2009-wear-Slurry and cavitation erosion resistance of thermal spray coatings!

Wear267(2009)160–167Contents lists available at ScienceDirectWearj o u r n a l h o m e p a g e:w w w.e l s e v i e r.c o m/l o c a t e/w e arSlurry and cavitation erosion resistance of thermal spray coatingsJ.F.Santa∗,L.A.Espitia,J.A.Blanco,S.A.Romo,A.ToroTribology and Surfaces Group,National University of Colombia,Medellín,Colombiaa r t i c l e i n f oArticle history:Received9September2008Received in revised form9January2009 Accepted10January2009Keywords:Cavitation erosionSlurry erosionThermal spray coatingsWear mechanisms a b s t r a c tThe slurry and cavitation erosion resistance of six thermal spray coatings were studied in laboratory and compared to that of an uncoated martensitic stainless steel.Nickel,chromium oxide and tungsten carbide coatings were applied by oxy fuel powder(OFP)process and chromium and tungsten carbide coatings were obtained by high velocity oxy fuel(HVOF)process.The microstructure of the coatings was analyzed by light optical microscopy(LOM)and scanning electron microscopy(SEM),as well as by X-ray diffraction (XRD).The cavitation erosion resistance of the coatings was measured in a vibratory apparatus according to ASTM G32standard and the slurry erosion tests were carried out in a modified centrifugal pump in which the samples were conveniently placed to guarantee grazing incidence conditions,as well as in a high velocity jet erosion testing machine.The results showed that the slurry erosion resistance of the steel can be improved up to16times by the application of the thermally sprayed coatings.On the other hand,none of the coated specimens showed better cavitation resistance than the uncoated steel in the experiments.The main mass removal mechanisms observed in all the coatings submitted to slurry erosion were micro-cutting and micro-ploughing as well as detachment of hard particles.In cavitation erosion,OFP coatings showed brittle fracture and microcracking,and in nickel-based coatings some ductile deformation was also observed.In HVOF coatings,detachment of small particles led to coalescence of pores in WC/Co coatings while in CrC coatings the main wear mechanism was brittle fracture of particles.Published by Elsevier B.V.1.IntroductionMartensitic stainless steels are widely used in hydroelectric power plants due to their good properties such as corrosion resis-tance,weldability and moderate cavitation erosion and slurry erosion resistance[1].Slurry erosion problems are especially important during rainy seasons in hydroelectric power plants due to the increase in the number of solid particles impacting the surfaces,especially in sys-tems where the installation of an exhaustivefiltration process is not possible[2].Cavitation erosion problems are found in Pelton and Francis runners as well as in other parts of the turbine system as a result of rapid changes in pressure,which lead to the formation of small bubbles or cavities in the liquid.These bubbles collapse near the component’s surface at a high frequency in such a way that the elastic shock wave created is able to cause erosion of the material straight afterwards[3].It is desirable that reparation of the worn components can be accomplished without changing significantly the microstructure of∗Corresponding author.E-mail address:jfsanta@(J.F.Santa).the steel and the shape of the part,in addition to the expected pro-tective effect of the new surfaces against cavitation and erosion.A good way to improve the slurry and cavitation erosion resistance of the components is by the application of thermally sprayed coat-ings[4,5].The term thermal spray describes a family of processes that use chemical or electrical energy to melt(or soften)and accel-erate particles of a material which is then deposited on a surface [6].The quality of the coatings obtained by thermal spray tech-niques is related to the nature of the process and the processing parameters.In some countries hydroelectric power plants are the most important sources of energy and,in order to operate them effi-ciently,the parts in contact withfluids usually require special maintenance procedures including welding repairs and time-consuming heat treatments.On the other hand,thermal spray coatings are a good option to repair components and prevent excessive wear because during the deposition process no significant changes to the microstructure of substrates or excessive deformation are promoted.In this work,six thermal spray coatings were studied in labo-ratory in order to evaluate their resistance to cavitation and slurry erosion.The evaluation was done in order to decide if the coatings