See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/233519186
Laser cladding repair of turbine blades in power plants: From research to
commercialisation
Article in International Heat Treatment & Surface Engineering · September 2009
DOI: 10.1179/174951409X12542264513843
CITATIONS
READS
27
1,728
5 authors, including:
Milan Brandt
Shoujin Sun
RMIT University
Griffith University
299 PUBLICATIONS 4,266 CITATIONS
95 PUBLICATIONS 2,342 CITATIONS
SEE PROFILE
Philip Bendeich
Australian Nuclear Science and Technology Organisation
43 PUBLICATIONS 419 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
3D Printed(SLM) Titanium Scaffold for Improved Bone Tissue Regeneration View project
Development of Ti64 TPMS Gyroid Structures via SLM for Orthopaedic Implants View project
All content following this page was uploaded by Milan Brandt on 08 July 2015.
The user has requested enhancement of the downloaded file.
SEE PROFILE
TECHNICAL PAPER
Laser cladding repair of turbine blades in
power plants: from research to
commercialisation
M. Brandt*1, S. Sun1, N. Alam2, P. Bendeich3 and A. Bishop4
Reliable and efficient power generation is a major global issue due to both political and
environmental concerns. Nevertheless, many critical components, particularly the blades of the
low pressure (LP) side of power generating steam turbines, are subject to failure due to severe
erosion at the leading edges. Since taking machines offline for maintenance and removal of
damaged blade for repair is extremely expensive, increasing the service life of these critical
components offers significant economic and political benefits. Conventional techniques to
increase service life include brazing of an erosion shield at the leading edge of the turbine blade,
open arc hardfacing, and cladding with erosion resistant materials using gas tungsten, manual
metal or plasma transferred arc welding. The authors have been investigating since 2001 the use
of laser cladding technology to deposit a high quality and erosion resistant protection shield on
the leading edge of LP blades. The project has demonstrated the feasibility of in situ repair of
turbine blades in trials conducted at a power station using a fibre delivered diode laser and a
robot. A company, Hardwear Pty Ltd, was established in late 2005 to commercialise this
technology and has to date carried out successfully several commercial contracts involving the
repair of 340 LP blades.
Keywords: Steam turbine erosion, Diode laser, In situ repair, Laser cladding, Low pressure steam blades
Introduction
In the conventional generation of electricity from fossil
fuelled power stations, a boiler is used to heat water to
produce steam. This steam is superheated and then
enters a turbine where the stored energy is used to turn
the turbine shaft which then turns a generator.
Superheated steam is very dry and causes no mechanical
damage to the blades. In a typical boiler, the superheated steam enters the high pressure stage of the
turbine at y545uC and 16?5 MPa pressure. The same
steam is returned to the boiler through the hot reheat
system, after which it enters the intermediate pressure
stage of the turbine, again at y545uC but only at
4?5 MPa pressure. The steam then goes directly to the
low pressure (LP) stage by which time the inlet temperature has dropped to y215uC and the outlet pressure
is basically below atmospheric (5?7 kPa) as the steam
enters the condenser.
A typical LP stage is shown in Fig. 1.
1
Industrial Laser Applications Laboratory, IRIS, Swinburne University of
Technology, Melbourne, Vic., Australia
CSIRO Materials Science and Engineering, Melbourne, Vic., Australia
3
Australian Nuclear Science and Technology Organisation, Lucas Heights,
Sydney, NSW, Australia
4
Welding Technology Institute of Australia, Melbourne, Vic., Australia
2
*Corresponding author, email mbrandt@swin.edu.au
ß 2009 IHTSE Partnership
Published by Maney on behalf of the Partnership
DOI 10.1179/174951409X12542264513843
As the steam exits the turbine, the pressure drop may
be enough to start the condensation of water droplets.
This is a function of the turbine design and the
temperature–pressure relationship at the exit. Since
these droplets have greater mass and inertia than the
vapour phase, they do not attain the same velocity as the
expanding steam. Their velocity relative to the leading
edge of the last row blade, commonly manufactured
from martensitic stainless steel, UNS 42000, is very high
and oriented directly towards the leading edge leading to
water droplet erosion.1,2 One of the methods the
manufacturers have used to minimise the erosion is to
add a shield to the blades in the area where there is
erosion. The materials tried as a shield include Stellite 6,
a cobalt based alloy and tool steels. The shields are often
used as pressed and sintered part and are attached to the
blade using a silver brazing alloy. This has proven to
extend the life of the blades considerably, but over time
shields also erode as shown in Fig. 2a. On some blades,
the shields are too short so that the blade material itself
is subject to severe erosion after only a few years in
service as shown in Fig. 2b.
