numerical approach and optimization of the combustion and ...combustion chamber and in the nozzle,...

Numerical Approach and Optimization of the Combustion and Injection Techniques in High Velocity Suspension Flame Spraying (HVSFS)

E. Dongmo, R. Gadow, A. Killinger, M. Wenzelburger, Institut für Fertigungstechnologie keramischer Bauteile (IFKB), University of Stuttgart, Stuttgart, Germany E. Dongmo, Graduate School of Excellence advanced Manufacturing Engineering (GSaME) Stuttgart, Germany

ABSTRACT The interest in submicron and nano-structured layers applied by thermal spray coating technologies on mechanical components has been significantly increased during the last decade. Conventional HVOF spraying processes are not suitable to achieve submicron and nano-particles. Therefore, High Velocity Suspension Flame Spraying (HVSFS) has been developed for the processing of nano-structured spray material to achieve dense surface layers in supersonic mode with a refined structure, from which superior mechanical and physical properties are expected. However, the chemical and thermodynamic phenomena occurring in HVSFS reacting flow field are a challenging, multidisciplinary issue. This study is intended to analyze and understand the HVSFS combustion and flow dynamic system on the basis of a CFD model. The model considers following phenomena: combustion of the fuel gas (premixed), energies transfer between the flame and the suspension droplets (organic solvent and particles), and the injection, vaporization and combustion of the suspension organic solvent (non-premixed). 1. INTRODUCTION One of the great challenges in thermal spray technologies to date is the manufacturing of nanostructured coatings and functional surfaces. This development is motivated by the discovery that properties of such materials are superior to those of conventional metal or ceramic coating materials, including greater strength, hardness, ductility, and sinterability [1]. Otherwise, reducing the grain sizes of coating materials from conventional levels (i.e., <10 µm in metals and <1-2 µm in ceramics) to nanostructured levels (i.e., <100nm) can considerably enhanced its mechanical strength and hardness [2, 3]. Powders composed of nanosized particles (< 100 nm) are extremely critical in handling and processing. The potential of these materials to pose risks to health is not yet fully examined and well understood. Nanoscaled particles easily distribute in air, penetrate the human skin or pass through the respiratory tract and the lungs, finally entering the blood circuit. In resume, there is no way to feed submicrometer or even nanopowders with conventional powder feeder techniques, not only that safety aspect would forbid any refilling, cleaning, etc. under habitual conditions, but mechanical feeding is virtually impossible due to the strong

EASC 20094th European Automotive Simulation Conference

Munich, Germany6-7 July 2009

Copyright ANSYS, Inc.