were a good option to repair worn parts of hydroelectric turbines0043-1648/$–see front matter.Published by Elsevier B.V. doi:10.1016/j.wear.2009.01.018J.F.Santa et al./Wear 267(2009)160–167161Table 1Processing parameters of coatings a .CoatingFeedstock materialsGas flow (10−4m 3s −1)(cfh)Air pressure (kPa)(psi)Coating rate (10−4kg s −1)(Lbs h −1)Standoff distance (10−3m)(in.)Travel velocity (10−2m s −1)(in.s −1)Density (kg m −3)Ni-1Ni (95%)Al (5%)particles O 2=3.7(35)CH 2=6.6(70)207(30)11(9)200(8)10(3.9)7.7Ni-2Cladded Ni,Cr,Mo,TiO 2=4.3(40)CH 2=5.8(60)241(35)7(10)200(8)13(5.1)7.7WC/Co–Ni WC/Co particles (46%)and Ni–Fe–Cr (54%)O 2=4.1(38)CH 2=5.8(60)207(30)6(13)75(3)10(3.9)13.2CrChromium oxide (100%)O 2=4.3(40)CH 2=5.8(60)76(11)4(10)200(8)8(3.1)5.6aEnglish units are shown along with SI units because all the indications in the equipment controls are given in English system.Fig.1.(a)Cavitation vibratory apparatus according to ASTM G32and (b)modified centrifugal pump for slurry erosion tests.which are submitted to cavitation and slurry erosion under grazing incidence.2.Experimental procedure 2.1.MaterialsThe coatings were applied onto ASTM A743grade CA6NM martensitic stainless steel (from now on,13-4steel),commonly used in turbine components,by OFP (Terodyn 2000gun)and HVOF processes.The deposition parameters chosen for OFP coatings are described in Table 1,while HVOF coatings were applied by a com-mercial manufacturer following proprietary deposition conditions.The 13-4steel was received in the as-cast condition and it was homogenized at 1050◦C for 1h and then air-cooled to room tem-perature.After that,the specimens were tempered at 620◦C for 1h and cooled down in air.The surfaces were prepared by sand blasting to a surface finish-ing of R a =5.4±0.5␮m.In some cases a nickel-rich bond coat was applied to improve the adhesive strength of the substrate-coating system.The coatings were applied on cylinders with 9.65mm in diameter and 9.65mm in height for slurry erosion testing (modi-fied centrifugal pump)and cylinders with 15.9mm in diameter and 6mm in height for cavitation erosion testing.The jet erosion tests were carried out using test samples with 45mm ×17mm ×5mm.The coated surfaces were machined in order to obtain samples with the same thickness in all cases.2.2.Microstructure and chemical characterizationThe microstructure of the coatings,feedstock materials and worn surfaces was analyzed in a thermoionic scanning electron microscope (SEM)and a light optical microscope (LOM).The char-acterization included measuring of volume fraction of pores and phases by digital image processing (DIP),localized chemical analy-sis with an EDS spectrometer coupled to the SEM,Vickers hardness and micro-hardness measurements and X-ray diffraction analysis(XRD)in a diffractometer with Cu K ␣radiation equipped with a 2D solid-state detector.The XRD patterns were used to identify the phases in the coatings and to understand the phase formation during coatings’build-up.2.3.Wear measurements2.3.1.Cavitation erosion testsCavitation erosion tests were performed using a vibratory appa-ratus according to ASTM G32shown in Fig.1a.Mass losses were measured every hour using a scale with 0.01mg resolution and each test had a duration of 6h.The results were converted to vol-ume losses in order to effectively compare the tested materials.The conditions of the test were as follows (Table 2).2.3.2.Slurry erosion testsThe slurry erosion tests were carried out in a modified centrifu-gal pump described elsewhere [7]in which the specimens were placed at the outlet of the device to ensure grazing incidence of the slurry (Fig.1b).The slurry was composed of distilled water and quartz particles with a mean diameter between 212and 300␮m (AFS 50/70)and the solids content was 10wt%.The mean impact velocity of the slurry was 10m s −1and the erosion resistance was determined from the mass loss results.Mass losses were measured every 30min by using a scale with 0.01mg resolution.The total duration of each test was 120min.Additional tests were performed in a jet erosion testing machine in which slurry composed of water and 0.4wt%of SiO 2particlesTable 2Testing parameters in cavitation erosion tests.Frequency 20±0.2kHz Amplitude 50±2.5␮m FluidDistilled water Temperature22±1◦CMass loss measurements Cleaning,drying and mass measurements every hour Total testing time6h162J.F.Santa et al./Wear 267(2009)160–167Fig.2.Schematic of jet erosion tests setup.