Over the years, researchers have investigated the
potential of repairing erosion damaged turbine blades
by laser cladding. Laser cladding is a laser surfacing
process in which the objective is to cover a particular
part of the substrate with material which has superior
International Heat Treatment and Surface Engineering
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:10:13
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
2009
VOL
3
NO
3
1
Brandt et al.
;
Laser cladding repair of turbine blades in power plants
costs involved in replacing blades are substantial. Each
LP blade is in the order of AU$5000 and since there are
typically 90 blades per row and two rows affected per
tribune, blade replacement alone is some A$1m. Added
to this is the production loss of electricity of y30 days
in the range from A$100 000 to A$200 000 per day
adding up to another A$3m to A$6m.
In 2001, the Cooperative Research Centre for Welded
Structures and the Welding Technology Institute of
Australia arranged funding from a group of power
stations in Australia to develop a suite of research
projects that could benefit the group. One of the
research projects was to investigate the possibility of in
situ laser repair of LP blades, that is, on site repair and
without removing the blades from the rotor. The vertical
blade surface and restricted space in which the clad had
to be applied presented some challenges from the
practical application of the technology (Fig. 3). At the
start of the project, laser cladding a surface that was not
horizontal created complexity due to the effects of
gravity on the coaxial powder delivery to the melt pool.9
In such nozzles, the powder biased towards the bottom
of the coaxial delivery annulus resulting in an uneven
distribution of powder across the melt pool and a
potential for clogging of powder during delivery within
the nozzle. Experimental results showed that with a
biasing of powder flow, a maximum tilt angle of y20u
from the vertical was possible before there were
significant effects within the clad layer.
The design of the powder delivery nozzle was therefore a critical element of this project. It had to have a
coaxial powder delivery, operate in a horizontal direction, be compact enough to fit the space between the
blades and be robust enough to work outside the
laboratory. This was achieved by developing a distributed type nozzle with multiple powder delivery apertures
rather than a single annulus. By measuring the laser
beam intensity profiles at and around the focus, it was
possible to arrange the powder feed to as closely match
the laser spot on the surface of the blade.10 By limiting
the laser power used to 2 kW at the workpiece and using
standoff distances greater than 10 mm, high reliability
clads up to 1 mm in thickness with powder efficiencies
typically y50% were obtained. The cladding head was
also constructed in a modular form to allow rapid
replacement of components should failure occur and to
facilitate rapid change to cladding conditions or powder.
ONLINE
COLOUR
ONLY
1 Typical low pressure stage from 200 MW turbine showing the blades
properties, producing a fusion bond between the two
with minimal mixing (dilution) of the clad by the
substrate. The process has received a lot of attention
over the years and is now applied commercially in a
range of industries such as the automotive, mining and
aerospace. For turbine blade repair, the process offers
controlled dilution and bonding with the substrate, low
heat input, hence low level of distortion and fine
microstructure due to rapid melt cooling rates.3–6 In
addition, the developments in laser technology itself
such as the fibre coupled diode lasers and fibre lasers7,8
open the possibility of bringing the laser to the part.
Laser cladding technology has been successfully
applied to the repair of leading edge of turbine blades
with the increased complexity of cladding blades, while
they are still attached to the LP rotor (total mass: 27 t)
and while located at the power station, that is, in situ.
The technology involves a portable fibre coupled diode
laser, a robot and a specially developed powder delivery
nozzle. Presented and discussed in this paper are some of
the issues examined in achieving this outcome.
Powder delivery and cladding nozzle
Power station operators when faced with blade erosion
issues have the choice of completely replacing worn
blades with new blades or fitting new shields. In either
event, the blades have to be removed from the rotor. The
ONLINE
COLOUR
ONLY
2 Erosion of a shield and b blade past shield after some 10 and 8 years in service respectively
2
International Heat Treatment and Surface Engineering
2009
VOL
3
NO
3
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:10:41
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
Brandt et al.
Laser cladding repair of turbine blades in power plants
;
ONLINE
COLOUR
ONLY
3 a position of blades on rotor showing limited access to leading edge: spacing between blades is y40 mm; shown is
compact powder delivery nozzle which was developed for application, and b powder flows from nozzle with and without shielding gas
In service, the cladding head was found to be
remarkably robust in its performance being able to
provide reliable clads over significant changes in nozzle
workpiece distance, powder types and surfaces angled
with respect to the incident laser beam. An international
patent application on the nozzle design has been
lodged.11
Experimental programme
Initially, a series of experiments were conducted in the
laboratory using a fibre coupled Nd:YAG laser.
