You are on page 1of 10

Design characteristics and development of a nozzle for coaxial laser cladding

J. Lin and W. M. Steen

Citation: Journal of Laser Applications 10, 55 (1998); doi: 10.2351/1.521821


View online: http://dx.doi.org/10.2351/1.521821
View Table of Contents: http://scitation.aip.org/content/lia/journal/jla/10/2?ver=pdfcov
Published by the Laser Institute of America

Articles you may be interested in


Numerical simulation of heat transfer and fluid flow in coaxial laser cladding process for direct metal deposition
J. Appl. Phys. 100, 024903 (2006); 10.1063/1.2209807

Laser beam analysis in direct metal deposition process


J. Laser Appl. 17, 136 (2005); 10.2351/1.1896965

Dimensional and material characteristics of direct deposited H13 tool steel by C O 2 laser
J. Laser Appl. 16, 245 (2004); 10.2351/1.1809638

Energy redistribution during CO 2 laser cladding


J. Laser Appl. 14, 78 (2002); 10.2351/1.1471565

Laser attenuation of the focused powder streams in coaxial laser cladding


J. Laser Appl. 12, 28 (2000); 10.2351/1.521910

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
JOURNAL OF LASER APPLICATIONS VOLUME 10, NUMBER 2 APRIL 1998

Design characteristics and development of a nozzle for coaxial laser


cladding
J. Lin and W. M. Steen
Department of Engineering, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, United Kingdom
~Received 10 June 1997; accepted for publication 23 December 1997!
Laser cladding by blowing powder into a laser generated melt pool is finding greater use in industry
particularly with the growing interest in direct build and repair techniques currently being
investigated. The process is traditionally done with a side blown powder feed. This has the
limitation of not being omnidirectional. To overcome this, a coaxial powder feed system has been
developed. This article describes a particular system by both mathematical models and experiment.
The system has so far achieved a catchment efficiency of up to 40% with a 1 kW CO2 laser beam.
The parameters controlling this important commercial parameter are described. They include the
flow structure of the powder in the beam. The control of this structure is discussed with reference
to this particular coaxial flow system. © 1998 Laser Institute of America.
@S1042-346X~98!00402-1#

Key words: coaxial nozzle, laser cladding, catchment efficiency, powder flow

INTRODUCTION shielding gas on the jet flow and the catchment efficiency
was analyzed by both theory and experiments. Flow visuali-
The laser cladding process has been investigated over sation with image analysis is used to check this simulation
the past decades. It is one of the advanced techniques for and some initial studies are reported on the effects of flow
surface treatment possible with a laser.1 In laser cladding the parameters on the catchment efficiency and degree of powder
clad layer can be applied as a powder, foil, or wire with the focus. The powder used in the tests was stainless steel 304L
objective of improving the corrosion, wear, or cosmetic powder. Clad trials were performed on a mild steel substrate
properties of the object being clad as well as applications with a 1.0 kW cw CO2 laser with a constant table speed of 10
such as repair and rebuild. In powder systems the powder mm/s.
can be preplaced or blown into the laser generated melt pool.
The blown powder route is one of the rare processes which is
capable of giving a fused bond with low dilution over large COAXIAL CLADDING NOZZLE
or small areas. It has thus found applications in industry and
is the subject of much research around the world. Most work The design of the coaxial nozzle structure is illustrated
to date has used an off axis powder feed system to avoid the in Fig. 1. The design specification included compact size,
powder melting and sticking to the feed nozzle or oxidizing efficient energy interaction between the laser beam and the
prior to cladding. However a coaxial feed would make this powder outside the nozzle assembly, and shrouded gas flow
promising process omnidirectional and therefore easier to which would tend to focus the powder stream.
use. This would open up new possibilities for repair and for The nozzle is composed of three separate parts: the cen-
the newly developed laser direct casting process ~LDC!.2–4 tral hole through which the laser beam and shield gases pass;
In this new technique, three-dimensional ~3D! metal compo- the surrounding annular zone into which the powder flows
nents can be directly built from a computer aided design and is blown out by its carrier gas as a cone into the laser
~CAD! drawing, suggesting a new route for low volume beam outside the nozzle; and an outer annular gas jet. The
manufacturing. beam focus is set at the nozzle exit so that the powder inter-
Some attempts have been made to operate coaxial pow- acts with an expanding beam whose focal length is 150 mm
der feeders in which the powder interacts with the beam just ~;8° divergence!. An inert gas, such as argon, is blown
outside the nozzle or to feed the powder from symmetrically through the inner and the outer nozzles to protect the focus-
placed external pipes creating a central powder flow.5,6 ing lens and prevent powder oxidation. A water jacket was
These systems gave a coaxial feed but with poor powder used to remove the heat from the nozzle for a prolonged
focus or some solidification within the nozzle assembly. working life.
There was also a problem with the heat balance during pro-
longed periods of processing.
In this article a novel coaxial design is studied. The de- FLOW VISUALIZATION
sign has been successfully used for LDC and behaves as a
‘‘metal pencil,’’ capable of making uniform deposits in any Two methods were used to visualize the powder stream
direction. In this article the nozzle characteristics are deter- coming from the coaxial nozzle. The first method filmed the
mined through a numerical simulation. The influence of the light scattered from an incident 2D light beam which illumi-