(AFS 50/70)reached the surface of the samples with an impact velocity of 75m s −1at an incidence angle of 15◦(Fig.2).Only fresh slurry was used for each sample and the tests lasted 4min with measurement of mass losses every 1min.In both wear tests the coatings were tested under grazing incidence using the same abrasive particles at different impact velocities.The testing conditions are similar to those observed in some components of hydraulic turbines such as liners and runners.2.3.2.1.Analysis of worn surfaces.The worn surfaces were analyzed by LOM and SEM in order to identify the wear mechanisms and relate them to the mass loss results.3.Results and discussion3.1.Microstructure of coatings and substrateFig.3shows the microstructure of 13-4steel as well as of the coatings and feedstock materials.Fig.4shows the XRD patterns of the studied coatings,where all the crystalline phases were identified.The microstructure of the steel is composed of martensite (aver-age hardness 280HV 62,5)with 15%retained austenite and 4%delta ferrite (␦)[8].Fig.3.Microstructure of 13-4stainless steel,thermal spray coatings and feedstock materials.J.F.Santa et al./Wear267(2009)160–167163Fig.4.XRD patterns of the thermal spray coatings used in this work.Ni-1coating is composed of14.6±1.2%aluminum oxide particles(1533HV25gf,15s)and Ni splats(191HV300g,15s).DIP mea-surements reported2.9±1.2%porosity.XRD patterns revealed that the oxides do not have a crystalline system as a consequence of the high cooling rates during coatings build-up.[9].The average thickness of the as-sprayed coatings was650␮m.Ni-2coating is composed of Ni–Fe–Mo–W–Si particles(385 HV50gf,15s)and titanium oxide particles(701HV50gf,15s).DIP mea-surements reported4.4±1.5%porosity.As described for Ni-1,the oxides in the coatings do not have a crystalline system although the XRD pattern of feedstock powder revealed the presence of a crystalline titanium oxide(TiO2).The average thickness of the as-sprayed coatings was760␮m.WC/Co–Ni coating is composed of46±2%Ni–Fe–Cr(639 HV300g,15s)particles and54±3%of WC/Co and WC particles(1211 HV HV300g,15s)as reinforcement.The volume fraction of pores was estimated to13.7±3.8%by DIP of SEM images.The coating has partially crystalline Co4W2C carbides which were identified by a convex zone near40◦in XRD patterns(Fig.4).The average thickness of the as-sprayed coatings was820␮m.Cr coating is composed mainly of Chromium oxide Cr2O3(1853 HV300g,15s)with a porosity of30%.The as-sprayed surfaces had some cracks and the morphology of splats can be described asflower shaped[7].It is noteworthy that these splats reported very low values of cohesion among lamellas.The average thickness of the as-sprayed coatings was450␮m.The microstructure of WC/Co HVOF coating is composed offine tungsten carbides(WC)in a Co matrix with a porosity of2%,which is considerably lower than that measured in OFP ceramic coatings. The average thickness of the as-sprayed coatings was85␮m.CrC HVOF coatings showed a very heterogeneous thickness, which caused that in some areas the uncoated steel was actually exposed.The coatings microstructure is composed of Cr3C2and metallic Ni.Other peaks identified in the XRD pattern were related to the substrate as a consequence of X-ray penetration.The average thickness of the as-sprayed coatings was40␮m.Generally speaking,the microstructure of the OFP coatings is related to what is commonly expected for this deposition process as the particles impact the surfaces at low temperatures and low velocities,which cause defects such as pores,partially melted par-ticles and others.Regarding the HVOF coatings,porosity levels are significantly lower than those observed in OFP samples even with presence of elevated volume fractions of ceramic phases with high melting points.This is coherent with the nature of HVOF process, in which a high amount of kinetic energy is given to the particles to reach velocities up to800m s−1.3.2.Cavitation erosion testsFig.5shows the volume losses measured after cavitation erosion tests,as a function of testing time.It can be seen that the uncoated stainless steel(13-4)showed the best cavitation resistance of all the samples tested.Among the coatings,the best behavior was reported by CrC(HVOF)coating followed by WC/Co–Ni.Although none of the tested coatings showed incubation period(mass losses were mea-sured from the beginning of the tests),the Ni-2and CrC coatings had uniform wear rate during the entire test.In the specific case of the CrC HVOF coating,it is believed that its good behavior is the result of insufficient coating thickness as SEM observations revealed that the surface of the stainless steel was reached soon after the begin-ning of the tests.In all the other cases,SEM examination of worn surfaces confirmed that the substrate was not exposed.The volume losses reported by Cr coating(omitted