Following the purchase of a fibre coupled diode laser,
further experiments were conducted. The experiments
investigated the optimum parameters for producing
metallurgically sound clad layers with low dilution,
and no cracks or porosity.12,13 In addition, experiments
were also carried out on a range of alloy powders to
determine their erosion resistance. Other issues were
anticipated. These included the development of residual
stresses during cladding and the effect the distortion
would have on the blades. Also considered was the issue
of robot programming for a relatively complex surface
profile of the LP blades.
This experimental programme took some 3 years to
complete before embarking on field trials.
Materials
The powder used was Stellite 6 with the particle size in
the range of 50–120 mm. The substrate initially was a
plate 10 mm thick of SS 420 grade stainless steel, which
has a similar chemical composition to that of LP steam
turbine blades. Following this, experiments were conducted on spare worn blades to better optimise the
parameters and procedures. The chemical composition
of clad powder and substrate is listed in Table 1.
Table 1 Nominal chemical composition (wt-%) of Stellite
6 powder and substrate
C
Si
Mn
Cr
Co
W
Fe
Mo Ni
V
Stellite 6 1.57 1.0 0.3 28.7 Bal. 3.9 0.4 0.6 1.6 …
SS420
0.41 0.7 0.45 14.3 … … Bal. 0.6 0.2 0.2
Laser cladding
The cladding was performed with a Laserline fibre
coupled diode laser with maximum output power of
3?5 kW and a Motoman robot. The cladding powder,
delivered at a given mass flowrate, was injected into the
laser spot on the workpiece through the coaxial nozzle
which was positioned horizontally relative to the
stationary workpiece. The robot scan rate was selected
to represent a range of speeds up to 2000 mm min21.
Both single track and multitrack clads were prepared to
investigate the dilution, distortion and cracking. The
multitrack was achieved by moving the laser beam at a
range of increments from 0?75 to 2?0 mm.
Metallurgical analysis
The microstructure was examined under optical and
scanning electron microscopy. The clad samples were
cross-sectioned, mounted, polished and etched with
Marble’s reagent to reveal the heat affected zone. The
thicknesses of the clad layer and heat affected zone
(HAZ) were measured on the optical microscopy
images. The clad layer was etched electrolytically in
4% nital at 10 V direct current to reveal its structure.
The variation of clad microstructure along its length and
height was observed by optical microscopy.
The elemental analysis in the unetched clad layer was
carried out by energy dispersive X-ray analyser. The
dilution was calculated based on both the elementary
analysis and geometry measurement for the multitrack
clad. Microhardness across the clad and heat affected
zone was measured in accordance with the ASTM E9282 using a Buehler Micromet 2100 microhardness tester.
Blade distortion
Before embarking on the field trials, it was envisaged
that there would be an issue with residual stress.
Laboratory trials were undertaken in an attempt to
create the anticipated residual stress and measure the
subsequent amount of distortion. Test coupons of
dimensions 100632 mm having a tapered profile similar
to a blade were machined from the blade material.
Illustrated in Fig. 4 is the experimental set-up for
measuring distortion.
International Heat Treatment and Surface Engineering
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:10:46
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
=
2009
VOL
3
NO
3
3
Brandt et al.
;
Laser cladding repair of turbine blades in power plants
cavitation erosion apparatus. The apparatus consisted
of a power supply, piezoelectric transducer, amplifier
and horn. Erosion samples having a diameter of
24?8 mm and a mass between 18 and 21 g polished to
a 1 mm finish, were preweighed to ¡0?1 mg and
attached to the end of the horn. The erosion tests were
conducted at vibration frequency of around 20 kHz and
amplitude of 50 mm. Most of the tests were run for 34 h.
Tests were interrupted at regular intervals to examine
changes occurring to the erosion surface using an optical
microscope and to measure weight loss. The weight loss
was plotted against time. Erosion damaged surfaces
were also analysed to characterise the erosion process
and to establish correlations with microstructures.
Detailed description of the tests and results is presented
in Ref. 14.
ONLINE
COLOUR
ONLY
Stress measurements
The deposition of a Stellite 6 clad layer on a steel
substrate results in the generation of residual stresses in
the clad region due to a number of complex mechanisms.
These include differences in thermal expansion
between the Stellite 6 (coefficient of thermal expansion
from room temperature to 900uC is 15?56
1026 mm mm21 uC21) and parent metal (coefficient of
thermal expansion from room temperature to 900uC is
11?761026 mm mm21 uC21) during cooling, thermal
strains resulting from differences in heating/cooling at
differing locations and the directional stiffness within
the sections present. Both test coupons and a section of a
LP blade were prepared for stress measurements.
Residual stress measurements were carried out using
the neutron diffraction method so that strains could be
measured at depth. Detailed description of the measurements is given in Ref. 15.