agglomeration of the powder particles that prevents a good flowability. Processing the powder in the form of a suspension (in aqueous or organic solvent) solves those problems, the handling is facilitated and feeding of the liquid can be realized with quite simple thermal spray techniques [1]. The focus of research on nanostructured materials is now shifting from powder synthesis to processing of nanostructured coatings using supersonic thermal spray processes. This new approach, based on the conventional high velocity oxide fuel (HVOF) thermal spray, consists to spray submicron or nanoparticles with hypersonic speed with the aim to form thin and dense coating layers [1, 13,14]. The HVOF spraying is a long established technique besides flame, arc wire or atmospheric plasma spraying. The HVOF thermal spray process is characterized by very high gas and particle velocities, followed by relatively low gas temperatures, as compared to plasma spray processes. Mainly ceramic coatings applied via HVOF have been extensively improved as corrosion-resistant, anti-corrosion, anti-adhesion, wear-resistant surface layers, or combinations of these to extend product life, to increase performance and to reduce time and costs. However, the conventional HVOF thermal spray system is only suitable for the processing of micro-scale powders and has distinct limitations in powder size due to powder supply and several difficulties mentioned above [15, 16]. This means that some adaptations of the HVOF process are needed for the processing of nano-scale powders. High velocity suspension flame spraying (HVSFS) compared to SPS and SPPS also use a liquid solvent as carrier fluid to process nano-scale materials. In those techniques (SPS, SPPS, HVSFS) the coatings material is processing in the form of a suspension. Suspension is a heterogeneous mixture containing solid particles and a solution. In our institute (IFKB) various suspensions were fabricated using aqueous and organic solvents (ethanol, isopropanol). This consist to mix the powders with the solvent and stirred together with appropriate dispersing agents (i. e. as Trusan 471 from Trukem) and stabilizers (citric acid, acetic acid) to ensure stable dispersions. The solid contents were in the range of 10–15 wt.%. Distilled water, ethanol and isopropanol were used as solvents. Subsequently the suspensions were ground using an attrition mill (Dispermat SL, Getzmann) to prevent any formation of agglomerates. After this procedure all suspensions showed adequate stability and were ready for processing [14]. By HVSFS, The suspension is injected directly (centered and in axial direction) into the combustion chamber of the operation Torch with the liquid contributing to the reaction enthalpy of the combustion process [14]. The suspension has to be delivered against the combustion chamber pressure, just like in standard powder feeding but with a special feeder system. A pressure vessel (maximum load 10 bar) holding the suspension that is operated with compressed nitrogen gas is use at IFKB for experimental processing of HVSFS. This feeder technique has been made to transport suspension and to prevent it from sedimentation in less stable case by using i.e. a stirring device. A valve allows switching between suspension and a flushing gas or solvent to prevent clogging of the suspension line when starting or terminating spray operation [1, 13, 14]. This injection process shows some difficulties compared to conventional powder feeding in terms of pressure distribution and also the possibility of composition variations and in-situ change of the composition. Therefore, an optimization of the injection behavior of suspensions was done in earlier work by modification of the nozzle geometry [1, 13, 14]. Using a solution as carrier fluid for nanostructured materials processing in thermal spray systems features some thermophysical and thermochemical changes of the new HVSFS process compared to the standard HVOF process. These include the existence of a third phase (liquid), its evaporation and combustion including a cooling effect in the spray torch, and also the resulting particle morphologies (agglomeration) are different from conventional




HVOF processes [9]. The chemical, thermal, thermophysical, and morphological states of the suspension and the particles during the process ultimately determine the coating microstructure and its macroscopic properties. Parameters such as droplet size, injection velocity, the location of solution evaporation and initial combustion, flame temperature and velocity fields in the combustion chamber and expansion nozzle can all have significant influences on the final outcome in terms of the coating structure and its properties. Regarding the thermophysical and thermochemical properties of HVSFS, the following phenomena have to be considered during analysis of the processes in the combustion chamber: - combustion of the fuel gas (premixed oxygen/propane), - heat, momentum and mass transfer between the flame (propane flame) and the

suspension droplets (organic solvent and particles), and - evaporation and combustion of the suspension organic solvent (non-premixed ethanol

combustion) In this work, 3D modeling and analysis of the combustion and gas dynamic phenomena of the HVSFS process, including the modeling of ethanol evaporation and analysis of the interaction mechanisms between gas and liquid as well as between gas and particles, is performed at the example of an industrial TopGun-G torch. The thermal and flow fields of gas are solved by an Eulerian approach and the particle velocity, temperature and degree of melting by a Lagrangian approach. Regarding the gas dynamics, the Reynolds and Favre-averaged Navier-Stokes equations implemented on the commercial CFD software ANSYS-CFX are solved. Due to high Reynolds numbers and large pressure gradients in the combustion chamber and in the nozzle, the κε-turbulence model is used with the scalable turbulent wall function in order to implement wall friction and heat transfer into the model. The eddy dissipation model [15, 17, 18], which assumes that the reaction rate is limited by the turbulent mixing rate, is employed to model the chemistry reaction occurring during premixed combustion (propane). An Antoine equation assumption is utilized in the liquid evaporation model [9] for determining the liquid droplet properties (temperature, velocity, diameter, etc.) during the energy transfer in the combustion chamber. After liquid evaporation, the finite rate chemistry model [19] is used to solve the rate of progress of elementary reactions taking place during diffusion combustion (non-premixed ethanol vapor). Particle breakup is modeled using the Blob method for disperse droplets and the ETAP method for disperse solids according to the Weber and Reynolds number of each particle. The Blob method is one of the simplest and most popular approaches to define the injection conditions of droplets. In this approach, it is assumed that a detailed description of the atomization and breakup processes within the primary breakup zone of the spray is not required. Spherical droplets with uniform size, dp=dnozzle are injected that are subject to aerodynamic induced secondary breakup. The spray angle is either known or can be determined from empirical correlations. The "blob-method" does not require any special settings but is the default injection approach in this process. A uniform diameter model was used for the ethanol droplets and a discrete diameter distribution model was used for the particle size distribution of the solid particles (titania powder in micro and nano scale). This parameter enables verification of the powder size distribution in the model. Two-way coupling was used for the energy and mass transfer between gas and liquid, while one-way coupling was used during modeling the thermal and kinetic energy transfer from gas to solid particles, assuming that the particles have no distinct influence on the gas properties [15, 16, 17]. 2. THEORY AND MODELING OF THE HVSFS PROCESS Numerical modeling of the HVSFS thermal spray process is a challenging, multidisciplinary issue. Regarding its physical description, this involves two phase combustion, liquid evaporation, turbulence, compressible flow, multi components, multiphase interactions,