in Fig.5)were extremely high and prevented any possibility for this coating to be considered a good option to protect against cavitation.How-ever,a reduction of porosity could probably result in an increase of164J.F.Santa et al./Wear267(2009)160–167Fig.5.Cumulative volume losses in cavitation erosion tests performed according to ASTM G32standard.the cavitation resistance of these coatings due to their high hard-ness.3.2.1.Analysis of worn surfaces and identification of wear mechanisms in cavitation erosionThe examination of worn surfaces in stereoscopic microscope revealed different features for each material(Fig.6).WC/Co–Ni coating showed detachment of large particles of WC/Co from the surface,which caused high volume losses and was related to pre-vious cracks formed during the growth of the coating as a result of high cooling rates and differences among the thermal expansion coefficients of the phases in the microstructure(Fig.7a).In Ni-rich coatings,on the other hand,the surfaces showed typical plas-tic deformation marks and no detachment effects were observed although SEM inspection of the as-sprayed surfaces showed that aluminum oxides were also cracked from the deposition process (Fig.7b).The main mass removal mechanism for13-4stainless steel was fatigue delamination due to accumulation of plastic deformation, which was revealed by the presence of deformation bands in highly deformed areas(Fig.8a).Ni-1and Ni-2coatings showed fracture of oxides and the formation of small cavities in Ni splats(Fig.8b and c).WC/Co particles of WC/Co–Ni coatings showed sharp surfaces as consequence of propagation of cracks(Fig.8d).In WC/Co HVOF coatings the removal mechanism was coales-cence of pores,which were created as a result of a process that began with the erosion of the Co binder followed by detachment of small particles of WC(Fig.8e).In CrC HVOF coatings,portions of the coating detached from substrate due to adhesion problems and the remaining areas showed ductile deformation of Ni particles and detachment of CrC(Fig.8f).3.3.Slurry and jet erosion testsFigs.9and10show the cumulative volume loss after slurry ero-sion tests carried out in modified centrifugal pump and jet erosion device,respectively.Generally speaking,no running-in period was observed probably as a consequence of the high severity of the tests.A significant improvement in the erosion resistance of the 13-4stainless steel was obtained by application of the OFPcoatings, Fig.6.Worn surface observed in stereoscopicmicroscope.Fig.7.Surfaces of as-sprayed coatings showing cracks formed during the deposition of the layers.J.F.Santa et al./Wear 267(2009)160–167165Fig.8.Worn surfaces of thermal spray coating tested in cavitation erosion.Ni coatings showed pits formation in splats.Hard phases in WC/Co and CrC coatings showed brittlebehavior.Fig.9.Cumulative volume losses in slurry erosion in modified centrifugal pump.Impact velocity 10m s −1.Solids content 10wt%SiO 2particles.Fig.10.Cumulative volume losses in jet erosion tests.Impact velocity 75m s −1.Solids content 0.4wt%SiO 2particles.being the WC/Co–Ni the best option in all cases.By considering the total volume loss measured at the end of the tests,under low veloc-ity impingement conditions (centrifugal pump)the improvement factor was 5:1while in the jet erosion tests it increased to 16:1.3.3.1.Analysis of worn surfaces and identification of wear mechanisms in slurry and jet erosion testsThe surface of the coated specimens tested in the modified cen-trifugal pump showed two distinct regions when observed in the stereoscopic microscope (Fig.11):the first one was typical of severe erosion conditions with deep and abundant grooves in metallic areas and microcracks in ceramic phases,although the erosive par-ticles did not reach the steel substrate in any case.In the second region mild wear was predominant since no significant changes were observed at the worn surfaces with respect to the as-sprayed condition.In both zones,the wear patterns were typical of erosion under grazing incidence (with micro-cutting and micro-ploughing marks)and also beach marks were evident (Fig.11a).In all cases,during the testing time the substrate was not reached by the erodent particles so the wear resistance of the coatings could be effectively compared.In some regions of the Cr coatings the lack of adhesion among lamellaes caused detachment of the coating (Fig.11c).This behavior is supposed to be a consequence of the low energy deliv-ered to the particles during coating formation and it is a limitation of the OFP technique.In jet erosion tests,on the other hand,the wear patterns observed in the stereoscopic microscope