4 a experimental set-up for measuring distortion and b
blade sample shape and dimensions
These experiments investigated the effect a range of
laser and powder parameters had on distortion. After
each layer of clad, the test coupons were cooled to room
temperature and measured for distortion relative to a
datum point. Also investigated was the effect of stress
relieving and grinding of the coating on distortion.
Results and discussion
Erosion testing
Effect of processing parameters
An important aspect of this research was the material
that would exhibit high resistance to erosion damage. In
the early stages of this programme, both powder and
wire feedstock were considered as a possible deposition
material. It was shown early in the project that wire
cladding was not as forgiving as powder and the project
focused on powder cladding. Erosion tests, however,
were conducted with both types of samples. Samples
were machined from a laser clad steel bar. The overlays
were deposited using Stellite 6 and Stellite 21 in both
wire and powder forms and nickel based wire. The
diameter of the wire was 1?6 mm, while powder particles
varied in size between 53 and 150 mm. The chemical
composition of these alloys is given in Tables 2 and 3.
Cavitation erosion tests were performed in distilled
water in accordance with ASTM G32 using a vibratory
The effects of diode laser power on clad thickness,
penetration into the substrate, HAZ and dilution for
constant mass flowrate and increment are illustrated in
Fig. 5. It can be seen that with increasing laser power
from 750 to 1500 W, the clad layer thickness increases
from y0?55 to 0?74 mm. The increase in clad thickness
is due to the greater amount of powder being melted at
higher laser power and fused to the substrate. At the
same time, the increased penetration, i.e. melting of the
substrate material, is reflected in the significant rise in
dilution.
The effects of powder mass flowrate on clad layer
thickness, penetration into substrate, HAZ and dilution
for constant laser power and increment are shown in
Fig. 6. The powder mass flowrate was varied from 11 to
29 g min21. The clad thickness increased from 0?24 to
1?24 mm, while the dilution decreased from 35 to 1%.
During laser cladding, the laser energy is distributed into
Table 2 Nominal composition (wt-%)
alloy used for laser cladding
Stellite
Stellite
Stellite
Stellite
6*
6{
21*
21{
of
cobalt
based
C
Si
Mn Cr
Co
W
Fe Mo Ni
1.10
1.21
0.22
0.25
1.0
1.3
0.5
2
0.6
0.4
0.5
1.0
62.8
59.9
61.4
58.8
4.5
4.9
…
…
2
2
3
2
28.0
28.3
27.1
28.0
…
0.1
5
5.5
Table 3 Nominal composition (wt-%) of nickel based wire
used for laser cladding
…
2.1
2.7
2.5
*Wire form.
{Powder form.
4
International Heat Treatment and Surface Engineering
2009
VOL
C
Si
Mn
Cr
Ni
W
Fe
Mo
Other
6/40
C0
0.35
0.08
4.5
…
…
…
22
16
R
R
1.5
4.5
4.5
4.5
…
16
625
0.05
…
0.50
22
R
…
4.5
9
B51.6
Co52.5
V50.35
Nb53.5
3
NO
3
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:10:49
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
Brandt et al.
Laser cladding repair of turbine blades in power plants
;
ONLINE
ONLINE
COLOUR
COLOUR
ONLY
ONLY
5 Effects of laser power on clad layer thickness, penetration into substrate, HAZ and dilution with substrate
three parts: absorption by powder, absorption by
substrate and reflection by powder and substrate.16
The laser energy incident on the substrate is the energy
remaining after attenuation by the powder jet by either
ONLINE
COLOUR
ONLY
6 Effects of laser power on clad layer thickness, penetration into substrate, HAZ and dilution with substrate:
laser power 1500 W and increment 1?0 mm
7 Effect of scan increment on clad layer thickness, penetration into substrate, HAZ and dilution with substrate:
scan
velocity
800 mm min21,
powder
flowrate
20 g min21 and laser power 1500 W
reflection or absorption. The energy attenuated by the
powder is directly proportional to the powder mass
flowrate. Therefore, with increasing powder mass
flowrate, a greater fraction of the laser energy will be
absorbed by the powder rather than the substrate,
leading to an increase in clad height and a decrease in
penetration into the substrate. At too high mass flowrate
for a constant laser power, insufficient energy is
available to melt the powder and fuse it to the substrate.
The increment or centre to centre spacing between the
tracks is an important parameter in laser cladding as it
can significantly affect the integrity of the clad layer at
the interface with substrate material,12 in particular the
formation of inter track porosity which can lead to crack
initiation in fatigue type situations. Shown in Fig. 7 are
the effects of increasing increment on clad layer
thickness, penetration, HAZ and dilution. It is clear
that with increasing increment from 0?75 to 2 mm, the
clad layer thickness decreases from y1?1 to 0?55 mm.