subsonithe conprocessliquid foThe adamixing torder tocombusin flight 2.1 Geo The geoconsistigeometgas (inchannepremixethe preaxially iGTV-Gmparticles

Figu

As waschambepremixepremixeEthanolparticlesdiameteparameas well

Tab Mass fTempePressu

ic/supersonnventional Hsing using aor a continuoaptation of cthe nanostro obtain astion chambduring heat

ometrical m

ometrical deng of a com

trical model this casels into a mix

ed oxy/fuel paration annjected intombH) as shs are molte

ure 1: The c

s mentioneder and exped combused combust is injecteds (dropletser of 0.3 meters at the as at the To

ble 1: Param

flow [kg/s] erature [K] ure [Pa]

nic transitionHVOF procea standard Hous, stable conventionaructured powa suspensiober to form sting and ac

modeling an

esign for mombustion chwas used i

e propane) xing chambmixture. Th

nd processo the combuhown in fig.n and acce

computation

d before, thpansion notion of protion that od into the

s). It was mm is gene

inlet of theopGun-G to

meters for th

Inlet C3H8/O0.189e-3 300 ----

n, droplet dess show tHVOF systeparticle tra

al HVOF spwder with aon as expsingle droplceleration.

nd process

odeling of thhamber andin earlier woand oxyge

ber or mixinghe major d

sing of the ustion cham1, where thlerated by t

nal domain

he HVSFS zzle. The

opane and ccur in thecombustio

assumed terated (diame combustioorch outlet f

he HVSFS b

O2 Inlet C0.917e300 ----

deformationthat single em [15], andnsport and

praying for tan organic splain abovelets that wil

s paramete

he HVSFS d a convergork for simuen are sepg case befoifference becoating ma

mber of the he solution ithe expandi

(3D symme

model conmajor intethe evapo

e combustion chamberthat duringmeter of thon chamberfor this study

boundary co

2H6O Lq I-3 0

3-

n and solidnano-scaled thereforeacceleratiohe processsolvent suc

e that can l dry and re

ers

process is bgent-straightulation of theparately inj

ore enteringetween HVaterial. Susoperation Tis evaporatng gas jet o

etrical segm

nsists of twrest was toration of on chamber, atomized atomizatioe injection

r and the boy are listed

onditions us

nlet TiO2 W0.917e-3 -300 4--- -

ification. Sie particles a, an organicn in the spring of nano

ch as ethanobe injecte

elease the n

based on tht expansione HVOF projected thro the combu

VOF and HVspension dTorch (TopGed and buronto the sub

ment of the H

o major sethe analysiethanol fol

er and the d, and formon, an initi

nozzle). Toundary cobelow in ta

sing oxy/fue

Wall Free---- ---- 460 298 ---- 1.05e

imulation reare not suitc solution aray torch is -particles col or isoproed into thenano-scale p

he TopGun-n nozzle. Thocess [15].

ough variouustion chamVSFS proceroplets areGun-G systrned, and thbstrate surfa

HVSFS proc

ections: comis of the tlowed by expansion

ms disperseial uniform

The initial monditions at able 1.

el ratio 0.8

e Jet Sub---- 298

e5 ----

esults of table for

as carrier needed. onsist of

opanol in e HVOF powders

-G torch, he same The fuel

us small mber as a

esses is directly em from

he nano-ace.

cess).

mbustion turbulent its non-nozzle.