varied depending on the tested material (Fig.12).The worn surface of WC/Co–Ni coatings exhibited both severe wear (marked by the arrow in Fig.12a)and mild wear regions,while Ni-2coating and 13-4stainless steel (Fig.13d)showed homogeneous patterns along the entire worn area.The examination of the worn surfaces by SEM revealed details on the wear mechanisms and their relationship with the microstruc-ture for both erosion tests.Titanium oxide particles cracked and detached from Ni-2coating’s surface (Fig.13a and f)leaving the Ni matrix unprotected.As a consequence of detachment of hard phases,the erodent particles caused micro-cutting and micro-ploughing.Although this mechanism was similar in both erosion166J.F.Santa et al./Wear267(2009)160–167Fig.11.Wear mechanisms in slurry erosion tests(centrifugal pump).Fig.12.Wear mechanisms in jet erosion tests.Fig.13.Worn surfaces of thermal spray coatings tested in slurry and jet erosion.J.F.Santa et al./Wear267(2009)160–167167tests,the degree of plastic deformation of single splats was signifi-cantly higher in the jet erosion tests.Cr coatings showed high mass losses due to brittle fracture of lamellas as a result of intersection of cracks and delamination effects,with no evidence of plastic deformation or ploughing marks (Fig.13c).Very low values of adhesion/cohesion strength were found(4.9±0.6MPa)after measurements carried out according to ASTM C633standard,which verified the prejudicial effect of the brittle behavior of the microstructure.WC/Co coatings showed micro-cutting of Ni splats as well as brittle fracture of WC/Co lamellas(Fig.13b and e).Also,a number of small cavities,probably formed by a synergistic corrosion–erosion mechanism[10],were observed in WC/Co particles.However,since Ni particles are expected to be more sensitive to corrosion than WC/Co particles,this hypothesis needs to be studied in detail and verified with the aid of critical experiments.4.ConclusionsSlurry and cavitation erosion tests of six thermal spray coat-ings and a bare13-4stainless steel were performed in laboratory in order to determine their wear resistance and identify the main wear mechanisms acting on the surfaces.The main conclusions of this work are-The cavitation erosion resistance of the studied coatings was lower than that of uncoated stainless steel in all cases.This was attributed to high porosity,low cohesion between splats and pre-existent cracks in the coatings.To overcome these drawbacks a number of options are available,i.e.changing the application technique,optimizing some process parameters,among others. -The slurry erosion resistance of the13-4stainless steel was improved by deposition of thermally sprayed coatings with differ-ent microstructures and chemical compositions,being obtained the best results with a WC/Co–Ni coating.-The main mass removal mechanisms observed in thermal spray coatings after cavitation tests were brittle fracture of hard phases and fatigue of ductile areas.In slurry and jet erosion tests the dom-inant wear mechanisms were micro-cutting and micro-ploughing of softer phases and,in some cases,detachment of hard particles.AcknowledgmentsThe authors thank the Empresas Públicas de Medellín E.S.P for their technical support.Financial support provided by Colciencias-EPM-UNAL project no.20201005975is also acknowledged. References[1]H.J.Niederau,in:K.H.Miska(Ed.),State of Development of Soft MartensiticStainless Chromium Nickel Steels,Climax Molybdenum Company,1997.[2]T.Kachele,Recent research results on predicting and preventing silt erosion,in:Proceedings of the First International Conference on Silting Problems in Hydropower Plants,October1999,India,1999.[3]S.C.Li,Cavitation of Hydraulic Machinery,Imperial College Press,2000.[4]K.Sugiyama,et al.,Slurry wear and cavitation erosion of thermal-sprayed cer-mets,Wear Volume258(2005)768–775.[5]J.S.Boy,et al.Cavitation-and Erosion-Resistant Thermal Spray Coatings USAC-ERL Technical Report97/118,July1997.[6]L.Pawlowski,The Science and Engineering of Thermal Spray Coatings,JohnWiley&Sons Ltd.,London,1995.[7]J.F.Santa,J.C.Baena,A.Toro,Slurry erosion of thermal spray coatings and stain-less steels for hydraulic machinery,Wear263(2007)258–264.[8]H.Pacheco,Phase Transformation Caused by a Post Weld Heat Treatment inASTM A743CA6NM Stainless Steel,National University of Colombia,2008. [9]F.Longo,Coating operations,in:Handbook of Thermal Spray Technology,ASMInternational,2004,pp.120–127.[10]M.M.Stack,T.M.Abd El-Badia,Some comments on mapping the combinedeffects of slurry concentration,impact velocity and electrochemical potential on the erosion–corrosion of WC/Co–Cr coatings,Wear264(2008)826–837.。