The decrease in layer thickness with increasing increment is due to the decreasing overlap between the
individual tracks.
From these results and those of simulated blade
distortion experiments described in the section on ‘Blade
distortion and cracking’, the operating conditions
selected were chosen to minimise any distortion while
producing a metallurgically sound clad layer of less than
10% dilution at maximum scan velocity on a laboratory
repaired blade. The blades were clad on the rear surface
near leading edge, the front surface and the leading edge
after the original worn edge shields where removed. The
sequence and direction of cladding used facilitated
minimal distortion while providing a wide enough
International Heat Treatment and Surface Engineering
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:10:55
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
2009
VOL
3
NO
3
5
Brandt et al.
;
Laser cladding repair of turbine blades in power plants
with each successive clad layer, the distortion changes. It
increases significantly from the first layer to the second
followed by a decrease with the deposition of the
backface layers. Cladding on the leading edge does little
to change the initial distortion. It can also be seen that
the test coupons tend to return towards their original
shape as the clad material is ground away.
The main mechanisms that drive blade distortion are,
first, the differential of thermal expansion between the
martensitic stainless steel (11?761026 mm mm21 uC21)
and Stellite 6 (15?561026 mm mm21 uC21) which
causes the blade to bend towards material with the
greater thermal expansion during cooling and second,
the input of heat into the surface. It is clear from
Fig. 10b that both of these play a significant role in
blade distortion with the layer thickness being the
dominant factor.
The results indicate that to keep blade distortion to a
minimum heat input should be kept to a minimum using
low laser powers and avoiding heat build-up at edges
and that the clad thickness should be kept to a
minimum. This was followed in practice.
Apart from minimal blade distortion, the quality of
the clad layer itself needs to be maintained. It was shown
in laboratory trials that with variation of the powder
feed rate, laser power, clad increment and clad speed, the
clad on the blade material which initially appeared good
quality was in fact porous or reduced in wear resistance
between the clad tracks, resulting in increased wear rates
and surface flaws when ground. This results from
insufficient energy to melt completely the clad track as
a result of the high clad height to clad width ratio.
Typically, intertrack porosity can be seen in crosssections as a small regular pore at the clad to substrate
interface or on occasion by a shallow surface flaw that
appears at the surface during grinding. It is believed that
the surface flaw results from the failure of the weakened
surface which is under tensile stress as a result of the
differential expansion between the coating and substrate
which leads to blade distortion.
Stellite 6 clad tracks on the blades can also crack. Two
modes of cracking were observed, first, cracks that
appear within a clad track itself and second, cracks that
appear across the clad tracks. These crack types are
shown in Fig. 11. Cracking within the clad track such as
that of Fig. 11a occurs where dilution rates are greater
than 50% and clad temperatures are excessive. Reducing
ONLINE
COLOUR
ONLY
8 a surface of Stellite 6 clad layer, b cross-section and c
cross-section of coating around leading edge at higher
magnification
build-up to facilitate the reconstruction of the leading
edge. Typical appearance of a clad layer is shown in
Fig. 8.
Microstructure of clad layers
The microstructure of the Stellite 6 clad layer is shown in
Fig. 9. In general, it shows a fine dendritic structure with
well developed secondary arms. The dendrites consist of
a face centred cubic cobalt based solid solution and are
surrounded by a eutectic formed of M7C3 type carbide
and face centred cubic cobalt based solid solution. The
microstructure at the interface (Fig. 9b) exhibits a plane
solidification front, while that near the surface of the
clad layer is equiaxial. This is consistent with a decrease
in the G/R ratio (temperature gradient/solidification
speed) from the interface to the top of the clad layer.
This type of microstructure is well suited to the erosion
type environment found in the LP stage of a turbine.17
The hardness of the first clad layer was y520 HV
which reduced to y500 HV after the application of
second layer.
Blade distortion and cracking
The distortion of the coupons as a function of the
different procedures for a fixed laser power of 1000 W,
scan speed 1000 mm min21 and mass flowrate of
13?9 g min21 is shown in Fig. 10a. Here the distortion
is measured along the leading edge. It can be seen that
9 Microstructure of clad layer a near surface and b at interface with substrate
6
International Heat Treatment and Surface Engineering
2009
VOL
3
NO
3
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:11:08
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
Brandt et al.
Laser cladding repair of turbine blades in power plants
;
ONLINE
COLOUR
ONLY
10 Distortion of test coupons during a multiple layer cladding of Stellite 6 and b different laser powers, scan speeds
and mass flowrates
the dilution by decreasing raster increment, laser power
and/or change in scanning speed eliminated this.