ed liquid droplet

modeling the wall

bstrate




Propane combustion in an HVSFS combustion chamber leads to high gas velocities with subsonic to supersonic transition in the expansion nozzle. Furthermore, flow instabilities in the form of supersonic shock waves during gas expansion in the free jet area occur, which are discussed elsewhere [15, 16]. Modeling of the open (free) boundary region that represents the flame expansion area was simplified in this work. 2.2 Analysis of reaction mechanisms for the HVSFS process Chemical reaction between fuel gas and oxygen occurs in the combustion chamber, leading to heat generation and reaction products. The thermal energy is converted into kinetic energy by gas expansion through the convergent-straight nozzle, accompanied by transfer of momentum and heat from the gas to the particles, whereby the particles are ethanol droplets with dissolved titania powder content. This leads to further atomization of the droplets as well as ethanol evaporation and combustion at the point where saturation temperature is reached in the HVSFS torch, and finally, melting and acceleration of the solid particles (titania). Therefore, the combustion process for HVSFS consists of two major phases: - a primary, premixed oxygen-propane reaction, and - a secondary, non-premixed oxygen-ethanol reaction. The premixed combustion was modeled in analogy to the HVOF process described in earlier work, using the eddy dissipation model for turbulent reacting flows, which was implemented into ANSYS CFX 11 [15, 18]. However, for the HVSFS process, the oxygen mass fraction is hyperstoichiometric in relation to complete ethanol combustion, because the oxygen remnant will be used for ethanol combustion as diffusion flame. Modeling of the secondary, non-premixed combustion associated with a premixed combustion system is a challenging task. The main difficulty is the ethanol phase change from liquid to gaseous, which must be modeled in a first step as ethanol mass transfer from dispersed liquid droplets to continuous gas [20]. This is the precondition for combustion modeling as diffusion flame. Evaporation of the ethanol droplets is modeled using the liquid evaporation model [21], which combines heat and mass transfer to and from (liquid) particles with the premise that the continuous gas phase is at a higher temperature than the particles. The model uses two mass transfer correlations depending on whether the droplet is above or below the boiling point. This is determined through the Antoine equation and is given by:

⎟⎠⎞

⎜⎝⎛

+−=

CTBAPPvap exp

where A (7.5K), B (1741 K) and C (238.1K) are user-supplied coefficients using in this work for ethanol. The particle is boiling if the vapor pressure, Pvap, is greater than the pressure P of the gas phase. The boiling point of ethanol is at about 351.6 K and the temperature range T in the HVSFS combustion chamber is in the range of 3,200 K. On this account, the mass transfer from the continuous gas to the ethanol droplet is determined by the convective heat transfer Qc:

VQ

dtdm Cp −=

In this equation, mp is the mass of the ethanol droplet and V is the temperature dependent latent heat of ethanol vaporization.