1042-346X/98/10(2)/55/9/$15.00 55 © 1998 Laser Institute of America


This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
56 J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen

FIG. 2. Photograph from the CCD camera by plane light illumination at a


stand-off distance S520 mm for U 0 54 m/s and U 1 52 m/s.

The concentration of the cold powder stream in the


transverse direction was analyzed from the image of the
CCD camera with a 45° projection angle to the stream center
with a stand-off distance of 20 mm. The variation of the
luminance intensity along the center line of the image was
analyzed by the OPTIMAS software. From Eq. ~1!, the lumi-
nance intensity will correlate to the powder concentration
distribution due to the light scattering at the low powder
concentration considered here. Figure 4 shows volumetric
distributions of the powder concentration at different gas set-
tings. It is seen that the powder distribution was significantly
FIG. 1. Coaxial powder feed nozzle for laser cladding. affected by the change in the inner and outer jet velocities.
The second method used the scanning device shown in
Fig. 2. A scanning powder sensor device based on a photo-
nated a single section across the flow; the second measured detector with He/Ne laser ~3 mW! was designed to measure
the powder stream opacity by scanning a He/Ne laser beam the spatial distribution of the powder concentration. The
across the stream. The objective was to measure the powder powder sensor received the radiation energy generated by the
stream divergence and the powder distribution within that He/Ne laser beam through the powder stream. A voltage sig-
stream. nal drop from the silicon photodiode due to the light extinc-
The first method used a charge coupled device ~CCD! tion by the powder stream could be detected. Since the out-
video camera with a framing speed of 50 frames/s. This cam-
era was focused onto the particles flowing from the nozzle
which were illuminated by a 2D sheet of light from a fiber
optic and a tungsten lamp formed by passing the beam
through a cylindrical lens. The sheet thickness was 2
60.5 mm in the framing area. A typical image from the
CCD camera is shown in Fig. 3. Image analysis gave the
variation in intensity across the powder stream.
The luminance intensity L is in proportion to the number
of particles illustrated per unit area. This is equal to the par-
ticle density at each radial location and the path distance
through the stream as well as the image size being analyzed.
Assuming a Gaussian powder distribution within the stream,
then the intensity would be ~meaning of symbols given in
nomenclature!
r 2p n 2/s2
L} e 22y . ~1!
s
Measurements were made for various inner and outer gas FIG. 3. Luminance intensity distribution from the image analysis on the
velocities varying from 2 to 8 m/s. The particles were stain- cross section of the powder stream under different compound jet velocity at
less steel 75630 m m with an average flow rate of 0.05 g/s. S520 mm.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen 57

FIG. 5. Signal output from powder scanning sensor at the inner jet velocity
of ~a! U 0 52 m/s, U 1 52 m/s; ~b! U 0 54 m/s, U 1 52 m/s.