Cracking across the clad tracks is believed to result
from high stresses that result from differential contraction after cladding. These cracks were eliminated by
reducing the clad length.
Erosion
Erosion results for clad layers are plotted as a function
of testing time in Fig. 12. It is evident from this figure
that the erosion resistance of the Stellite 6 powder clad
layer was similar to that of the Stellite 6 wire clad layer.
However, there are slight differences in the erosion
resistance for the Stellite 21 powder and wire clad layers
after 20 h of testing. For this alloy, the powder clad
layer was found to be less resistant than the wire clad
layer. In all cases, the rate of erosion loss was
ONLINE
COLOUR
ONLY
ONLINE
COLOUR
ONLY
11 Cracking of Stellite 6 clad a within clad track and b
across clad tracks
12 Plot of erosion loss for stellite clad layers deposited
on steel substrate
International Heat Treatment and Surface Engineering
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:11:15
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
2009
VOL
3
NO
3
7
Brandt et al.
;
Laser cladding repair of turbine blades in power plants
ONLINE
COLOUR
ONLINE
ONLY
COLOUR
ONLY
13 Comparison of cavitation erosion loss of typical low
pressure (LP) blade and various cobalt and nickel
based alloys: numbers above bars indicates test
duration
14 Coupon stress measurements in parent metal: depth
under Stellite 6 layer is indicated
the Stellite 6 and HAZ are simultaneously quenched by
the bulk of the substrate. As this occurs, the stellite and
HAZ layers begin to contract, generating a global tensile
stress balanced by compressive stresses in the substrate.
As the Stellite 6 has a higher coefficient of thermal
expansion, there is also a local stress gradient generated
at the interface between the stellite and the HAZ.
approximately constant over the first 12 h period, and
then increased significantly.
The nickel-based clad layers were also tested in a
similar way. However, due to their high rate of erosion,
these samples were only tested for 10 h. Additional
testing was carried out for turbine blade to compare the
erosion resistance of both cobalt and nickel based alloys
with that of an LP turbine blade. The results are plotted
in Fig. 13, and illustrate that the erosion resistance of
the blade is substantially less than that of the laser clad
layers. It also reveals that the stellite clad layers have a
higher erosion resistance than the nickel base alloys. The
wear mechanism in the cobalt based alloys has been
studied by others18–20 and phase transformation induced
by strain during the collapse of bubbles has been
identified as the main reason for the increased erosion
resistance. Based on these tests, it was decided to use
Stellite 6 powder alloy for the trials.
Field trials
The stresses in the LP blade were a factor of 2 higher
than those in the test coupons; nevertheless, the trend in
the result is consistent with the results of the coupon
measurements. A comparison of transverse stresses in
the clad and post-weld heat treated (PWHT) LP blades
clearly indicates the benefit of the heat treatment in
minimising tensile stress. The significance of the PWHT
results is that there are not the stress gradients through
the parent metal that were present in the clad test
coupons and clad LP blade samples. This indicates that
the PWHT was effective in minimising the stresses
imparted by the cladding process, thus reducing the
probability of any crack generation in the clad layer.
One of the power stations supporting the research,
TRUEnergy at the time, now AGL Torrens Island in
Adelaide, SA, Australia, agreed to participate in a field
trial, and made a LP section of a 200 MW turbine
available. The field trial was carried out in September
2004 during a scheduled outage. The objective was to
clad a few blades and then operate the unit followed by
detailed inspection and a further trial. It involved
bringing all the processing equipment and personnel
from Swinburne University to AGL. Repair was carried
out on six blades selected by AGL staff. The trial
showed that in situ laser cladding of turbine blades was
feasible and practical. Shown in Fig. 15 is the robotic
cladding operation at the power station. It is believed
that this was the first time in the world that such an in
situ application has been successfully carried out.
Although successful, the trial also highlighted a
number of issues such as blade preparation, blade
distortion during cladding and cladding path. The
leading edge of the actual blades was thinner than that
of the laboratory blades which in some repaired blades
resulted in distortion. Although it was of little significance to the blade performance and only in the order
Stress measurements
In test coupons, the stresses calculated from the strains
in the parent metal (Fig. 14), within y0?5 mm of the
Stellite 6 interface, appear to be in compression. A steep
stress gradient occurs towards a tensile stress peaking at
y1?3 mm below the interface before reversing back into
compression through the majority of the thickness
before translating again to a strong tensile stress near
the backface of the coupon. The relatively high tensile
stresses near the backface of the coupons reflect the
dominance of the global bending force generated by the
contraction of the stellite and HAZ at the coupon/clad
interface.
