austenitic/ferritic
a MSc in lightweight materials at the Manchester
dissimilar steel joints, Journal of Materials Science
Materials Science Centre in 2005. He is currently a
Letters 13(11), 802-805.
research student, investigating superalloy metallurgy
in the DLD process.
[13] Sireesha, M., Shankar, V., Albert, S.K. &
Sundaresan, S. (2000) Microstructural features of Professor Lin Li is Director of the Laser Processing
dissimilar welds between 316LN austenitic stainless Research Centre at the University of Manchester with
steel and alloy 800, Materials Science and over 400 publications in laser materials processing
Engineering A 292(1), 74-82.
Dr Michael Preuss is a lecturer at The University of
[14] Chatterjee, S., Abinandanan, T.A. & Manchester. His work focuses on metallurgical and
Chattopadhyay, K. (2006) Microstructure residual stress aspects in high temperature materials
development during dissimilar welding: Case of laser for aero engines and nuclear application and
welding of Ti with Ni involving intermetallic phase advanced welding techniques.
formation, Journal of Materials Science V41 (3), 643.

[15] Roy, G.G., Elmer, J.W. & DebRoy, T. (2006)

Mathematical modeling of heat transfer, fluid flow,
and solidification during linear welding with a pulsed
laser beam, Journal of Applied Physics 100(3),
034903.

[16] Henderson, M.B., Arrell, D., Larsson, R.,

Heobel, M. & Marchant, G. (2004) Nickel based
superalloy welding practices for industrial gas turbine
applications, Science and Technology of Welding &
Joining 9, 13-21.

[17] Moat R, Pinkerton A, Hughes D, Li L, Preuss M

et al (29 October 2007 ) Stress Distributions in
Multilayer Laser Deposited Waspaloy Parts
Measured using Neutron Diffraction, in 26th
International Congress on Applications of Lasers and
Electro-optics (ICALEO), USA.