the nonuniform powder distribution along the He/Ne laser


FIG. 4. Experimental setup for powder concentration measurement.
beam direction as discussed previously. Figure 6 shows a
Gaussian curve for the concentration measurements made by
put signal linearly depends on the light intensity falling in the photosensor and the CCD camera. The stream structures
the sensor, this intensity drop could be evaluated by with different shielding gas settings were also analyzed. The
Bouguer–Beer law as expressed in Eq. ~2!7 results of the luminance intensity of the powder stream mea-
sured by the CCD camera could be interpreted as the volu-
I5I o exp~ 2nAEl s ! . ~2! metric concentration of the powder distribution from which
Scanning results were collected for a cold powder stream the stream divergence angle could be measured. Both the
with various shielding gas velocity. The inner and outer gas spraying angle of the cold and hot powder stream were mea-
velocities were set by controlling the gas flow rate into the sured. Here the hot stream is defined as the inner region of
coaxial nozzle. The sensor was located at variable stand-off powder stream heated by the laser beam and is usually lo-
distances and scanned along the transverse direction. The cated at the center of the stream. According to the image
receiving intensity drop due to the light transmitted through analysis of the powder stream and its concentration distribu-
the powder stream was very clear in Fig. 5. The powder tion, the powder spraying angles with various settings of the
distribution along the longitudinal and transverse axes was gas velocity in the inner and outer jets were measured and
affected by the gas velocity. shown in Fig. 7. The spraying angle increases with inner and
outer gas velocities.
To realize the reasons for the shield gas effects, a nu-
INFLUENCE OF THE SHIELDING GAS ON THE
merical simulation was used in this study. A 2D axisymetri-
POWDER STREAM STRUCTURE
cal powder stream problem in the cylindrical coordinate sys-
The powder distribution was evaluated by image analy- tem was simulated by the commercial FLUENT package. The
sis of the results of flow visualization. Images from the CCD governing equations used here are the equations of mass con-
camera of the scattered intensity of the powder stream indi- servation and momentum conservation for both argon gas
cate the powder concentration and powder stream width, as and dispersed stainless steel particles. The standard k2 e tur-
shown in Fig. 4 made with different jet velocity settings. The bulence flow model was adapted in this computation in solv-
results of the scattered intensity and the intensity drop on the ing the compound jet problem. Without heat radiation from
photosensor show similar distributions of the powder con- the laser, typical computational results ~Fig. 8! show the
centration for the transverse sections. The data from the pho- change of the powder concentration along the longitudinal
todetector need to be weighted by a volumetric factor due to axis from the nozzle exit. The distribution of powder con-
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
58 J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen

FIG. 6. Normalized concentration distribution from ~a! scanning powder FIG. 7. The stream angle change with variable velocity of the compound
sensor, ~b! CCD camera to assure a Gaussian mode of the powder stream jets for ~a! cold stream, ~b! hot stream.
spraying.

centration at different radial position is shown in Fig. 9. The


distribution is similar to a Gaussian distribution with a re-
duction of the powder concentration due to an increased ve-
locity of shield gas. The faster the stream the less density per
unit volume for a given mass flow rate. It should be noted
that the peak value of the powder concentration is near the
jet center and the nozzle exit. The concentration is reduced
further downstream due to the divergence of the powder
stream which is affected by the inner and outer jets.
The intruding effect of shield gas on the powder stream
structure was very clear from the experiments and the nu-
merical simulation. The most important parameter of com-
pound jets is the relative velocity between the inner and
outer streams or the ratio between the initial velocity of the
compound jets and the surrounding gas.8 Depending on the
velocity ratio, compound jets may be referred to as either
weak or strong. Generally speaking, strong jets mix much
faster than weak ones due to the propagation of the flow
shear zone caused by the different velocities between the
jets. In the case of a jet with high-velocity gradients in the
fringes to the center, the more shear stress is created, the
faster the shear layer propagates toward the center and the
outer region. The increase of the inner jet momentum will
further diffuse the turbulent flow, and the particle dispersion
rate is significantly enhanced by increasing the mean veloc-
FIG. 8. The influence of the compound jets velocity on the distribution of
ity of the turbulent flow. Typical results are the particle con- powder concentration along the beam axial direction at center of powder
centration drop and the flow field expansion.9 stream.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen 59

FIG. 9. The influence of the compound jets velocity on the distribution of powder concentration along the stream radial direction.