Examination of the test coupon results shows clearly
the significance of cladding a dissimilar material, Stellite
6, onto a steel substrate at the interface region, namely,
large stress gradients are generated at the interface
which is a result of the dissimilar thermal expansion
coefficients of the two materials. This is particularly
marked between 600uC and room temperature where the
Stellite 6 has thermal expansion values ranging between
1?4 and 6?5 times that of the parent metal respectively.
During heating, the Stellite 6 layer behaves as a soft
compliant material. As heating continues, the substrate
softens in the HAZ relieving the strain. During cooling,
8
International Heat Treatment and Surface Engineering
2009
VOL
3
NO
3
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:11:27
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
Brandt et al.
Laser cladding repair of turbine blades in power plants
;
in processing parameters and processing technology
such as the real time clad layer thickness measurement.
These are currently being addressed. Shown in Fig. 16
are the as clad and ground blades from the repair
activity carried out in 2008.
ONLINE
COLOUR
Conclusions
ONLY
15 In situ laser repair of eroded low pressure (LP) blade
of 1?0 mm, it still warranted further investigation. The
robot path also required modifications to better follow
the profile of the blade and keep the distance between
the cladding nozzle and surface of the blade within
designed limits.
Following the inspection of the laser clad blades and
demonstrated improvement in erosion resistance of laser
repaired blades, it was decided to conduct another trial
in 2005 again at AGL. The issues of blade distortion and
robot path programming were addressed in laboratory
and implemented in the 2005 trial. In this trial, an
additional seventeen blades were repaired. The distortion was almost eliminated being on the order of 0?3 mm
in some blades. The clad path was also improved with
better blade surface following path and a reduction in
the speed of processing. This in particular is important when considering the repair of the full blade
complement.
In parallel with this, a company, Hardwear Pty Ltd,
was established to commercialise this technology. The
company has to date carried out laser repair at AGL in
2007 and 2008 repairing 340 LP blades in total. Further
repair is planned at AGL and blade repair enquiries
have been received from a number of local and overseas
utilities. Similar to field trials, the large scale repair
activity has also demonstrated the need for improvement
The in situ laser cladding repair of steam turbine blades
has been achieved using a compact coaxial laser
cladding head coupled to a fibre delivered diode laser.
This head facilitates even powder delivery at any angle
of presentation and is modular in construction for quick
repair or changing cladding conditions.
To successfully clad repair the blades with Stellite 6
alloy requires controlling the amount of material
deposited and care with the substrate temperature
especially where edges are involved. The elimination of
cracks in the clad layer is achieved by reducing dilution
with substrate, reducing the length of clad and optimising cladding conditions. The clad layer thickness was
shown to change as a result of the thickness of the
substrate showing it also to be a parameter when
cladding.
The 2004 field trial showed that in situ laser cladding
of turbine blades is feasible and practical. This was the
first time in the world that such an application had been
successfully demonstrated. In May 2005, the turbine was
inspected and the laser repaired blades demonstrated
good performance and no sign of damage. The October
2005 trial increased the level of confidence further in the
technology through increased number of repaired blades
and improved processing.
A company, Hardwear Pty Ltd, was formed to
commercialise the technology and has successfully
delivered two commercial contracts in 2007 and 2008
respectively. It is now examining other opportunities.
Finally, this project has clearly demonstrated the
benefits of a partnership between industry, research
providers and government in delivering a cost effective
solution to a major industry problem. It also highlights
the importance of having an industry champion such as
the AGL Torrens Island Power Station in this project
without which the commercialisation path of this
technology would have been more difficult.
ONLINE
COLOUR
ONLY
16 Laser clad blade shown in front and rear view a and b after cladding and c and d after grinding
International Heat Treatment and Surface Engineering
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:11:33
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
2009
VOL
3
NO
3
9
Brandt et al.
;
Laser cladding repair of turbine blades in power plants
10. M. Brandt, J. Harris, S. Sun, B. Dempster, N. Alam and A. Bishop:
‘In-situ cladding of leading edge of LP turbine blades using fibre
delivered diode laser’, Proc. Conf. ICALEO 2005, Miami, FL,
USA, October–November 2005, LIA, Paper 1103.
11. J. Harris and M. Brandt: ‘Powder delivery nozzle’, Australian
Provisional Patent Application PA2005904580, 2005.
12. S. Sun, M. Brandt, J. Harris and Y. Durandet: ‘The influence of
Stellite 6 particles size on the inter-track porosity in multi-track
cladding’, Surf. Coat. Technol., 2006, 201, 998–1005.