For the combustion process of ethanol as diffusion flame, different models and approaches are available, which were tested and compared with respect to their accuracy. Furthermore, the finite rate chemistry model [19, 20] in Arrhenius form that was implemented into ANSYS CFX 11 for the prediction of non-premixed hydrocarbon fuel oxidation was modified using FORTRAN subroutine for the modeling and the computation of ethanol combustion. 2.3 Conservation and transport equations for the HVSFS process The HVSFS thermal spray process is an inhomogeneous, multiphase flow process in which the continuous gas dynamics and particles (ethanol droplets and Titania powder) dynamics are coupled to each other. The solid particle (titania) loading in the suspension is about 10 % and is typically less than 4 % over the entire HVSFS thermal spray process and therefore, the assumption of one way coupling is usually made [16, 22]. With this assumption, the existence of particles has a neglectable influence on the gas dynamics, while the particle velocity and temperature can be determined based on the two-phase momentum and heat transfer equations. However, because of the presumed cooling effect of ethanol in the HVSFS torch during evaporation, a two-way coupling assumption was modeled for the energy and mass transfer between the continuous gas and the disperse ethanol droplets. The governing equations applied to describe the thermal spray process are the conservation of mass, momentum, energy, species transport, turbulent kinetic energy and dissipation rate [15, 22]. The set of equations describing this process was discretized and solved numerically by a finite volume technique implemented in commercial CFD software. The computational domain is shown in fig. 1. The combustion and transport process in this system can be assumed as a rotational symmetric problem. Due to the symmetric geometry of the torch and the process boundary conditions [1], only a segment of the unstructured 3-D grid had to be modeled and the flow was described as radially symmetric to the centerline. This enabled the reduction of computation time [15]. The momentum transfer from the continuous gas to the disperse particles by different forces (drag, pressure, lift, thermal differential, virtual) based on the Lagrange assumption and the heat transfer by convection and radiation are discussed in detail in earlier work [15]. 3. NUMERICAL RESULTS AND OPTIMIZATION 3.1 Reactions mechanisms An overview of kinetic modeling of all common hydrocarbon combustion given by Westbrook and Dryer contains chain-branching steps that produce CO and H2 as primary products of hydrocarbon oxidation reaction on one side, and on the other, contain the oxy-hydrogen (H2O) reaction mechanism and CO oxidation (CO2) in subsequent, slow secondary reaction processes [23]. The mass fraction of each species taking part in the combustion of premixed propane fuel with an excess of oxygen and non-premixed combustion of ethanol fuel with the oxygen residues after complete propane combustion are shown in fig. 2. It is interesting to observe the spray breakup of ethanol droplets in the combustion chamber, where the mass fraction of C2H6O gas is close to zero after injection and increases slowly at an axial position of approx. 0.035 m (inside the nozzle), where ethanol begins to evaporate until complete evaporation and phase change (liquid to gas). The point of total evaporation of ethanol also locates the highest value of the ethanol gas mass fraction (mf = 0.35) in the process, before gaseous ethanol is burned and the mass fraction decreases along the centerline. At the point of ethanol evaporation, the presence of excess amounts of oxygen leads to combustion of gaseous ethanol as diffusion flame (non-premixed combustion of ethanol vapor). The particle (droplet) size tends towards zero at the time of evaporation and phase change of ethanol, followed directly by the combustion of the gaseous ethanol. It is noticeable that evaporation of ethanol mainly occurs in the area behind the nozzle entrance. This is due to the high-speed suspension injection in combination with the heat transfer from the flame to the cold particles (293 K at the entrance).




Figuprod

3.2 FBecausmultiphacontinuoand of tThe heavelocityforce onMach ncompre The conevaporafig. 3 rethe exhcombusthe gas

ure 2: Massducts derive

Fluid and pe of the fase systemous gas phthe particlesat and mom

y and tempen the particlumber [21]ssibility effe

ntour plot ofation taking egarding stahaust gasesstion chambs velocity in

s fractions oed from prop

particles prfact, that

m in which hase. This es in-flight bementum tranerature diffeles is not on. The heat ects.

f the gas flaplace at th

atic pressurs expand thber and a sncreases co

of C2H6O gapane and e

roperties the HVSFSthe ethano

enables sepehaviors. nsfer from gerence betwnly a functiotransfer coe

ame tempere outside ore and Machrough the straight profontinuously

as componeethanol (gas

S thermal ol droplets parated illu

gas to suspween each pon of the Reefficient bet

rature, see of the combch number inozzle, whfile close to

y. At the en

ent and of oxs) oxidation

spray promove with

stration of

pension parphase. At hieynolds numtween gas a

fig. 3, showustion chamin the internich has a co the outletntrance to t

xygen, propin the HVS

ocess is anh the relatithe continu

ticles strongigh relative mber but aland particle

ws the coolimber. The cnal and extconvergent t, the pressthe nozzle,

pane, and reSFS torch

n inhomogive velocity

uous gas pr

gly dependvelocities, so a functio

es also dep

ng effect ofcenterline pernal flow fprofile clos

sure decreathe Mach

eaction

geneous, y as the roperties

s on the the drag on of the pends on

f ethanol rofiles in fields As se to the ases and

number




increases. The gas is accelerated during expansion in the nozzle and reaches supersonic velocity (Mach number of 2.3) at the nozzle outlet as show in fig. 3.