THERMAL PROFILE IN THE POWDER STREAM ing heating. This implies an infinite thermal conductiv-
ity, k, such that k@h, where h is the particle to gas heat
The process to be modeled consists of a stream of pow- transfer coefficient.
der particles falling in an expanding laser beam. The power ~3! Latent heat effect is neglected in this analysis.
distribution within the laser beam will be one of the specified
structures, typically Gaussian mode, and thus a given particle In general, the motion of the particle is determined by the
may be subjected to a changing power density as it falls in fluid forces acting on it as described by Eq. ~3!. However this
the beam. On impact with the substrate to be clad, the par- ignores the heating effects of the gas adjacent to the particle
ticle will either adhere and form a clad or bounce off. The which would in practice cause significant gas expansion af-
mechanism determining which event occurs is not clearly fecting its velocity and direction
understood but is expected to include whether the particle or dV 3 m C D Re
the substrate or both are molten. It is felt that a solid/solid 5 ~ U2V ! 1g. ~3!
interaction will most probably result in a ricochet. The one dt 4 r p D 2p
exception could be that the substrate is sufficiently hot to The drag coefficient C D of the flow is expressed in Eq. ~4!
melt the particle during the impact time. Thus the points of for Re,1000:
interest in this system for laser cladding are the distance from
the nozzle at which the powder melts, the temperature of the 24
C D5 ~ 110.15 Re1.687! . ~4!
particle and substrate at the impact point, and the velocity of Re
the particle upon impact.
Some modeling of such problems has been done for the The Reynolds number, Re, is defined in the usual way in Eq.
plasma spraying process10 and little for the laser cladding. ~5!:
Some of this analytic work included solving the equations of r u U2V u D p
conservation for mass, momentum, and energy at high tem- Re5 . ~5!
m
peratures. This level of analysis was not considered neces-
sary for this problem. Here a simplified approach is dis- The change in the particle temperature is obtained from an
cussed to solve the thermal profile of the coaxial powder energy balance on the particle as expressed in Eq. ~6!:
stream based on the following assumptions:
dT p 6Nuk 3eI
~1! The particle is spherical with constant absorptivity to 5 2 ~ T g 2T p ! 1 . ~6!
dt D pr pC p 2 r pD pC p
10.6 mm radiation and is moving in a uniform jet flow
and a uniform radiation field. The Nusselt number for a spherical particle moving in a gas
~2! The internal temperature of the particle is uniform dur- stream is expressed as Nu5210.6Re0.5Pr0.33. 10 The geomet-
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
60 J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen

tion. The significant cooling effect by convection is seen


from the powder temperature in the downstream of the flow
for various gas flow velocities. At the start of the laser inter-
action, the temperature of the particle will rise rapidly near
the nozzle exit due to the strong laser radiation. After that the
temperature discrepancy between the powder and gas will
induce a significant convection cooling and the particle tem-
perature will drop with decreasing laser intensity on the par-
ticle.

TEMPERATURE MEASUREMENT OF THE HOT


FIG. 10. The particle temperature distribution along the beam axis under PARTICLES
variable jet velocity with an initial particle velocity of 2 m/s.
An experimental arrangement for the temperature mea-
surement of the dispersed particles was developed in this
ric relationship among the beam intensity I, beam power P, study. Figure 11 illustrates the basic concept of the tempera-
the stand off distance S, and beam divergence angle w could ture detector used. A photodiode with pin-hole tube ~diam-
be expressed as Eq. ~7!: eter of 100 mm and view angle about 4°! was designed to
catch the radiation from the particles in front of a hot powder
P stream. When a solid particle is injected into a laser beam, it