13. S. Sun, Y. Durandet and Y. Brandt: ‘Parametric investigation of
pulsed Nd:YAG laser cladding of stellite 6 on stainless steel’, Surf.
Coat. Technol., 2005, 194, 281–291.
14. A. Nazmul, H. Damien and S. Akif: ‘Erosion assessment of laser
clad overlay deposited using stellite in powder and wire forms’,
Proc. Conf. PICALO 2004, Melbourne, Vic., Australia, April 2004,
LIA, Paper 301.
15. P. Bendeich, N. Alam, M. Brandt, D. Carr, K. Short, R. Blevins,
C. Curfs, O. Kirstein, G. Atkinson, T. Holden and R. Roggee:
‘Residual stress measurements in laser clad repaired low pressure
turbine blades for the power industry’, Mater. Sci. Eng. A, 2006,
A437, 70–74.
16. H. Gedda, J. Powell, G. Wahlström,W.-B. Li, H. Engströmand
C. Magnusson: ‘Energy redistribution during CO2 laser cladding’,
J. Laser Appl., 2002, 14, (2), 78–82.
17. J. Mello, M. Durand-Charre and T. Mathia: ‘A sclerometric study
of unidirectionally solidified Cr–Mo white cast irons’, Wear, 1986,
111, 203–215.
18. Z. Xiaojun, L. A. J. Procopiak, N. C. Souza and A. S. C. M.
d’Oliveira: ‘Phase transformation during cavitation erosion of a Co
stainless steel’, Mater. Sci. Eng. A, 2003, A358, 199–204.
19. B. Vyas and C. Preece: ‘Cavitation erosion of face centred cubic
metals’, Metall. Trans. A, 1977, 8A, 915–923.
20. C. Heathcock and A. Ball: ‘Cavitation erosion of cobalt-based
stellite alloys, cemented carbides and surface treated low alloy
steels’, Wear, 1981, 74, 11–26.
Acknowledgements
M. Brandt would like to acknowledge the significant
input into this project from Dr J. Harris who has left
Swinburne, B. Dempster, Dr R. Deam, A. Moore and
T. Waterman, during the 2007 and 2008 Hardwear
Repair Contracts. The authors also acknowledge the
support of the CRC for Welded Structures and WTIA
Powergen Group during the research phase of this work.
References
<
10
1. B. Stanisa, Z. Schauperl and K. Grilec: ‘Erosion behavior of
turbine rotor blades installed in the Krsko nuclear power plant’,
Wear, 2003, 254, 735–183.
2. L. Shepeleva, B. Medres, W. D. Kaplan, M. Bamburger and
A. Weisheit: ‘Laser cladding of turbine blades’, Surf. Coat.
Technol., 2000, 125, 45–48.
3. Y. P. Kathuria: ‘Some aspects of laser surface cladding in the
turbine industry’, Surf. Coat. Technol., 2000, 132, 262–269.
4. T. Peters and W. Jahnen: ‘Steam turbine leading edge repair by
stellite laser cladding’, Proc. Conf. EPRI ’02, 2002, EPRI, Paper
ST7.
5. W. M. Steen: ‘Laser material processing’, 3rd edn; 2003, London,
Springer.
6. R. Vilar: ‘Laser cladding’, J. Laser Appl., 1999, 11, (2), 64–79.
7. Available at: http://www.laserline.de
8. Available at: http://www.de.trumpf.com
9. A. Weisheit, G. Backes, R. Stromeyer, A. Gasser, R. Wissenbach
and R. Poprawe: ‘Powder injection: the key to reconditioning and
generating components using laser cladding’, Proc. Int. Cong. on
‘Advanced materials and processes, Munich, Germany, October
2001, ASM International, 1–7.
International Heat Treatment and Surface Engineering
2009
VOL
3
NO
3
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:11:40
The Charlesworth Group, Wakefield +44(0)1924 369598 -
Rev 7.51n/W (Jan 20 2003)
Authors Queries
Journal: International Heat Treatment and Surface Engineering
Paper: 84
Title: Laser cladding repair of turbine blades in power plants: from research to commercialisation
Dear Author
During the preparation of your manuscript for publication, the questions listed below have arisen. Please attend to
these matters and return this form with your proof. Many thanks for your assistance
Query
Reference
Query
Remarks
1
Author: Please confirm the running head.
2
Author: Please supply city,
country and month of the conference.
3
Figures 3b and 4a are low
quality, please supply a higher
resolution version if possible.
International Heat Treatment and Surface Engineering iht84.3d 19/10/09 22:11:41
The Charlesworth Group, Wakefield +44(0)1924 369598 -
View publication stats
Rev 7.51n/W (Jan 20 2003)