Figure 3: Contour plot of temperature and centerline profiles for static pressure and Mach’s number for the continuous gas phase (from combustion chamber to substrate)

In the combustion chamber, reaction of the premixed reactants (oxygen/propane) results in an increase of gas temperature to values above 3,200 K and a pressure in the region of 3 bar (approx. 3.65 bar at the centerline) in the combustion chamber. Fig. 4 left shows the temperatures profiles of the continuous gas and the disperse particles (ethanol liquid until vaporization and titania particles with different size range). The centerline profile of the continuous gas phase clearly shows the cooling effect due to ethanol vaporization, which leads to a temperature loss of 700 K in the calculation. However, due to ethanol combustion, the reaction gases quickly increase in temperature up to the maximum values of approx. 3,200 K along the centerline. Furthermore, the gas temperature decreases continuously along the centerline in the free jet region due to mixing with ambient gas.

free jet area atmospheric conditions (air ideal gas) 20 °C, 1 bar




The specific heat, density and melting point of the particles are important properties to be considered for the choice of the appropriate fuel gas. In this study, titania powder with a discrete diameter distribution of 0.5-50 µm was applied with a density of 4,230 kg/m³, a specific heat capacity of 3.78 kJ/(kg*K), a thermal conductivity of 10.4 W/(m*K), and a melting point of approx. 2,100 K. As mentioned before, in an inhomogeneous, multiphase flow, various components move at different velocities in the system space. The comparison of different particle sizes (5 nm, 10 nm, 50 nm, 20 µm and 25µm) shows that with decreasing particle size, particle temperature and velocity increase, see fig. 4. The maximum velocity of the combustion gas was calculated as approx. 2,050 m/s close to the nozzle exit in the free jet region due to under-expansion effects. However, due to their mass inertia, the spray particles do not follow this high gas acceleration, but are more or less separated from the gas depending on their size. Thereby, large particles are more inert than small particles. This leads to higher final velocities for smaller particles. Generally, the particle velocity remains below 1,000 m/s, whereas for particles with 5 nm in diameter, even higher velocities up to 1,200 m/s were calculated. For small particles, a higher acceleration in the gas flow not always leads to higher impact velocities on the substrate, because for smaller mass inertias, the particle velocity also experiences a higher decrease prior to impact in the impinging flow region where the axial velocity of the spray jet is decelerated due to flow deflection at the substrate surface.

Figure 4: Temperature (left) and axial velocity (right) profiles of gas, ethanol and titania particles with different size along the axis from the combustion chamber to the substrate

3.3 Modification of the injection angle The simulation results of the reaction mechanism and the flow transport properties show that the evaporation of ethanol occurs outside of the combustion chamber. This is based on the fact that the residence time of ethanol droplets is relatively short in the combustion chamber. This also involves a cooling effect in the nozzle as shown in fig. 4 and fig. 5, which influences the energy balance of the HVSFS system. Analysis of the mass fraction distribution in the expansion nozzle, see fig. 2, shows a high amount of O2, C2H6O, CO and H2 over the whole length of the expansion nozzle, which leads to the assumption that the combustion of ethanol gas is incomplete.

axis [m]

combustion chamber & expansion nozzle

C2H6O liquid

continuous gas

5 nm

10 nm

50 nm

20 µm

25 µm

free jet area: Gas effect of the particles and the description of the flow instabilities are neglected

tem

pera

ture

[K]

titania particles

axis [m]


C2H6O liquid

continuous gas

5 nm

10 nm

50 nm

20 µm

25 µm


tem

pera

ture

[K]

titania particles

axis [m]

velo

city

[m/s

]

C2H6O liquid

continuous gas

5 nm

10 nm

50 nm20 µm

25 µm



titania particles

axis [m]

velo

city

[m/s

]

C2H6O liquid

continuous gas

5 nm

10 nm

50 nm20 µm

25 µm



titania particles




In order to avoid this cooling effect, some optimizations experiments were made with the aim to reduce the evaporation length of ethanol droplets, so that evaporation will take place in the combustion chamber. The principal modification in this work is the variation of the suspension injection angle (ethanol droplets and titania) from 0° (axial injection) to 30° off axis (see fig. 5 for comparison).