F SD G
I5 . ~7!
w d8 2 is heated by the optical energy. Its temperature can reach as
p tan S1 high as a few thousand degree Kelvin depending on the op-
2 2
erating conditions, the thermal properties, the size and the
As an example, the beam power of 1 kW with absorptivity of velocity of the particle, and the gas velocity in the jet flow.
0.4 on a stainless steel particle of 75 mm and initial velocity According to Planck’s law,11 the heated surface will radiate
V 0 52 m/s under varying inner jet velocity was used. The over a wide range of wavelengths depending on the surface
calculated results of a moving particle under laser radiation temperature and emissivity. Here we assume the particle be-
are shown in Fig. 10. It could be found that particle heating haves as a grey body. The radiation energy detected by the
is mainly determined by the beam focus of the laser radia- photodiode on the particle surface will correlate to its

FIG. 11. ~a! Particle temperature measured by a pin-hole photo detector. ~b! Calculated temperature of the tungsten lamp with various electric power input.
~c! The voltage response of the photo diode with various lamp temperatures.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen 61

axis was measured and plotted to compare with the calcu-


lated data as shown in Fig. 10.
It can be seen from the experimental and the analytical
results that the insertion point of the powder stream into the
beam is critical, since it determines the maximum possible
temperature. The melting distance of the powder flow can
also be evaluated from Eqs. ~3!–~7!. According to the results
in Fig. 10, a proper selection of the stand off distance is
based on the stream temperature which should be up to the
melting point of the particle to trap on the substrate. This
distance is obviously a function of the laser power and flow
rate.

POWDER CATCHMENT EFFICIENCY

The performance of the coaxial nozzle for cladding is


mainly based on its ability to deliver powder which forms a
clad with the correct metallurgical qualities. The main per-
formance criterion is the catchment efficiency, h. This is
defined as the ratio of the powder trapped in the formed clad
to the powder delivered. It is considered probable that the
powder which is caught has either impacted with a liquid/
solid interaction or it melted during the very brief time it was
FIG. 12. Typical signal output and temperature distribution from the pin- in contact with the surface. For cladding of stainless steel on
hole sensor on the powder stream under 1 kW laser interaction with U 0 mild steel, the melting points of both materials are similar
54 m/s, U 1 50 m/s at S510 mm. and hence it is unlikely that a solid particle arriving on a
solid surface will melt during impact. This leaves the only
option that clad would form if either the powder particle or
temperature. The location of the pin-hole sensor on the par- the surface or both were molten at the point of impact. With
ticle surface decides the type of the heat source. In our case, the coaxial nozzle all three alternatives are possible. For clad
the heating area of the particle is similar to the target area of forming before the particles have melted the catchment effi-
detector if the sensor head is located near the flying particle ciency will be mainly determined by the ratio of the laser
stream. In such a condition, a single particle may be treated generated melt area to the impact area of the powder
as an extended heat source to the detector and the response stream.12
of the sensor is independent of the background radiant According to the results from flow visualization, the
energy.11 Gaussian distribution of the powder concentration will pro-
The calibration of the detector involves finding a corre- vide sufficient information to estimate the theoretical catch-
lation between the particle temperature and the pulse ampli- ment efficiency from the area relationship between the melt
tude for a fixed particle diameter. In practice, particles with a pool and the powder stream. Some assumptions should be
known temperature and diameter and a suitable velocity are made in this analysis. The diameter of powder stream and the
not available. Therefore they have to be simulated. The sys- melt pool are assumed to be linear functions of the spraying
tem employed to simulate the stream of particles comprises a angle and the stand-off distance. The ratio of the clad width
pin-hole, a tungsten ribbon lamp, and a chopper. The pin- to the stream diameter is expressed in Eq. ~8!. The clad width
hole, imitating the particle, is placed over the sensor tube is treated as the same size as the beam diameter and in-
head. The lamp with a known relationship between the fila- creases with stand-off distance. This is reasonable for a short
ment temperature and the input power is substituted as the stand-off distance (S,203R 0 ), after which the pool size is
source of radiation. The chopper with a constant speed is smaller due to the falling power intensity
placed between the sensor and lamp to imitate a hot particle d ~ d 8 1S tan w !
in flight for temperature calibration. The relationship of the 5 . ~8!
D ~ D 8 1S tan u !
tungsten filament temperature and the input power of the
lamp is calculated by the Stefan–Boltzmann law of the According to the Gaussian distribution of the dispersed pow-
blackbody radiation.11 The calibration curves of the lamp der on the spraying area, the catchment efficiency could be
temperature and the voltage response of the photodiode are evaluated by integrating the volumetric powder concentra-
plotted in Fig. 11. tion on the clad area to the spraying area. It could be ex-
Figure 12 shows a typical measurement on the powder pressed as Eqs. ~9! and ~10!:
stream taken for a sample of 400 acceptable particle tem-
1 2
peratures in a powder stream with 1 kW laser interaction. f ~ z !5 e 22 z , ~9!
The mean temperatures of the powder stream along the beam A2 p
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
62 J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen

angle. It is seen that the spraying angle and the initial diam-
eters of the powder and laser beam will significantly affect
the catchment efficiency. In our case, the reduction of the
spraying angle, which is controlled by the velocity of the
compound jets, will increase the catchment efficiency up to
40% at the spraying angle of about 10° for both inner and
outer gas velocities of 2 m/s.

CONCLUSIONS
This study demonstrates a nozzle design to improve jet
spraying on clad surfaces. The omnidirectional ability of the
coaxial nozzle will provide a possible technique for rapid
FIG. 13. The relationship between the powder catchment efficiency and the
prototyping applications. The interaction between the laser
diameter ratio d/D of the clad to powder stream for Gaussian mode distri- beam and the powder concentration should be strictly con-
bution. trolled to achieve an efficient cladding process. The control-
ling of the exit jet of the powder stream is seen as a possible
way to improve its efficiency.

h SD
d
D
5
* z d f ~ z ! pz d z
* z D f ~ z ! pz d z
, ~10!
The Gaussian mode of the powder distribution in coaxial
cladding jet flow was obtained by both experimental and
numerical approaches. Further evidence from the photosen-
where z d 5d/ s and z D 5D/ s , s is the stream diameter cov- sor and plane beam image analysis supported the same con-
ering a portion of 68% of the concentration in the section clusion. This provided a basic distribution of the powder
plane. stream to estimate the catchment efficiency as a function of
Equation ~9! can be rearranged as Eq. ~10! if a stream the shield gas effects.
diameter is selected as D53 s ( z D 53). This new definition The theoretical study of the single particle heating in a
of the stream diameter will cover 99.8% of the concentration laser beam was verified by the temperature measurement ob-
in the section plane and it will be much more accurate in the tained using a pin-hole sensor. The thermal profile of the
calculation of catchment efficiency using the Gaussian dis- powder stream determines the stand-off distance for a proper
tribution mode clad setting. This melting distance is a strong function of the
2 laser power and flow rate.
h 512e 22 z d . ~11! The velocity of the compound jet will determine the
The numerical relationship of Eq. ~11! between the catch- stream shape and hence the catchment efficiency. The melt
ment efficiency and the diameter ratio (d/D) of the clad to pool arising from the laser spot will have a diameter propor-
the powder stream is shown in Fig. 13. It can be found that tional to the beam diameter up to the point where the power
the catchment efficiency will be increased with d/D and ap- density fails to melt. This will give a peak value in the pow-
proaches 80% at d/D51.5. In our optics setting and nozzle der catchment as a function of the stand-off distance for
size the d 8 /D 8 is about 0.26 and it could be substituted in small spraying angles. Considering the Gaussian distribu-
Eq. ~8! to calculate the value of d/D through the beam axis. tions of the powder streams, it gives an approximated for-
An important relationship between the catchment efficiency mula for the catchment efficiency h up to 40% in our nozzle
and geometric parameters of stand-off distance and spraying design. This formula was verified by experimental measure-
angle was established in Fig. 14. The catchment efficiency is ments.
compared to the experimental data with varying spraying
ACKNOWLEDGMENTS
Many thanks are due to Dr. W. O’Neill and Dr. G.
Shannon of the University of Liverpool for their invaluable
suggestions to this work, which is under the LEMA ~Laser
Engineering for Manufacturing Applications! project spon-
sored by EPSRC, UK.