Figure 5: Comparison of the C2H6O evaporation point and of the gas and particles temperature profiles for 0° (left) and 30° (right) suspension injection angle

Modification of the injection angle led to good results regarding the position of the ethanol vaporization point as well as prevention of a distinct cooling effect during ethanol vaporization, see fig. 5. Due to the longer flow path and therefore residence time in the combustion chamber for an injection angle of 30° off axis, the evaporation of ethanol still occurs in the nozzle, however, the evaporation point is much closer to the combustion chamber. The cooling effect in the expansion nozzle almost disappears, as we can observe at the temperature centerline profiles of the gas and particles comparing the simulation results for the two different angles. 4. SUMMARY AND CONCLUSIONS Using a solution as carrier fluid for nano particles processing in thermal spray systems is a novel approach that features some thermophysical and thermochemical changes of the new HVSFS process compared to the standard HVOF process. 3D modeling and analysis of the combustion and gas dynamic phenomena of the inhomogeneous, three-phase HVSFS process based on a numerical model for a conventional HVOF process was performed in this work at the example of an industrial




TopGuncombusbetweenEulerianapproac The simcombusexpansioptimizato somechambenozzle drawbaccombusresiden

Figuusin

Further system,betweenFurthermcombusThe aimof the determithe des Referen

[1]

[2]

n-G torch, instion, and an gas and pn approachch using the

mulation restion occur ion nozzle wation procese improvemer and therecompletelycks in the

stion chamce time in th

ure 6: Instabng a 30° inje

steps in th, aiming forn the torchmore, somestion chambm of this woHVSFS thene the effecign of HVSF

nces

R. GadowSpraying C.C. KocNoyes Pu

ncluding theanalysis of tparticles. Thh, whereas e commerci

esults for thoutside of

which has ss consistin

ments such efore, reducy disappeare process ber walls, he combust

bilities of titaection angle

his work wilr the accurah outlet ane studies o

ber have to ork is to devermal spract of individFS torches.

w, A. Killing(HVSFS). S

ch, Nanostrublications,

e modeling the interacthe thermal athe particleal CFD soft

he HVSFSf the comban influenc

ng of an injeas the inc

ction of the rs. Howevedue to por particle

tion chambe

ania particlee for the sus

l include anate and detnd the subon the instbe accompvelop an any system’s

dual parame

ger, J. RaucSurf. Coat. Tructured MWilliam And

of ethanol tion mechanand flow fiee in-flight btware ANSY

process sbustion chace on the enection anglecrease of et

evaporatioer, this moarticle agg

e oxidation er, as show

es’ flow pathspension

n optimizatiailed descrstrate, whiabilities of lished.

nalytical mes capability eters on the

ch, New reTechnol. 20aterials—Pdrew Publis

evaporationnisms betw

elds of the rebehavior waYS-CFX 11

show that amber, invonergy balane modificatiothanol resid

on length, soodified injecglomeration

because ow in fig. 6.

h in the HV

ion processription of thech was simthe particle

ethodology funder var

e process, a

sults in Hig02 (2008) rocessing,

shing, Norw

n followed bween gas an

eaction gasas simulate.

the ethanoolving a conce of the Hon for suspdence time o that the cction angle, particle of their lon

VSFS combu

s of the HVe flow field mplified in es’ flow pa

for the anarious operaand finally,

gh Velocity

Properties,wich, NY, 20

by its non-pnd liquid ass were solveed by a Lag

ol evaporatooling effecHVSFS sysension injecin the com

cooling effe can also deposition

ng flow pa

ustion cham

VSFS thermin the free the presen

aths in the

lysis and prating conditto aid and

Suspensio

, and Appl002

premixed s well as ed by an grangian

tion and ct in the stem. An ction led

mbustion ect in the

lead to on the

aths and

mber by

mal spray jet area

nt work. HVSFS

rediction tions, to promote

n Flame

ications,




[3] Y. Lu and P.K. Liaw, The Mechanical Properties of NanostructuredMaterials, J. Metals, 2001, 53(3), p 31-35

[4] R.S. Lima and B.R. Marple, Superior Performance of High-Velocity Oxyfuel-Sprayed Nanostructured TiO2 in Comparison to Air Plasma-Sprayed Conventional Al2O3-13TiO2, J. Thermal Spray Technol., 2005, 14(3), p 397-404

[5] L.L. Shaw, D. Goberman, R. Ren, M. Gell, S. Jiang, Y. Wang, T.D. Xiao, and P.R. Strutt, The Dependency of Microstructure and Properties of Nanostructured Coatings on Plasma Spray Conditions, Surface Coatings Technol., 2000, 130, p 1-8

[6] R.S. Lima, and C.C. Berndt, Integrity of Nanostructured Partially Stabilized Zirconia After Plasma Spray Processing, Mater. Sci. Eng. A, 2001, 313, p 75-82

[7] H. Kassner, R. Siegert, D. Hathiramani, R. Vassen, and D. Stoever, Application of SPS for Manufacture of Ceramic Coatings, J. Th. Spray Tech., 2008, 17, p 115-123

[8] M. Gell, E.H. Jordan, Thermal Barrier Coatings Made by the Solution Precursor Plasma Spray Process, J. Thermal Spray Technol., 2008, 17, p 124-135.