NOMENCLATURE
A: mean particle cross section area of the particles
CD : drag coefficient of the flow
cp : particle specific heat
Dp : particle diameter
d: laser beam diameter
D: powder stream diameter
FIG. 14. The effect of the stream angle on the catchment efficiency for both d 8: effective initial beam diameter
theoretical and experimental approaches. D 8: effective initial stream diameter
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23
J. Laser Appl., Vol. 10, No. 2, April 1998 J. Lin and W. M. Steen 63

E: extinction coefficient z: ratio of laser beam diameter to powder stream


f ( j ): Gaussian normal error function diameter
g: acceleration due to gravity s: stream diameter to the e 21/2 position of the powder
h: heat convection coefficient distribution
I: intensity of the laser light beam r: density of the gas flow
Io : initial intensity of the laser light beam e: absorptivity of the particle surface
w laser beam divergence angle
k: thermal conductivity of the gas
u spraying angle of the powder stream
L: luminance intensity
h powder catchment efficiency
ls : distance of the light passing
n: average number of the particles per unit volume
Nu: Nusselt number 1
W. M. Steen, Laser Material Processing ~Springer, Berlin, 1993!.
2
J. L. Kock and J. Mazumder, Rapid Prototyping by Laser Cladding, Pro-
P: laser beam power ceedings of ICALEO, 1993, pp. 556–559.
Pr: Prandtl number 3
M. Murphy, C. Lee, and W. M. Steen, ‘‘Studies In Rapid Prototyping By
Re: Reynold number Laser Surface Cladding,’’ Proceedings of ICALEO 1993, pp. 882–891.
4
M. A. McLean, G. J. Shannon, and W. M. Steen, ‘‘Laser Generating
rp : radius of the area analyzed
Metallic Components,’’ Proc. SPIE, Edinburgh, UK, Aug. 1996, paper
S: stand-off distance GLC/HPL.
5
t: time P.-A. Vetter, Th. Engel, and J. Fontaine, Laser Cladding: The Relative
Parameters for Process Control ~SPIE, Bellingham, WA, 1994! Vol.
Tp : particle temperature
2207, pp. 452–462.
Tg : temperature of the surrounding gas away from the 6
S. Carty, et al. ‘‘Catchment Efficiency For Novel Nozzle Designs Used In
boundary layer Laser Cladding and Alloying,’’ in Laser Processing: Surface Treatment
and Film Deposition, edited by J. Mazumder et al. ~Kluwer Academic,
U: velocity of the gas flow
Dordrecht, The Netherlands, 1996!, pp. 395–410.
U0 : inner shield gas velocity 7
C. N. Davies, Aerosol Science ~Academic, New York, 1966!.
8
U1 : outer shield gas velocity M. Jankovie and J. Mostaghimi, ‘‘A new nozzle for dc plasma spray
V: particle velocity guns,’’ Plasma Chem. Plasma Process. 15, 607–628 ~1995!.
9
A. A. Mostafa, ‘‘Turbulent diffusion of heavy-particle in turbulent jets,’’
V0 : initial particle velocity ASME Trans. J. Fluid Engineer. 114, 657–671 ~1992!.
10
y: distance along the y axis from the center of the R. M. Young and E. Pfender, ‘‘Generation and behaviour of fine particles
stream in thermal plasma-A review,’’ Plasma Chem. Plasma Process. 5, 1 ~1985!.
11
R. D. Hudson, Infrared System Engineering ~Wiley, New York, 1969!.
m: viscosity of the gas flow 12
V. M. Weerasinghe, ‘‘Laser cladding of flat plates,’’ PhD dissertation,
rp : particle density Imperial College of Science and Technology, 1984.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
137.195.150.201 On: Fri, 02 Jan 2015 05:23:23

You might also like