[9] P. Fauchais, R. Etchart-Salas, Parameters Controlling Liquid Plasma Spraying: Solutions, Sols, or Suspensions, J. Therm. Spray Technol., 2008, 17, p 31–59.

[10] F-L. Toma, D. Sokolov, G. Bertrand, and C. Meunier, Comparison of the Photocatalytic Behavior of TiO2 Coatings Elaborated by Different Thermal Spraying Processes, J. Therm. Spray Technol., 2006, 15, p 576-581.

[11] F-L. Toma, G. Bertrand, and C. Meunier, Nanostructured Photocatalytic Titania Coatings Formed by SPS, J. Therm. Spray Technol., 2006, 15, p 587–592.

[12] R. Etchart-Salas, V. Rat, J.F. Coudert, and P. Fauchais, Influence of Plasma Instabilities in Ceramic Suspension Plasma Spraying, Thermal Spray 2007: Global Coating Solutions, B.R. Marple, M.M. Hyland, Y.-C. Lau, C.-J. Li, R.S. Lima, and G. Montavon Ed., 2007 (Beijing, China), ASM International, 2007, p 621-626.

[13] G. Bolelli, J. Rauch, V. Cannillo, A. Killinger, L. Lusvarghi, R. Gadow, Investigation of High-Velocity Suspension Flame Sprayed (HVSFS) glass coatings, Mater. Lett., 2008, 62, p 2772–2775.

[14] A. Killinger, M. Kuhn, R. Gadow, High Velocity Suspension Flame Spraying (HVSFS), a new approach for spraying nanoparticles with hypersonic speed. Surf. Coat. Technol. 201 (2006), p 1922-1929

[15] E. Dongmo, M. Wenzelburger, R. Gadow, Analysis and optimization of the HVOF process by combined experimental and numerical approaches. Surf. Coat. Technol. 202 (2008) 4470–4478

[16] Mingheng, L., Panagiotis D., Christofides. Multi-scale modeling and analysis of an HVOF thermal spray process. Chem. Eng. Sci. 60 (2005), 3649– 3669

[17] Yang, X., Eidelman, S., 1996. Numerical analysis of a high-velocity oxygen-fuel thermal spray system. J. Therm. Spray Technol. 5, 175–184

[18] T. Furuhata, S. Tanno, T. Miura, Y. Ikeda, T. Nakajima, Performance of numerical spray combustion simulation, En. Convers. Manage. 38 (10–13) (1997) 1111–1122

[19] A.R. Lopez, B. Hassen, W.L. Oberkampf, R.A. Neiser, T.J. Roemer, Computational fluid dynamics analysis of a wire-feed, HVOF thermal spray torch, J. Therm. Spray Technol. 7 (1998) 374–382

[20] Li, M., Shi, D., Christofides, P.D., Diamond jet hybrid HVOF thermal spray: Gas phase and particle behavior modeling and feedback control design. Ind. Eng. Chem. Res. 43 (2004)

[21] S. Prakash and W.A. Sirigano, Theory of convective droplet vaporisation with unsteady heat transfer in the circulating liquid phase, Intl. J. Heat Mass Transfer 23 (1980), 253-268

[22] Lih, W.C., Yang, S.H., Su, C.Y., Huang, S.C., Hsu, I.C., Leu, M.S., 2000. Effects of process parameters on molten particle speed and surface temperature and the properties of HVOF CrC/NiCr coatings. Surf. Coat. Technol. 133

[23] C. K. Westbrook and F.L. Dryer, chemicah kinetic modeling of hydrocarbon combustion, Prog. Energy Combust. Sci. 10 (1984), 1-57




numerical approach and optimization of the combustion and ...combustion chamber and in the nozzle,...

Documents